SKELLIG MICHAEL, CO. KERRY:
THE MONASTERY AND SOUTH PEAK
Archaeological stratigraphic report: excavations 1986–2010
Edward Bourke, Alan R. Hayden, Ann Lynch
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Skellig Michael, Co. Kerry: the monastery and South Peak
Archaeological stratigraphic report:
excavations 1986–2010
(E338; 90E34; 93E195)
Edward Bourke, Alan R. Hayden, Ann Lynch
With contributions by:
Ryan Allen, Teresa Bolger, Claire Cotter, Julie Franklin, Sheila Hamilton-Dyer, Michael Kenny, Linda
Lynch, Clare McCutcheon, Rosanne Meenan, Emily Murray, Joseph Norton, Michael O’Connell,
Lorna O’Donnell, Michael O’Sullivan, Anette Overland, Eileen Reilly, Grellan D. Rourke
Drawings by Patricia Johnson (artefacts) and Conor McHale (site drawings)
Department of Arts, Heritage and the Gaeltacht
© Government of Ireland 2011, subject to the moral rights of the individual authors
as established under the Copyright and Related Rights Act 2000.
All rights reserved. No part of this book may be reprinted or reproduced or utilised
in any electronic, mechanical or other means known or hereafter invented, including
photocopying or recording, or otherwise without either the prior written consent
of the copyright holders or a licence permitting restricted copying in Ireland issued
by the Irish Copyright Licensing Agency Ltd, The Writers’ Centre, 19 Parnell
Square, Dublin 1.
General editor: Ann Lynch
Designed and typeset by Wordwell Ltd
Copy-edited by Emer Condit (Wordwell Ltd)
CONTENTS
Acknowledgements
List of tables
Section 1. INTRODUCTION
1.1 Introduction
1.2 Summary background information
1.3
The early history of Skellig Michael
1.4 A summary of the later history
1.5 Background to the conservation works
Section 2. THE EXCAVATIONS
2.1 Introduction
2.2 The monastery
2.2.1 The small oratory terrace and ledge
2.2.2 The large oratory
2.2.3 South entrance 1 (inner enclosure) and leacht area
2.2.4 South entrance 2 (inner enclosure)
2.2.5 East entrance, phase 3 (inner enclosure)
2.2.6 Cistern 3
2.2.7 The lower monks’ garden
2.2.8 The uppper monks’ garden
2.2.9 Structure at base of east steps
2.3
2.3.1
2.3.2
2.3.3
2.3.4
2.3.5
2.3.6
2.3.7
The South Peak
Introduction
The influence of the geology of the South Peak
The lower part of the route
The primary route to the South Peak
The final route to to the South Peak
The outer terrace
Areas where stone was quarried and won
2.4
2.4.1
2.4.2
2.4.3
Newly discovered steps and access routes
New features on the east steps
The steps to the south-east landing
A new route to the monastery
Section 3. THE FINDS
3.1 Introduction
3.2 The medieval pottery
3.3 The post-medieval pottery
3.4 The clay pipes
3.5 Crucifixes
3.6 Coins
3.7 The small finds
3.7.1 Copper alloy
3.7.2 Lead
3.7.3
3.7.4
3.7.5
3.7.6
3.7.7
3.7.8
Iron
Bone and antler
Stone
Glass
Leather
Miscellaneous
Section 4. THE RADIOCARBON DATES
Section 5. THE HUMAN REMAINS
Section 6. PALAEOENVIRONMENTAL ANALYSES
6.1 Introduction
6.2 Plant macrofossils
6.3 Charcoal and wood
6.4 The insect remains
6.5 Past environment and land use on Skellig Michael in the monastic period—the pollen analytical evidence
Section 7. THE FAUNAL REMAINS
7.1 The mammal bones
7.2 Bird and fish bones
7.3 The marine molluscs
Section 8. INTERIM CONCLUSIONS
Bibliography
Appendix I. Finds catalogue
Appendix II. Concordance of context numbers
Appendix III. Catalogue of disarticulated human remains
ACKNOWLEDGEMENTS
Edward Bourke and Ann Lynch gratefully acknowledge
the assistance of the following:
Grellan D. Rourke, senior conservation architect, has
been in charge of the conservation works on Skellig
Michael for over 30 years. His ongoing commitment to
and assistance with all aspects of the archaeological work
have ensured an integrated approach to the project. His
friendship and collegiality are much appreciated.
Working conditions on Skellig Michael can be
challenging in the extreme, and the workmen from the
National Monuments Service depot in Killarney are the
true heroes of the project. Sincere thanks to all of those
who have toiled on the site since the early 1980s. Patrick
O’Shea, site foreman, deserves special mention for his
unfailing assistance in all matters. The District Works
Manager of the Killarney district has overall
responsibility for servicing the site and ensuring that
everything
from
excavation
equipment
to
accommodation is provided for the archaeologists. Our
grateful thanks to Michael Keane†, Pat Cremin† and the
present incumbent,Terry Murphy. Over the years several
archaeologists have worked on the excavations and their
efforts, which have contributed to this report, are
gratefully acknowledged: Claire Cotter, Paddy O’Leary†,
Lee Snodgrass, Linda Lynch and Anke Halmschlag. Joss
Lynam†, mountaineer and engineer par excellence, not
only dealt with the many engineering challenges
presented by the site but also helped to resolve logistical
problems relating to the excavations. Our thanks also to
the guides who work and reside on the island and who
have provided assistance over the years.
Accessing Skellig Michael can be treacherous at
times and we are grateful to the local boatmen who have
ensured our safe transport to and from the island. In
particular we wish to thank Dermot† and Owen Walsh,
who have been the official project boatmen for many
years and who with their boat, the Agnes Olibhéar, have
never let us down. The Irish Helicopter pilots have also
transported personnel and equipment to and from the
island and we are grateful to the Commissioners of Irish
Lights for allowing us to tie into their schedule of flights.
In the years before automation of the lighthouse took
place (April 1987) the lighthouse-keepers provided
valuable assistance to the project as well as muchappreciated company when inclement weather
conditions cut us off from the mainland for days at a
time. In the days before mobile phones, their radio
contact with the outside world also provided a vital safety
net. Richard Foran, attendant lightkeeper, Skellig Rock,
continues to provide vital assistance in many ways.
The precipitous terrain on Skellig Michael requires
stringent safety measures to be put in place for all
workers. The Kerry Mountain Rescue Team provided
the safety regime for the excavations carried out in the
early 1980s and this role was subsequently taken over by
Irish Rope Access and Safety Consultants Ltd. It is thanks
to these groups that no worker has suffered injury in the
course of the excavations.
We are grateful to the specialists who have
contributed to this report and who have discussed their
results at length with us. The artefact drawings were
completed by Patricia Johnson (AHG) and the site
drawings by Conor McHale, and we are grateful to both
for their patience and forbearance. Many of the
photographs, especially the aerial shots, were taken by
Con Brogan of the Photographic Section of DAHG,
who has been recording the project since the early days
and to whom we are most grateful.
Alan Hayden wishes to acknowledge the assistance of
the following:
The works on the South Peak were undertaken under
the overall direction of Grellan D. Rourke. Because of
the nature of the site and the work, excavation and
conservation were undertaken in close cooperation at all
times. Grellan at all times unselfishly and generously
shared his great knowledge and experience of Skellig and
he can justifiably be proud of what was achieved in such
a difficult working environment. Many of the
conclusions reached were only arrived at as a result of
our constant discussions about the site. The most recent
OPW crew on the island—Pat O’Shea, Micheál
O’Sullivan, Tom Kerrisk, Mike Connors and Johnny
Lyne—also cannot be thanked enough. They were one
of the most hard-working, resilient, dedicated and
cheerful crews with whom I have had the fortune to
work. The job could not have been done without their
skill, knowledge and fortitude. Very special thanks to
Colin McGorlick and the other Irish Rope Access
supervisors, who were responsible for our safety at all
times and who also helped in innumerable ways. The
knowledge and observations of Michael O’Sullivan, the
project geologist, were always generously shared and
were often vital to the interpretation of the remains on
the Peak. Thanks to Edward Bourke (who began the
excavations on the South Peak) and Dr Ann Lynch of
the National Monuments Service for their support and
help. Con Brogan’s (DAHG) photographs were very
important to the work on the Peak, and he and Tony
Roche (DAHG) are also due thanks for helping to source
some of the illustrations used.Terry Murphy and the staff
of the OPW depot in Killarney are due thanks for their
logistical support.Thanks also to the various cameramen
and soundmen who helped to record the works on site.
Owen Walsh brought us and our gear to and from the
island safely, but not always dry, on the Agnes Olibhéar.
Seán Obida undertook the three-day EDM survey of
part of the Peak that allowed us to tie in some of the
excavated areas. Thanks also to the pilots of Irish
Helicopters and Richard Foran of the Lighthouse
Service for help and transport. On land, Conor McHale
compiled most of the finished plans used in this report.
They were based on the site surveys and plans made by
Alan Hayden and Grellan Rourke. I am also grateful for
the support of Ann Hayden, Brian Hayden and Robin
Hayden, which allowed me to spend time on Skellig.
LIST OF TABLES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Clerics of Skellig Michael recorded in the annals.
Excavations carried out on Skellig Michael.
Medieval pottery identifications.
Post-medieval pottery identifications.
Radiocarbon dates from the lower monks’ garden.
Radiocarbon dates from the leacht area, the east
entrance and the small oratory terrace.
Distribution of human skeletal remains.
Minimum number of individuals.
Comparison of average male stature from Skellig
Michael with contemporary populations.
Prevalence of observed dental diseases in all
observable human teeth.
Adult cranial measurements.
Adult post-cranial measurements.
Juvenile skeletal measurements.
Plant remains recorded from Skellig Michael.
Samples analysed for insect remains.
Basic habitat statistics for insect remains—MNI
counts and percentage presence of each habitat
group.
List of sites used in ordination of insect
assemblages.
Charcoal and wood identifications.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
Insect species list.
Pollen analytical data.
Distribution of mammal species (NISP) by phase.
Distribution of mammal species (MNI) by phase.
Distribution of sheep/goat and goat elements by
phase.
Element distribution for seal bones.
Metrical data for cattle, goat, sheep/goat and sheep.
Tooth eruption and tooth wear data for sheep/goat
and goat.
Sample numbers, location and phasing for
features/contexts that produced animal bone
and/or shell.
Bird and fish bone condition: counts and
percentages.
Bird species: context totals.
Manx shearwater and auk: element counts.
Other birds: element counts.
Butchery: species anatomy.
Fish species: context totals.
Fish species: element counts.
Range and number of marine molluscs by context
and phase.
1. INTRODUCTION
1.1 INTRODUCTION
The ongoing conservation works programme at the early medieval monastic site on Skellig Michael
commenced in 1978 and has continued each summer season since then. The first season’s work was in
response to the collapse of a section of retaining wall to the west of St Michael’s Church within the
monastery, and shortly thereafter work focused on the repair of the south steps, the main access route to
the monastery. Survey work began at this time and the first archaeological intervention took place in
1980, with excavations proper commencing in 1986 and continuing almost every season until 2010 (see
Section 2 for full details). Over the years the archaeological work ranged from monitoring and supervision
to full excavation and, because the scope of the archaeological work was determined by and large by the
conservation works programme, investigations were focused on the monastery and associated structures
and the South Peak. In 2010 survey and conservation works commenced on the lighthouse road; once
this is completed, a programme of conservation works will be carried out at the old (disused) lighthouse.
This report (which is to be published in pdf format on the website www.worldheritageireland.ie
and which will be lodged in the state archaeological archives) is the full account of all the archaeological
works carried out within the monastery and South Peak structures. It contains stratigraphic details with
supporting plans, sections and photographs; a finds catalogue with descriptions and interpretations by
relevant specialists; the raw data and results of palaeoenvironmental analyses and the osteoarchaeological
analyses of human remains; a full account of the faunal remains; and the radiocarbon dates. The report
concludes with an interim statement of the site’s significance. The purpose of this report is to make
available the details of the archaeological works undertaken to date, in advance of the planned
multidisciplinary publication programme.
It is intended to produce a series of publications that will integrate the archaeological results with
those of the conservation works programmes and related projects. Work has commenced on these
publications and it is envisaged that a number of separate volumes will incorporate the following:
• The historical and documentary background of Skellig Michael. This will cover accounts from the
early historic period up to the surveys carried out in the 1950s. A research project looking in detail
at the lighthouse period of occupation will be published at a later stage in conjunction with the
results of the works that have yet to be carried out to these later structures.
• The architecture, archaeology and conservation of the monastery and associated structures.This will
include detailed surveys and descriptions of the monastic structures and a full account of the
engineering, structural and conservation works undertaken. The results of the archaeological
excavations will be incorporated with those of other related projects, e.g. water management.
• The architecture, archaeology and conservation of the South Peak structures. This will incorporate
a full record of all structures discovered and their conservation, together with the results of excavations
carried out.
• A study of the geology of the island and its influence on the form of the monastic settlement.
• An in-depth discussion of the monastery, its role in early medieval Ireland and its European context.
All excavations described in this report have been carried out by, or on behalf of, the National
Monuments Service (NMS).The NMS was part of the Office of Public Works (OPW) until 1996, when
1
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Tralee
N
Dingle Peninsula
Blasket
Islands
Killarney
Dingle Bay
0
10
20
30
40
50
kilometres
Valentia
Island
Iveragh
Peninsula
Ballinskelligs
Skellig
Michael
re
ma
n
Ke
er
Riv
Beara Peninsula
ay
ry B
t
Ban
Skellig Michael
N
Ill. 1.1—
Location of
Skellig Michael
and contour
map of island,
showing main
features.
Blind Man's Cove
Boat Landing
Blue Cove
North
Steps
186 m
North
Landing
East Steps
Monastery
South Peak
218 m
ad
Christ's
Saddle
e
us
Ro
tho
gh
Li
Cross Cove
Lighthouse Ruins
South Steps
South Landing
Seal Cove
Blue Man’s Rock
Lighthouse
0
100
200
metres
2
300
INTRODUCTION
an integrated Heritage Service was established in the Department of Arts and Culture. In 2003, however,
the different elements of the Heritage Service were reassigned to various government departments; at
the time of writing (2011) the architectural component of the NMS (responsible for the conservation
and management of all monuments in state care) is in the OPW and the archaeological service is part of
the Department of Arts, Heritage and the Gaeltacht (DAHG).
1.2 SUMMARY BACKGROUND INFORMATION
12.1 LOCATION AND
LEGAL STATUS
The island of Skellig Michael (townland: Sceilg Mhichíl) lies 11.6km off Bolus Head, the westernmost tip
of the Iveragh peninsula, Co. Kerry (SMR: KE 104A-001; National Grid Reference 024812 060654)
(Ill. 1.1). The island, which is approximately 21.9 hectares in area, is owned by the Minister for Arts,
Heritage and the Gaeltacht on behalf of the Irish people, with the exception of the lower (working)
lighthouse and its curtilage, the helipad and adjacent store. Skellig Michael is a National Monument in
state ownership, the preservation of which is a matter of national importance by reason of its historical,
architectural, artistic or archaeological interest.
The entire island was inscribed on the UNESCO World Heritage List in 1996 in recognition of the
outstanding universal significance of its cultural landscape and the importance of its protection to the
highest international standards.
1.2.2 GEOLOGY AND TOPOGRAPHY
Michael O’Sullivan
Skellig Michael represents one of the most westerly exposures of Devonian (Old Red Sandstone) rocks
in western Europe. These rocks are sedimentary in origin and were deposited during the upper part of
the Devonian period, between 360 and 374 million years ago. During this time Ireland, as part of a larger
continental land mass, was situated south of the equator. A crustal depression or trough, known as the
Munster Basin, existed in southern Ireland at this time, allowing the accumulation of a great thickness
of sediment.The basin was bordered by mountainous or upland areas to the north and south.The 200m
of sediments exposed on Skellig Michael were deposited in alluvial and fluvial environments in this basin.
The present outcrop pattern of the Devonian (Old Red Sandstone) is due to subsequent structural
deformation.This compression or folding of the rocks occurred during the Hercynian Mountain-building
period, approximately 300 million years ago.The structure of Skellig Michael is characterised by a major
open trough-shaped fold (syncline) developed about an axis that plunges to the east. Conjugate joint
sets (criss-crossing joint sets resulting in a diamond fracture pattern) are generated symmetrically about
this axis, while an intensive cleavage fabric parallels the axis orientation. It is this relationship between
fine-grained sedimentary rocks and planes of weakness such as bedding, cleavage and jointing that allows
for small tool workability in the main.
The topography of Skellig Michael (Ills 1.1, 1.2), its iconic twin peaks and intervening valley (Christ’s
Saddle) is entirely controlled by bedrock geology. A major north–south-trending fault (running from
Blue Cove in the north to Washerwoman Rock in the south) is expressed today as the valley between
the peaks. The bedrock adjacent to this fault zone is dislocated, brittle and friable and erodes out more
easily than the surrounding bedrock. It is this geological feature that underlies the saddle-like island
outline we see today.
3
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Landing
ps
t
North Steps
s
Ea
South Peak
on
M
Ch
ris
t’s
Sa
dd
Lighthouse
(Disused)
le
South Steps
Lighthouse
Ill. 1.2—LiDAR image of the island of Skellig Michael.
4
y
er
t
as
e
St
INTRODUCTION
1.2.3 THE
MONASTIC SETTLEMENT
Introduction
There are two separate elements to the monastic settlement on Skellig Michael: an extensive and wellpreserved monastery constructed just below the top of a high, sloping rock platform on the east side of
the island and a range of structures constructed on ledges high on the South Peak.
Three long flights of steps lead up to the monastery from three different landing places. The
monastery consists of an inner enclosure containing two oratories, a mortared church, seven beehive
cells and the remains of a ‘latrine’, water cisterns, a cemetery, leachta, crosses and cross-slabs. Two large
terraces, referred to as the upper and lower monks’ gardens, comprise the outer enclosure. High retaining
walls support all the terracing upon which everything is constructed (Ills 1.3a, 1.3b).
On the other side of the island, rock-cut steps and ledges lead up to the structures on the South
Peak. They comprise a series of platforms, traverses, enclosures and terraces daringly constructed on
quarried ledges just below the peak (Ill. 1.4). The oratory terrace still retains its original features: an
oratory, altar, leacht, bench, water cisterns and a possible shrine. Crosses and a cross-slab were also found
on the South Peak.
Access to the monastery
There are three long flights of steps (east, south and north steps) that lead to the monastery and are part
of a possible succession of routes, with traces of other, possibly earlier steps being discovered in places as
survey work proceeds (Ill. 1.1).The monks used three different landing places, depending on the prevailing
weather and sea conditions at the time of voyage.The basal sections of the three stairways were rock-cut,
with the steps constructed of drystone masonry once they reached a level where stormy seas could no
longer reach them and cause damage.
The base of the east steps was blasted away by the lighthouse-builders when they constructed the
pier and the lighthouse road in the 1820s. Above this level the steps have been conserved and are in very
good condition. The remains of a structure, which may have sheltered a boat or housed provisions, are
located adjacent to these steps.
The north steps were used extensively by the lighthouse-builders. The lower rock-cut section has
been very eroded by the action of the sea and a parapet was added at the lowest section of the drystone
steps, which are in one long, continuous flight. These steps have been repaired, but owing to collapse on
the very steep ground it has not been possible to recover all of the steps and some sections have been
ramped. Neither the east nor the north steps are accessible to the public.
The south steps are used by visitors today and are accessed from the lighthouse road. They join the
north steps at Christ’s Saddle and continue as one flight up to the monastery. There are traces of other
structures associated with these steps, including the remains of walling which may belong to terracing, a
substantial prayer or pilgrimage station and lighthouse features.
The outer enclosure
Access to the monastery is via the outer enclosure, which comprises two large terraces known as the
upper and lower monks’ gardens. The present entrance into the upper monks’ garden is not original and
was probably constructed in the early nineteenth century. This terrace would have provided a suitable
area for cultivation as it is south-facing and well shielded from the elements. Excavation in the lower
monks’ garden revealed the remains of an early cell (Cell G) surrounded by paving, indicating that at
some stage in its history it could not have been used solely for cultivation (see Section 2.2 below). Part
of this terrace is very steep but it is possible that small cultivation terraces were located here.
5
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 1.3a—LiDAR image of the monastery using a sky-view factor processing algorithm.
6
INTRODUCTION
18
1
15
16
14
13
9 4
2
12
3
19
11
17
7
5
6
10
20
8
1. Small oratory terrace
2. Large oratory
3. South entrance 1, inner enclosure
and Leacht area
4. East entrance, inner enclosure
5. South entrance 2, inner enclosure
6. Cistern 3
7. Lower monks’ garden
8. Upper monks’ garden
9. Monks’ graveyard
10. East steps
21
Ill. 1.3b—
Annotated
LiDAR image of
the monastery.
11. Cell A
12. Cell B
13. Cell C
14. Cell D
15. Cell E
16. Cell F
17. Cell G
18. Small oratory
19. Saint Michael’s Church
20. South entrance, outer enclosure
21. Guesthouse
7
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Outer Terrace
Upper Traverse
y
ll
u
G
Small Enclosure
Needle’s Eye
Lower Traverse
Garden
Terrace
Oratory Terrace
mC
p fro
u
Way
ddle
’s Sa
hrist
Ill. 1.4—Aerial photograph of south peak indicating principal features.
8
INTRODUCTION
The enclosure walls
One of the most imposing elements of the monastery is its retaining walls.The monks used these massive
walls to create terraces upon which to build and to provide shelter from the prevailing winds. There is a
long history of collapse of these walls, even during the monastic occupation.
The east retaining wall has three phases of construction representing repeated collapse and rebuilding.
The retaining wall of the small oratory terrace is largely original in spite of considerable movement over
the centuries.The long, south-facing wall of the inner enclosure comprises at least two phases of monastic
construction, repair during the lighthouse occupation in the early nineteenth century (including the
construction of a dry toilet for the construction workers) and a late nineteenth-century construction.
Minor conservation work was carried out on this wall in the 1970s. The long, south-facing outer
enclosure wall has also suffered repeated collapse. Most of the retaining wall holding the upper monks’
garden had collapsed and was rebuilt during the current conservation works programme. The retaining
wall of the lower monks’ garden is almost fully original at its western end, but the eastern section had
collapsed and was reconstructed during the current programme of conservation.
The entrances
There are two entrances into the outer enclosure. The earlier one (south entrance 1, outer enclosure), at
the top of the east steps and leading into the lower monks’ garden, has been repaired but is no longer in
use.The later one, leading into the upper monks’ garden, probably dates from the early nineteenth century
and is the one currently used by visitors.
There are three entrances into the inner enclosure. The earliest (east entrance, phase 3), leading
through the east retaining wall, is a short distance from the east steps. This was abandoned after serious
collapse at an early stage in the history of the monastery. The monks then constructed an entrance on
the south side (south entrance 1, inner enclosure) from the lower monks’ garden, which is no longer in
use and which pre-dates the large oratory. The last entrance to have been built by the monks (and
modified in the nineteenth century) is accessed from the upper monks’ garden and is the one still in use
today (south entrance 2, inner enclosure).
The inner enclosure
The large oratory
This oratory is of the usual inverted-boat shape, with the door in the west wall (Ill. 1.5). It is built of
stones of moderate size laid in horizontal courses. Internally it is rectangular (measuring 3.45m by 2.35m)
up to a height of 2.3m and then develops into an elongated dome. The internal walls show signs of later
limewashing. There is a small rectangular window in the east wall. On either side of the door, which has
inclined jambs, the wall is faced inside with two large vertical slabs. The walls of the oratory are about
1.2m thick. Externally, there is a cross of white quartz inserted into the stones of the wall above the
doorway, but this is a later addition.
The small oratory
This oratory is built on a small artificial terrace at the eastern end of the inner enclosure (Ill. 1.6). The
interior is of superior construction to the large oratory and measures 2.4m by 1.8m by 2.4m in height.
There is a small rectangular east window. The door has inclined jambs. The corners are articulated
throughout, both inside and out. This structure is of later date than the large oratory.
Nearby are the remains of a beehive cell which has been interpreted as a latrine—possibly a unique
structure in the context of Irish monasteries of this date.
9
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 1.5—The
monks’
graveyard with
the large
oratory behind
(Con Brogan,
DAHG).
Ill. 1.6—The
small oratory
terrace viewed
from south
(Con Brogan,
DAHG).
The dwelling cells (Ill. 1.7)
Cell A is the largest and clearly had a communal function. It is particularly imposing and has very thick
walls, with several offsets to facilitate construction.The walls, 1.8m thick at the base, are built of small flat
stones and there are a number of projecting stones on the outside at the upper level, again a construction
feature. The door has a double lintel, inside and out, and the floor retains most of its original paving.
The internal space measures 4.6m by 3.8m and the cell is 5m high. It is subrectangular at floor level,
10
INTRODUCTION
Ill. 1.7—The
cells viewed
from west
(Con Brogan,
DAHG).
evolving into an ovoid shape about 1m above ground level. There are two cupboards and projecting
stone pegs (probably used for storage purposes). Uniquely, there are window openings, one of which
focuses on the South Peak and another on the Little Skellig.
Cell B, to the east of Cell A, is built of very carefully worked large stones but there are no projecting
stones inside or outside. The interior is square in plan, measuring 2.75m by 2.75m by 3m high. The
corners are well defined and the corbelling leads to a large, square capstone. The cell is paved but much
of this dates from the nineteenth century. There are two cupboard niches in the walls.
Cell C lies to the east of Cell B and is very similar to it. It is probably the last cell to have been
constructed. It is also built of large stones but not as refined. It is almost square in plan, measuring 2.75m
by 2.6m, and again the corners are well defined, with the corbelling leading to a rectangular capstone.
There are no cupboards or special features. This cell retains most of its original paving and there is a
drain running down the centre of the floor, under the paving.
Cell D, to the east of Cell C, is no longer intact. It is probably the earliest surviving cell within the
inner enclosure and was already ruined when Cell C was constructed. It is D-shaped in plan and its
shape is defined by orthostats. There is still some original paving visible at the entrance.
Cell E lies to the north-east of Cell D and at a higher level than the other cells. It is built of relatively
small stones and has stepped plinths and projecting stones at the upper level. It is quite spacious inside,
3.65m by 3.58m, and is almost 4m high.Though almost square in plan, it becomes circular in shape as it
reaches the roof. Most of its paving dates from the nineteenth century. There are some projecting stone
pegs on the interior and there are holes in the walls to take wooden beams for a loft.
Cell F, to the south-east of Cell E, is built of moderately sized stones. It measures 2.98m by 2.75m,
with the corners defined up to about 60cm above the lintel, after which it becomes circular. A stone
with a circular opening completes the dome; it is unclear whether this is an original feature. There are
three cupboards in the walls and projecting stone pegs. The floor is irregularly paved and includes some
upright slabs that define a raised section on three sides, which may have been where the monks slept.
11
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
St Michael’s Church
St Michael’s Church lies to the west of the large oratory (Ill. 1.8). Part of this mortared church collapsed
in the late nineteenth century when the enclosure retaining wall to the south collapsed. The church is
rectangular in plan and has straight sides with sharp, rectangular corners. It would have had a timber roof
structure. There is an east window and a doorway on the north side. There is evidence of two distinct
phases of construction: an earlier, smaller, mortared church, constructed of large blocks of stone, was later
expanded to provide a larger church, the masonry of which is primarily oblong stones laid horizontally
with definite coursing. Sandstone for the door and window of this later church was sourced on nearby
Valentia Island. There is evidence that this later church was rendered externally.
Ill. 1.8—The
monastery
viewed from
south-east, with
the remains of
St Michael’s
Church in
centre
foreground
(Con Brogan,
DAHG).
The leachta
Leachta (plural; leacht singular) are common features of many early medieval monastic sites: they are
rectilinear drystone platforms and may have been prayer stations associated with pilgrimage rounds within
the monasteries.
There are two leachta within the monastery, one on the south side of the large oratory, which was
fully excavated, and another close to the entrance of the small oratory.
Burial platform
A stone-built platform (2.3m by 3.6m), sometimes described as a leacht, lies on the north side of the
large oratory. This was constructed in two phases, with the earlier phase pre-dating the construction of
the oratory. Although not excavated, this feature is described as a burial platform because of its similarity
in form to the nearby monks’ graveyard, which does contain human remains.
The monks’ graveyard
This is located to the east of the large oratory: it takes the form of a subrectangular platform (constructed
in two phases) measuring 7.3m by 3.2m (Ill. 1.5). Its base is defined by large, long orthostats, against
which a row of crosses and cross-slabs are placed vertically on the west side. The graveyard was bigger
12
INTRODUCTION
originally but the east side fell away when part of the adjacent enclosure retaining wall collapsed.
Paving
The area of the inner enclosure where the cells and oratories are located was fully paved throughout.
Large white quartz flags are used to define a symbolic area in front of the large oratory.The paving dates
from the monastic period and gives the monastery an almost urban quality.The paved area to the east of
the monks’ graveyard is a nineteenth-century repair.
Crosses/cross-slabs
There are two large, decorated crosses near the large oratory, one on either side, each of which is
accompanied by smaller cross-slabs. There are many crosses and cross-slabs behind the large oratory in
the monks’ graveyard. On the island there are over 90 crosses/cross-slabs, most of them roughly shaped.
Quarrying and water collection
The sloping rock above the monastery provided the major source of building stone for the monks. The
area beneath was also quarried. Once the sloping bedrock was exposed, the monks used it to collect
rainwater, cutting channels into it to direct the water into the cisterns below.
The cisterns
There are two cisterns within the core area of the monastery. They are constructed on the exposed
sloping bedrock and incorporated within the stone-built plinths beneath the cells. Their sides are
constructed of orthostats and drystone walling.Together these cisterns can hold about 450 litres of water.
An additional cistern (Cistern 3) to the west of Cell A was identified incorrectly as a souterrain prior to
its excavation. A fourth cistern is located outside the monastery, below the lower monks’ garden by the
east steps. This is probably associated with the structure referred to as a ‘guesthouse’.
The ‘guesthouse’
At the top of the east steps, just before the entrance into the lower monks’ garden, is a drystone structure
on the left-hand side. It is elongated in shape, with a doorway on the east side, directly opposite the steps.
It has one cupboard.This was most probably a guesthouse, where visitors could be accommodated outside
the monastery. Associated with this place is a large, undecorated stone cross known as Dunraven’s Cross.
The South Peak
The first mention of the possible existence of the structures on the South Peak was made by the Ordnance
Survey of 1841, but it was not until a visit to the island by the antiquarian Lord Dunraven in the 1850s
that reference was first made to a probable oratory: ‘Near the highest point of the island, which is called
the Spit, I found the remains of a little building which appears to have been quadrangular, probably an
oratory’ (Dunraven 1875–7, vol. 1, 34).
Liam de Paor, who studied the monastic remains in the early 1950s, made a vague reference to
structures on the South Peak. Subsequently, noted academic Françoise Henry, although unable to ascend
the Peak herself, received a plan of the ruins of the structure from one of the lighthouse-keepers, which
confirmed Lord Dunraven’s observations.
The lack of information about the South Peak structure prompted the Office of Public Works to
conduct its own detailed study, which began in the mid-1980s. This investigation discovered what has
been described as a hermitage constructed on the steep slopes of the peak. The ‘hermitage’ consists of
three separate terraces, labelled garden/dwelling terrace, oratory terrace and outer terrace. The
garden/dwelling and oratory terraces are located near each other, on the two best natural ledges of the
13
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
peak.Their spatial proximity is reinforced by the construction of the two routes between them, suggesting
that they had an important functional relationship. The outer terrace, in contrast, is set very much apart
from the other terraces and is also the most difficult to reach.The recent works carried out on the South
Peak suggest that the structures constitute an elaborate pilgrimage station and perhaps use of the term
‘hermitage’ is not wholly accurate in this context.
Access to the South Peak
The climb to the South Peak starts at Christ’s Saddle and follows rock-cut steps and handholds to the
lower traverse, below the Needle’s Eye—a point where the monks cut almost vertical rock-cut steps
though a cleft in the rock. Above this is a small enclosure which may have been a contemplation or
prayer station (Ill. 1.9). A further steep climb leads up to the first of three terraces. At a point halfway up
this climb is a separate, more basic route, which leads right up to the summit. The rock-cut steps near
the top of this route run under the upper traverse, evidence that this was part of the original route used
by the monks to get to the summit prior to the construction of the ‘hermitage’.
The garden terrace
This kidney-shaped terrace is 13m long and varies in width from 2m to 4m. Its long axis runs roughly
from north-west to south-east. The retaining wall, 1.5m high at the north-western end of the platform,
Ill 1.9—The small enclosure, possible
prayer station, just above the Needle’s
Eye on the South Peak (Con Brogan,
DAHG).
14
INTRODUCTION
is built on firm bedrock and is in impeccable condition. Much of the remaining terrace has collapsed,
the current ground level now being below the original level. It is possible that a small dwelling cell may
have existed here.
The oratory terrace
The second and most important of the terraces on the South Peak, the oratory terrace lies at right angles
to the garden/dwelling terrace and 4m above it (Ill. 1.10).The main structure is a corbelled oratory with
a narrow entry midway along the west wall and its east wall partially built on a stone slab bridging a cleft
in the rock. Internally it measures approximately 2.3m by 2m. Against the east wall are the remains of an
altar. Two small, interconnected rock-cut basins are located beside the church. These hold water—a vital
resource for a monk in this inhospitable place.The monks had cut channels into the near-vertical exposed
rock faces above this terrace to channel the rainwater down into the basins below.
At the western end of the terrace, approximately 1m east of the rock face, are the remains of a
rectangular leacht, 1.1m by 1.6m.This is most likely to have been an external altar. A low drystone bench
runs along the rock face, looking towards the oratory. This terrace appears to have been fully paved
originally. It is constructed in quite a complex way to allow for circumnavigation of the oratory itself. A
similar arrangement can be seen on the small oratory terrace within the monastery, indicating that this
must have been an important liturgical requirement. To the east of the oratory terrace is a long, narrow,
Ill. 1.10—The oratory terrace on the
South Peak, with Little Skellig and the
Kerry coastline in the background (Con
Brogan, DAHG).
15
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
tapering terrace upon which are the possible remains of a shrine.
The upper traverse
This is located above the oratory terrace and leads across to the final rock-cut climb to the summit and
the outer terrace. It was originally paved and had a parapet wall.
The outer terrace
The outer terrace is the most isolated of the three terraces on the South Peak. It is structurally dissimilar
to the others as the masonry remains consist of a 17m-long perimeter wall enclosing a series of stepped
ledges. The function of this terrace is unclear, as the location of a dwelling cell here would have been
impossible. It may have been used as a shelter or as a place of contemplation. Indeed, it may never have
been completed.
Quarrying on the South Peak
There is clear evidence of quarrying below the South Peak and on the peak itself.The area between the
Needle’s Eye and the garden/dwelling terrace displays evidence of major quarrying. Below the southern
retaining wall of the oratory terrace there is also evidence of quarrying, with platforms constructed to
store the stone prior to bringing it up. At the base of this retaining wall is a small raised platform that
would appear to have been the place from which the monks winched up the quarried stone from below.
Quarrying on the South Peak is currently the subject of further investigation and research.
1.2.4 THE
NATURAL HERITAGE
Birds
Skellig Michael, together with nearby Little Skellig (townland: Sceilg Bheag) are two of Ireland’s most
important sites for breeding seabirds. Both the size of the seabird colonies and the diversity of species
present make these islands highly significant on a national and international scale. Skellig Michael, together
with the Blasket Island Group and Puffin Island, supports some of the biggest breeding populations of
Manx shearwater and storm petrel in the world. Other seabird species breeding on Skellig Michael
include fulmar, kittiwake, guillemot and puffin (OPW 2008, 14).
Skellig Michael is known as a traditional eyrie for peregrine falcon, although the birds do not breed
on the island every year. Other birds that have been recorded as breeding in small numbers are chough,
raven, rock pipit and wheatear.
Owing to its ornithological importance, Skellig Michael is designated as a Statutory Nature Reserve
and a Special Protection Area, and is a proposed Natural Heritage Area.
Mammals
A small number of mammals have been recorded on Skellig Michael over the years. Grey seal haul out
on rocky ledges around the island and, while the numbers are not significant on a national scale, they
add to the diversity of the island’s fauna.This species is listed under Annex II of the EU Habitats Directive
and the Irish population is monitored on a regular basis. Other mammals recorded are rabbit and house
mouse. Rabbit is a relatively recent introduction and was probably brought to the island by the lighthouse
personnel in the early nineteenth century (ibid., 15).
Vegetation
Much of Skellig Michael is composed of poorly vegetated habitats such as rocky sea cliffs and exposed
16
INTRODUCTION
rock. The vegetation that does occur is typical of highly exposed maritime conditions limited by thin
soil, steep ground, salt spray and high winds. Common plant species include thrift, sea campion and rock
sea-spurrey, with patches of red fescue, dock and sea mayweed occurring frequently (ibid., 15). Lavelle
(1977) records 38 species of higher plant.
1.3 THE EARLY HISTORY OF SKELLIG MICHAEL
Teresa Bolger
1.3.1 INTRODUCTION
Most published studies of the site have focused on the archaeology and architecture of the monastic
settlement (e.g. de Paor 1955; Henry 1957; Horn et al. 1990). Examination of the historical references
and material in relation to the site has largely consisted of recitation of the key references with minimal
analysis.
The source material
This study of the early history of the site has focused on references to the site dating from before AD
1300. A variety of historical sources, both primary and secondary, have been consulted; details of the
published editions of these sources are included in the bibliography. References to the site occur in the
annals, martyrologies and a series of prose texts dating from between the eighth and thirteenth centuries.
These references are generally sparse and largely incidental.
References to the island (as opposed to the monastic foundation) occur in a number of prose texts:
• Conall Corc and the Corcu Loigde
• Lebar Gabála Érenn
• Cath Finntrága
References to the monastic foundation occur in three of the surviving sets of annals:
• Annals of Ulster (AU)
• Annals of Inisfallen (AI)
• Annals of the Four Masters (AFM)
in two martyrologies:
• Feilire Úa Ghormáin
• Martyrology of Tallaght
and in a number of narrative prose texts:
•
•
•
•
Cogadh Gaedhil re Gallaibh
Caithréim Cellacháin Chaisil
Giraldus Cambrensis’s Topographia Hibernica
Libellus de fundacione ecciesie Consecrati Petri
17
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The Augustinian abbey at Ballinskelligs is listed in the ecclesiastical taxation of 1302–4, which was the
latest record included within the scope of this study.
1.3.2 SCEILLEC AND ASSOCIATED
PLACENAME EVIDENCE
The term sceillec in Old Irish is generally translated as meaning a small piece of rock or a steep rock or
crag (Quin 1953; Toner et al. 2007). It is not a common element in Irish placenames and can only be
confirmed in use at three locations—at the western end of the Iveragh peninsula, Co. Kerry (in the area
around the Skelligs themselves), at Bunskellig, Co. Cork, and at Templenaskellig in Glendalough, Co.
Wicklow (Irish Placenames Commission 2010). It has been suggested that a number of other placenames,
such as Skelgagh, Co. Tyrone, and Spellickanee, Co. Louth, also derive from sceillec or include it as an
element (Joyce 1875, 421–2).
Debate over the etymology of sceillec had suggested a possible Old Norse origin for the term. This
argument was first put forward by Oftedal (1976, 128–9), who questioned the reliability of the earliest
annalistic reference to the site (AU, AD 824) and suggested that sceillec might be a Hibernicisation of the
Old Norse skellingar (‘the resounding ones’), which also occurs as a placename element in Norway.
Initially this argument was accepted (e.g. Fellows-Jenson 1992, 31), but doubts about its validity have
more recently been raised (Ó Corráin 1999).
Ó Corráin notes that Oftedal’s original argument does not adequately account for the well-attested
occurrences of sceillec in a variety of different texts (such as Aislinge Meic Congline, Bretha Coemgen or
Sanas Cormaic) in its ordinary substantive meaning (Quin 1953; Ó Corráin 1999, 311; Toner et al. 2007).
Further, while the annalistic references and that in the Martyrology of Tallaght may be the earliest definite
references to the monastic foundation, the placename sceillec (referring to the Skelligs themselves) occurs
in the text ‘Conall Corc and the Corco Luigde’ (Meyer 1910, 60), which can be dated to c. 700, thus
pre-dating any influence from Old Norse (Ó Corráin 1999, 311).
While sceillec does appear to be the original or early name, there are occasional hints of an alternative
name for the island—‘Glascarraig’ (lit. ‘the green rock’). Cath Finntrágha relates how a fleet of ships
belonging to the high king of the world made harbour at ‘the green rock that is called Sgellig Michil
today’ (Meyer 1885, 4; O’Rahilly 1962, 3). O’Rahilly (1962, 78n.) notes that ‘Glascarraig’ occurs as an
alternative name for Skellig Michael in late manuscripts, and an article on the Skelligs in the Kerry
Archaeological Magazine notes that the island is sometimes known as ‘Green Skellig’ (S.M. 1913, 164).
1.3.3 ÍARMUMU—WEST MUNSTER
POLITICS IN THE EARLY MEDIEVAL PERIOD
For a large part of the early medieval period Mumu or Munster was divided into two principal
territories—Aurmumu (east Munster, centred on and controlled from Cashel) and Íarmumu (west
Munster)—although the king of Cashel was nominally recognised as the king of all Munster. Until the
emergence of the Dál Cais in the tenth century, the dominant political force within Munster were the
Eóganachta. With a small number of exceptions, the kings of Cashel down to the mid-tenth century
were drawn from branches of the Eóganachta, primarily the Eóganachta Chaisil and the Eóganachta
Glendamnach.
Íarmumu was controlled by the Eóganachta Locha Léin, who were based in and around Killarney, Co.
Kerry.This region was also commonly referred to as Íarluachair, referring to Sliabh Luachra, which formed
the eastern boundary of the territory. There are indications, however, that, perhaps at an early period, the
influence of the rulers of Íarmumu extended into south Clare and as far east as a line between the modern
18
INTRODUCTION
cities of Limerick and Cork (Ó Buachalla 1952, 80–1).Very few of the Eóganachta Locha Léin succeeded
in obtaining control of Munster as a whole; the last to be acknowledged as king of Cashel was Ólchobur
m. Cináed († 851), who succeeded Feidlimid m. Crimthainn and was previously abbot of Emly.
West Munster contained a number of significant subkingdoms; the kingdom of Corcu Duibne
occupied all of the Iveragh peninsula, the southern and western portions of the Dingle peninsula and
the linking lands at the head of Dingle Bay (MacCotter 2008). Corcu Duibne itself was further divided
into three main territories, Irrus Tuaiscirt, Áed Conchinn and Áes Irruis Deisceirt; analysis of the Corcu
Duibne genealogies suggests that these latter three divisions existed as political entities by at least 900
(ibid.). Áes Irruis Deisceirt occupied most of the Iveragh peninsula and the reference to Skellig Michael
in the Lebor Gabála expressedly states that it was located within Áes Irruis Deisceirt (Best et al. 1954, 48).
It would appear that by the eighth/ninth centuries AD the control of the Eóganachta Locha Léin
over the subkingdoms of Íarluachair was waning. The title accorded to the rulers of the dynasty is
increasingly given as ‘ríg Locha Léin’ rather than ‘rí Íarluachair’ in the annals (Byrne 2001, 218). Further
evidence of dissent comes from the so-called ‘West Munster Synod’ (Meyer 1912, 315–16); the present
form of the text appears to date from the ninth century (Ó Buachalla 1952, 80; Ó Cróinín 2005, 224)
and is an account of an alleged mid-sixth-century gathering of west Munster clerics and saints at the
behest of the king of Ciarraige Luachra. It is unlikely that this is an account of a real event—many of the
alleged attendees were not contemporaries (Ó Buachalla 1952, 78); it is more likely a propagandist text
developed to support the claims of subkingdoms of west Munster (in particular the Ciarraige Luachra)
for greater independence from the Eóganachta Locha Léin. A number of ‘prophecies’ within the text
clearly correspond to late eighth-century personages and events (ibid., 80), indicating that the political
landscape described within the text is likely to be an accurate representation of the status quo in the late
eighth and early ninth centuries.
1.3.4 EARLY
MEDIEVAL FOUNDATION
Foundation
It is not clear when a monastic settlement was established on Skellig Michael. The earliest definite
reference to an ecclesiastical foundation at the site is the annal entry of AD 824 which records a Norse
raid on the island.
The next reference to the site, only slightly post-dating the first, is in the martyrologies deriving
from the monastery at Tallaght, Co. Dublin. The earliest in the sequence of martyrologies is the
Martyrology of Tallaght (MT) itself, which includes an entry under 28 April referring to ‘Suibni in Scelig’
(Best and Lawlor 1931, 37). Suibne of Skellig is also included in the Féilire Uí Ghormáin, which was
compiled between 1166 and 1174 at Knock, Co. Louth (Stokes 1993[1895], 86–7); this later martyrology
refers to the original Martyrology of Tallaght by name and would appear to have derived its list of saints
from a copy of MT that no longer survives.The earliest surviving copy of MT is the incomplete example
included in the Book of Leinster, which dates from the mid-twelfth century, but Ó Riain’s (1990, 37–8)
analysis of the surviving corpus of the Tallaght martyrologies points to the original composition of MT
no earlier than 828. Interestingly, the Féilire Óengusso, also composed in the ninth century and deriving
from a redaction of MT (Ó Riain 1990), does not include Suibne.
What is certain is that the site was well established by the time these earliest surviving references
occur, and most studies of the site would place its foundation sometime before 700 (e.g. de Paor 1955;
Henry 1957; Horn et al. 1990). If we accept the association of the site with St Fionán (see below), it
could well be a sixth-century foundation.
As previously mentioned, there is a reference to the Skelligs in the c. 700 text ‘Conall Corc and the
19
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Corco Luigde’ (Meyer 1910); this is a reference to the island itself—the existence of a monastic settlement
is not indicated. The relevant section of the text describes a dispute between the king of Cashel and the
king of Íarmumu, which results in the king of Íarmumu fleeing to ‘Scellec’ and ‘Gaur mac Maugo’ (ibid.,
60). Neither is expressly described in this text as an ecclesiastical site, but both are islands off the Kerry
coast. Gaur mac Maugo (recte Gair Mic Moga) is recorded as the site of an ecclesiastical foundation and
is commonly identified with the modern island of Garinis (e.g. Hogan 1910, 435), though more recent
studies suggest Scarriff Island (Ó Carragáin and Sheehan 2008). The text states that both islands are the
property of the Eóganachta Locha Léin.
The events recorded in ‘Conall Corc and the Corco Luigde’ are set in the mid-fifth century. This
does not mean, however, that it is an accurate record of the period; the events and political relationships
that it describes are far more relevant to its period of writing (c. 700). While the Eóganachta Locha Léin
would have been overlords of Corcu Duibne (and thus by extension its offshore islands, such as the
Skelligs or Scarriff Island) c. 700, the reference in this text seems to imply a more direct relationship.The
Eóganachta Locha Léin may have directly held these island territories or, perhaps more likely, they may
have been the main patrons of the ecclesiastical foundations that came to be established on both islands.
This association of the Skelligs with the Eóganachta Locha Léin contrasts strongly to the later
reference to Skellig Michael in the twelfth-century text Caithréim Cellacháin Chaisil (Brugge 1905, 38).
Like the Cogadh Gaedhil re Gallaibh, the Caithréim is a propagandist text (in this instance in favour of the
MacCarthaig dynasty, the ruling dynasty of the Eóganachta); far from detracting from the reference to
Skellig Michael, however, this fact enhances it. In the relevant section of the text various subject kings of
Munster explain their rationale for fighting against the Norse. The three kings of Corcu Duibne justify
their involvement as a reaction to an attack on Skellig Michael, placing the site on an equal footing with
such regionally important sites as Scattery Island (Inis Cathaig) and suggesting a link between the site
and the ruling dynasties of Corcu Duibne.This lends weight to recent arguments regarding the potential
regional importance of the site (Ó Carragáin 2008), while also suggesting that patronage and support
for the site by the eleventh or twelfth century derived more from the local dynasties of the Corcu Duibne
than from the Eóganachta Locha Léin, as may once have been the case.
Dedication
There are many questions surrounding the dedication of the site. As the name of the site indicates, the
foundation came to be dedicated to St Michael, but the association of this saint with the site is commonly
believed to have occurred quite late in its history.The first reference to the dedication is the latest in the
sequence of annal entries relating to the site (AD 1044) and occurs only in AFM. The corresponding
entry in AI simply refers to ‘Sceilic’. AFM was compiled during the seventeenth century using a variety
of surviving annalistic texts as its sources, including redactions of many surviving sets of annals (such as
AU and AI) in addition to material that has not survived to the present. It is presumed that the reference
to the site as ‘Sgellic-Mhicil’ derives from an earlier set of annals used by the compilers and is not an
interpolation by them (certainly the earlier reference to the site in AFM at AD 950 refers just to ‘Sgeillic’),
but the divergence with AI is worth considering. The entry in AI is quite long and by this period the
surviving manuscript of AI is likely to be a contemporary record of events. It could even be argued that
the ‘newness’ of the dedication is reflected in its omission from AI.
Aside from St Michael, the other saint commonly associated with the site is Fionán, an important
local saint who was a member of the Corcu Duibne and founded Inisfallen; though much of his
ecclesiastical career was spent in the midlands, the surviving Latin Life emphasises his connections to
south Kerry (Ó Riain 2009). No evidence for this association with Fionán has thus far been identified
in early medieval texts, however. Skellig Michael is not recorded as one of the sites founded by Fionán,
which is a curious omission, given the likely proposed regional significance of the site (Ó Carragáin
20
INTRODUCTION
2008).The earliest identified instance of the association is the assertion by Smith (1756, 61) in his account
of the site that it was originally founded by Fionán. Subsequent scholarly work on the site derives the
association with Fionán from Smith’s account.There are three holy wells in the townland of Kinard West
(Tobar Muire, Tobar Fionain and Tobar Michil) on the south side of the Dingle peninsula; there is a folk
tale that the wells were created by Saints Fionán and Michael when they landed there from Skellig
Michael (Ó Danachair 1960, 75). On that basis it is likely that Smith’s association of Fionán with the site
derives from local folk tradition. The antiquity of this tradition is difficult to determine.
As already noted, one saint is mentioned in connection with the site in the early medieval documents.
The Martyrology of Tallaght and the Feilire Uí Ghormáin both record a Suibne of Skellig with a feast-day
of 28 April. It is tempting to suggest that Suibne could be the name of the original founder of Skellig
Michael. Ó Riain’s (1990, 26–35) schema for dating the original compilation of the martyrologies relies
heavily, however, on the annalistic obits of late eighth- and ninth-century clerics who were accorded
sainthood in the martyrologies. In such a context, Suibne may well have been a prominent cleric at
Skellig Michael during this period; given the paucity of recorded information about the site it is difficult
to be more definitive.
Clerical succession
Only four clerics are recorded in the annals (Table 1) and only one of the four is expressly titled abbot
(Flann m. Cellach, †882); it is possible that the others were also abbots, though Aed (†1044) is described
as a priest.Though the site is sometimes considered a hermetical foundation, none of the clerics is noted
as being an anchorite. Possibly we could add Suibne to this list also, bringing the total of recorded clerics
to five; while there is a potential that Suibne could be the name of the original founder of the site, many
of the saints listed in the Tallaght martyrologies can be identified with clerical office-holders whose obits
occur in the annals (for examples see Ó Riain 1990, 26–35).
Table 1—Clerics of Skellig Michael recorded in the annals.
Year Name
Annal
Entry
824
AU
Étgal
882
Flann
AI
Étgal of Scelec was carried off by the heathens, and died shortly afterwards
of hunger and thirst.
Scelec was plundered by the heathens and Étgal was carried off into
captivity, and he died of hunger in their hands.
Repose of Flann son of Cellach, abbot of Scelec.
950
Blathmac
AFM
Blathmhac of Sgeillic died.
AI
Aed Sceilic, the noble priest, the celibate, and the chief of the Gaedil in
piety, rested in Christ.
AFM
Aedh of Sgelic-Mhichil.
AI
1044 Aed
A single Norse raid is recorded in annals, in 824, resulting in the death of Étgal, possibly the abbot
but certainly an important cleric. A record of a further raid is preserved in the Cogad Gáedel re Gallaib
(CGG). CGG is a problematical text, as it is primarily a work of political propaganda (though early
scholars mistakenly viewed it as a historical treatise), but it does draw directly on the contemporary
annalistic records (Ní Mhaonaigh 1996), many of which do not survive to the present. So it is likely that
the reference to the additional raid on Skellig Michael is authentic. There is an identifiable stratum of
unique material within CGG relating to the territories of the Ciarraige and the Eóganachta Locha Léin,
which suggests that the compiler of CGG may have had access to a west Munster source, though equally
21
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
the material may reflect strong ties between the Ciarriage and Lismore, the possible source of the unique
material for south-east Munster and the Deisí (ibid., 120–1).This second raid is listed amongst a series of
undatable entries within CGG which cannot be correlated with the surviving sets of annals (Todd 1867,
16–17). The position of the entry within the text of CGG suggests a general ninth-century date; in the
redaction of CGG preserved in the Book of Leinster the raid on Skellig Michael is described as
originating ‘from Limerick’ (ibid., 228), which might suggest a mid-ninth-century date, after the
establishment of the longphort but before its destruction in 887 (Valente 2008, 50).
1.3.5 A TIME
OF TRANSITION: TWELFTH-CENTURY REFORM
While the tenth and eleventh centuries were significant periods in the political history of Munster as a
result of changes wrought by the emergence and expansion of the Dál Cais, the twelfth century was an
important watershed in the history of the Irish church. By the end of the eleventh century religious and
ecclesiastical life in Ireland had diverged sharply from the contemporary norms of mainland European
Christianity. As a result, the twelfth century witnessed a growing movement within the Irish church in
favour of ecclesiastical reform. The reform movement was characterised by a series of major church
synods that set out and developed a new diocesan system and began the process of bringing the Irish
church closer to European norms.
The impact of this reform movement can be seen in the history of many prominent ecclesiastical
sites within Ireland, sometimes through the loss of episcopal standing, sometimes through the introduction
of the new religious orders or the adoption of their rule.
The exact standing of the ecclesiastical foundation at Skellig Michael prior to the reform movement
is difficult to ascertain. There are no indications that it was ever an episcopal foundation; nevertheless,
the reference in Caithréim Cellacháin Chaisil does suggest that it was a significant site within Corcu Duibne
and that, while it may originally have drawn support from the Eóganachta Locha Léin, its main patrons
by the twelfth century are likely to have been the local Corcu Duibne dynasties.
This period broadly coincides with the time-frame during which year-round settlement at Skellig
Michael may have been abandoned and a new foundation established on the mainland at Ballinskelligs.
In light of the existence of a vibrant medieval tradition of pilgrimage to the monastery (see below), it is
likely that the site continued to be occupied at least seasonally.
As with the foundation of the original monastic settlement at Skellig Michael, the exact date at
which the Augustinian priory was established at Ballinskelligs remains elusive. The priory is listed in the
ecclesiastical taxation of Ireland of 1302–4 (Sweetman and Handcock 1886, 298); the prior is described
as the ‘collector’ for the deanery of Agadoe (ibid., 295), so it was clearly well established by the late
thirteenth century.
It is possible that the establishment of the priory at Ballinskelligs pre-dates the arrival of the AngloNormans. One of the characteristics of the twelfth-century reform movement was the introduction of
new religious orders, such as the Cistercians and the Augustinians (notably the Arroasian rule). Though
this led to the establishment of new religious foundations, the new monastic rules, in particular the
Augustinian rule, came to be adopted at existing foundations also. The initial introduction of the new
religious orders is generally credited to Malachy Ó Moirgair—new foundations such as the Cistercian
abbey at Mellifont and the Arroasian abbey at Louth can be attributed to his direct patronage. During
his term as papal legate (1140–8) Malachy was known to have been active in Munster, and the adoption
of the Augustinian rule (commonly the Arroasian form of the rule) at existing ecclesiastical foundations
within Munster, such as Roscrea, has been attributed to that period and to his influence (Dunning 1945,
303–4). O’Sullivan and Sheehan (1996, 347) have stated that the priory was founded in or shortly after
22
INTRODUCTION
1210, following the date suggested by Gwynn and Hadcock (1970, 192).
The Life of Malachy written by St Bernard of Clairvaux records that Malachy established a new
foundation at ‘Ibracanese’ in Munster; this placename has been described as deriving from Íbh Rathach
(Iveragh), suggesting a location on the Iveragh peninsula, leading to the suggestion that this foundation was
the abbey at Ballinskelligs (Fenton, The Kerryman, 13/11/1948).There are a number of difficulties with this
(Gwynn 1992, 207–8), not least of which is the assertion in the text of the Life that Cormac MacCarthaig,
the king of Cashel, who had endowed the foundation, could have had daily access to the new priory.
1.3.6 TWELFTH- AND THIRTEENTH-CENTURY
RECORDS
References to both the island monastery of Skellig Michael and the Augustinian priory at Ballinskelligs
are sparse during the twelfth and thirteenth centuries, the best-known references being the account of
Giraldus Cambrensis and the ecclesiastical taxation of 1302–4, previously mentioned.
Giraldus Cambrensis’s account of the site is not particularly informative. He does not name Skellig
Michael, but describes how a hollow stone situated outside a church miraculously produces wine for the
celebration of the Eucharist each day.This miracle occurs ‘in the south of Munster near Cork’ on a ‘certain
island which has within it a church of Saint Michael, revered for its true holiness since ancient times’
(O’Meara 1982, 80). Despite the dubious geography of the reference it is unlikely that any location other
than Skellig Michael would fit the broader description; in addition, the miraculous tale recounted occurs
in a later document, which definitely confirms Skellig Michael as the location (see below).
A more interesting reference to the site occurs in the Libellus de fundacione ecclesie Consecrati Petri,
commonly referred to as the Regensburger Schottenlegende.This is a mid-thirteenth-century Latin text giving
an account of the foundation of the Irish Benedictine monasteries in Germany, at Weih Sankt Peter and
St James in Regensburg, St James in Würzburg and St Nicholas in Memmingen.The text appears to have
been composed at Ratisbon by an Irish monk, potentially originally from Kerry (Breatnach 1977–8, 58).
The initial section of the text offers an account of the career of St Patrick which includes a version of
Patrick’s expulsion of the demons from Ireland, featuring an intervention by St Michael that occurs on a
rock off the coast of Ireland.This rock (according to the text) is known as ‘Silex Sancti Michaelis’ as a result.
The text then includes a detailed description of the rock, its setting and various miraculous tales concerning
the site (ibid., 59). One of these miracles is identical to that described by Giraldus Cambrensis.
Breatnach (1977–8) has argued that this section of the Libellus functions much like dindsenchas (and
may have been intended as such), providing a rationale and origin-tale for Skellig Michael and possibly
for another prominent local landmark, the Saint’s Road, which leads to the summit of Mount Brandon
on the Dingle peninsula. The text also points to a well-established tradition of pilgrimage to the site by
the thirteenth century; the abbey at Ballinskelligs is likely to have provided a useful base for pilgrimage
to the island. It may well be that the development of Skellig Michael as a place of pilgrimage provided
impetus for the establishment of the Augustinian abbey on the mainland, to provide a controlled ‘gateway’
to the island.
1.3.7 CONCLUSIONS
The surviving evidence for the early history of Skellig Michael is very sparse; references to the site in texts
prior to the thirteenth century tend to be brief and incidental. Some of these references are to the
ecclesiastical foundation while others are to the island itself, as a prominent landmark off the south-west
coast.
23
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The earliest surviving reference to the island is contained in the c. 700 narrative text ‘Conall Corc
and the Corcu Loigde’, while the earliest surviving references to the monastic foundation occur in AU,
AI and the Martyrology of Tallaght, all dating from the ninth century .
By the eleventh century the site was clearly dedicated to St Michael and it is likely that a tradition
of pilgrimage to the site had begun to evolve. There is a slight possibility that the site was originally
dedicated to St Michael, though at present the balance of evidence suggests a rededication either from
an unnamed local saint or from St Fionán (though the association of Fionán with the site has not been
sourced in any medieval documentation).
There is no strong historical information that would establish a foundation date for the site, though
a foundation before 700 seems indicated.
Historical sources do suggest that the original patrons of the site could have been the Eóganachta
Locha Léin, with patronage shifting to local Corcu Duibne dynasties by the twelfth century. This would
tally with the diminution in power of Eóganachta Locha Léin from the late eighth/early ninth century.
Again, uncertainty surrounds the date of the establishment of the Augustinian abbey at Ballinskelligs,
though its background context would probably have been the influence of the twelfth-century reform
movement in combination with a growing tradition of pilgrimage.
1.4 A SUMMARY OF THE LATER HISTORY
In the early thirteenth century a general climatic deterioration resulted in colder weather and increased
storms on the seas around the south-west coast.This, along with a shift in the Irish church from a monastic
to a diocesan structure, signalled the end of Irish eremitic island settlements, with the result that the
community of Skellig Michael eventually moved to the mainland at Ballinskelligs. This is likely to have
happened over a period of time in the later thirteenth century and possibly the early fourteenth century
(see Section 8 below).
The island probably continued to be used as a dependency of the Augustinian abbey at Ballinskelligs,
being occupied by some monks in the summer months. The prior of Ballinskelligs was still addressed in
papal letters as ‘Augustinian prior of St Michael’s, Roche (de Rupe)’. The Augustinians must also have
been actively involved in promoting and managing pilgrimages to the island and maintaining the
structures there.
Skellig Michael remained in the hands of the Augustinian monks until 1578, when, as a result of the
Desmond Rebellion, Queen Elizabeth I dissolved certain monasteries that were under the protection of
the earl of Desmond. The Skellig islands thus passed into secular hands, to the Butler family. Although
the monastery ceased to exist, the island continued to be used as a place of pilgrimage.
In the early 1820s the Corporation for Preserving and Improving the Port of Dublin (the predecessor
of the Commissioners of Irish Lights) purchased Skellig Michael from John Butler of Waterville under a
compulsory purchase order for the purpose of erecting two lighthouses on the Atlantic side. These were
made accessible by an improved landing on the east side and a road that was blasted out on the precipitous
southern and western sides of the island. During the period of construction, the lighthouse-builders
occupied many of the beehive cells within the monastery, and the structural modifications carried out
at this time have had a significant impact on areas of the monastic settlement. Both lighthouses and their
associated domestic quarters were completed by 1826.
In 1880 the OPW took the monastic remains into guardianship and commenced a project for the
repair of collapsed structures. Since that time, the OPW has continued in its efforts to maintain and
preserve the monastic remains.
24
INTRODUCTION
1.5 BACKGROUND TO THE CONSERVATION WORKS
Grellan D. Rourke
1.5.1 INTRODUCTION
Skellig Michael had been abandoned for a long time, during which much deterioration had taken place.
Given the topography, the pressures of retained material and the lack of any maintenance over centuries,
considerable loss must have taken place. During the lighthouse occupation in the 1820s considerable
works were undertaken to render the place safe for use. This necessitated the creation of level platforms,
improved paving and, most notably, works to the retaining walls. Before the builders departed, additional
works were undertaken to clean up the site and to conceal the large amount of building debris created
over a period of six years. Later on in the century the OPW took over the care of the monument and
repair works were undertaken to walls, most notably to St Michael’s Church, which had partially collapsed
in the intervening period.
The condition of the site in the late 1970s was such that there were considerable structural problems
requiring attention. Some were very serious in scale with potentially grave consequences, while others
were more localised. With the consultant engineer Joss Lynam, a strategy was developed. The structural
problems were rated and the most serious were tackled first.The works detailed in this report were driven
by the need to resolve structural issues of all kinds in order to preserve the unique remains. Great care
was taken to retain as much of any original material as possible, and this was the primary guiding principle.
In addition to the preservation of the site, it was necessary to undertake the works to ensure the safety
of visitors now and into the future.
The following is a summary of the rationale behind decisions taken regarding conservation works
and the resultant archaeological excavations. The information is presented under the same headings as
the archaeological excavations for ease of reference. Additional information is supplied for the north and
south steps, where archaeological monitoring took place.
1.5.2 THE
SMALL ORATORY TERRACE AND LEDGE
The terrace retaining wall
This was the first major work to have been undertaken on Skellig Michael. This retaining wall is an
original monastic structure and retains a terrace upon which the small oratory is constructed (Ill. 2.26).
Over time it had moved considerably owing to the pressure of the retained material above. The terrace
itself had filled with a peaty soil and much water is channelled onto it from the rock above. The wet,
peaty fill added to the weight this wall was required to support. Structural drystone construction is very
flexible and the wall simply adjusted to accommodate the pressure of the retained material. The most
vulnerable walling is the centre section of the south wall, which is not constructed on rock. This section
began to move outwards at the base and, as it did so, the wall migrated both downwards and inwards
above. This movement was slow and considerable; as the wall dropped, additional stonework had to be
added to the top to maintain an enclosing wall. In the end a retaining wall had also to be constructed on
the inside of part of the terrace to stop the upper parapet wall from falling in onto the terrace.
The deformation was considerable, with the base of the wall 2.6m out from the top, although there
would have been some slight batter to the wall originally. The large base stones in the centre had fallen
away, exposing a large hole (Ill. 2.3). Failure was not far off and it was remarkable that this wall had not
collapsed, given the intense pressure it was under. It was not possible to correct the deformation in any
25
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
way but it was essential to halt any further movement by providing a counterbalance secured from the
rock well below.This was effectively a reinforced concrete arch put in at an angle to the wall and secured
into the rock at either end. From the crown of the arch a buttress, faced in drystone, was built under the
centre of the retaining wall, impeding any further slippage. Once the engineering works were completed,
the wall was given additional strength in situ by tightening the existing drystone work.
The small oratory terrace
The material that had accumulated on top of the paving level on the terrace had added to the structural
problems, and when this material was excavated the weight on the terrace was reduced to a more
acceptable level. Some of the original paving was still intact. There are three major structures associated
with this terrace: the small oratory, the monks’ ‘latrine’ and a leacht.
The excavation exposed the original level of the paving on top of which the oratory was constructed.
This building had suffered over time as the built-up ground consolidated under the weight. The northwest corner was partially built on exposed rock but the other walls were mostly constructed on paving
which lay on fill. The settlement of the building varied and was most evident at the south-east corner,
where most subsidence had taken place. Here the damage to the stonework was worst. The outer stone
skin had suffered in the usual way with loss of stone at the upper level. As a consequence, water was
coming into the building from the roof and was causing erosion of material from the interior, which is
not paved.The south-east corner and the upper section of the outer stone layer were repaired to preserve
the structure.
The monks’ ‘latrine’ may pre-date the construction of the terrace and could have been accessed from
a rock-cut ledge on the upper side.The leacht was partially constructed on this sloping bedrock.The base
of the ‘latrine’ and substantial walling on the west side were fully exposed in excavation. The walling on
the interior was corbelled, demonstrating that this had been a small beehive structure. All its features
were intact and were repaired.
1.5.3 THE
LARGE ORATORY
This structure has had a long history of movement, some happening very slowly over time, so that the
building was able to deform considerably without collapse, notably in the north-west corner. There is
one area (the north-east corner) that had collapsed and had been repaired during the monastic
occupation. The north-east corner collapsed and was repaired when the annexe to the burial platform
was built on its east side.
The main issue with the oratory was water ingress. This building had been used again from the
nineteenth century, first as an office during the lighthouse construction in the 1820s and later as a church
by the lighthouse families stationed on the island.There is still evidence of limewash on the interior.The
outer skin of the dome had been repaired very roughly, with the stones placed in a very haphazard way.
In time this facilitated the ingress of water into the building and the erosion of the internal floor. It was
necessary to halt this cause of the deterioration of the structure.The rough work to repair the dome had
been undertaken during the lighthouse occupation and a pipe bowl dating from this period was found
when this repair was taken down.The upper part of the external skin on the dome was reconstructed in
such a way as to throw the water away from the roof and walls. The interior masonry was in such good
order that no work was deemed necessary.
26
INTRODUCTION
1.5.4 SOUTH
ENTRANCE
1 (INNER
ENCLOSURE) AND LEACHT AREA
This entrance is located in an area not now easily accessible and in a wall of curious construction. A
length of the original walling remains on the west side, while that on the east side had collapsed down
onto the ‘garden’ below. It was this collapse that probably damaged the circular cell and the lower retaining
wall in this location. The west jamb was reasonably intact, although it had started to deform under the
weight and pressure of the wall above and to the side.
This is the second entrance into the inner enclosure.The seriousness of the collapse of this important
feature needed to be addressed. Left unattended, the west jamb would have collapsed with additional
loss of the surrounding upper retaining wall. This wall is quite unique on Skellig Michael. While it gives
the appearance of a collapsed structure, it was constructed intentionally this way with a very rough face,
the large, long stones built into the wall with the short ends protruding out in a haphazard and irregular
fashion. There is a parallel on the mainland in the enclosure wall of the early medieval monastic site of
Killrelig. This form of construction is very solid but has the disadvantage that it can be scaled very easily
in the event of an attack, which might explain why it went out of fashion.
The main thrust of the work here was engineering, and it was possible to jack back the very large
collapsing stones in the west jamb. The walling on top was taken down and the area was excavated. This
revealed the original innermost lintel. It was possible from the collapse to reinstate the east jamb, and the
wall to the east was built in a similar style to retain the structures above. A facing wall was constructed
within the entrance opening at the back to support the retained material behind. On the exterior the
base of a platform was revealed; this gave access to the entrance, which is at a higher level.
The large oratory was constructed after this entrance had been abandoned. It was not possible to
excavate the area inside the wall fully, as the foundations for the oratory are constructed on the fill. It
was possible to reach a sufficiently low level to show the very interesting stepped foundation of the
oratory. The ground to the west was excavated in order to undertake repairs to the inside face of the
upper retaining wall, and in the course of this excavation a number of skeletons were uncovered. Also
excavated was a leacht with a large vertical stone at the west end.This stone proved to be the base of one
of the two major decorative crosses in the inner enclosure. Where soil had built up against St Michael’s
Church a patch of original external render was uncovered. This was conserved in situ. The second phase
of St Michael’s Church had been rendered externally and probably limewashed, creating quite a stark
contrast to all the earlier structures on the site.
1.5.5 SOUTH
ENTRANCE
2 (INNER
ENCLOSURE)
While repairs had taken place on either side of this entrance, repair to the actual entrance had only taken
place above the lintel level. There were two structural problems that manifested themselves by the early
1980s and both required immediate attention, as this is the only way for visitors to access the inner enclosure.
The nineteenth-century wall running from the entrance on the interior was giving cause for
concern. It was effectively a retaining wall, holding back material to a height of a little under 2m. The
forces at play were causing the wall to bow inwards, and it would have had a catastrophic effect had this
collapsed when visitors were accessing the monastery. In addition, one of the lintels on the inside of the
entrance had cracked and greater support was required, as most visitors also pass overhead.
During excavation the nineteenth-century wall was taken down; some of the rubble behind the wall
was taken back and a better-quality wall constructed in its place. In exposing the inner corner of the
west jamb of the entrance it became clear why this intervention had occurred in the first place: the inner
face of the upper retaining wall had collapsed inwards and there was very little of the inner face left.This
27
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
original corner was consolidated and the new retaining wall constructed back from the face of the
previous wall so as to expose the inner corner of the entrance, showing the great thickness of the original
upper retaining wall.
The area above the lintel on the inside was excavated and support was put in place to take the stress
off the original cracked lintel, which was left in situ. In examination of this entrance it became clear that
the stepped paving within the entrance was not original. There is a set of paving beneath, on top of
which the entrance was built. The state of this paving was much dilapidated by the time the secondary
paving was built. Interestingly, it does not turn west towards the current entrance into the upper monks’
garden but east to make its way down to the entrance in the outer enclosure.
1.5.6 EAST
ENTRANCE, PHASE
3 (INNER
ENCLOSURE)
The existing east wall of the inner enclosure had collapsed numerous times and was repaired on three
different occasions. Each repair had been set back from the previous one, at a higher level.The last repair
was undertaken in the early nineteenth century, when considerable enabling works were undertaken by
the lighthouse-builders.This length of repair coincided with the long dimension of the monks’ graveyard.
It had become dangerous, and this is an area where visitors congregate. The initial project had been to
take down this later section of walling and reconstruct it, but the walling beneath was also quite unstable
and it was necessary to reduce the wall further. In the process another wall with an intact entrance came
to light on the interior, and its condition required additional investigation to consolidate sufficiently to
facilitate the upper wall repair.
This was one of the most significant structural finds and brings the history of the monastery back
considerably in time. It was possible in the excavation to reveal the full extent of this early entrance and
then investigate the interior in a limited way. It was soon evident why this entrance had been abandoned.
There was significant evidence of a collapsed structure within, collapsing in front of the entrance on the
inside.The fallen stones were long and flat with smooth faces and were aligned east–west.These remains
are most likely those of an earlier well-built structure. It was possible to consolidate this entrance and to
repair the later wall in such a way that visitors can now view the feature from above.
1.5.7 CISTERN 3
This structure had been considered a souterrain. It was possible to crawl into the entrance and see that
the interior chamber was located on the west side, but most of the low, long chamber was filled with
debris from the collapsed roof structure. Visitors could walk on the ground above and this was
exacerbating the structural failure. Excavation established the scale of the structure and that it was in fact
another cistern rather than a souterrain.The side walls, which had collapsed, were repaired and the lintels
put back into position again.
1.5.8 THE
LOWER MONKS’ GARDEN AND ITS STRUCTURES
This was the most challenging part of the work undertaken on the monastery. There had been at least
two considerable collapses of the upper retaining wall into this area, with rebuilding by the lighthousebuilders and repairs by the OPW in the nineteenth century. The rebuilding was focused on the east end
of the terrace; walls were built to enclose the steep and vulnerable terrain and to stop further erosion. A
28
INTRODUCTION
cross-wall had been built across the terrace, cordoning off the east end, which was very dangerous. In
this area an additional wall had been constructed, set back from the line of the lower retaining wall. In
time this latter wall had been undermined from underneath and had begun to fail.The overall structural
stability of the east end of this terrace was considerably compromised, and any collapse would have
cascaded down the steep south slope onto the lighthouse road below.
Owing to the topography it was necessary to have a different strategy for this terrace. Because of the
seriousness of the situation at the east end, remedial works and excavation began here. Considerable
safety precautions were required before the failing nineteenth-century interventions could be removed.
Once the area was made safe, excavation began in earnest. This revealed the base of the original lower
retaining wall, which must have collapsed most spectacularly. Set back from this wall were the remains
of an early circular cell, with one door jamb still intact. The area surrounding this cell on the north and
west was paved, so it clearly could not have been part of a garden.The condition of the drystone walling
of the cell was good and this was tightened. Originally this cell must have been built against and partially
into the lower retaining wall.
Near the entrance to the cell and at a lower level, and against the lower retaining wall, the remains
of an indeterminate structure were revealed, possibly a storage chamber. The retaining wall, which had
fallen away at this end of the terrace, was reinstated to retain all the features above. In doing so it was
possible to present the full remains of this small chamber.
Only when this end of the terrace had been consolidated was it possible to work on the central
section. Owing to the fall across the terrace it was not possible to fully excavate in this area; indeed, to
reveal enough for structural consolidation it was necessary to construct a retaining wall midway along
the garden to provide support to the upper retaining wall. This facilitated the exposure of more of the
original south retaining wall. It showed that a series of large rocks, most likely from the base of the upper
retaining wall, had come crashing down, hitting the lower retaining wall with such force as to cause it to
deform. The major force of the collapse hit the base, causing a reaction which saw the upper section fall
inwards. This dramatic deformation had to be addressed, and very slowly this long length of walling was
carefully jacked up on the inside to increase its stability.
At the upper west end no works were undertaken other than those recorded when the outer entrance
was excavated and repaired.
South entrance 1 (outer enclosure)
This is the only entrance into the lower monks’ garden, and the north, south and east steps lead to this
point. The entrance had been used by the lighthouse-builders, who before they left built a drystone
facing wall on the exterior of the entrance, where most of the collapse had occurred. The structural
problems relating to this entrance are inextricably bound up with the upper monks’ garden. The west
jamb is one of the ‘pillars’ underpinning the south-east corner of this garden. In addition, stonework at
the top of the blocked-up ope was beginning to unravel.
Excavation exposed a very complete doorway with pillar jambs that had fallen outwards considerably.
Much of the west jamb on the exterior had fallen away. The inner lintel had survived thanks to its
protected location; it was still in its original position. A certain amount of jacking back was required for
the pillar stones. The entrance had been designed to take a strong door and there were holes in the wall
on the interior to accommodate a large draw-bar.The outer west jamb had to be repaired and new lintels
found for the outer section of the entrance.
On the interior of this entrance a curved wall on the west side follows the curved steps that lead up
into the terrace above.This wall had begun to fail and had fallen in and partially collapsed at the top and
the end where it abutted the entrance. It was necessary to take down the upper section of this wall. The
wall that was partly dismantled would appear to date from the nineteenth century.
29
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
1.5.9 UPPER
MONKS’ GARDEN
This area has suffered a series of collapses. To the left of the inner enclosure entrance the upper wall had
collapsed and been repaired considerably, in a stepped fashion, by the lighthouse-builders, when they
inhabited the monastery and constructed their latrine.The outer retaining wall had dramatically collapsed,
most likely after the abandonment of the monastery. Most of it had fallen over onto the steep southern
slope of the island and been lost. The lighthouse-builders needed to enclose this garden area and make
it safe, so they built a new retaining wall. This was constructed in quite a different style to the original.
One of the problems facing them was the dearth of building stone and it was only possible to build a
narrow wall. In order to increase its strength they curved the wall in and out, which significantly changed
the appearance of this feature.The third intervention had been to build a narrow wall on the diagonal at
the south-east corner, where a section of the original walling had fallen away as the south-east corner of
the garden had collapsed.
There were two structural issues, both relating to nineteenth-century interventions which were
beginning to fail. The outer retaining wall had been pressured by some of the retaining material, and in
two places the wall had failed where it abutted the remains of the original monastic walling. This failure
manifested itself in partial collapse, creating holes that deteriorated each year. The ground on which the
diagonal wall had been built had considerably eroded under the exposed exterior face, undermining the
base of the wall. Had either of these structures failed there would have been fatalities, as visitors to the
monastery congregate in this area before entering the inner enclosure.
In reconstructing the outer retaining wall it was possible to straighten out the footprint somewhat
so that it better resembled the original line of the wall. Excavation revealed portions of the straight line
of the inner face upon which the repair was constructed. Excavation at the south-east corner revealed
why this part of the garden had failed, the perimeter walls here being built on a layer of large boulders
sloping steeply to the south-east. In the end, the pressure became so great that this section collapsed,
taking with it the west jamb of the outer entrance below. It was possible to provide sufficient support to
this part of the repair so that the original line of the corner could be reinstated.
1.5.10 STRUCTURE AT THE
BASE OF THE EAST STEPS
For the most part the east steps were reasonably intact, although many had slipped, which can represent
the first phase of failure. In a number of locations this slippage had begun to accelerate and, given the
steep topography on this side of the island, this would have resulted in loss of individual steps and a
further unravelling of the staircase. Given the sensitive location of this staircase right above the landing
pier, there was also a serious safety risk to those landing on the island. In all cases these steps, with attendant
masonry, were secured back into their original positions. All slipped material was retrieved.
An elevated stretch near the end of the long flight leading down from the south entrance 1 (outer
enclosure) had collapsed; both steps and masonry had fallen down the very steep ground to the south,
some of it lost. If this length of structure had been left untouched further collapse would have occurred.
This could have had serious impact far below, where the access road from the pier is located. In addition,
given the height of the staircase above the ground at this location (this length of staircase has the highest
supporting exposed masonry on the island), it would have been very difficult to access the remainder of
the east steps to undertake works and effect ongoing maintenance. As much as possible was retrieved
from the steep slope beneath and this section of the staircase was repaired using this material; some
additional material was also required to complete the work.
At the base of the east steps lie the remains of a drystone structure. Below this location the steps had
30
INTRODUCTION
been dynamited away when the lighthouse-builders constructed the pier and road. This action
unfortunately destroyed half of the long, rectangular structure, and what remained was very vulnerable
to erosion as the ground falls away very steeply at this point. It is probable that more of this structure had
survived the blasting and that it had fallen away in the intervening years. It was essential to consolidate
and conserve what was left of this unique structure, possibly a boat-house. It is situated at a level at which
the housing of a boat would have ensured its protection from winter storms. It was only necessary to
tighten the existing masonry and to support the end wall where the ground fell away. One jamb of the
entrance was revealed during excavation and this was secured.
This area was cleaned of vegetation to reveal a whole series of rock-cut steps. As these became worn,
new steps were cut out alongside. This is the only location on the island where this has been noticed.
Further up the east steps there is evidence of rock-cut steps running parallel with the later masonry steps.
This can be seen elsewhere, and it is clear that the monks created their routes to the monastery in a
simple straightforward way at first. Once they had established the preferred routes and had more time at
their disposal, they created the major drystone staircases.
1.5.11 THE
NORTH STEPS
The north steps are made up of two long flights leading down to the water from Christ’s Saddle. These
steps had been repaired and much used by the lighthouse-builders in the nineteenth century. The lower
flight is made up entirely of rock-cut steps, many of which had been widened by the lighthouse men
and since that time have become considerable eroded by the harsh seas on the north side. The upper,
longer flight is of drystone construction. Given how they were constructed into the steeply sloping
ground along their length, they were particularly vulnerable to erosion and collapse. In addition, on this
side of the island there are continual falls of stone from the cliffs above and these would have exacerbated
the situation. This stone fall continues and each year there is damage done to these steps, often quite
considerable.
Unlike the east steps, when the side support gave way the individual steps fell away quite dramatically.
The topography meant that they travelled easily down the steep slope, gaining considerable momentum;
some were halted by large rocks below but many fell to the bottom, either falling into the sea or crashing
onto the bedrock below and breaking up. The rate of loss to this staircase was accelerating as the steps
unravelled, and the intervention required was quite different. The goal was to retain as much of the
original as possible.
Where the steps had fallen away, a residue of masonry had remained. From examination of the slope
and of the construction that remained it was clear that this was a long, straight flight of steps with the
occasional flat landing. It was possible to set up ropes from the rock above to scour the steep ground
below and retrieve what had been caught in the fall.This was a very slow process and the retrieved steps
had to be pulled back up to the locations whence they fell. All that was possible to retrieve was taken
back up but gaps remained; in these areas the side retaining walls were reinstated and the ground above
consolidated and left sloping. It is now possible to reach the north landing safely again.
1.5.12 THE
SOUTH STEPS
The south steps have been in continual use and were no doubt repaired by the lighthouse-builders. Only
one section of steps, that situated on the lower section of the flight above Christ’s Saddle, posed a structural
problem. It had begun to fall away although it had not yet failed. It was a relatively easy task to put the
31
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
steps back into their original positions, thus consolidating the staircase again. The south steps regularly
become loose from visitors travelling over them, and there is an ongoing daily maintenance programme
in place to secure them and to avoid the development of more serious problems.
32
2. THE EXCAVATIONS
2.1 INTRODUCTION
Conservation works at the early medieval monastic settlement on Skellig Michael commenced in 1978.
The initial focus was on the reinstatement of a section of the inner enclosure retaining wall, which had
collapsed, and on repair of the south steps. In the early 1980s two early nineteenth-century revetments
that were in a state of collapse were removed, as was the waterlogged peat that covered much of the
inner enclosure.
The first archaeological intervention took place in August 1980, when Conleth Manning supervised
the removal of peat on the north side of Cell A and the removal of the sod layer that partly overlay the
plinth of the small oratory. Excavations proper commenced in 1986/7, when the small oratory terrace,
south entrance 2 (inner enclosure) and the large oratory were investigated (Table 2). Works, including
archaeological excavations, continued throughout the 1990s and early 2000s, focused mainly on the
retaining walls, entrances and the upper and lower monks’ gardens. Repair and consolidation of the east
and north steps and associated structures also took place. In 2004 conservation and excavation works
commenced at the South Peak, where survey work in the mid-1980s had identified a range of structures.
These works, including detailed post-conservation recording, were completed in 2010 (Ill. 2.1).
The scope of the archaeological work on Skellig Michael was determined by and large by the
conservation needs. This strategy was deemed the most appropriate, given the limited area actually
available for excavation on this precipitous island and the intact nature of the structures, in the monastery
in particular, which were left undisturbed. Within these parameters, then, the role of the archaeologist
on Skellig Michael was to exploit fully the opportunities provided by the interventions required to
conserve elements of the site, in order to provide new insights into this unique monastic settlement.
2.1.1 SURVEY AND
RECORDING
Detailed pre-works surveys commenced in the late 1970s and have continued throughout the duration of
the works programme. Both measured surveys and photographic surveys are carried out, and since 1994
the works have been recorded professionally on film. Plans, sectional profiles and elevations are recorded at
various scales during excavation; following conservation, all structures are again recorded in detail.
Surveying on Skellig Michael presents many challenges, not least of which is the vertiginous nature
of the terrain, with its attendant health and safety requirements. Plane table surveys were used extensively
and in 1982 a photogrammetric survey (1:1000) of the island was commissioned which provided detailed
contours and allowed the individual monastic structures to be correctly located on the island.This survey
was subsequently tied into the National Grid. As survey and recording of the South Peak structures
progressed, however, it became clear that the level of locational detail on the contour map was insufficient
for accurate recording in this precipitous terrain. Consequently a three-dimensional geometric survey
of the island was carried out in 2007, using aerial LiDAR.
LiDAR (Light Detection And Ranging) is a method for determining the three-dimensional position
of a point using a laser pulse. Traditionally, fixed-wing aircraft are utilised in the collection of these data,
but the accuracy and resolution required in the effective recording of Skellig Michael (ground point
density of 60pts/m2) necessitated a different approach. To achieve such a high resolution the FLI-MAP
33
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Table 2—Excavations carried out on Skellig Michael.
Year
Licence
No.
1986–7 E338
Director
Details
Ann Lynch
1986–7 E338
Ann Lynch
Excavation of the small oratory terrace and south entrance 2, inner
enclosure.
Excavation of the large oratory, inner enclosure.
1988
Paddy O’Leary Supervision of works to south entrance to the lower monks’
garden.
Claire Cotter
Excavation of western end of upper monks’ garden.
1990
90E0034
1993–6 93E0195
Edward Bourke Excavation of lower monks’ garden.
1996–7 93E0195
Edward Bourke Excavation of Cistern 3, west end of inner enclosure.
1998–
2000
2000
2001–2 93E0195
Edward Bourke Excavation of south entrance 1, inner enclosure and the burials
and leacht, next to the large oratory.
Edward Bourke Excavation of east entrance, phase 3, inner enclosure and
supervision of conservation works to monks’ graveyard.
Edward Bourke Supervision of conservation works to east and north Steps.
2003
93E0195
Edward Bourke Excavation of early structure found at base of east steps.
2004
93E0195
2005
93E0195
Edward Bourke Excavation of features on the South Peak, including the lower
traverse, upper and lower platforms, the small enclosure above the
Needle’s Eye (the ‘prayer station’) and a test-trench on the ‘garden’
terrace.
Alan Hayden
Excavation of the ‘garden’ terrace and part of the oratory terrace.
2006
93E0195
Alan Hayden
Completion of the oratory terrace, excavation of the upper
traverse and ledges beneath the oratory terrace and a sondage in
the outer terrace.
2007
93E0195
Alan Hayden
2008
93E0195
Alan Hayden
2009
93E0195
Alan Hayden
Excavation of the outer terrace, ledges below the outer terrace,
southern traverse of the northwest passage and a platform and
steps at the base of the Needle’s Eye.
Excavation of the northern traverse and stairs of the northwest
passage, northern cliff ledges, the lower gully, and from the blind
corner to the cliffs below the Needle’s Eye.
Surveys on the South Peak.
2010
93E0195
Alan Hayden
93E0195
93E0195
Excavation of the ledge with fallen stones and the broad ledge,
both on the South Peak.
400 system was used. This aerial LiDAR survey system, which utilises a helicopter sensing platform, was
initially designed to survey infrastructural assets such as roads, railways and electricity supply networks.
The sensor system mounted beneath the main helicopter fuselage consists of:
• three 150kHz LiDAR sensors (forward, nadir and aft);
• two RTK GPS receivers, which provide accurate location when used in conjunction with RTK
base stations;
• an Inertial Navigation System (INS) to continuously track the position, orientation and velocity of
the helicopter;
• digital imaging (11 megapixel) and digital video capture.
34
THE EXCAVATIONS
- A. Lynch
- E. Bourke
- C. Cotter
Ill. 2.1—Plan of
the monastery,
showing areas
excavated.
N
0
10
20
30
40
metres
This technology has several advantages, the first being that it is helicopter-mounted, allowing for relatively
slow air speeds and low-altitude flight paths which result in the collection of extremely high-resolution
height data.The second is that instead of the conventional single laser this system has three 150kHz laser
scanners (forward–nadir–aft) with a range accuracy of 1cm (1 sigma). The result is extremely accurate
point data, with an absolute accuracy of ±8cm in horizontal and ±5cm in vertical position. Several
35
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
imaging devices, including a high-resolution mapping camera and three video cameras, accompany each
laser scanner.
Initial processing by Fugro-BKS included:
•
•
•
•
the merging of the three laser scanner data sets into a seamless point cloud;
geo-referencing of all point cloud data;
removal of outlying information caused by objects such as birds;
interpolation of areas where there was a distinct lack of survey points.
The final data set, provided in four tiled data sets, included approximately 2.75 million x,y,z height points
stored within an ASCII file structure.
Additional processing was carried out by the Discovery Programme to improve the final visualisation
of the DSM, including:
•
•
•
•
merging of tiles into single DSM;
removal of holes within the data specifically at the interface between the terrain and the shoreline;
generation of gridded data model DSM (12.5cm resolution);
visualisation of the DSM using several algorithms:
—directional hill-shade,
—composite hill-shade,
—slope modelling,
—principal component analysis,
—sky-view factor.
These visualisation data sets are now being used to create scaled plans and sections for the whole island
and areas of archaeological significance.
Traditional instrument surveys are also used to fix spatially the various features/structures as they are
exposed and recorded. As the window of opportunity for work on Skellig Michael is so limited (midMay to end of September), however, and with the deterioration in climatic conditions over the past
number of years, a backlog has developed in this area of survey work. As a result, a number of the recently
discovered features on the island, e.g. parts of the primary access route to the summit of the South Peak,
although recorded, have not been surveyed in detail. This work will continue, as conditions allow, over
the coming seasons. It is also envisaged that terrestrial LiDAR will be used for the detailed recording of
structures in the future.
2.1.2 FIRE
DAMAGE
In August 1996 a gas explosion followed by a fire took place in the accommodation huts on the lighthouse
road, which completely destroyed the archaeologists’ accommodation. Unfortunately many of the
excavation site plans from the 1993–5 seasons of excavation were in the hut at the time, as well as the
site notebooks from 1993–4 and part of 1995. Some of the nineteenth-century finds were also in the
hut, together with at least five sherds of medieval pottery and fragments of a number of stone crosses.
Luckily the section drawings from the 1993 season and part of the 1994 season were in Dublin at the
time, being drawn up, and all of the finds had been numbered and described so that their contexts are
known. No site photographs were lost in the fire.
Following the fire, the wreckage was thoroughly searched; much of the pottery and glass was
36
THE EXCAVATIONS
retrieved, and all the stone was recovered and subsequently conserved. Wherever possible, plans were
redrawn—all of those for the 1995 season were redone and a full report on that year’s work has been
compiled. The area for which most information has been lost is Cell G and its immediate surrounds in
the lower monks’ garden: no plans, sections or notebooks survive for this area. Cell G was in use during
the lighthouse period and nineteenth-century pottery and clay pipe fragments had been recovered from
its floor surface.
In this report, all features described as F1000 or greater refer to features for which the site notebooks,
plans and/or sections were destroyed in the fire.
2.2 THE MONASTERY
2.2.1 THE
Ann Lynch
SMALL ORATORY TERRACE AND LEDGE
(E338)
The small oratory terrace is located c. 10m to the north-east of the core monastic area and is supported
by a high retaining wall that forms the north-eastern limit of the inner monastic enclosure (Ill. 2.26).
This artificially created terrace, although within the inner enclosure, has a feeling of isolation and
remoteness by virtue of its location at a lower level, out of sight of the other monastic structures and
hemmed in by the steeply rising rock face to the west. Its massive drystone retaining wall (5m high
externally), which appears to be substantially original, had developed serious structural problems, with
the top section falling inwards and a
significant bulge at its base where the
stonework had also begun to unravel
(Ills2.2, 2.3). The collapse of this wall
not only would have had serious
consequences for the monastic
structures on the terrace behind but
also, owing to its location c. 160m
directly above the main landing stage
for the island, would have posed serious
safety concerns for visitors.
The work to preserve the small
oratory terrace retaining wall was the
first major structural intervention to be
undertaken during the current works
programme. The engineering solution
proposed for the consolidation of the
base of this wall involved the
construction of supporting structures
on the cliff ledge on which the wall was
founded. This in turn necessitated
archaeological excavation to be
undertaken on the ledge to test the
archaeological potential of the soil
cover which had to be removed. This
work was undertaken in June 1986.
37
Ill. 2.2—The
small oratory
terrace
retaining wall,
looking north,
before works
(A. Lynch).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.3—
Unravelling at
base of small
oratory terrace
retaining wall,
before works
(A. Lynch).
The small oratory ledge
The ledge in question is irregular in outline
but measures c. 6m wide (max.) and slopes at
an angle of c. 40º before a sheer drop of c.
160m to the sea below (Ill. 2.4). The logistics
of carrying out excavations and complex
conservation works in such an exposed,
dangerous location were formidable. Before
any works commenced, a safety platform was
erected on the cliff face to provide a relatively
secure work space, and workers on the ledge
were harnessed to safety ropes at all times.
The safety aspects of this particular excavation
were devised and overseen by the Kerry
Mountain Rescue Team.
Initially, two cuttings were opened which
extended from close to the base of the
retaining wall across the width of the ledge
(Ill. 2.5). For safety reasons these were located
on either side of the unstable bulging section
of the wall. Once excavation of these cuttings
was completed, a small extension was opened
at the western end of each to examine the
foundations of the wall. Two small trial-pits
were also excavated to record the depth of
bedrock for engineering purposes.
Ill. 2.4—The
small oratory
ledge, looking
north, before
excavation (A.
Lynch).
38
THE EXCAVATIONS
0
5
metres
N
retaining wall F216
G1
batter on retaining
wall F216
CUTTING 1
H1
ledge
TRIAL-PIT 1
CUTTING 2
TRIAL-PIT 2
G2
H2
safety platform
155-metre steep drop to sea level
The northern cutting (Cutting 1) measured 5.4m by c. 2.7m (max.), with rising bedrock forming its
irregular northern edge.The southern cutting (Cutting 2) measured 4.5m by c. 2.1m (max.), with bedrock
forming its irregular southern edge.
The stratigraphy exposed was generally consistent across the ledge, and as bedrock shelved steeply
from north to south the deposits were considerably deeper in Cutting 2. Bedrock was exposed at a
maximum depth of c. 0.5m below the sod throughout Cutting 1 (Ill. 2.6). A moist, black humic clay with
high gravel content (F106) filled the crevices in the bedrock and was present up to a maximum depth
of 0.12m at the interface between the bedrock and the overlying sandy soil and rubble (F102/104).
Bedrock was not reached within Cutting 2, where a compact, grey sandy boulder clay (F105) formed
the limit of excavation (Ill. 2.6). The rockface that formed the southern edge of the cutting was shown
to drop vertically and the undisturbed boulder clay lay against it. Overlying the bedrock in Cutting 1
and the boulder clay in Cutting 2 was a layer of rubble (F104) intermixed with and overlain by a brown
sandy soil (F102), which also contained lenses of burnt clay (F103). The burnt clay was more extensive
in Cutting 2, where it was up to 0.03m thick in places. The rubble (F104) in Cutting 1 consisted of
stones ranging from 0.15m by 0.08m to 0.7m by 0.35m, whereas in Cutting 2 the average stone size was
0.6m by 0.25m, with many measuring over 1m in length. The artefacts recovered from F102 include
nineteenth-century ceramics and a T-shaped piece of iron (E338:19) which may be part of a strap hinge.
A rectangular whetstone (E338:20) was found in the bottom of the rubble (F104) in Trial-pit 1. Fragments
of brick, coal and cinders were also recovered from the upper levels of F102, close to the base of the
retaining wall. The entire ledge was covered with a sod layer which supported a dense growth of sea
campion.This averaged 0.1m in thickness and produced modern finds, including plastics and drinks cans
(Ill. 2.7).
39
Ill. 2.5—Plan
of the small
oratory ledge,
showing areas
excavated.
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
G2
G1
160
lowest courses of
retaining wall F216
159
bedrock
F106
F102/F104
158
F101
157
M.O.D.
bedrock
safety
platform
Ill. 2.6—The
small oratory
ledge, sectional
profiles.
H1
H2
F101- sod/campion layer
lowest courses of
retaining wall F216
160
F102- brown sandy soil
F104- rubble layer
F103- red/brown burnt soil
F105- boulder clay
159
F102
F103
F101
F104
bedrock
F101
F102
safety 157
platform
M.O.D.
158
F105
The extensions to the cuttings, which allowed examination of the foundations of the retaining wall,
revealed that the wall was sitting directly on at least 0.5m of loose rubble (F104), with brown sandy soil
(F102) filling the interstices (the lower levels of the rubble were not exposed for safety reasons). The
rubble stones averaged 0.2m in length and the coursing of the drystone wall proper only commenced at
present ground level on the ledge (Ill. 2.8).
40
THE EXCAVATIONS
Ill. 2.7—The
small oratory
ledge during
excavation,
cutting 1 in
foreground (A.
Lynch).
Interpretation
It is difficult to say how much, if any, of the clay
and rubble was brought onto the ledge for the
construction of the terrace retaining wall, but
since the wall sits on this deposit some of it may
have been in situ. The area from which the
construction stone was quarried has not been
identified, and while much of the rubble
remaining on the ledge could be the result of
stones falling from the upper levels of the wall
over the centuries, some of it could also be the
residue left following the construction. The
lenses of burnt clay as well as the small number
of artefacts and debris recovered from the ledge
must be the result of casual dumping over the
terrace wall in modern times, while the
whetstone, found in the lower levels of the
rubble, may have been lost or discarded during
the construction of the terrace wall, when there
must have been considerable activity on the
ledge.
Ill. 2.8—Rubble
foundations of
the small
oratory terrace
retaining wall
(A. Lynch).
The small oratory terrace
Before excavation the small oratory terrace was
densely overgrown with sea campion supported
by a covering of moist peat. This peat masked
41
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
all structural features with the exception
of the oratory (from the level of the plinth
up), the remains of a small beehive cell
interpreted as a latrine, fragments of
drystone walling close to the rock face, the
retaining wall of the terrace and a
stockpile of stone and brick at the
southern end of the terrace (Ills 2.9, 2.10).
A substantial stone cross found lying close
to the oratory had been erected close to
the rock face almost directly opposite the
doorway of the oratory. The enclosed
nature of the terrace, accentuated by the
steep rock face bordering it on the west
side, combined with poor drainage had
resulted in the development of
waterlogged conditions. This was
responsible not only for the growth of
peat but also for ongoing seepage of water
into the small oratory and the build-up of
pressure on the retaining wall to the east
of the oratory, where serious structural
Ill. 2.9—The
small oratory
terrace, viewed
from west, in
the 1950s
(DAHG
archive).
Ill. 2.10—Plan
of the small
oratory terrace
before works
(survey and
drawing: G.D.
Rourke).
42
THE EXCAVATIONS
problems had developed (see above). It was decided that, in tandem with the conservation of the retaining
wall and prior to any consolidation of the oratory, the peat cover would have to be removed from the entire
terrace, thereby removing the source of much of the moisture and allowing improved drainage to be put
in place. This work was carried out by the author in June/July 1986 and July 1987, while Claire Cotter
supervised additional work in August 1986.
The removal of peat exposed features including paving, steps, terracing and a leacht, all part of a
highly ordered space centred on the small oratory. Once these features were uncovered, they were fully
recorded, consolidated and left in situ. Further excavation was carried out to investigate and interpret
features on the north and east sides of the oratory, and a narrow cutting was opened on the south side,
where the terrace stratigraphy was fully investigated (Ill. 2.11).
The peat
In 1980 Conleth Manning removed the sod layer from around the perimeter of the small oratory in
order to expose fully the plinth, which had become partly buried, and to allow easier access to the
doorway at the west end of the structure.This slight reduction in ground level also helped to reduce the
amount of water seeping into the oratory. The plinth, which varies from 0.2m to 0.32m in width, is at
the level of the doorway lintel on the west wall but drops by c. 0.4m below this on the south wall and
by c. 0.3m on the north wall, and continues at this level around the east wall. During this removal of sod,
four stone crosses or fragments thereof were found.
In 1986–7 the peat cover on the terrace was fully removed.The upper levels (F201), which supported
the dense growth of sea campion, were moist, mid-light brown in colour and of moderate humification,
with extensive root penetration.The depth of this peat varied from 0.18m on the uphill, west side of the
‘latrine’ to 0.3m against the rock face opposite the oratory entrance, while it was shallower (averaging
0.12m) on the floor of the terrace to the south of the oratory (Ill. 2.17, section C1–C2). An underlying
deposit of peat (F202) was present in the area of the ‘latrine’ and extended downhill, flanking the rock
face, to the area immediately west of the oratory.This highly humified peat, dark brown/black in colour
and containing small leached spalls of sandstone, became progressively wetter as one moved downhill
and had accumulated to a maximum depth of c. 0.3m on the uphill side of the slab set on edge (F210),
close to the ‘latrine’. Three sherds of transfer-printed pearlware, dating from about 1820, were found at
the interface of the two peat deposits adjacent to the ‘latrine’ wall, indicating peat growth of c. 0.2m in
this area since the mid-nineteenth century.
The western area
A narrow trial-trench was opened, extending from the rock face to the ‘latrine’ wall (c. 2m long by c.
0.5m wide), to assess the depth of deposits on the uphill side of the terrace (Ill. 2.11). Here it was revealed
that overlying the bedrock was 0.06–0.08m of moist black sandy clay (F205), which in turn was overlain
by a loose stony layer (F206) comprising 80% flat spalls of sandstone intermixed with grey/brown sandy
soil and with campion root penetration to the bottom of the layer. At the base of the rock face there was
an accumulation of small fragments of broken sandstone and fine moist brown clay (F203). These latter
deposits represent weathering from the adjacent rock face and clays and humic material washed down
from further uphill and from the overlying peat deposits (F201 and F202). A concentration of loose stone
within F204 at the eastern end of the trench must represent collapse from the adjacent ‘latrine’ structure.
No finds were recovered from these deposits.
Terracing
The remains of a series of small terraces, built against and partially incorporating projecting sections of
the rock face, were revealed along the western side of the small oratory terrace (Ill. 2.12).
43
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
‘latrine’
F206
trial-trench
0
4
metres
scale 1:50
B1
retaining wall F216
bedrock
area excavated by
removal of peat
B2
C2
C1
Ill. 2.11—Plan
of the small
oratory terrace,
showing areas
excavated.
retaining wall
F216
E1
F1
D2
area
excavated
to boulder
clay
D1
wall
F234
F236
area excavated by
removal of peat
N
E2
bedrock
F235
F2
A drystone retaining wall (F207) curves north-eastwards from the rock face to the retaining wall
(F216) at the entrance to the ‘latrine’ (Ill. 2.13).This wall, constructed of sandstone, survives to an average
height of 0.2m and averages 0.35m in width. At its northern end, stones on edge define its inner face,
but elsewhere a stone and clay fill (F214) lies directly behind its rubble core. This retaining wall defines
one side of a triangular terrace, the other two sides of which were formed by the rock face on the west
44
THE EXCAVATIONS
‘latrine’
F206
trial trench
168.15
0
4
metres
scale 1:50
166.83
B1
retaining wall F216
F210
bedrock
F214
166.72
166.81
F207
drain
F226
F208
B2
C2
C1
31
F2
Ill. 2.12—Plan
of the small
oratory terrace,
showing
features
exposed during
excavation.
166.25
retaining wall
F216
E1
F1
F208
166.28
F220
D2
leacht
F233
N
collapse
D1
F229
wall
F234
165.89
166.57
bones
under
F236
F238
166.32
- step treads
E2
bedrock
- quartz
- metres OD
F235
collapse
F2
and by a large slab on edge (F210) on the north. This slab (2.16m long by 0.17m thick by 0.65m high),
which had been prised from the bedrock and placed on edge, also served as a type of dam, preventing
the seepage of water into the entrance area of the oratory below. The build-up of wet peat deposits
(F202) behind the slab bears testament to this. The make-up of the terrace (F214) consists of a dark
brown moist clay with 70% stone content.The stones range from pebbles to the more common sandstone
45
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.13—The
low retaining
wall (F207) and
paving stones
(F208) during
excavation (A.
Lynch).
spall averaging c. 0.1m in length. Pockets of light brown clay are also present, and fragments of charcoal
and a tiny sliver of wood were recorded on the surface.
The northern end of the retaining wall (F207) is roughly bonded with the lower (primary) courses
of the small oratory terrace retaining wall (F216), suggesting contemporaneity.The relationship with the
adjacent ‘latrine’ is less clear, however, as the junction of both retaining walls corresponds with the entrance
to the ‘latrine’, which is c. 0.2m above. Access to the ‘latrine’ must have been via a step up from the paving
to the top of the retaining wall (c. 0.2m) and, from here, another step of 0.2m to the threshold of the
‘latrine’, although no flagstone survived on top of the retaining wall. Such an arrangement of steps would
be in line with the three steps uncovered in the lower peat deposit (F202) and underlying stony clay
(F204) which lead uphill, skirting the west wall of the ‘latrine’ (Ill. 2.11).While these three steps probably
belong to a period later than the construction of the ‘latrine’, they do indicate an access route uphill to
the crest of the rock above the monastery at this point.
The space between the rock face and the west end of the small oratory averages just 2.5m in width.
It is, however, a space that was highly organised and that constituted the only route to and from the
oratory (Ill. 2.12).
Traces of a drystone revetment wall (F231) were uncovered running north/south opposite the
entrance to the oratory and 0.75m distant from it. Just a single course high, it takes the form of large
stones filling the gaps and crevices in the bedrock, with smaller packing stones filling the interstices. In
places, the bedrock forms the eastern face of the revetment. A single paving slab (now split in two) covers
the area between the wall and the entrance to the oratory, and eleven paving slabs still survive between
this and the north-west corner of the oratory. The exposed bedrock rises to the north, and the natural
changes in level were exploited by careful positioning of flagstones to provide four steps up to the higher
level on the north side of the oratory (Ills 2.16, 2.18).The difference in level between the oratory entrance
and the paving at the north-west corner is c. 0.74m (Ill. 2.12).
Just to the west of the leacht and 0.6m distant from it, a drystone wall (F234) was uncovered linking
two projecting outcrops of the rock face, which resulted in the creation of a small terrace behind. The
46
THE EXCAVATIONS
Ill. 2.14—The
small oratory
terrace during
excavation,
showing
retaining wall
(F234) in
background, the
leacht and the
remains of a
possible
retaining wall
(F238) in
foreground.
(A. Lynch)
wall, which was loosely constructed of small/medium-sized stones, is c. 2.5m long and, when exposed,
ranged in height from 0.95m in the south to 0.45m in the north, where some unravelling and collapse
had taken place. The base of the wall was not exposed as it lies below the level of the paving, but it is
probably founded, at least in part, on bedrock. The make-up of the terrace behind is a dark brown
silty/peaty soil with a high stone content, some of which appear to be spalls of sandstone weathered
from the rock face behind (Ill. 2.14).
Further south, another stretch of drystone walling (F235) was uncovered running east–west, linking
the exposed rock face with the main terrace retaining wall (F236). Only the inner (north) face of this
wall was exposed but it must also have been a retaining wall, given the difference in ground level between
the paved surface of the small oratory terrace and the higher ground of the monastery to the south.This
wall (maximum height 0.95m) was in a poor state of preservation and considerable collapse had taken
place, particularly at its eastern end close to its junction with the terrace retaining wall. Here the remains
of two, possibly three, steps were revealed, leading down from the core area of the monastery to the small
oratory terrace. A large slab (1.6m long) that had been laid on edge at the western end of the wall had
slipped forward and the wall had partially collapsed over it.
The removal of peat also revealed a discontinuous line of stones (F238) running from the south-east
corner of the leacht towards the retaining wall F235 (Ill. 2.14). The deposit of stony soil behind these
stones was not excavated, but this rather ephemeral feature may be the remains of a secondary, low terrace.
The leacht
Just over 1m from the south-west corner of the oratory, a rectangular drystone-built leacht (F233) was
revealed under a shallow (0.17m) layer of peat (Ills 2.15. 2.16). The leacht, which measures 1.65m by
0.85m by 0.75m (max.) high, was loosely constructed of sandstone and quartzite blocks, with a number
of quartz ‘nodules’ lying on top. It was constructed on sloping bedrock, which meant that it is just two
to three courses high on the western side and up to seven courses high on the downhill, eastern side (Ill.
2.17, section D1–D2). Just south of the entrance to the oratory, a step (0.28m) in the bedrock led up to
47
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.15—The
leacht after
excavation (A.
Lynch).
Ill. 2.16—The
small oratory
terrace, viewed
from north,
during
excavation (A.
Lynch).
a paved surface (F208) that flanked the northern and western sides of the leacht, with some of the paving
stones running under the leacht itself. The southern face of the leacht, and in particular its south-west
corner, had unravelled, with some of the stones slumping outwards. Following excavation, these stones
were set back in position and the masonry of the entire structure was tightened. Otherwise the leacht
was left undisturbed.
48
THE EXCAVATIONS
B1
B2
167
modern
addition
small
oratory
F216
retaining
wall
F201
F217
F227
F228
166
bedrock
F225
drain wall
F224 flagged drain
base
M.O.D.
F232
168
C1
C2
167
Ill. 2.17—The
small oratory
terrace,
sectional
profiles.
F201
small
oratory
bedrock
D1
166
M.O.D.
F209
D2
167
quartz
leacht
F233
small
oratory
166
paving slab
M.O.D.
wall
F234
paving slab
F229
F201- campion/light brown peat
F232- dark brown gritty clay
F209- dark brown/black silty peat
F227- dark brown stony clay
F217- rubble with sandy soil
F229- brown sandy clay
F224- rubble with fine silty clay
F228- boulder clay
49
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Paving
In addition to the paving stones adjacent to the leacht (described above), eleven large paving slabs (F208)
survive in situ between the retaining wall F207 and the north-west corner of the oratory; while they
appear to lead towards the ‘latrine’, they do not line up with its entrance (Ills 2.12, 2.18). All the paving
slabs are sandstone, with the exception of a large quartz block at the southern end, and all are closely set,
with smaller stones filling the gaps, particularly adjacent to the revetment wall. From the quartz block
there is a step down of 0.12m to three paving stones and a band of bedrock, from which there is a further
step of 0.33m to a ledge of bedrock. The next step, of 0.2m, is onto a wedge-shaped stone that has been
set on a ledge of bedrock, and the final step of 0.18m leads down to the paving stone at the entrance to
the oratory.
Ill. 2.18—The
small oratory
terrace, viewed
from west,
during
excavation (A.
Lynch).
The ‘latrine’
The beehive cell (1.3m by 1m internally) built on the northern edge of the terrace and projecting from
the end of the terrace retaining wall is interpreted as a latrine (Ills 2.19, 2.20). A slab set on edge c. 0.2m
from its eastern wall forms the edge of an opening to a narrow gully leading down the cliff face below.
The heavy growth of sea campion which had encroached on the structure was removed during the
excavation. The western half of the cell is best preserved and here the drystone wall is founded on the
bedrock, which rises steeply to the north. The cell wall averages 0.64m in width and survives to a
maximum height (internally) of 0.82m on its western side, where traces of corbelling still survived.
The northern area
The oratory is sited close to the northern edge of the terrace, leaving a narrow passage between its north
50
THE EXCAVATIONS
Ill. 2.19—The
‘latrine’ after
removal of
campion
growth (A.
Lynch).
Ill. 2.20—The
east wall of the
‘latrine’ and
opening to gully
below, after
removal of
campion
growth (A.
Lynch).
51
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
wall and the retaining wall (Ill.
2.12). This passage narrows from
1.6m at its western end to just 0.5m
at its eastern end. Peat cover
reached a maximum depth of 0.3m
in this confined space and
underlying the peat there was a
layer of loose stone (F217), which
extended under the paving (F208)
at the north-west corner of the
oratory. This deposit of loose stone
averaged just one stone deep close
to the terrace retaining wall but
Ill. 2.21—The
stone-lined
deepened to over 0.5m where it
drain (F226),
abutted the plinth of the oratory
viewed from
wall. Removal of this stone deposit
west, after
excavation (A.
revealed a stone-lined drain (F226)
Lynch).
skirting and partly underlying the
north wall of the oratory (Ills 2.21,
2.17, section B1–B2).
The floor of the drain was
paved with flagstones, some of
which extended under the north
wall of the oratory. At its western
end, the outer edge of the drain was
delimited by three stones set on
edge on the flagged floor. Further
east there was no well-defined
outer edge, but here the narrowing
of the passage meant that the
terrace retaining wall could serve as the side of the drain. The inner (southern) edge of the drain was
defined by the plinth of the oratory wall for a distance of c. 1m from its western end, but further east the
edge of the drain comprised two to three courses of drystone walling set back on average 0.15m under
the edge of the plinth. The depth of the drain averaged 0.25m under the plinth and 0.4m outside the
oratory.The drain appeared to swing slightly northwards just beyond the north-east corner of the oratory,
but this area could not be fully excavated owing to the unstable condition of the adjacent retaining wall.
It seems likely, however, that the drain would have debouched through the terrace wall in this area and,
while no outlet could be identified, the masonry of the wall was significantly looser and more ‘porous’
in this area.The floor of the drain is relatively level throughout most of its length, but over the final 1.6m
it slopes downwards west to east, dropping 0.28m over that distance.
The fill of the drain (F224) varied throughout its length. The westernmost 0.9m consisted of loose
rubble with fine silt, while further east, for a distance of c. 2.3m (to the break in slope), the fill consisted
of a dark brown loose sandy soil with stones in the upper levels and a very fine silty clay of sticky
consistency in the lower levels.The final section of the drain, where it sloped downwards, was filled with
a rubble deposit containing stones up to 0.4m in length. This rubble extends beyond the limits of the
drain under the east wall of the oratory. The only artefact recovered from the drain fill was a slate with
the remains of a perforation at one end (E338:36), found in the upper levels of the sandy soil, under the
52
THE EXCAVATIONS
plinth of the oratory (Ill. 3.10). This slate, which is ovoid in shape (1.78m by 0.89m max.), was initially
thought to be a roofing slate, with the notch representing a broken peg hole.This presented problems of
interpretation since it was clearly placed in the fill of the drain at the time of construction, which was
coeval with the construction of the oratory. As St Michael’s Church is the only building within the
monastery which may have had a slated roof, this would imply that the small oratory post-dates the
construction of St Michael’s Church, which is generally considered to be a late tenth/early eleventhcentury foundation.The suggested late seventh/early ninth-century date for the construction of the small
oratory (see further discussion below) would negate such an interpretation.
The drain was clearly constructed at the same time as the oratory and was planned as an integral
part of the terrace structure. The flow of surface water from the steep rock faces on the western edge of
the terrace was seen as a potential problem at the time of construction (as it still is today), and the inclusion
of this drain in the planning of the terrace was an effort to divert these waters away from the oratory.
The drain must have functioned much like a modern-day French drain, with loose stones and soil used
to fill the structure.The finer silty clay deposits are the result of water percolating through the drain.The
layer of loose stone (F217) that covered the drain seems to have been used to level up the ground surface,
and although it is likely that the paving which is evident elsewhere on the terrace also extended along
the north side of the oratory there are no flagstones surviving.
The eastern area
The removal of peat (0.1–0.15m
deep) on the east side of the oratory
revealed a series of eleven large
flagstones projecting from beneath
the wall of the oratory (Ills 2.12,
2.22). These served the dual
purpose of providing a footing for
the oratory wall and a paved surface
for this corner of the terrace. The
two flagstones closest to the southeast corner of the oratory also
extended under the adjacent
terrace retaining wall. Limited
investigation underneath these
flagstones revealed that they were
sitting on a layer of loose rubble
and clay (F240) and in several areas
the clay had washed through the
retaining wall to the east, leaving
large voids under the stones. Several
of the flagstones sloped down to
the east owing to movement and
subsidence of the adjacent retaining
wall.
The gap between the oratory
and the retaining wall (F216)
narrows to c. 0.3m at the south-east
corner of the oratory (Ill. 2.12).The
Ill. 2.22—The
flagstones
projecting from
beneath the
east wall of the
small oratory
(A. Lynch).
53
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
E1
E2
167
F223
small
oratory
F201
F218
166
F219
F230
F229
165
Ill. 2.23—The
small oratory
terrace,
sectional
profiles.
F201
F1
M.O.D.
F230
paving slabs F220
F2
F201
F18
small
oratory
F202
face of retaining wall F216
F219
steps
steps
F201- campion / light brown peat
F18- nineteenth-century rubble
F202/218- dark brown silty peat
F229- brown sandy clay, primary construction
layer of terrace
F219- dark brown silty clay with rubble
F230- undisturbed boulder clay
F223- brown/orange clay with charcoal
and charred vegetation
movement of the retaining wall is responsible for the narrowing of this passageway, however, because the
upper section of wall had slumped inwards as the lower section bulged outwards.
The southern area
The main body of the terrace lies to the south of the oratory, and the full depth of stratigraphy was
investigated in a 1m-wide cutting that extended for 7m from the south wall of the oratory (Ill. 2.11).
The top layer of peat (F201) was quite shallow on this part of the terrace (average depth was 0.12m),
54
THE EXCAVATIONS
and towards the southern end of the terrace it contained a dump of stone, brick, modern glass bottles, a
piece of cast iron and fragments of mortar (F211). The lower, more humified, moist peat (F218/202),
which had a high silt content and was extremely plastic in consistency, averaged 0.2–0.3m in depth.This
earlier peat petered out c. 0.7m from the oratory wall, where the underlying deposit of dark brown silty
clay (F219) was exposed (Ill. 2.23, section E1–E2). At the interface between the peat and F219, at a
distance of c. 2.1m from the oratory, a deposit of silty brown/orange clay together with charred vegetation
and charcoal (F223) was revealed. This deposit measured c. 0.2m by 0.3m in extent and represents the
remains of a fire built during the early stages of peat formation.
The clay deposit (F219) contained a large number of sandstone blocks with occasional quartzite
blocks and gravel-type inclusions. A few lenses of light brown fibrous vegetation, resembling decayed
roots, were also noted. Finds included part of a rhyolite cross (E338:27), a tiny flint flake (E338:26) and
two pieces of burnt coal or cinder, the latter indicating a nineteenth-century date for the layer.
Underlying the clay and rubble deposit (F219), the remains of paving (F220) were uncovered (Ill.
2.24). With the exception of three quartzite blocks, all surviving paving stones were sandstone, four of
which extended under the south wall of the oratory. The paving stones to the east of the cutting all
sloped down to the east as a result of movement and subsidence of the terrace retaining wall.
Ill. 2.24—Paving
stones (F220)
on south side
of the small
oratory (A.
Lynch).
The paving stones rested on a mid-brown sandy clay, fine in texture and with flecks of charcoal and
lenses of darker humic material throughout (F229). This is the primary construction layer of the terrace
and was c. 0.08m deep under the paving at the oratory wall but deepened considerably to the south (as
the underlying boulder clay dipped down), where it reached a maximum depth of 0.35m at the edge of
the cutting, thereby providing a relatively level surface for the paving.Three fragments of moist, soft bone
(not identifiable) were recovered from the upper levels of this deposit, and a small collection of animal
bone (three sheep/goat mandibles, one cattle horn), bird bones (Manx shearwater and puffin), a vertebra
from a cod and one scallop shell were found in the bottom of the layer, just above boulder clay. A date
of cal. AD 672–869 was obtained from the cattle horn (see Section 4 below).
At a distance of c. 2m south of the south-west corner of the oratory, a rectangular-shaped stone
(0.74m east–west by 0.36m north–south and 0.04m thick, maximum dimensions) was uncovered on the
55
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.25—Stone
under which
human remains
were found and
cross fragment
(E338:27),
which is ex situ
(A. Lynch).
surface of F229 just under the clay
and rubble layer (F219) and at the
same level as the paving stones (Ill.
2.25). This stone was lifted and a
small collection of human bone was
found to lie directly underneath, in
the top of the mid-brown sandy
clay F229. The human remains
consist of five fragments of long
bone, a portion of the right side of
a mandible, some loose teeth, small
slivers of long bone and other
unidentifiable fragments. The
remains are very poorly preserved
and much decalcification and postmortem decomposition has taken
place. Most of the teeth are
represented only by shells of
enamel. Analysis suggests that the
remains belong to one individual, a
juvenile who was probably aged
about 9–12 years at the time of
death (see Section 5 below). The
bones were not found in any
anatomical order, suggesting that
they had either been disturbed and
regrouped or, more likely, had been
redeposited from elsewhere. A
crudely shaped stone cross
(E338:41) with a broken shaft was
found lying on edge in the bottom of the clay/rubble layer F219, immediately west of the stone that
covered the human bones (Ill. 3.12). Its locational association with the human remains may be
coincidental, but it is also possible that the cross once stood at the west end of the stone covering the
remains of the burial.
As far as could be ascertained, the south-west corner of the oratory is sitting on undisturbed boulder
clay (F230), but the remainder of the wall is founded on the paving slabs.
The terrace retaining wall
Viewed from the seaward side, the drystone retaining wall (F216) for the small oratory terrace is the
most impressive section of the monastic enclosure (Ill. 2.26). It is also one of the few surviving stretches
of original walling that has not been significantly altered or modified since construction. The wall was
well constructed using flat sandstone slabs laid horizontally and capped with large, lintel-type slabs.
On the north side of the oratory, adjacent to the ‘latrine’, the retaining wall is built against the rising
bedrock. Further east, as the bedrock dips, the inner face of the wall is founded on boulder clay.The wall
here averages 1.1m in width and, where exposed, the inner face survives to a maximum height of c. 1.1m.
The original height of the wall above the walking surface of this part of the terrace may have been as
low as 0.5m; at some later stage, presumably for safety reasons as the wall slumped eastwards, it was
56
THE EXCAVATIONS
Ill. 2.26—Aerial
view of the
small oratory
terrace, from
east, after
works. (Con
Brogan,
DAHG).
heightened by the addition of an extra 0.7m.This later wall was just 0.8m wide and was easily discernible
from the original, as it was loosely constructed with suggestions of clay bonding and lacked the capping
used on the earlier wall.
As the retaining wall swings south, just to the east of the oratory, only one or two courses of the
upper (later) wall survive. The height of the original wall above the foundation/paving stones averaged
0.3–0.4m, although some distortion of levels has occurred here owing to the considerable subsidence in
this area. At a distance of 0.5m south of the (internal) corner, a recess (0.77m north–south and averaging
0.3m in depth) has been incorporated into the original retaining wall (Ills 2.12, 2.27). A large, lintel-type
stone that covers the back wall of the recess is at the same level as the large slabs forming the top of the
primary retaining wall, reinforcing the suggestion that this is in fact the original height of the retaining
wall. A large block of sandstone was exposed sitting loosely in the recess at the level of the rubble and
clay layer (F240) that underlies the paving stones; as excavation did not extend below this level, the full
depth of the recess is not known. The function of this recess is also not clear. Originally it was thought
that perhaps it contained the outlet for the drain, but there is no sign of such an outlet in the exposed
back wall of the recess or in the outer face of the retaining wall. In addition, the drain does not quite line
up with the recess and what little evidence has survived suggests that the flow of the drain was diverted
northwards (see above).
The width of this north–south stretch of retaining wall varies from 0.6m to 1.1m, but at a point c.
1.6m south of the oratory it widens significantly to a maximum of 1.6m (Ill. 2.12). This is the result of
the addition of a revetment on the internal face to provide extra stability when the original wall began
to show signs of collapse. Both the northern and southern limits of this wide stretch of wall were badly
collapsed when exposed, but at its northern end, close to the oratory, two stones seemed to form
rudimentary steps that would have given access from the area to the south of the oratory to the narrow
passage between the east end of the oratory and the retaining wall, which was c. 0.3m higher.These ‘steps’
lay on the nineteenth-century stony clay deposit F219, which in turn overlay the paving associated with
the oratory (Ills 2.23, section F1–F2, and 2.28). At its southern end, rubble (F18) containing lighthouse-
57
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.27—Recess
in the retaining
wall (A. Lynch).
Ill. 2.28—
Rudimentary
steps and
revetment to
retaining wall
visible on right
(A. Lynch).
period debris (clay pipes and coal) had been dumped on the retaining wall. The maximum exposed
height of this wider wall was 0.95m and its foundations were not uncovered. The rather tenuous
relationship between the wall and the steps would suggest that the widening of the wall took place some
time in the nineteenth century, presumably by the lighthouse personnel who occupied the monastery at
that time. The additional height added to the north wall of the terrace, and to a lesser extent to the east
wall, probably also dates from this time.
58
THE EXCAVATIONS
Interpretation
The small oratory terrace is a highly ordered enclosed space within the inner monastic enclosure.
Although located close to the core area of the monastery, it is somewhat out on a limb at the northern
edge of the enclosure and consequently retains a feeling of shelter, separation and isolation. When one
stands on the terrace, the other monastic structures are hidden from view and the eye is drawn eastwards
to the Little Skellig and the mountains of Kerry on the horizon.The terrace, which is roughly triangular
in shape (max. c. 9m east–west by c. 9m north–south), has as its focal point a small drystone oratory. A
small beehive cell, interpreted as a latrine, was built into the terminus of the terrace retaining wall where
it meets the rising bedrock, just north of the oratory. Removal of peat from the terrace has revealed a
carefully laid-out range of structures, including terracing, steps, paving, a leacht and a drain.
The full depth of stratigraphy on the terrace was only revealed in a very narrow cutting but it suggests
that the primary construction layer was deposited on undisturbed boulder clay some time in the
eighth/early ninth century AD. The date of cal. AD 685–869 obtained from the cattle horn in this
construction layer may relate to the actual construction period, but the possibility that the bone
assemblage was redeposited with the clay used to build up the terrace should also be considered. This
would mean that the eighth/early ninth century would be the terminus post quem for the construction of
the terrace.
The massive drystone retaining wall, measuring up to 5m high externally, may only have been c.
0.5m in height above the early medieval walking surface on the terrace. The indications are that the
entire terrace surface was paved with sandstone (and some quartz) flagstones, some of which also served
as foundation stones for the oratory. The rising bedrock on the western side of the terrace was skilfully
incorporated in the arrangement of steps and paving leading to the oratory entrance. The sense of order
and enclosure was accentuated by the construction of short stretches of drystone walling linking
projecting sections of rock face and bedrock along the western edge of the terrace. Before excavation, a
substantial stone cross that had been found lying on the terrace was erected close to the rock face opposite
the oratory entrance, but owing to its suffering ongoing damage during the tourist season it was later
moved to a less vulnerable position on the terrace, just to the north-west of the oratory. It is likely that
these small, artificially created terraces were used as prayer stations during the lifetime of the monastery,
and they may originally have held some of the small stone crosses that were found just under the sod
close to the oratory and elsewhere throughout the monastery. The leacht, which is also sitting on paving
stones, was constructed on sloping bedrock and is an integral element of the suite of features on the
western side of the terrace. Its probable function as a pilgrimage station would fit in with the
interpretation of the adjacent terraces as prayer stations.
The layout of the small oratory terrace as a whole is strongly suggestive of a predetermined ‘round’
or perambulation, whereby the monk entered the terrace by stepping down from the core monastic area
and made his way to the entrance of the oratory via the first low terrace or prayer station and the leacht.
On exiting the oratory, he would have continued his round via the largest prayer station in the northwest corner, along the narrow passage on the north side of the oratory, returning along the eastern side
of the oratory and stepping down again to the main body of the terrace before returning to the
monastery. It is tempting to suggest that the niche in the terrace retaining wall could have held a large
stone cross. Close to the sharply angled corner of the terrace, this is the most remote point within the
monastery, where solitude would be conducive to contemplation and prayer whilst gazing eastwards to
sea.
The sheer rock face that borders the terrace to the west, and the steeply sloping ground between it
and the ‘latrine’, results in rainwater flowing directly down onto the terrace. In an effort to direct this
surface water away from the oratory, a substantial stone-built drain was incorporated in the construction
59
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
N
K3
J2
K4
Ill. 2.29—Plan
of the large
oratory,
showing areas
excavated.
excavated area
excavated
area
K2
K1
J1
0
5
metres
scale 1:50
of the north wall of the oratory.
The only evidence of human remains recovered was the small assemblage of bone found underneath
the slab on the south side of the terrace and the three fragments found in a similar stratigraphic context
nearby. As most of the monastic deposits on the terrace remain unexcavated, however, it is not possible
to say whether further human remains have been interred in this area.
During the nineteenth century a substantial layer of clay and rubble was deposited on the south side
of the terrace. This must be attributed to the period when the lighthouse-builders inhabited the
monastery and clay and stone were cleared from adjacent areas and spread onto the terrace. The
subsequent growth of up to 0.75m of peat indicates extremely waterlogged conditions on this lower
section of the terrace. Peat growth had commenced prior to the nineteenth century to the north-west
of the oratory, close to the rock face, but the recovery of early nineteenth-century ceramics at a depth
of c. 0.18m in this area also indicates rapid growth in the past two centuries.
60
THE EXCAVATIONS
2.2.2 THE LARGE
Ann Lynch
ORATORY
(E338)
Prior to excavation, most of the floor of the oratory was roughly and unevenly paved with flagstones. A
stone-built altar, incorporating some brick, had been constructed at the east end of the oratory, presumably
by the occupants of the lighthouse, some time in the nineteenth/twentieth century. This altar had been
dismantled before the excavation commenced in order to fully expose the small east window and to
provide room for the excavation to proceed. Fragments of an almost complete wooden crucifix were
retrieved from the rubble fill of the altar (Ill. 3.3) and pennies from the 1920s onwards were found
inserted between the stones.
The northern drystone jamb of the doorway in the west wall was in urgent need of repair and
consolidation, and it was also decided that the rough and uneven paving on the floor should be renewed
in view of the increasing number of visitors entering the oratory. Consequently, two cuttings were opened
within the oratory: a small cutting (0.6m by 0.9m) immediately inside the doorway, abutting the northern
jamb, and a larger cutting (1.31m by 1.71m) encompassing the south-east quadrant of the structure (Ill.
2.29). The latter was opened to examine the subfloor levels under the paving and to allow examination
of the two large slabs set on edge against the south wall of the oratory and the two smaller stones set on
edge close to the east wall, which were exposed when the altar structure was removed.
The south-east quadrant
The sandstone slab (F311) set on edge against the south wall of the oratory was exposed to its full depth
in a narrow (0.35m-wide) cutting at the western edge of the quadrant (Ills 2.30, 2.31). Elsewhere, the
quadrant was excavated to the level of the top of the redeposited boulder clay (F315). The slab on edge
(F311) measured 1.7m in length, averaged 0.16m in thickness and, where fully exposed, was c. 0.9m in
height. It had been set on edge in the top of a rich brown, greasy, humic layer (F318), which averaged
0.09–0.1m in thickness and must represent the sod layer of an old ground surface (Ill. 2.31, section J1–J2).
Pollen analysis carried out on samples from this layer has provided insights into the contemporary local
vegetation (see Section 6 below). The underlying undisturbed boulder clay (F319) was a moist, hardpacked stony clay. Once the slab had been placed in position, boulder clay was deposited behind it to a
maximum depth of c. 0.5m, to hold it in position while the drystone masonry of the south wall of the
oratory was constructed against its south face. Immediately east of F311, a second slab on edge (F312),
0.9m long, was partially exposed and presumably served the same function as F311 (see interpretation
below).
In the centre of the cutting, a thin lens of reddish-brown peaty soil (F316) lay on top of the
redeposited boulder clay (F315). This peaty soil contained what appeared to be decayed vegetation that
still retained its rich golden-brown colour. On analysis, this was identified as matted organic material,
possibly grasses and rushes, with very small stones and tiny pieces of charcoal embedded in it.This material
had been criss-crossed, possibly to form a type of matting or floor covering laid down when the oratory
was first constructed (grateful thanks to Ellen O’Carroll for identifying this organic material). A stakehole (F317) was recorded cut into the redeposited boulder clay under the peaty soil (Ill. 2.30, level 4). It
measured 40mm in diameter, 80mm in depth and appeared as a void with no evidence of decayed wood
in situ, suggesting that the stake was removed shortly before the peaty deposit was laid down, which then
sealed the top of the void and prevented soil from overlying deposits from filtering through.
The deposits that accumulated during the early medieval and later periods of usage of the oratory
average 0.25m in depth. They comprise the main deposit of crumbly brown clay with small gravel-size
stones, charcoal flecks and smears of burnt clay (F306) and three distinct areas of burning (F304, F310
and F314) (Ill. 2.30). A lens of sticky beige clay (F309) and lenses of greasy humic brown clay (F313)
61
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
N
F322
burnt clay
& charcoal
LEVEL 1
F304
F323
F321
F305
F306
F306
F303
F311
F312
N
LEVEL 2
F327
F304
F305
F306
Ill. 2.30—Plan,
at different
levels, of
features
exposed during
excavation of
the large
oratory.
Level 4 is the
lowest level.
F309
F310
F311
F312
N
find-spot of
lignite bead
LEVEL 3
F304
F306
F314
F310
F311
F312
N
excavated to
lower level
LEVEL 4
F315
stake-hole
F317
F316
F306
312
0
311
5
metres
62
find-spot of
knife handle
THE EXCAVATIONS
J1
J2
F312
oratory
wall
F313/F316
F313 F310
F303
compact
stony soil
F302
164
F304
F306
F313/F316
F320
F315
F319
pollen sample
(fine)
163
pollen sample
(coarse)
F302- dark brown sandy clay
F320- dark brown humic clay
F306- brown clay
F315- redeposited boulder clay
F304- brown clay
F318- brown greasy humic layer
F310- black/grey sandy clay
F319- boulder clay
F313/316- red brown humic/peaty clay
Ill. 2.31—Eastfacing section of
south-east
quadrant of the
large oratory.
M.O.D.
F318
- burnt clay
were also recorded. The burnt deposits, which averaged 0.1m in depth, were remarkably similar in
composition—yellow/orange burnt clay and black charred material in distinct layers, indicating that the
burning took place in situ. Traces of burnt shell were noted in the burnt clays of F304. The date and
purpose of this burning are difficult to determine—with the exception of a flint waste flake (E338:31)
found in F304, no artefacts were recovered from the deposits.The edges of the burnt areas were not well
defined and they did not have the appearance of formal hearths. The lowest level of F304 also
incorporated some matted organic material similar to that found on the surface of the underlying
redeposited boulder clay.
The two parallel stones set on edge at right angles to the east wall of the oratory (F305), which were
visible once the modern altar was removed, were set in the brown clay (F306). Each of the two stones
was 0.61m long, 0.03–0.05m thick and 0.31–0.33m in height, and they were set 0.03–0.06m apart (Ill.
2.30). While no sockets as such could be identified, smaller flat stones had been set on edge and used as
support/packing stones on the south side, while similar but looser stones on the north side may have
served a similar function. About eight smaller stones were loosely set on edge in the top of the brown
clay (F306) and these seemed to form a rough arc enclosing the larger slabs and a concentration of loose
stones. Without further excavation in the north-east quadrant of the oratory it is difficult to interpret
this arrangement of stones, but its location, centred on the east window, could suggest association with
an earlier altar structure.
The main clay deposit (F306) was quite homogeneous throughout the cutting: brown, sandy texture
63
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
with gravel-sized inclusions, charcoal flecking and lenses of greasy clay (F313)—in general a fairly mixed
appearance. An irregular-shaped pocket of F306 was uncovered in the top of the underlying boulder
clay (F315), adjacent to the wall slab (F311) on the south side of the oratory (Ill. 2.30, level 4). The
remains of an iron knife with part of a wooden handle attached (Ill. 3.6, E338:30), which must have been
lost or discarded during the construction of the oratory, was recovered from this pocket of clay.The only
other artefacts recovered from F306 were a lignite disc bead (Ills 3.10, 3.15, E338:28) found close to the
east wall and a small waste flint flake (E338:32).
A band of compacted stony soil with small water-rolled pebbles (F303) was uncovered running
across the centre of the cutting on top of the clay deposit (F306) and directly underneath the modern
clay floor surface. This sterile layer may have been deposited to improve the floor surface at some late
stage in the use of the oratory.
The clay deposit (F302) that underlay the paving slabs and that formed the floor surface where
exposed between the slabs consisted of a brown sandy soil, dry and powdery in texture except along the
southern side of the cutting, against the wall of the oratory, where it was moist and sticky owing to
ingress of rainwater. This layer was very shallow (only 20mm deep in places) at the western end of the
cutting but deepened considerably towards the east. The range of artefacts recovered indicates a very
mixed and disturbed layer, which
is to be expected for what was a
floor surface over several
centuries. The material recovered
from this layer includes small coal
and cinder pieces, fragments of
decayed wood, mortar fragments,
an iron stud, twentieth-century
pennies,
sherds
of
early
nineteenth-century ceramics and
modern bottle glass. Three
artefacts belonging to the
medieval period were found in
this layer—a silver long-cross
Ill. 2.32—Southpenny of mid-thirteenth-century
east quadrant
date (E338:23), part of a lignite
of the large
oratory after
ring (Ills 3.10, 3.14, E338:24) and
excavation
a flint flake with retouch
(Con Brogan,
(E338:25).
DAHG).
The entrance cutting
A small cutting (0.9m north–
south by 0.6m max. east–west)
was opened immediately inside
the oratory entrance to facilitate
repair of the northern jamb of the
doorway (Ill. 2.29).
The stratigraphy revealed was
broadly similar to that in the
south-east quadrant, although
shallower (Ill. 2.33). The old
64
THE EXCAVATIONS
164.5
K2
K1
F302
F321
black humic lens
F328
164
M.O.D.
threshold stone
F329
boulder clay
K4
K3
F326
K3
Ill. 2.33—The
large oratory,
sectional
profiles.
164.5
164.5
K2
164
M.O.D.
F324
F328
F325
F328
164
M.O.D.
F302
F302
F329
F329
boulder clay
F302- dark brown sandy clay
F325- mixed layer similar to F324
F321- compact stony,
F328- redeposited boulder clay
F324- sandy clay with charcoal flecks
F329- brown greasy humic layer
ground surface (F329–F318) was no more than 0.01–0.02m thick, covered by 0.21–0.23m of redeposited
boulder clay (F328–F315).The vertical slate slab (F326) that forms the face of the west wall of the oratory,
just north of the doorway, was set into the redeposited boulder clay, indicating that at least some of this
clay was in position before the oratory wall was constructed.
The ‘activity levels’ relating to the use of the oratory comprise a single mixed deposit (F321–F306),
0.2m in maximum depth. This is a compact, stony, brown sandy soil with a lens of mottled black and
orange clay with traces of charcoal and burnt shell. A roughly circular post-hole (F327), c. 0.13m in
diameter and 0.17m deep, had been cut through this deposit (Ill. 2.30, level 2).This was a well-constructed
feature, with four stones on end lining its sides and two stones forming its base. It appears to have been
deliberately backfilled with a stony brown sandy soil (F327A), similar to F321, once it went out of use.
Its location inside the doorway and almost 0.2m from the nearest door jamb would suggest that it did
not function as part of the door furniture. Even if, as was the case with many drystone churches, the
wooden door was secured through the use of a perforated slab that projected from above the lintel on
the interior of the doorway (Ó Carragáin 2010, 52), this post-hole is sufficiently out of line with the
northern door jamb that it could not have held one of the stationary posts for the door frame. Further
excavation on the south side of the entrance could reveal associated features that would help in its
interpretation.
An irregular-shaped cut (F322) was revealed centred on the northern door jamb (Ill. 2.30, level 1).
This feature had a maximum depth of 0.15m and was filled with loose, moist, dark brown clay (F323).
The bottom of the cut coincided with the base of the drystone door jamb and this feature must represent
an earlier effort at accessing the jamb for repair.
65
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Interpretation
Even though just a small area of the oratory was excavated, details have come to light about how the
structure was constructed and about its subsequent usage.
A number of large sandstone slabs were set on edge on the line of the inner face of the oratory walls
and boulder clay was deposited behind them, in the space which would form the interior of the oratory.
The drystone walls were constructed against the outer faces of the slabs. The stake-hole recorded in the
boulder clay probably belongs to the construction phase of the oratory.
Excavation in the area south of the oratory revealed a stepped plinth on the south wall of the oratory
(see Section 2.2.3 below). The line of the lower step of the plinth deviates from that of the oratory wall,
suggesting that it may perhaps relate to an earlier structure. Excavation within the oratory did not reveal
any evidence to support this hypothesis, however, and what is interpreted as undisturbed boulder clay
(F319) was exposed at a level considerably higher than the putative earlier structure.
Over the centuries, clay deposits accumulated to depths of up to 0.3m within the oratory—
presumably these clays were brought in to cover the floor at different times, although the level of mixing
and disturbance of deposits did not allow for any clear stratification to be discerned. In the area excavated,
there were no traces of structures or features that could be related to ecclesiastical activities, with the
possible exception of the ephemeral features exposed below the east window, which relate to an early
altar structure. Extensive fires were lit within the oratory at some unknown date—these could, in fact,
have been as late as the nineteenth century. The large altar structure that had been built against the east
wall, partially obstructing the east window, may have been constructed for use by pilgrims during the
post-medieval period but it more likely belongs to the nineteenth century, when the lighthouse personnel
made use of the oratory for religious practice.
2.2.3 SOUTH ENTRANCE 1 (INNER
Edward Bourke
ENCLOSURE) AND LEACHT AREA
(93E0195)
Introduction
In 1993 a test-trench was excavated in the area south of the monks’ garden to establish stratigraphic
relationships that would allow the south-east corner of the inner enclosure to be presented in a coherent
way (Ill. 2.34).The test-trench measured 5m by 1.5m and ran in an north–south direction (Ill. 2.35) from
the wall of the inner enclosure (F513) to a low retaining wall based on bedrock above the east end of
the lower monks’ garden (F1006). It was excavated to determine the state of conservation of the eastern
end of the inner enclosure wall, which was not visible on the surface, and there was a slope of campion
running down over it from the inner enclosure to the south.
It became clear that the early portion of the inner enclosure wall (F513) (Ill. 2.36), which ran
eastwards from the south wall of St Michael’s Church, terminated at its eastern end in a jamb with the
inner lintel of an entrance surviving. This lintel was visible on the surface to the west of the test-trench.
This feature appeared to be an early entrance (south entrance 1). The exterior face of this wall did not
survive east of this jamb and the jamb itself was in urgent need of conservation(Ill. 2.37).
West of the area excavated lay part of the inner enclosure wall, characterised by being built of stones
laid at right angles to the line of the wall. This presents an extremely jagged front elevation facing out
onto the lower monks’ garden. This is the only part of the monastery with such a wall still in position
and it was felt that it probably represented an enclosing element from an early phase of the monastery. It
was hoped that the excavation would reveal whether the front footings of this wall could be found
continuing eastwards across this area.
It was decided at this point that the conservation of this portion of the inner enclosure wall and of
66
THE EXCAVATIONS
Ill. 2.34—The
eastern end of
the lower
monks’ garden
and the inner
enclosure from
the south-east
in the 1970s
(DAHG
Archive).
the entrance was going to be a major operation, and that this work could not be undertaken until
excavation and conservation of the lower monks’ garden was completed.
Work recommenced briefly in 1994, when an attempt was made to find the front face of the base
of this wall. The front face did not in fact survive east of the entrance, and it was decided that further
work involving the excavation of the entrance would be necessary. This work did not take place until
1998, when the area of the entrance was excavated and the completion of a new wall, east of the entrance,
allowed the jacking back into position of the exposed jamb stones.
The 1993 excavation
The test-trench ran from the edge of the paved area of the inner enclosure down to a ledge revetted by
a single-faced drystone wall (F1006) that stands on bedrock overlooking the lower monks’ garden (Ill.
2.38). This wall ran east–west along the edge of the ledge for c. 4.8m.
A layer of sea campion roots and dark brown organic material (F1001)1 sloped downwards from the
paving of the monastery to the ledge above the lower monks’ garden at an angle of c. 40º. This layer
varied in thickness between 0.3m and 0.7m and represented a recent build-up of soil over this area (Ill.
2.39). It overlay a small east–west retaining wall (F1007) and a layer of dark, loose organic soil with small
stones (F1002). South of these features, F1001 overlay a thin layer of decayed sea pink roots (F1003), up
to 0.1m thick, which represented an old sod level in this area. This in turn overlay a layer of dark brown
to black organic soil with small stones (F1004), which was very similar to F1002. This overlay the
revetment wall (F1006), which retained this material from falling into the lower monks’ garden, and the
sloping bedrock behind it.This wall was constructed of relatively small stones and was only faced on the
1
All features numbered F1000 or higher are from contexts for which the notebooks, finds registers, plans and some of the
sections were destroyed in a fire in the archaeologists’ accommodation in 1995 (see Introduction for details).
67
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.35—South entrance
1 from the south, showing area of entrance and
area east of entrance in
1994 (Con Brogan,
DAHG).
outward or southern side. The ground was levelled up behind it by backfilling with rough, loose, large
stones and brown earth, between it and the bedrock. All this material appeared to date from the period
of lighthouse occupation, as creamware, coal and brick were found in the wall (F1006) and the backfilling
behind it (F1005). String, cigarette packets and nylon twine were also incorporated into the layer (F1005),
probably by burrowing sea birds.
The small upper retaining wall (F1007) was on the line of the original enclosing wall (F513) but
was clearly built to retain the collapse of that wall at a later date. It was built as a low revetment wall with
no rear face and none of the stones were larger than 0.45m by 0.3m by 0.25m.
Excavation of the entrance
The excavation of the early entrance took place in 1998.The lintel, F1008, which measured 1.24m long
by 0.28m deep by 0.18m high, lay in position and was keyed into the inner face of the western jamb of
the entrance. It did not appear to be attached to an eastern jamb, which appeared to be missing. It was
decided to excavate the entrance and record how much of it survived, and to determine its position in
relation to the large oratory. It was hoped that some trace of the original exterior of the wall to the east
of the entrance would survive, as five large building stones were exposed, slumped out of position, in the
1993 season (F1010). These stones, which measured up to 1.8m by 0.3m by 0.4m, were similar in size
68
THE EXCAVATIONS
Ill. 2.36—Western jamb and
lintel of south
entrance 1 visible on surface,
1994 (E.
Bourke).
Ill. 2.37—Inner
and outer enclosure walls
prior to excavation (Con Brogan, DAHG).
to the larger building stones of the wall west of the entrance. Excavation in this area showed that these
appeared to be the last of the eastern jamb stones, thrown out of position in the collapse of the wall.
Further work indicated that the front face of the wall did not survive east of the entrance.
Description of entrance
The western jamb of this entrance survived to a height of 3.05m above the bedrock that formed the
base of the entrance.The wall was c. 1.82m thick at its top and c. 2.27m wide at its base.The interior face
69
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
M2
M1
unexcavated
F1001 - dark brown rubble / campion
F1001
F1002 - dark organic
F1002
F1003 - decayed roots
F1004 - dark organic
F1005 - dark organic
wall
F1007
F1003
Ill. 2.38—
Section of 1993
cutting.
unexcavated
F1001
bedrock
F1004
wall
F1006
bedrock
F1005
0
5
metres
was almost vertical and the exterior face displayed a slight batter. The jamb stones were aligned north–
south, parallel with the entrance, and this caused a weakness in the structure. A large stone (up to 2.4m
long by 0.61m wide and 0.4m thick) at the base had been forced out of the wall and lay at an angle of
c. 30º across the entrance. Above this were a series of large stones, also aligned north–south, the largest of
which was 1.72m long by 0.49m wide by 0.27m deep. All of the lower stones had moved in antiquity
and the side stones presented a bowed shape, being most severely displaced halfway up the jamb. The
inner face of the wall had stones that lay parallel to the wall. Thus the alignment of the inner face was at
right angles to the stones of the outer face and entrance jamb. This inner face survived west of the jamb
but could not be discerned east of the entrance at this time. It was encountered intact in the area of the
east entrance (see Section 2.2.6 for details).
The eastern jamb had entirely disappeared in the collapse and no trace of it could be discerned, nor
of the outer face of the wall to the east of the entrance.The inner lintel (F1008) survived in position and
measured 1.24m wide, 0.28m deep and 0.18m high, which would indicate an entrance width of c. 1m
at its top.
70
THE EXCAVATIONS
Ill. 2.39—
General view
showing trench
locations in
1994 (Con
Brogan,
DAHG).
The excavation
A cutting, 4.3m by 2.5m, was opened through the entrance in order to examine the stratigraphy and the
relationship between the entrance and the plinth of the large oratory (Ill. 2.40).
Monastic steps (phase 1)
Above the bedrock within the entrance area lay a very thin (0.02–0.05m) layer of pale grey clay (F554).
This ran under the inner enclosure wall (F513) and underlay a set of low steps (F553) that ran up through
the entrance from midway through the wall. This series of three steps ran under the plinth of the large
oratory and represent a time when the layout of the monastery must have been quite different. Two
further steps also run under the plinth (F559A), but it is not clear whether these represent part of this
phase of the entrance or the later phase (see below). These steps protrude through later layers, but it is
unclear whether they had already been displaced when the phase 2 steps were inserted or whether they
form part of that second phase.
Monastic steps (phase 2)
Above the earlier steps and bedrock lay a layer of dark reddish-brown compacted material with large
stones (F552).This layer was up to 0.46m thick and contained the large displaced stone from the base of
the west jamb described above. This layer and the stone appear to represent a failure in the structural
integrity of the wall, and a second level of steps was then added to the entrance above the level of the
large displaced stone. Above this lay a layer of grey shattered grit with medium-sized stones (F551) that
ranged in thickness from 0.02m to 0.16m. Above this was a layer of medium-sized stones and spalls
71
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
X2
X1
puffin
burrow
F544
F536
F553
lintel
large
oratory
plinth
F534
F535
W2
W1
F537
F540
monastery
wall
F555
F541
F545
F542
F543
F546
F547
F547
F548
projecting
stone
Ill. 2.40—Section through
south entrance
1.
F549
F550
steps F553
rubble 556
F552
F551
unexcavated
F552
displaced
stone
pale grey clay
unexcavated
bedrock
0
5
metres
F553 - campion / topsoil
F541 - dark brown clay
F550 / F549 with medium stones
F535 - black stony soil
F543 - dark grey / brown sandy clay
F551 - grey shattered grit
F536 / F540 / F542- yellow / brown ash
F546 - sandy soil, ash & small stones
F552 - compact dark red / brown
F537 - dark grey stony clay / yellow ash
F548 - dark brown loose soil
F547 - dark brown / grey silty soil
F549 - bluish-brown silty soil
(F550), up to 0.2m thick, which overlay both F551 and F552. Within this layer (F550) were two steps
(Ill. 2.41) that had been displaced by pressure from the weight of the plinth of the large oratory. It is not
clear whether these two steps were in situ when F550 was deposited, and therefore would have belonged
to phase 1, or whether they form part of the phase 2 steps.The drystone risers had fallen away when the
steps were displaced and it is therefore not clear whether they were steps of phase 1 or new steps inserted
as part of the less formal stairway of phase 2.
Above this was a layer of almost bluish-brown clay (F549), between 0.01m and 0.09m thick. This
was overlain by a layer (up to 0.22m thick) of dark brown loose soil with stones (F548). This in turn
underlay a layer of dark grey/brown clayey silt (F547), up to 0.13m thick, which was quite waterlogged
72
THE EXCAVATIONS
Ill. 2.41—Upper
level of steps,
south entrance
1, from east (E.
Bourke).
and had been cut in half by the overlying layer, F546, a beige mixture of sandy soil, ash and small stones
between 0.14m and 0.38m thick. This underlay a flat slab angled at 45º to the line of the earlier steps
and appears to be where the phase 2 steps turned westwards, respecting the line of the plinth of the large
oratory.
West of this slab there was a line of flat slabs, only three of which could be called steps, which roughly
filled the area between the earlier plinth and the south wall of the inner enclosure (F513).Three slabs lay
on the surface and there were two slabs under which the remnants of drystone risers remained.
Features post-dating phase 2 steps
Overlying the flat slab that lay on the surface of F546 (Ill. 2.42) was a layer of loose rubble (F544), up to
0.22m thick, which had some of the overlying layer (F543), a dark brown silty clay up to 0.08m thick,
washed down into it.
Above this was a layer, up to 0.36m thick, of yellow ash with lenses of brown clay and some charcoal
(F542). It in turn underlay a layer of dark brown clay and small stones (F541) up to 0.2m thick. All these
layers (F542–F544) had been deposited into the entrance behind the blocking wall (F557, F544 and
F545), which closed off the entrance (see below). Further west in the cutting the layers associated with
the backfilling of the phase 2 steps continued, but in the area of the entrance itself later (nineteenthcentury) work overlay F542, where the lighthouse-builders dug out some of these layers while laying a
foundation for widening the plinth (see below).
The area west of the blocking (F541) was overlain by a layer of yellow ash with some small stones
(F540), up to 0.2m thick, which in turn was overlain by a layer of dark earth and small stones (F539).
This layer only survived up against the northern edge of the cutting. South of F539 was a layer of
brownish-yellow ash (F538), 0.02–0.07m thick, which overlay F539, F540 and F537. A small patch of
dark grey clay with stones (F537), 0.11–0.28m thick, lay at the south end of the cutting below F536, a
layer of yellow ash up to 0.06m deep.
These layers represent the backfilling of the phase 2 steps after the entrance was blocked up. The
73
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.42—Upper
level of steps,
south entrance
1, from north
(E. Bourke).
layers associated with the phase 2 steps had eroded out into the entrance, which was then blocked with
a rubble-filled wall. This wall had an inner face (F557) consisting of ten courses of well-built drystone
walling and a rubble core filled with loose rubble and black organic material (F545). Above F545 was a
layer of cleaner rubble without organic material (F532), which also formed the filling layer behind the
inner face of the wall (F557).The outer side of this blocking wall had eroded away and slumped out into
the entrance. This feature stood on top of F546, upon which one of the stones of the phase 2 steps lay.
All of the backfill layers (F544–F536) had been dumped inside this blocking wall. Above the topmost
slab of this stairway (F559) lay F583 (see ‘The leacht area’, below).
Nineteenth-century deposits
At the east end of the cutting there was a nineteenth-century intervention where the flagstones forming
the top of the plinth of the large oratory were added to in order to provide a wider paved surface. A
layer of rubble underpinning (F531), up to 0.55m deep, was inserted in order to provide a foundation
for the newly widened paved area. This rubble underpinning had cut through a layer of black loose soil
with stones and campion roots (F530), which was retained by the blocking wall (F577) and overlay the
ash layer (F542).This latter deposit contained brick, some fragments of corroded iron and a stone (green
conglomerate) from St Michael’s Church.
West of this disturbance and overlying the backfilled layers was a layer of black, very stony soil with
few roots (F535). This layer, which was 0.2–0.3m thick, ran over the backfilled deposits, over F583 and
the eastern ends of the layers associated with the burials in the leacht area. Above this was a layer of loose
black clay with stones and campion roots (F334), 0.3–0.58m deep. This directly underlay the campion
layer that covered the whole area up to the leacht and measured between 0.08m and 0.12m in depth to
the east of the leacht. The northern edge of the layers described above (F535–F533) was revetted with a
single-faced drystone wall abutting the paving surrounding the large oratory, and on its southern edge
layers F533 and F534 tapered out over the centre of the enclosing wall (F513).
74
THE EXCAVATIONS
Ill. 2.43—Lower
plinth beneath
large oratory,
from east (E.
Bourke).
Interpretation
It is not clear what the layout of the inner enclosure was when south entrance 1 was in use, as it predates the building of the large oratory by some considerable time. The steps of phase 1 run north–south
and continue under the plinth of the large oratory (Ill. 2.43), so the layout of the interior of the inner
enclosure must have been different at that time.The second phase of activity occurred when the entrance
had been partially blocked up with soil and the large oratory and its possible predecessor had been built.
At this stage, steps were built between the edge of the oratory plinth and the enclosure wall, leading
down to the partly blocked entrance. It appears that the phase 2 entrance had gone out of use either at
the time of or before the arrival of the builders of the lighthouse on the island in the 1820s. It is hoped
to obtain a radiocarbon date from material from F545, which may resolve whether the blocking of the
entrance belongs to the medieval period or the nineteenth century.What is definite is that the lighthousebuilders widened the southern edge of the plinth of the large oratory by paving alongside it. At some
later stage there was a catastrophic failure of the wall between the entrance and the south-east corner of
the inner enclosure, and the outer face of the wall collapsed into the lower monks’ garden, burying Cell
G (see below, Section 2.2.7).
Excavation to the east of the entrance showed that while the outer face of the wall had fallen away
much of the core remained, with the inner face intact up to the level of the ground inside the inner
enclosure.The rubble core of this wall to the east of the entrance was left in situ and no further excavation
was carried out here. A retaining wall was built on the outside to hold back the surviving features of the
wall to the east and to allow the entrance to be blocked (Ills 2.44 and 2.54).
The leacht area
Introduction
Excavation commenced in the interior of the enclosure between St Michael’s Church and west of
entrance 1 in 1999. There was a leacht (F560) visible in the north-west corner of this area, and a puffin
75
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
had been observed ‘excavating’
human remains from within the
area to the north-east of the leacht.
Pre-leacht features
Below the pavement associated
with the leacht the remains of a cist
(F599) were revealed (Ill. 2.45).This
structure consisted of a long stone
(1.32m by 0.12m) set on edge,
which formed the eastern side of
the feature, but its western and
southern sides had been truncated
Ill. 2.44—South
by the building of St Michael’s
entrance 1 after
Church. The structure had been
conservation
(C. Brogan,
divided into two compartments by
DAHG).
the insertion of two stones set on
edge midway within it.These were
aligned east–west and measured
0.22m by 0.03m and 0.35m by
0.04m.The northern compartment
appeared to survive to almost its full
width,
but
the
southern
compartment
was
severely
truncated by the foundations of the
later church.
Within the cist, a 0.38m-deep
layer of mid-brown silt with some
charcoal (F588) was revealed. It
contained a fragment of a human
14
skull and seven water-rolled pebbles. A C sample from this layer yielded a date of between AD 779 and
970. This layer formed the fill of both compartments of the cist, while outside to the south was a layer
of gritty mid-brown clay with very little charcoal (F589).Within the fill (F588) there was a vertical stone
(F600), 0.12m by 0.48m and aligned north–south, with propping stones driven in around it. This stone
was in fact the in situ base of a large cross (Ill. 2.46). The cross, which had been broken from its base, was
recorded in the mid-nineteenth century standing in the structure known as the ‘guesthouse’ (Dunraven
1875–7). Subsequently it had been moved and placed upside down within the area of excavation.
Originally it had been set up within the cist, and later the leacht was added up against its eastern face.
When originally set up this cross would have matched a similar cross mounted in the burial platform on
the northern side of the large oratory (Ill. 2.47).
The paving
An area of paving extended south of the leacht and directly underlay it (F582).This arrangement of large
flat slabs, which varied from 0.08m to 0.19m in thickness, was laid out west of the top of the secondary
steps (F559) coming up from east entrance 1 and was roughly contemporary with them. At its western
edge the paving continued beyond the leacht but the stones here sloped downwards towards the west
(F582A). It became clear that this was because they rested at their eastern end on the upright slabs of the
76
THE EXCAVATIONS
large oratory
N
position of
lintel
ry
l
ar
e
29 for
F5 inth
pl
Ill. 2.45—Plan
of F599 and
earliest levels
around
cist/shrine.
to
ra
o
ier
ps
te
s
59 2
F5 ase
ph
r
de
F600
ain
un
su
re
82
F588
re
F5
lo
nc
F5
F599
east entrance 1
13
dr
ne
n
ll i
st
wa
ea
F589
0
5
metres
Ill. 2.46—F599
with base of
cross and
dividing stones
of cist/shrine (E.
Bourke).
77
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.47—Leacht
and cross after
conservation
(E. Bourke).
cist (F599). It was also clear that the digging of the foundations for St Michael’s Church and the
subsequent burial of Skeleton 3 (see below) had caused their subsidence to the west. A line of only three
slabs of F582A survived, and these averaged 0.6m by 0.4m by 0.1m.The northernmost of these slabs had
been rebated to fit around the base of the large cross (F600).
The drain
Prior to the laying of the paving, an L-shaped drain was constructed to carry water away from the area
(Ill. 2.48). This ran east–west for c. 1m and then turned to run north–south, continuing on to meet the
wall of the inner enclosure (F513). The drain was 0.35–0.5m wide and up to 0.2m deep. The lintels that
covered it formed part of the pavement (F582).
The base of the drain at its southern end consisted of a layer of extremely gritty mid/dark brown
sandy silt (F597), which underlay the side stones of the drain. Above this was a layer of compact dark
brown sandy silt (F596), which ran almost the full length of the drain and was up to 0.15m thick in
places. A small pocket of material (F595), only 0.05m thick, lay between the flagstones that formed the
base of the drain at its western end. Above this there was a layer of compact dark brown clayey silt (F594),
which became progressively stonier towards its base; this layer was confined to the western half of the
drain. Above this was a layer of compact dark brown clayey silt with a high content of grit (F593), which
ran the entire length of the drain and averaged 0.06m in depth. Samples were taken from F593 and F594
for analysis, but they had been contaminated by nesting storm petrels (see insect remains report). Above
this lay F592, a thin layer of brown-black silt that was only 0.01m deep. This was overlain by F591, an
almost identical layer containing slightly more stones. The uppermost layer in the drain was a fairly
compact brown sandy silt, 0.03m deep (F590), which contained a piece of flint.
The leacht
The leacht consisted of a rectangular drystone construction, 2.95m by 0.99m by 0.65m high, wedged
between the base of the large cross at its western end and an upright slab at its eastern end. It was taken
down, stone by stone, and planned, photographed and numbered at each stage down to the paving. No
78
THE EXCAVATIONS
Ill. 2.48—Drain
and paving (E.
Bourke).
large oratory
N
position of
lintel
Ill. 2.49—Plan
of burials, phase
1a.
leacht
F560
98
w
F5
re
ton
ele
sk
rro
3
ton
le
ske
1
ffin
F5
east entrance 1
13
su
lo
bu
pu
er
c
en
nn
st
li
al
w
ea
0
5
metres
79
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
large oratory
N
W1
position of
lintel
S1
X1
Ill. 2.50—Plan
of burials, phase
1b.
leacht
F560
east entrance 1
98
on
let
e
sk
w
rro
2
n
ffi
pu
bu
F5
re
S2
F5
13
X2
W2
su
lo
nc
e
er
nn
ll i
a
tw
s
ea
0
5
metres
soil or other features were noted within the leacht. Above the drystone construction was a layer of silty
sand with some quartz pebbles and some mortar (F573). This layer extended slightly beyond the leacht
to the south, where it overlay F576 and therefore post-dates the burials to the south. Overlying this was
a similar layer of silty sand with some quartz pebbles and angular stones (F572) and which also contained
some bird and fish bone.This layer averaged 0.3m in depth.The topmost layer on the surface of the leacht
was a layer of sea pink roots (F571), which averaged 0.2m in depth.
Monastic features post-dating the leacht
Directly above the pavement (F582) lay a brown silty deposit with some grit and charcoal (F581), which
produced a radiocarbon date of between AD 775 and 941. North of F582 and against the leacht was a
mottled redeposited thick layer of sand (F583), which had been dug through to position a stone associated
with the later burials.This post-dates the leacht itself and appears to be roughly contemporary with F581,
as indicated by the 14C date of between AD 778 and 942, which is almost identical.
Above F581 was a patchy deposit of sand and clay to the south and east of the leacht (F580), which
contained some metal fragments and a water-rolled pebble. This deposit is all that remains undisturbed
by puffins at this level. The rest of the area at this level is occupied by F579, a 0.1m-thick layer almost
completely disturbed by puffins. Above F579 and F580 was a layer of loose, dark brown sandy silt (F577).
This contained a large quantity of bird bones and disarticulated human remains, together with an iron
nail, a piece of milky quartz, 32 roof slates, an incised slate cross, a stone disc and six water-rolled pebbles.
This layer in turn underlay a compact mottled yellow/red/pale brown clay with some minor ashy deposits
(F576), measuring 1.12m north–south by 0.8m east–west, which contained a water-rolled pebble and a
80
THE EXCAVATIONS
Ill. 2.51—
Skeleton 1,
showing puffin
disturbance (E.
Bourke).
S2
S1
0
0.5
1
large oratory
metres
Skeleton 2
Male 25–30
AD 1040–1280
south wall
inner
enclosure
Skeleton 1
Juvenile
AD 995–1180
Ill. 2.53—
North–south
section at
leacht.
leacht
unexcavated
piece of rock crystal. It in turn underlay a layer of brown earth with few stones (F564), the burial matrix
containing Skeletons 1 and 2 (Ills 2.49–50). This layer extended from a near-vertical stone beside the
leacht to the south wall of the enclosure (F513), but no grave-cut was observed at either its eastern or
western end.
Skeleton 1 was a juvenile (9–11 years) who was laid out in a supine position in a roughly southwest/north-east direction (Ill. 2.51).This burial had been disturbed from below by a puffin burrow (F598).
Skeleton 2 overlay Skeleton 1 and partly disturbed it. Skeleton 2 was a young adult male laid out in a
supine extended position, also aligned in a roughly south-west/north-east direction (Ill. 2.52). Skeleton
1 was dated to AD 995–1180 and Skeleton 2 to AD 1045–1280. Both skeletons had been disturbed by
a puffin burrow (F598) that contained animal and bird bone, brick, modern glass, mortar, nineteenthcentury pottery and a water-rolled pebble. Seventy-four fragments of human bone were found within
81
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
the burrow.
Above the burial matrix (F564) was
a layer of medium-sized stones and
organic material which appears to have
been laid down to protect the burials. It
extended from the leacht to the
enclosure wall and was up to 0.25m
deep, covering Skeleton 2. This layer
was truncated at its south-western edge
by nineteenth-century disturbance and
at its western edge by the burial of
Skeleton 3.
West of F564 (the layer containing
Skeletons 1 and 2), disturbance had
been caused by the digging of
foundations for St Michael’s Church,
and later to insert another burial
(Skeleton 3). These disturbances had
removed the western side of the cist
(see above) and appear to have removed
the edge of the grave-cuts for Skeletons
1 and 2.
Some time after the construction
of St Michael’s Church, a burial
Ill. 2.53—Skeleton 3, from north (E. Bourke).
(Skeleton 3) was interred against the exterior
of its east wall (Ill. 2.53).This burial had been
inserted into the church’s foundation trench,
with the lower body overlying some of the
foundation stones and the upper body
overlying a compact gritty silt (F575). The
fill of the grave consisted of a layer of angular
stones and brown organic material (F568)
and no grave-cut could be discerned.
Skeleton 3 was an adult, 50–60 years
old, and was laid out in a supine extended
position aligned south-east/north-west.
Only the leg bones, pelvis and vertebrae
were uncovered in situ.The rest of the bones
were redeposited in a pile next to the
skeleton, as they had been disturbed when a
wall footing was put in, possibly to protect
the leacht. This redeposited material (F578)
consisted of a dark brown, fairly loose, rich
humic soil which also contained bird bone.
Ill. 2.54—South
entrance 1 after
conservation
(Con Brogan,
DAHG).
82
THE EXCAVATIONS
The grave was not very formally laid out, as the feet and tibiae were at a higher position than the rest of
the body.The burial was dated to AD 990–1165. During the washing of the skeleton a tiny water-rolled
pebble was found.
It was noted during this phase of the excavation that a small area of the original render (F569)
covering St Michael’s Church had survived on the eastern exterior wall. This was conserved in situ.
Nineteenth-century features
In the south-west corner of the area excavated there was a layer of nineteenth-century material (F562),
1.25m by 2.5m in extent. This feature contained some wine bottle sherds and some pottery, all of
nineteenth-century date, together with eight water-rolled pebbles, an iron strip and an oyster shell.
Disturbance associated with this deposit had uncovered the skull of Skeleton 2, which had been backfilled
immediately.The skull was replaced facing the wrong way, with a doormat placed over it.This disturbance
had removed all of the burial matrix layers, F563 and F564, in this small area of the site.
Above this was a layer of campion, between 0.12m and 0.3m thick (F561), which covered the whole
area south of the plinth for the large oratory. This layer contained nineteenth-century pottery, a slate, a
stone cross fragment, eight water-rolled pebbles and a fragment of a writing lead.
Interpretation
The earliest feature in this area of the inner enclosure is the stone structure, which may be the remnants
of a cist similar to those recorded on nearby Illaunloughan and Church Island, which have been
interpreted as belonging to the tradition of the ‘translation’ of human remains dating from the late
seventh/eighth century.
The paving around the leacht appears to
be roughly contemporary with the second
phase of steps in the south entranceway (Ill.
2.55). Those steps respect F529, the lower
step of the large oratory plinth, which may
belong to an earlier oratory aligned parallel
with the south wall of the enclosure. This
paving overlies the cist (F599), which had a
cross inserted in its northern compartment,
possibly at the time that the leacht was
erected. The leacht was erected on the
paving, which follows the line of the earlier
plinth and may date from the same time as
the suggested earlier oratory. Subsequently
the large oratory was built. Also subsequent
to this, three burials were interred, two
between the leacht and the south wall of the
enclosure and one parallel to the foundations
of St Michael’s Church. There was some
evidence for small-scale nineteenth-century
disturbance of these features.
Unlike elsewhere on the site, the
radiocarbon dates relating to this area are
consistent with each other and can be tied
directly together in terms of stratigraphy.
83
Ill. 2.55—Leacht
after conservation, from east
(Con Brogan,
DAHG).
Campion layer
Skeleton 3:
F571
F533
F572
F561
F534
F573
F562
F535
F531
F563
Blocking of entrance:
F539
F574
Skeletons 1 and 2:
F580
Leacht:
Paving:
F600
F560
F582A
F581
F582
Drain:
2nd level steps:
F564
F540
F576
F541
F577
F542
F579
F532
F583
F545
F588
F591
F546
F592
F547
F593
F548
F594
F549
F544
F595
F551
F589
Enclosure wall:
F557
F559
F597
F599
F543
F590
F596
Cist:
F538
F550
F552
F513
F553
F554
Ill. 2.56
84
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
F575
F530
Ill. 2.56—Matrix for south entrance 1 and leacht area.
THE EXCAVATIONS
The dates for soils pre-dating and post-dating the leacht seem to centre on about AD 850, while the dates
for the burials are also consistent and seem to centre on about AD 1100. As we know the precise
stratigraphic relationship between all six dates, it is hoped to carry out Bayesian analysis on these six dates
in order to narrow down, if possible, the date ranges.
The layers and their stratigraphic relationships are shown in Ill. 2.56, with the earliest on the right
and the latest on the left, illustrating the consistency of the two phases of activity.
The finds are also consistent.The nineteenth-century finds, though not closely datable, indicate two
phases of disturbance. The first area of disturbance was in the south-west corner, disturbing Skeleton 2;
a second area of damage, where a wall was put in against the leacht, disturbed Skeleton 3. The monasticperiod finds include the writing lead, scallop shells, a nail, roof slates, crosses, discs and many water-rolled
pebbles. Water-rolled pebbles also occurred on the prayer station on the South Peak. It is not yet clear
whether water-rolled pebbles can occur naturally on Skellig. If not, these items may well have been
brought by pilgrims to the site. The deposition of water-rolled pebbles also occurs at the site of
Templeteenaun, Co.Wicklow, where water-rolled stones, rather than pebbles, seem to have been brought
onto the site from a river far below the site (A. O’Sullivan, pers. comm.).
2.2.4 SOUTH
Ann Lynch
ENTRANCE
2, INNER
ENCLOSURE
(E338)
At least three long flights of steps, built
by the monks, provided access to the
monastery.The most significant of these
are the south steps and the north steps,
which merge at Christ’s Saddle and
enter the outer enclosure from the west,
and the east steps, which enter the outer
enclosure from the south (Ills1.1, 1.3a).
All three routes meet at what in
modern times is the only entrance in
the inner enclosure wall (Ill. 2.57).
A series of seven drystone steps lead
upwards to this entrance from the
lower, outer enclosure. These steps are
flanked to the east by a drystone wall
(F14), which abuts the inner enclosure
wall and runs southwards from it for a
distance of 4.5m before swinging
south-south-east for a further 2m,
where it flanks the steps leading down
to the lower monks’ garden. The wall
survives to a maximum height of 1.9m
and is 0.5m wide at the top but widens
considerably towards its base, where a
distinct batter on its eastern face gives
it an overall thickness of c. 2m.The style
of construction (small stones randomly
Ill. 2.57—South
entrance 2
(inner
enclosure),
steps leading
from outer
enclosure (Con
Brogan,
DAHG).
85
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.58—South
entrance 2
(inner
enclosure)
viewed from
inside enclosure
and D-shaped
‘leacht’
(A. Lynch).
placed) would suggest that this wall was built possibly as late as the 1820s, when considerable modifications
were carried out to the monastic structures by the builders of the lighthouses. Several of the adjacent
steps may also be late in date. Remains of wall footings can be traced flanking the west side of these steps
and running westwards towards the entrance in the outer enclosure, suggesting a possible early terrace
on the north side of the pathway that leads from the outer to the inner enclosure entrance.
The entrance itself is a lintelled passage, c. 3m long, c. 1m wide and averaging 1.9m in height, running
through the thickness of the inner enclosure wall.The masonry of the passage walls is comprised of large
stones horizontally laid in distinct courses. A crudely built section of the east elevation indicated an area
of repair, possibly necessitated by shifts in the masonry caused by the cracking of the overhead lintel. A
series of eight steps ascend through the passage, and during repair/conservation works these were seen
to be replacements for earlier steps, the remains of which survive underneath. An earlier threshold also
survives under the step at the south face of the entrance (G.D. Rourke, pers. comm.).
Emerging from the entrance passage, three steps (incorporating exposed bedrock) continue upwards
for a distance of 5.65m before turning east, where the final two steps lead to fairly level ground in front
of Cell A. These steps are bounded on the west by a drystone retaining wall (F1) that abuts both the
inner enclosure wall and a similar retaining wall (F2), which forms the northern boundary of the passage
as it swings east (Ills 2.58, 2.59). Both walls are founded on exposed bedrock in the area where they
intersect. The tops of these retaining walls are flush with the ground surface behind, and the height of
F1 ranges from 2.55m at the entrance passage to 1.7m at its junction with F2.The steps are bounded on
the east by a D-shaped ‘leacht’-type structure (2.47m north/south by 2.1m max. east/west). This is
constructed of drystone walling enclosing a fill of clay. Four crudely shaped stone crosses are set upright
in the clay and a fifth lies on the surface. The height of the drystone walling ranges from 2.05m at its
junction with the entrance passage to 0.8m at its curved end and 0.15m where it abuts the enclosure
wall. These measurements illustrate clearly the terracing, which results in a 2m drop from the inner
monastic enclosure to the outer enclosure. The D-shaped structure is likely to be nineteenth-century in
86
THE EXCAVATIONS
- metres OD
Cell A
0
5
metres
scale 1:50
bulging section face
due to rubble F5
F2
k
oc
dr
be
rubble core
of wall F1
A3
Ill. 2.59—Plan
of south
entrance 2
(inner
enclosure),
showing area
excavated.
F15
F12
A2
N
‘leacht’
A1
167.51
166.85
lintel
slab
batter
F1
n
r o wall
tte
ba ning
tai
re
of
ll
wa sure
g
lo
nin enc
tai
re ner
in
date and must have been constructed when the adjacent entrance passage and steps were modified.
Before excavation, the drystone walling of the repaired section of the eastern elevation of the entrance
passage was badly ravelled and bulging inwards. Just inside the entrance, the western retaining wall (F1)
was also bowed and in danger of collapse.Timber shoring and props had been put in place to ensure safe
access for visitors in the short term, but it was clear that major repair and consolidation works would be
necessary to secure this focal point of the monastery. As these works would involve the dismantling and
rebuilding of wall F1, an area was excavated immediately behind it.The objectives of this small excavation
were twofold:
• to examine the nature of the deposits behind the wall as it was being dismantled and to gain some
information about the sequence of building activities in this area;
• to determine the relationship between walls F1 and F2.
87
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
While the outer face of F1 abutted F2 at the curve of the passage, there was some suggestion that F2
may have extended westward behind F1, thereby pre-dating it.
The excavation
A single cutting excavated in June 1986 encompassed the northern end of the passage and its curved
corner (Ill. 2.59). The dimensions of the cutting were irregular but it measured a maximum of 4.62m
east/west and 2.4m north/south. Ground level sloped steeply to the south, with a drop of 1.08m over a
distance of c. 6m from the northern end of the cutting to the inner enclosure wall to the south.
The western retaining wall (F1)
This wall was recorded and dismantled as the excavation proceeded. It was founded on bedrock and was
poorly built, with a facing of roughly coursed sandstone of varying sizes and a rubble core behind. The
rubble core averaged 0.8–1m in thickness and in places was difficult to distinguish from the rubble fill
(F5) behind (the stones within the wall core tended to dip down slightly to the west, while those within
the fill F5 lay horizontally). Pockets of dark brown sandy clay containing lumps of coal and brick
fragments were noted within the core of the wall. Following excavation, the upper levels of the wall to
the south of the cutting were dismantled under archaeological supervision, and a leather man’s shoe of
late nineteenth/early twentieth-century date (E338:29) was recovered from the core of the wall.
The junction of the upper levels of walls F1 and F2 was exposed at the northern end of the cutting
and it was revealed that the walls are contemporary.The facing of wall F2 did extend for c. 0.35m behind
wall F1, possibly to strengthen the walls at the point where the passage changed direction, but the rubble
core of one was indistinguishable from the other.
Deposits behind retaining walls
While both retaining walls were built on bedrock, undisturbed boulder clay (F19) was exposed rising
gradually westward, at a point c. 1m to the west of F1 (Ill. 2.60). This took the form of compact dark
brown/grey clay with decayed/shattered bedrock in places. Overlying both the bedrock and this
undisturbed boulder clay were deposits F18 and F17, which comprised compact dark brown clay with
large stones, charcoal flecking and three small fragments of bone (not identifiable). Lying immediately
behind wall F1, F18 also contained pockets of sandy clay and lenses of yellow clay. Both these layers are
interpreted as mixed and disturbed boulder clay associated with the building of wall F1.
A layer of loose, dark brown sandy soil (F12) containing a small number of stones, charcoal flecks
and coal had been deposited on top of the mixed/disturbed boulder clay. This layer averaged 0.2m in
thickness at the northern side of the cutting but was up to 0.5m thick on the southern, downhill side.
Also contained within F12 were lenses of yellow/brown clay (F13) with traces of reddened burnt clay,
charcoal, burnt stone, a pig incisor and small fragments of scallop shell. A flint flake (E338:10) showing
signs of retouch was recovered from this layer.This dump layer must have been put in place as the retaining
wall F1 was being constructed. A shallow gully (F15) had been cut into the surface of F12 and extended
from midway along the north baulk in a south-easterly direction for a distance of 1.1m to the loose
stones in the core of the retaining wall F1 (Ill. 2.59).The gully, which was U-shaped in section, averaged
0.1m in depth and 0.2–0.25m in width and was filled with black, moist, sandy clay (F16). Its function
was unclear but it could have been a drainage channel dug to direct surface water towards the retaining
wall, through which it could seep freely.
The upper surface of F12 was covered in places by a thin layer of loose sandy soil (F10) with a high
proportion (c. 80%) of small, friable sandstone fragments. This in turn was covered by a well-defined old
ground surface (F9), which comprised a layer of dark brown/black humic soil with patches of
orange/brown, less-humified material with roots and vegetative material clearly visible. It ranged from
88
THE EXCAVATIONS
A1
A2
A3
168
F3
F10
F5
F3
F4
F9
F5
F10
167
F9
F1 wall
F12
F17
F12
F17
F13
F13
F18
bedrock
F19
166
F3- sod/campion layer
F12- dark brown sandy clay
F4- dark brown sandy soil
F13- brown fine sandy clay
with burnt flecks
F5- rubble with dark brown sandy soil
F18- mixed silty peat/rubble/decayed
bedrock with F12 & F17 lenses
F9- old ground surface,
dark brown humic soil
F17- disturbed boulder clay,
grey/white compact clay
F10- grey/brown sandy clay
F19- boulder clay
M.O.D.
F19
0.05m to 0.15m in thickness and produced several pieces of coal, cinders and a single clay pipe stem
(E338:11) of early nineteenth-century date.The presence of such a well-defined old ground surface indicates
that the retaining wall was c. 0.6m lower than its present height for a period during the nineteenth century.
Later in the nineteenth century, or possibly the early twentieth century, further material was deposited
on top of the old ground surface and the retaining wall was increased in height. A distinctive deposit of
fine light brown soil with several lenses of orange/red burnt sandy clay (F11) was uncovered lying on
top of the old ground surface. Concentrated in the northern half of the cutting and up to 0.08m thick,
this must be the residue of burning that took place further uphill to the north. A substantial deposit of
rubble (F5) was deposited on top of the old ground surface and F11, ranging in thickness from 0.4m at
the western end of the cutting to a maximum of 0.9m where it abuts the core of the retaining wall (F1)
to the east.The stones in this deposit averaged 0.1–0.2m in length but could be up to 1m. A small quantity
of quartz and shattered ‘Portland Stone’, used extensively by the lighthouse-builders, was included
amongst the sandstone rubble.Within and partly overlying the rubble was a deposit of dark brown sandy
soil (F4), which was in turn covered by a layer of moist peaty soil (F3) that supported a dense growth of
sea campion (Silene uniflora). The sandy soil (F4) contained pieces of coal, brick, metal slag, a clay pipe
stem, golf tee and a range of ceramics of early/mid-nineteenth-century date (see Section 3 below).
Interpretation
The retaining walls (F1 and F2) that form the eastern and northern boundaries of the current entrance
passage in the inner enclosure wall are contemporary and post-medieval in date. The presence of coal
and brick fragments in the core of the lower portion of the western retaining wall (F1) indicates that it
was built by the lighthouse personnel in the early nineteenth century. It is likely that this area, at the
only entrance to the inner enclosure, was significantly modified during the nineteenth-century
occupation of the monastery. A further c. 0.6m was added to the height of the retaining wall in the late
nineteenth/early twentieth century, possibly by the Office of Public Works, shortly after the site came
into state care in 1880. The rubble and clay deposits exposed during the excavation (with the exception
of the lower boulder clay deposits) are the result of dumping or accumulation during the nineteenth
century.
No deposits or structures belonging to the monastic period were recorded, and consequently we do
not know how this area, just inside the entrance to the inner enclosure, was organised in the early
89
Ill. 2.60—South
entrance 2
(inner
enclosure),
sectional
profiles.
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
medieval period.The construction in post-medieval times of the lower levels of retaining wall F1 resulted
in the disturbance of deposits behind the wall, and it is possible that the remains of a pre-existing wall
were removed at that time. There was no evidence within the area excavated to suggest that a monastic
structure had existed to the west of wall F1, which would indicate that the original passage wall (if such
existed) was either on the line of the later wall or very close to it.
2.2.5 EAST ENTRANCE, PHASE 3 (INNER
Edward Bourke
ENCLOSURE)
(93E195)
Introduction
Inspection of the south-eastern corner of the inner enclosure revealed that it was in urgent need of
conservation. It was apparent that the two walls (southern and eastern) were not bonded at the base.The
east wall had collapsed and been rebuilt twice in antiquity at this point (Ills 2.61–62). A third repair,
which was high in the wall, might have been either of monastic or lighthouse period construction.
Phase 1
Ill. 2.61—Phases of
development in east
wall, inner enclosure,
from south (Con
Brogan, DAHG).
Phase 2
Phase 4
Phase 3
90
THE EXCAVATIONS
Phase 4
Phase 1
Phase 3
Ill. 2.62—Phases
of development
in east wall,
inner enclosure,
from east (Con
Brogan,
DAHG).
Entrance
Phase 2
Garderobe
Chute
Entrance
All traces of the earliest phase of the
eastern wall had disappeared in the southeast corner, though a small section survived
just east of Cell F. The second phase was
stepped back from the first by between 0.6m
and 1.5m. This phase 2 wall also partly
collapsed in the south-east corner and only
its base survives. A second rebuilding
occurred (phase 3/F501), which was in turn
stepped back from the phase 2 wall by up to
2m. This third phase abuts the south wall
forming the south-east corner of the inner
enclosure.
A large sloping stone, possibly a lintel,
was observed in the outer face of this wall
(Ill. 2.63), and it was important, for both
archaeological and structural reasons, to
establish whether this marked an early
entrance to the site. It was established that
the phase 2 wall was in need of urgent
conservation and that the phase 3 wall would
have to be reduced in height to relieve the
pressure on the earlier wall during
conservation work and to stop the base of
the phase 3 wall moving to the east. It was
Ill. 2.63—
Exterior of
lintel, showing
wall pushing
outwards (E.
Bourke).
91
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
decided to excavate behind the wall to establish the state of the phase 3 wall and possible entrance and
the nature of the fill behind the wall. A section of the wall above the lintel had been rebuilt on the same
line as the phase 3 wall.This phase 4 wall (F500) was quite high at its southern end and was putting a lot
of pressure on the phase 3 walling.
Pre-excavation works
Prior to excavation and after full recording, the entire phase 4 wall (F500) was removed, as was the upper
part of the phase 3 wall down to the level of the paving within the enclosure.This work was carried out
in 1999 in advance of the excavation of the interior of the wall, in consultation with the project engineer,
who also supervised the safety aspects of the work.
The upper part of the wall (F500) had been substantially rebuilt in the nineteenth century, and
fragments of glass, creamware, brick, coal, non-local stone and iron were found in the upper 0.6m.
Excavation
A single cutting was opened, measuring 0.9m by 2m at the top. It extended from the monks’ graveyard
in the north to the south wall of the inner enclosure (F513), the area excavated being limited by a desire
to cause as little disturbance as possible to the standing monuments within the enclosure. The exterior
of the south wall had been largely rebuilt as part of the conservation work associated with the excavation
of entrance 2, but on the interior the stonework survived to a height of 2.8m in the entrance itself.
Excavation established that the lower part of the inner face of the wall survived across the full width of
the cutting.
A short length (c. 1.5m) of the drystone wall at the south end of the monks’ graveyard had to be
removed for safety purposes, to ensure that none of the wall could collapse into the cutting during
excavation.The stratigraphy revealed that, behind the drystone outer kerb (F520), the fill (F521) consisted
of 0.45m of loose unsorted rubble with some coal and brick fragments in the matrix, as well as many
white quartz stones. This indicates that the south end of the monk’s graveyard was rebuilt post-1820,
which tallies with the information from the small excavation of the north end of the graveyard (see
below).
Description of entrance 3
The entrance (phase 3) consisted of a lintelled, drystone passage in the east wall at its junction with the
south wall. The entrance was only bonded into both walls at lintel level and it is clear that the east wall
post-dates the south wall. As described above, phase 3 of the east wall is the second rebuilding of this
wall, and the south wall continues eastward to meet the point where phase 1 of the east wall would have
joined it. The entrance jambs are of drystone construction. The north jamb forms the terminal of the
east wall and the south jamb is also of dry stone, poorly bonded into the south wall. The outer lintel is
bonded into the south wall, but a crudely built drystone pillar supported the inner lintels. The entrance
is 2.3m high at its outer end and 2.15m at the innermost lintel. It is 1.2m wide at the base on the inside
and 0.95m wide at the lintel on the inside, but the wall on the north side of the entrance is severely
distorted and has moved laterally by up to 0.48m (Ill. 2.64).
A series of at least nine drystone steps (F502) led up through the entrance and into the inner
enclosure. It was not possible to excavate down to the outermost steps for safety reasons, as this would
have entailed digging into the phase 2 wall and undermining the concrete beam described above. It is
possible to estimate that three or four more steps could have existed towards the outside of the entrance.
A second possibility is that there could have been a flat area at the outer side of the entrance, given that
the entrance is c. 3m above ground level at this point.The steps are regular on the inside of the entrance,
averaging between 0.2m and 0.35m for the tread and between 0.18m and 0.26m for the riser, but within
92
THE EXCAVATIONS
North/south
section through
entrance, showing
distortion
V1
V2
Ill. 2.64—Section showing
distortion.
0
3
metres
N
top of
east wall
F500
new
face of
east wall
F501
ining
reta
wall
monks'
graveyard
Ill. 2.65—Plan of entrance,
lower level.
up
ps
ste 502
F
overhang
face of south
wall F513
stones forced
out of wall
0
2
metres
the entrance itself the steps appear to have slumped when the walls moved (Ills 2.65–6).
The south wall appears to have moved outwards, leaving the south jamb sloping gently, while the
east wall appears to have slipped eastwards violently at its base, leaving the wall bulging at the front and
dropping the height of the outer lintel by 0.7m. At the same time the top of the wall, together with the
lintel, moved southwards, while the base of the north side of the entrance remained in place (Ills 2.67–8).
It is likely that the severe movement of the enclosing walls also caused the collapse of the stone structure
(F514) associated with F509 (see below).
Monastic period blocking of the entrance
Directly on top of the steps (F502) was a layer of broken stones and some large stones (F510), which
varied in thickness from 0.28m to 0.55m. This was overlain by a layer, up to 0.48m thick, of greyishyellow gritty soil with small stones (F509), which contained the remains of a collapsed stone structure
93
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.66—Inside
entrance, from
east, showing
steps (E.
Bourke).
east wall rebuilt C.19th
U2
U1
F505
504
lintel stones
F506
F500
modern
retaining wall
F507
F503
F509
F510
F508
Ill. 2.67—Section through east
entrance, inner enclosure
(phase 3) (C. McHale).
steps F502
east wall F501
phase 3
stones bulging
from wall
east wall F591 phase 2
F504- dark brown gritty soil
F509- grey/yellow gritty soil
F505- yellow mottled ash with charcoal
F510- loose aerated soil with large stones
F506- brown silty clay
F503- rubble dark earth and stones
F508- grey silty soil
0
5
metres
94
THE EXCAVATIONS
Ill. 2.68—Inner
face of
entrance during
excavation
(E. Bourke).
(F514) at its base. These consisted of
about twenty large slabs that had slumped
into the entranceway.They ranged in size
between 0.1m by 0.24m by 0.37m and
0.43m by 0.3m by 1.02m, and appear to
represent collapsed building stones from
a structure that would have underlain the
monks’ graveyard (Ill. 2.69). This layer of
stones sloped downwards towards both
the south and east. The stones were all
aligned east–west as they collapsed, which
adds to the impression that they may have
come from some form of drystone
structure to the north of the steps (Ill.
2.72). It is also likely that the soil of F509
may have been the fill behind the wall of
that structure, becoming partly
incorporated among the stones but
mainly overlying the stone collapse.These
layers sloped south and east, whereas the
layers above only sloped downwards
towards the east, the implication being
that these two layers are part of the
original collapse, while the layers above
represent the backfilling of the entrance.
A radiocarbon date of AD 778–948 was
obtained for F509.
Ill. 2.69—
Entrance from
the east,
showing stones
displaced from
north jamb and
steps
(E. Bourke).
95
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.70—
Nineteenthcentury
blocking stones
inserted into
inner face of
entrance under
lintels (E.
Bourke).
The blocking of the entrance commenced with the construction of a drystone wall (F516) across its
front face.This was a single-faced wall, its inner side being unfaced.The implication of such a construction
method is that the wall would have been built in stages, with the backfill behind it being gradually added.
This wall varied in thickness between 0.35m and 0.7m but is not shown on the section drawing, as for
safety reasons supports had to be placed in the entrance as the excavation progressed so that it was
impossible to record in section (Ill. 2.63).
Behind the drystone wall (F516) and overlying the original collapse (F509) was a layer of greyish
silt with limpet shells and one cattle bone (F508), which was up to 0.12m deep and survived only in the
middle of the cutting.Two radiocarbon dates were obtained from this layer; the sheep vertebra produced
a date of AD 691–876, while the limpet shells (after allowing for the marine reservoir effect) produced
a date of AD 675–871. This layer had been partially removed by nineteenth-century disturbance (see
below). Above F508 was a layer of mottled yellow ash (F507) with some charcoal, up to 0.46m deep.
This in turn underlay a thin layer of brown silty clay (F506), only 0.04m thick. Overlying this was a layer
of yellow mottled ash (F505) containing charcoal and small stones; it was 0.17m deep at its western end
and only 0.08m at its eastern end. This layer was more mottled than F507.
Nineteenth-century repairs and modifications to the entrance blocking
At some time early in the nineteenth century there seems to have been a movement of the phase 3 wall
that alarmed the lighthouse-builders enough to cause them to attempt remedial work on the blockedup entrance. The reason for this seems to have been to stop the movement of the south wall and the
slipping forward of phase 3 of the east wall of the inner enclosure. This movement post-dates the early
nineteenth-century paving of the area around the monks’ graveyard. Above the original blocking layers
was a layer of dark brown gritty soil with small stones and broken stone fragments (F504), 0.17m deep
at its western end and 0.44m at its eastern end. This feature contained some brick, a sherd of white
earthenware and some modern glass. It seems to have been laid down as a foundation for the paving in
this corner of the site.
96
THE EXCAVATIONS
A pit (F503) was dug into layer F504
in order to inspect and to repair the inner
lintel of the phase 3 wall by inserting a set
of blocking stones (F512) under the inner
lintel of the entranceway.These consisted
of well-chosen flat slabs, aligned east–
west, wedged into the entrance to stabilise
the area between the jambs and to
support the lintel (Ill. 2.70). Associated
with this inner propping, the upper part
of the phase 3 wall (F501) was taken
down to lintel level, the middle lintel was
removed and the area between the inner
blocking (F512) and the original outer
blocking (F516) was filled with clean
loose stones (F517); this implies that the
original backfilling of the entranceway
may have subsided somewhat in the
intervening period. The inner lintel
measured 1.36m by 0.4m, and the outer
lintel measured 1.75m by 0.61m. The
middle lintel was removed and then
replaced after the cavity had been filled
with F517 (Ill. 2.71).
As the work proceeded, this pit
(F503) was backfilled with a layer of rubble and large stones. F503 also cut the eastern edges of F506,
F507 and F508.
Dating
Three radiocarbon dates were obtained from this entrance.Two samples from F508 were dated, one from
a sheep vertebra and one from a deposit of limpet shells within the same layer. The animal bone was
dated to AD 691–876, while the limpet shells produced a date of AD 675–871. A third sample came
from a cattle tibia from F509 and was dated to between AD 778–948. These dates show, unfortunately,
that the material dumped into the entrance must come from soil redeposited from elsewhere on the site
and is not of any real use in dating the backfilling of the entrance. It is intended to try Bayesian analysis
on the three dates, as their stratigraphical relationship is known and all three dates overlap.
It is clear from this that the deposits dumped into the entrance came from midden deposits elsewhere
on the site. All that can be said for certain at the moment is that the material was backfilled at some time
after a period ranging from the late seventh century to the first half of the tenth century.
The east entrance is of crude construction compared to south entrance 1 of the outer enclosure,
with its orthostatic jambs and high-quality stonework. It betrays the fact that it was built as part of the
repair work associated with the building of phase 3 of the east wall.While this entrance forms part of the
phase 3 rebuild, there was also an entrance in phase 2 of the same wall. A lintel and two vertical breaks
in the masonry indicate that this entrance was at ground level on the exterior and c. 3m north of the
phase 3 entrance. No trace of a phase 1 entrance survives, but the relevant portion of wall did not survive
in this area (Ill. 2.73).
What is clear is that the phase 3 entrance pre-dates the paving of the interior of the south-east corner
97
Ill. 2.71—
Original lintels
in situ after
nineteenthcentury
blocking was
removed (E.
Bourke).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
N
wall
ining
reta
ern
mod
face of
east wall
F501
top of
east wall
F500
V2
monks'
graveyard
blocking
stones F512
Ill. 2.72—Plan
of entrance,
upper level.
U2
lintel
original
face of south
wall F513
lintel
V1
U1
collapsed
stones F514
Ill. 2.73—
Entrance area
from east,
showing wall
collapse (Con
Brogan,
DAHG).
98
th
ou
reb
ts
uil
ll
wa
15
F5
0
2
metres
THE EXCAVATIONS
of the monastery and pre-dates the monks’ graveyard. The steps associated with the entrance could have
led up as far as the level of the large oratory plinth. The precise relationship between these structures
cannot, unfortunately, be established, as this would entail further disturbing the monks’ graveyard and
possibly undermining the large oratory.
The evidence of collapse from a structure (F514) that clearly pre-dates the monks’ graveyard provides
an intriguing indication that the layout of this corner of the inner enclosure must have been very different
in the earlier years of the monastery. Although nothing can be inferred about the function of this
structure, the use of very high-quality building stone suggests the possibility that it was important. Its
stratigraphic position indicates also that it pre-dates the building of the large oratory plinth, and the
possibility that it represents an earlier oratory cannot be discounted.
An interesting issue raised by the excavation of this entrance is that the entrance leads out onto the
top of the phase 2 wall and that there are no steps visible below it. The phase 2 wall is c. 3m high at this
point, and either a wooden ladder or a rope hoist would have had to be used for access to the entrance
from below, although no structural evidence for this survives. This method of accessing is well known
from Greek and Egyptian monasteries (Nicol 1963). It should also be noted that no trace of steps, rockcut or drystone-built, has been observed beneath the wall at this point. It is possible that an access route
into the area of the outer enclosure around the end of the south wall could have provided an alternative
method of entry and exit. Recent discoveries of steps leading to the north from this area indicate that
the east steps may not have been the only ones accessed from this point.
In relation to the collapses of the south wall of the inner enclosure, it is interesting to note that the
slope of collapse of the phase 4 wall (F500) above the east entrance is steep and falls off towards the
south. This slope is reflected in the slope of the campion above the collapsed south wall (Ill. 2.73). It is
possible that this nineteenth-century repair (phase 4) pre-dates the collapse of the south wall which
demolished Cell G and continued on to destroy the south wall of the outer enclosure below it. This
would tie in with the fact that there is a lighthouse phase of occupation within Cell G and may show
that the rebuilding of the east wall and the collapse of the south wall both happened in the 1820s, with
the east wall rebuild happening first.
The monks’ graveyard
A small collapse at the north end of the monks’ graveyard was repaired in the late 1970s, revealing an
earlier faced edge to the graveyard (G.D. Rourke, pers. comm.).This was similar to the situation that was
encountered when part of the revetment wall was removed during excavation of the east entrance. It
was decided to monitor the removal of this revetment wall along its north and east sides in order to
establish whether there was an earlier footing which could be conserved (Ills 2.74–5).
The rebuilt stone revetment on the north side of the graveyard consisted of a drystone facing (F520).
The gap between this and the original edge of the graveyard (0.4–0.6m wide) was filled with loose
unsorted rubble and brick fragments (F521)—the same construction technique encountered during
excavation of the east entrance. As this was removed, the lowest stones of an earlier retaining wall were
revealed under the nineteenth-century backfill. This process of gradual removal of the wall continued
down the east side of the graveyard. It became clear that, behind the nineteenth-century drystone wall,
the monks’ graveyard had been constructed in two phases, with the northernmost phase added to a larger
southern burial platform. It is interesting to note that the nineteenth-century rebuilding of the revetment
walls was done before the area was paved, as the paving stones in this area respect the repair, not the
original burial platforms (Ills 2.76–7).
The backfill layer (F521) fell away from the material directly behind it, so there was no need to
interfere with the (possibly original) deposits behind it. A drystone wall was built upon the footings of
the original burial platforms, but this had to be estimated for the east face of the northern or later
99
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.74—
Monks’
graveyard from
the east during
excavation
(Con Brogan,
DAHG).
Ill. 2.75—
Monks’
graveyard
during
excavation,
from north
(Con Brogan,
DAHG).
100
THE EXCAVATIONS
Ill. 2.76—
Monks’
graveyard after
conservation,
from north
(Con Brogan,
DAHG).
Ill. 2.77—
Monks’
graveyard after
conservation,
from east (Con
Brogan,
DAHG).
101
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Phase 4 wall:
F500
Paving
Lintels
F503
F504
F512
F505
F517
F506
F507
F508
F509
Ill. 2.78—Matrix
for layers
associated with
East Entrance.
Phase 3
F516
F501
F510
F514
F502
F513
Phase 2 wall:
Phase 1 wall:
Key:
102
19th century repair, 2nd phase:
F500
19th century levelling up of the area, 1st phase :
F504
Medieval blocking layers:
F505
Phase 3 features:
F501
South wall (inner enclosure):
F513
THE EXCAVATIONS
platform, as the footings did not survive. All rebuilding was done using the stones from the nineteenthcentury repair. All human bone found in F521 was sent for analysis, although obviously from a context
disturbed in the nineteenth century. Very few bones were found in F521 and these were all in poor
condition, which may argue for the bones within the platforms being in similar condition.
There may well have been a remodelling of the graveyard, even in the early medieval period, as the
west wall at the back of the burial platforms is composed of orthostats, which include crosses.These were
left undisturbed during the conservation work and the question of their relationship with the southern
and western drystone kerb remains unresolved.
2.2.6 CISTERN 3 (93E195)
Edward Bourke
Introduction
Cistern 3 is one of four cisterns associated with the monastery that were constructed to gather and store
rainwater. Located c. 18m west of Cell A in the inner enclosure (Ill. 2.79), it had been described by
O’Sullivan and Sheehan (1996, 279, 286) as a souterrain-type structure. Cistern 3 differs from the other
two cisterns within the inner enclosure in that it is built above ground as opposed to being incorporated
within the base of drystone walling.
Ill. 2.79—
General view of
the area west
of Cell A in the
1950s (DAHG
Archive).
103
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The cistern had been surveyed in the 1980s and was recorded as having a lintelled opening, 0.6m
high and 0.38m wide; the interior was a roughly rectangular chamber (3.1m by 0.9m) running northeast–south-west. The rear (north) wall and floor of the structure were formed by the bedrock and it was
bounded on the south by a low, apparently drystone wall.The eastern terminus was a large upright stone
and the west end appeared to be blocked by rubble.The maximum height was 0.51m against the bedrock.
There were twelve visible lintels, which sloped downwards to the south.
There was also a low wall, c. 0.5m high, running along the south-eastern side of the structure, with
a small area of paving in front of it. Both of these features appeared to be of nineteenth-century date and
continued outside the area excavated.
The stratigraphy above the lintels was excavated in 1995. The lintels were then numbered, surveyed
and removed, with the exception of lintel 2, which was too heavy to be moved by hand. The site was
then covered for the winter and the excavation of the interior took place in 1996.
The excavation
Two cuttings were opened, one (3m by 2.5m) directly over the lintels, and a second (2m by 2.5m) further
west in order to check the stratigraphy beyond the known structure.The baulk between the two cuttings
was subsequently removed in order to show the method of construction of the cistern (Ill. 2.80), and for
the purposes of this report the two cuttings are treated as one.
The building of the cistern took advantage of a geological anomaly whereby on the north side of the
area excavated there was a layer of solid sloping bedrock that partially overlay another. Above the lower
layer of bedrock and beside the upper one was a layer of vertically cleaved bedrock, which was easily quarried
out. An area (c. 3.7m by at least 2m) was quarried from the vertically cleaved bedrock to accommodate the
cistern; the southern limit of the area quarried was not reached, as this would have required removing later
walling constructed by the lighthouse-builders. At the south-eastern end of this feature a crudely built
drystone facing wall (F621) was built. The north-eastern end of the cistern was blocked off by a large
orthostat (F613), which measured 0.9m wide and 0.7m high and whose edges were caulked with dark
red/orange/brown clay (F617) against the wall of bedrock and the base of the cistern. A drystone wall
(F618) was built on top of this orthostat on which the north-easternmost lintels rested (Ill. 2.82).
The south-eastern side of the cistern was constructed by digging a 0.1m-deep hollow into the
bedrock floor of the structure; five orthostats (F608) were set into this hollow and caulked with the same
clay (F617) used to caulk F613. Four of these orthostats measured (0.79m by 0.1m by 0.3m high
(F608.01); 1.14m by 0.08m by 0.35m high (F608.02); 0.31m by 0.06m by 0.37m high (F608.03); and
0.79m by 0.06 by 0.35m high (F608.04). They ran parallel to the north-eastern wall of the cistern and
about 0.7m distant from it.The fifth orthostat (F608.05) was angled inwards somewhat in order to meet
the orthostat (F613) that formed the north-eastern end of the cistern. It measured 0.61m by 0.1m by
0.36m high. All of these orthostats had collapsed inwards at an angle of about 45º. South-east of these
stones was a packing of beige/gray shaly earth (F611). The full extent of this layer was not determined,
as it ran under later lighthouse walls.
The cistern was roofed with 26 lintels (F605) running north–south (Ill. 281), which were comprised
of two types:
• large lintels such as 02, 04, 05, 15, 20 and 24, which spanned the whole width of the structure; the
largest of these (F605.02) measured 1.76m by 0.66m and was up to 0.16m thick;
• smaller lintels that were laid on top of and between the larger lintels to cover fully the space below.
These ranged from largish stones such as lintel 03, which measured 0.95m by 0.29m and was up to
0.1m thick, to quite small stones such as lintel 25, which measured 0.14m by 0.19m and was only
0.06m thick.
104
THE EXCAVATIONS
T2
N
bedrock
F622
T3
F6
09
lintels
F605
baulk
d
vertically cleave
bedrock
lighthouse
facing wall
F620
T1
Ill. 2.80—Plan
and section of
Cistern 3.
F610
lintels
F605
F603
T1
T2
T3
F601
F601
F603
F602
F602
north-west wall F614
F604
F604
bedrock
bedrock
F601- campion
cistern
west wall
F621
F602- stony brown humic soil
F604- brown humic soil
F615
rotted
F615
vegetation
F603- old sod line
F618 drystone wall
stones
F612
orthostat
F613
cistern
east wall
0
F619
F615
5
F615- black silt
metres
At the north-eastern end of the cistern the lintels ran north-east/south-west. They lay on top of
the drystone wall (F618) above the orthostat at the north-eastern end, and lay on F605.02 at their
south-western end. These lintels (F622.01–03) measured on average 0.65m long and varied in width
from 0.39m to 0.16m and in thickness from 0.05m to 0.11m.
These lintels covered the cistern as far as the south wall but stood proud of the top of the wall,
allowing direct access to the interior for the collection of water. Above the lintels was a layer of dark
brown humic soil with stones, mostly rock chips (F604). This seems to represent some of the material
105
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.81—The
lintels during
excavation (E.
Bourke).
Ill. 2.82—East
end of cistern
with most of
the lintels
removed (A.
Halmschlag).
106
THE EXCAVATIONS
quarried during the construction of the cistern.
It covered the whole area of the cutting, except
to the south-west; it was between 0.42m and
0.53m thick at the north-western baulk and
tapered away at its south-eastern end, where it
was in parts truncated by the building of the
lighthouse wall (F620). Above this layer was a
layer of old sod (F603), 0.08m thick. This too
had been truncated slightly by lighthouse walls
at the entrance to the cistern and ran out onto
the bedrock at its south-western end.
A slightly curved wall (F614) built on the
vertically cleaved bedrock was recorded to the
north-west of the cistern but was not fully
investigated (Ill. 2.83).
Ill. 2.83—
Vertically
cleaved rock
and west wall
of cistern (E.
Bourke).
Cistern fills
The fill of the cistern consisted of two layers.
The uppermost (F606) consisted of material
that had collapsed through the capstones during
their removal as part of the excavation. Beneath
this was a layer of dark brown silty material
(F615), the upper part of which contained
modern material such as a decayed diesel can,
plastic, cans, remnants of rubber gloves and some
polystyrene foam sheeting. Other finds in the fill
consisted of coal, brick, some ten sherds of
nineteenth-century pottery and a sherd of Ham
Green A/B transitional ware. A soil sample from
F615 produced some insect remains, which are
listed in the palaeoenvironmental report, but
given the diversity of the finds from this feature
not much more can be said about these remains
than that they indicate that the cistern may have
been used as a cesspit at some stage. Given the
proximity of the lighthouse toilet to the cistern,
this may not be so surprising. A layer of stones
at the base of this deposit (F616) may be the
remains of drystone walling from between the
orthostats and lintels, which would have fallen
inwards when the orthostats moved (Ill. 2.84).
Ill. 2.84—Base
of cistern, east
end, showing
orthostats
leaning over (A.
Halmschlag).
Lighthouse period alterations
The south-eastern edge of the cistern was
largely remodelled during the lighthouse
period. A long drystone wall (F620) was built
along the south-eastern side and drystone
107
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
lintel
N
bedrock
quartz
base
retaining
wall
ne
sto
d
rth
ly
an
y
dr
ea
a
sh
Ill. 2.85—Postconservation
plan.
baulk
lighthouse
facing wall
bedrock
lighthouse
facing wall
0
5
metres
108
THE EXCAVATIONS
Ill. 2.86—
Opening into
Cistern 2, inner
enclosure (Con
Brogan).
masonry was added to the north-eastern wall of the cistern (F619), with three more lintels added, all of
which resulted in the partial obscuring of the original opening.This made entering the cistern, or getting
water out of it, much more difficult. The last of the original lintels was directly above the orthostats that
formed the south-eastern wall of the cistern. The addition of the extra lintels and the new wall meant
that a new, narrower entrance had an opening 0.6m high and 0.38m wide that was 0.75m out from the
side wall of the cistern.
Above the new entrance a layer of brown soil and small stones (F602) was laid.This ranged in depth
from 0.42m to 0.58m at the north-west baulk and tapered away where it met the lighthouse alterations
to the south-eastern side (F619 and F620). Above this was a layer of campion and small stones (F601),
which varied in thickness from 0.12m to 0.39m.
Interpretation
Cistern 3 was built in an area where there was a junction between solid, sloping, sandstone bedrock and
an area of bedrock containing vertical lines of cleavage. This allowed the rainwater to filter down along
the sloping solid bedrock, letting the water become trapped by the vertical cleavage lines in the upper
layers and flowing down a gentle slope into the area excavated for the cistern.The orthostats to the south
and east were waterproofed with clay in order to retain water within the cistern. Following reconstruction
the cistern still retains water, even though no extra caulking was added (Ill. 2.85).
Examination of the location of all three cisterns within the monastery suggests that the same junction
in the geology was exploited for each of them, as all follow the same north-east/south-west alignment.
This implies a sophisticated understanding of the geology of the island and its potential for watergathering.
The other two cisterns within the monastery are built into the base of drystone walls (Ill. 2.86) and
are directly accessible from the front. Cistern 3 appears to have been an isolated structure, but the
lighthouse period alterations to the front of the structure make it more difficult to understand the original
access for drawing water from it.
109
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Cistern 3:
F601
F602
F614
F603
F619
F620
F604
Ill. 2.87—Matrix
for Cistern 3.
F621
F607
F605
F622
F606
F618
F615
F613
F616
F617
F608
F611
Bedrock
Key:
19th- century remodelling:
F500
Cistern fills
F504
Original cistern features:
F505
The cisterns in the monastery itself need to be used constantly to prevent the water from becoming
brackish. It was noted that those cisterns also needed to be emptied and cleaned from time to time for
the same reason. Once used and cleaned, the water was potable. The same would have been true of
Cistern 3, but the lighthouse additions made both emptying and cleaning the cistern much more difficult,
and it seems to have gone out of use soon after that. By the twentieth century it had become a dump
for old cans and used equipment.
2.2.7 THE
LOWER MONKS’ GARDEN
Lower monks’ garden (east end) (93E195)
Edward Bourke
Introduction
A test excavation carried out in 1993 at the east end of the lower monks’ garden (Ill. 2.88) revealed that
the material here all appeared to be from a collapse of the east end of the south wall of the inner enclosure
(F513). This collapse ran due south from the inner enclosing wall in the area to the south of the monks’
graveyard (Ill. 2.89). All the material recovered appeared to be of nineteenth-century date and the decision
110
THE EXCAVATIONS
Ill. 2.88—Lower
monks’ garden
from the south
(Con Brogan,
DAHG).
was taken that this was the correct area in which to begin conservation work in 1994. Unfortunately the
site notebooks, finds register, sample registers, the plans and all but one of the sections were on the island
when the fire took place in 1995 in the archaeologists’ accommodation, so the description here is from
memory and from the locations described on the finds and sample bags.
Initial inspection on the ground and from aerial photographs had shown that the later lighthouse
wall (F1023) diverged from the earlier monastic wall (F640) further west, and that the lighthouse wall
continued north-west of the original monastic wall (Ill. 2.91). A second wall (F1024) ran north-west
towards a bedrock face, cutting off the east end of the lower monks’ garden. The southern wall (F1023)
measured c. 0.6–0.8m in thickness and 0.5–0.7m in height above the original ground surface of the
garden. The north–south wall (F1024) varied in thickness between 0.4m and 0.8m, and was c. 0.6m on
average above the original ground surface of the garden.The northern side of this end of the garden was
delimited by a bedrock shelf, except at the east end, where a low, c. 1m-high wall (F1026) was built,
joining the bedrock shelf with an outcrop of bedrock at the extreme eastern end of the garden.
One of the main reasons for removing the lighthouse walls and excavating this end of the garden
was that the eastern end of the monastic wall (F640) was moving outwards at the top and the excavation
111
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
of the material behind it would help to
alleviate that pressure, while a new wall was
needed to the east of it and tied into it to
ensure that the surviving monastic wall
stayed in position.
The excavation
The area excavated extended for 10m from
a near-vertical slab of bedrock on the
north side to a second, lower bedrock
outcrop on the south side of the terrace,
and from the east end of the garden for a
distance of 15m in a westerly direction.
The site sloped sharply downwards to the
south, where the original monastic south
wall of the garden appeared to have
disappeared.
Ill. 2.89—
Collapse over
Cell G (E.
Bourke).
The southern wall of the outer enclosure
The earliest visible feature on the site was
the outer enclosing wall, which formed
the southern wall of the lower monks’
garden (F640).This survived to the west of
the area excavated, but only three steps of
the original wall-walk survived into the
area excavated in 1994 (see lower monk’s
Ill. 2.90—
Eastern end of
outer wall of
lower monks’
garden (Con
Brogan,
DAHG).
112
THE EXCAVATIONS
Ill. 2.91—
Divergence
between upper
and lower walls
at eastern end
of lower
monks’ garden
(Con Brogan,
DAHG).
Ill. 2.92—Wall-walk at western end
of the cutting after conservation
(E. Bourke).
garden (middle section)). The area above the bedrock ledge was excavated—removing the absolute
minimum quantity of soil—in order to find out whether any trace of the base of the original outer
enclosing wall (F640) had survived (Ill. 2.92). At the eastern end of the ledge three or four base stones
survived where they were protected by an adjoining piece of rock outcrop, so it was clear that the wall
(F640) continued to meet the outcrop at its eastern end.
Directly above the bedrock there was a layer of dark reddish-brown compacted material (F1028).
This appeared to be natural and a similar layer was found at the base of the southern wall of the inner
113
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
enclosure at entrance 1. It seems that both walls were built on this natural soil rather than removing the
soil and building directly onto bedrock. Above this layer (F1028) and beneath the campion there was a
layer of dark organic material (F1021).This feature was observed just inside the wall (F640) below F1022
and produced one small piece of leather.
Walls at the east end of the lower monks’ garden
There was no wall visible between the two parallel vertical rock outcrops that ran east–west at the southeastern corner of the garden. A steep slope covered by campion ran down between the faces of the rock,
which would have been extremely hazardous for visitors. A similar excavation to that for the east wall
was carried out here. The absolute minimum of stratigraphy was removed and revealed a few stones in a
line at the base. It was not absolutely certain that this represented the line of a wall, but some structure
was needed here in order to retain this corner of the site and prevent anyone falling into the space
between the two rock outcrops. A new wall was built here in order to retain the material behind it and
make the area safe (Ill.2.93).
N
ll
w
wa
roc
k fa
ce
ne
new
wall
Ill. 2.93—Plan
of eastern end.
rock face
cell G
large
stone
F1027
ce
ck
fa
ro
F1
earlier
structure
F1022
02
2/
F6
37
stone
setting
F1036
ll
w
ne
wa
0
5
metres
A drystone structure (F1022) with a small jamb was visible on the surface of the slope, protruding
from the campion (Ill. 2.94). This feature, which was the north wall of a small structure, measured 2.5m
east to west and had a small southward projection at its east side that terminated in a jamb. This jamb
measured 0.2m out from the south face of the wall and was slightly splayed.The surviving length of wall
114
THE EXCAVATIONS
Ill. 2.94—F1022
as excavated (E.
Bourke).
being so small, it was impossible to
interpret this feature. What can be said
of it is that it is at a higher level than the
surviving stones of the wall-walk of the
original enclosing wall (F640) and
therefore must date from a time when
that wall-walk was no longer in use. It
can also be shown that it lies beneath
the paving surrounding Cell G (see
below).Thus it can be shown that there
are at least two phases of early medieval
activity that post-date the original outer
enclosing wall.
Overlying F1022, a layer of paving
(F1027) survived both within and
surrounding Cell G and extended
eastwards to the break in slope, where
it was retained by an outcrop of
bedrock on its east side. This paving
(F1027) consisted of small to mediumsized stones laid directly on the
underlying soil, which was not
excavated (Ill 2.95). Within Cell G, an
east–west drain lay beneath the paving.
This drain was 2.1m long, about 0.15m
wide and 0.1m deep, and was overlain
by single small paving stones, c. 0.3m by
Ill. 2.95—
Pavement and
stone setting,
showing their
stratigraphic
relationship
with F640 (E.
Bourke).
115
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.96—Large
rock, part of
the collapse
that removed
Cell G (E.
Bourke).
0.2m by 0.02m (Ill. 2.96). This area of paving was separated from the western half of the lower monks’
garden by a low drystone north–south wall (F1029). This was 4.7m long and ran from the bedrock at
the northern side to a point 1.75m from the wall-walk of F640. It was 1.4m wide at its northern end
and 1m wide at its southern end.
In the area between F1029 and the wall-walk, a stone setting (F1030) was defined by four stones set
on edge enclosing a flat slab. This feature measured 1m east–west and 0.75m north–south, and the flat
slab in the centre measured 0.8m by 0.7m.This structure was not excavated. Directly west of this feature
the paving had been smashed by the earlier collapse of F638 (see below, lower monk’s garden (middle
section)), the twelfth- to fourteenth-century collapse on the western side of the lower monks’ garden.
Description of Cell G
The remains of a circular beehive cell (Cell G) were revealed in the centre of an area of paving at the
east end of the lower monks’ garden. The northern half of this cell survives to a height of up to 2m. The
cell would have been roughly circular in plan, measuring 3.9m in internal diameter from the northern
jamb of the entrance to the east wall.The wall appears to have been corbelled, but this effect is exaggerated
at the northern edge of the building, probably owing to movement of stones during its collapse. The
north jamb of a west-facing entrance survives, with a rebate at the base of its inner face to take a door.
The wall is 1.3m thick at its base at this point, and the rebate measures 0.2m by 0.2m. The wall is
completely built of dry stone, using thin, flat stones of a small size. The middle of the south wall is built
on a rock outcrop, and the outside face of the end of the wall to the east also lies on bedrock (Ill. 2.100).
The paving outside the cell (F1027) rises towards the back of the cell, where the exterior wall survives,
in places only 0.5m higher than the paving (Ill. 2.99).
The outline of the southern jamb of the entrance survives in negative form, where there is no paving
in an area corresponding exactly with the location of that jamb (Ill. 2.98). It is interesting to note that
Cell G is at a higher level than the surviving wall-walk of the monastic enclosing wall (F640), so there
must have been a remodelling of that wall in the early medieval period to incorporate the southern side
116
THE EXCAVATIONS
Ill. 2.97—Cell G
during
excavation (E.
Bourke).
Ill. 2.98—Cell G
during
conservation
(E. Bourke).
117
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.99—
Northern side
of Cell G,
showing height
of wall (Con
Brogan,
DAHG).
Ill. 2.100—Base
of Cell G at
south-eastern
end, where it
lay directly on
bedrock (E.
Bourke).
118
THE EXCAVATIONS
of Cell G into the top of the outer wall. All trace of this remodelling was removed by the collapse of the
south wall of the inner enclosure and the demolition of half of Cell G.
Nineteenth-century use of Cell G and its collapse
It can be shown that Cell G survived intact into the early nineteenth century. The bottom layer of
occupation soil within the cell was a layer of dark organic material (F1018).This layer, which was between
0.2m and 0.3m deep, lay directly on the paving inside the cell and contained ten sherds of nineteenthcentury pottery. Above F1018, within Cell G, there was a layer of stones with a small admixture of dark
soil at its base (F1015/F1017), c. 0.35m thick. This layer contained a clay pipe stem and seven sherds of
nineteenth-century pottery.
Outside Cell G and just to the west of it, just above the pavement, a layer of dark organic material
(F1016) contained a single sherd of nineteenth-century pottery, a clay pipe stem and a sherd of
nineteenth-century glass. Just to the east and north of the cell, in the lower level of collapse (F1013), a
clay pipe stem, a piece of coal, an oyster shell and ten sherds of nineteenth-century pottery were found.
Close to the east and north of the cell, almost directly on the pavement (F1020), a ceramic lining for a
furnace was found, which may be an early medieval survival.
It seems likely that Cell G was occupied by the lighthouse-builders in the early nineteenth century,
but at some time during this occupation there was a catastrophic collapse of the south wall of the inner
enclosure (F513) directly above the cell, which partially demolished the cell and went on downslope to
remove almost all traces of the outer enclosure wall (F640) south of the cell.
The upper part of the fill of the cell consisted of rubble (F1014) that had collapsed inwards, leaving
the stones jumbled and without any particular orientation. This layer was up to 1.2m deep on the north
side of the cell and tapered out about 1m north of the break in slope at the edge of the pavement. It
contained eight sherds of nineteenth-century pottery and one sherd of local medieval ware and appears
to represent the collapse inwards of the upper part of Cell G.
A layer of collapsed drystone masonry (F1025) overlay the remains of Cell G and extended
northwards to the bedrock. The lower part of this collapse against the bedrock consisted of stones with
no particular orientation. Once this collapse had reached a point where it formed a slope of about 30–
35º, its upper part was aligned with the angle of slope and ran southwards towards the break in slope
delimited by the paving (F1027).
Later reuse of the east end of the lower monks’ garden
Because its surface was sloping southward at an angle of 30–35º, this end of the terrace had now effectively
been halved in size, and two new walls, F1023 and F1024, were built directly onto the collapse from
F513 (see description above). Inside these there was a build-up of stones and campion roots (F1012)
which contained nine sherds of nineteenth-century pottery and a piece of slag.
Above these layers lay the sea campion and its roots, which covered the whole of the east end of the
lower monks’ garden. This layer (F1011) was up to 0.3m in depth and covered both the area inside walls
F1023 and F1024 and the slope that ran southward and eastwards to the edge of the lower ledge of
bedrock at the southern edge of the garden. Finds from this layer included a clay pipe bowl, dating from
between 1800 and 1830, 21 sherds of nineteenth-century pottery and one piece of slag.
Interpretation
The last phase of construction by the lighthouse-builders when they walled in the smaller area left
available to them at the east end of the terrace was poorly built, as the walls were constructed directly
on top of the collapse (F1025). It is interesting to note that the pottery and pipes from this later
intervention are roughly contemporary with the pottery found in Cell G, implying that the rewalling
119
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.101—
Eastern end of
lower monks’
garden after
conservation,
from northwest (Con
Brogan).
happened quickly after the collapse. This makes sense, given that the monastery was probably only
occupied by the lighthouse-builders in the 1820s.
It appears clear that Cell G and its surrounding pavement were used by the lighthouse-builders in
the 1820s (Ill. 2.101). It is also clear that the western end of the lower monks’ garden was already filled
with collapsed material and redeposited soil from an earlier collapse of the wall of the inner enclosure,
at some stage in the fourteenth century (see below, lower monk’s garden (middle section)). So access to
the east end of the lower monks’ garden from the fourteenth century onwards was through the middle
section of the lower monks’ garden. This begs the question as to whether this area of the monastery was
ever used as a garden. It can be seen that the pavement (F1027) was completely removed by the
fourteenth-century collapse (F638), and it is probable that the original function of this part of the
monastery was residential, or at least contained some buildings, prior to the fourteenth-century collapse.
The stone setting (F1030) in pavement F1027 is an enigmatic feature, and the pavement continues for a
small distance beyond the north–south wall (F1029) before being removed by the earlier collapse. This
feature was not excavated and its function is therefore unknown.The paving is stratigraphically later than
the wall-walk of the original south wall of the terrace (F640), which implies that there was an original
surface beneath the level of the wall-walk in this area.The occurrence of a stone feature (F1022) beneath
the paving associated with Cell G and also later than the wall-walk of F640 shows that there are at least
two phases of early medieval activity post-dating the enclosure of the lower monks’ garden (Ill. 2.102).
Unfortunately, it was not possible to excavate beneath the pavement, or to excavate a large enough
amount of material at the base of the enclosing wall (F640) to date its construction. The nineteenthcentury collapse was in a straight north–south direction. The fourteenth-century collapse was also in a
north–south direction but fanned out in a north-west to south-east direction towards the north–south
wall (F1029), where it was restrained from covering the paved area around Cell G.Thus the two areas of
collapse hardly overlap at all except at the western end of the pavement, which was crushed by the
fourteenth-century collapse (F638).
120
THE EXCAVATIONS
Campion:
F1011
L i g h th o u s e r e b u i l d :
F 1024
F 1012
C o llapse :
F 1023
F 1025
F1014
F1015
F1017
F i r s t p h a s e l i g h th o u s e :
F 1018
F 1020
F 1013
F 1016
Third phase monastic:
C e ll G
F 1027
F 1030
F 1029
Second phase monastic:
Ill. 2.102—
Matrix for east
end of lower
monk’s garden
F 1022
F1021
Earliest phase monastic:
F640
Natural:
F1028
Bedrock:
Key:
Collapse and rebuilding:
F 1025
First phase 19th century:
F 1018
Third phase monastic:
F 1027
Second phase monastic:
F 1022
Earliest phase monastic:
F 640
N a tu r a l:
F 1028
Lower monks’ garden (middle section) (93E195)
Edward Bourke
Introduction
The test excavation carried out in 1993 (Ill. 2.103) at the base of the wall east of St Michael’s Church
(Cutting A) revealed that this section of the inner enclosing wall had collapsed on two occasions in the
past.The second occasion was in the latter half of the nineteenth century and the first was at some time
in the medieval period. The northern end of the cutting started at the base of the exterior of the inner
enclosure wall and revealed that the nineteenth-century rebuild was on a layer of rubble (Ill. 2.104).
This was not what the workmen had been instructed to do, according to a surviving drawing from the
archives of the OPW that shows how the wall was to be rebuilt (Ill. 2.105). In 1995 a more detailed test
excavation took place (Cutting 2), which showed that the wall was built on a set of flat slabs lying directly
on top of unsorted rubble from the previous collapse.
It became clear, even before going into the lower layers of stratigraphy, that the excavation of any
121
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.103—The
middle part of
the lower
monks’ garden
in the 1950s,
from the east
(DAHG
Archive).
Ill. 2.104—Base
of wall of inner
enclosure,
directly below
St Michael’s
Church,
showing 1890s
rebuild directly
onto unsorted
rubble (E.
Bourke).
122
THE EXCAVATIONS
Ill. 2.105—
Drawing from
1890s showing
that the wall
was supposed
to be rebuilt
onto bedrock
(OPW).
N2
N1
wall
0
5
metres
rubble
Ill. 2.106—Testtrench section,
Cutting A
(1993).
656
631
632
F635
631
F631 - campion / topsoil
F633
F632 - dark brown organic with some stones
F656 - backfill over repair of walls, stones & organic
wall
unexcavated
F633 - fine loose dark brown soil with mortar
F635 - dark yellowish sandy soil with some charcoal
F635
636
F636 - boulders
more of the base of the wall could undermine it and cause another collapse. Excavation in 1993 (Ill.
2.106) continued to the south of this area and showed that beneath the nineteenth-century material
were three further layers of collapse which contained no nineteenth-century pottery and which overlay
a debris layer of large stones that lay up against the inner face of the monastic enclosing wall at the south
end of the lower monks’ garden.
123
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The nineteenth-century wall above
this was not directly founded on the
earlier wall and was in poor condition; it
was also placing a heavy load on the
monastic wall beneath. The monastic wall
had a stepped wall-walk on its inner face
and the lower collapse of stones had
crashed down onto this, causing the wallwalk to move so that its stones were
sloping inwards at an angle of c. 20º (Ill.
2.107).
It was decided, since it was most likely
that the entire nineteenth-century
addition to the south wall of the outer
enclosure was in the same condition, that
the preferred method would be to scaffold
the wall on the outside in order to
monitor any movement in it and to
facilitate removal of the nineteenthcentury wall (Ill. 2.108). It was also
decided to attempt to reveal the monastic
wall of the outer enclosure and to carry
out conservation works on it. This would
require the removal of the nineteenthcentury wall above it along its entire
length.
Ill. 2.107—
Boulders and
F640 from east,
showing the
boulders
collapsed
against the
interior of F640
and the
resulting
distortion of
the wall (E.
Bourke).
The outer enclosure wall
The earliest feature visible in the lower monks’ garden was the original monastic wall (F640). This ran
eastwards from the junction with the upper monks’ garden as far as the western edge of the 1994
excavation of the east end of the ‘garden’. Within the excavated area the base of the wall could not be
seen anywhere.There had been a collapse of the south wall of the inner enclosure and these large stones
obscured the base of the wall (see below).
The original top of the wall was up to 1.02m wide, with an internal wall-walk with steps in it.These
steps varied in width between 0.52m and 0.38m.The internal face of the wall-walk had been altered on
a few occasions and it was possible to associate these adaptations with attempts to reconfigure the wallwalk after the collapse that filled the garden (Ill. 2.109).
Phase 1
The first phase (F640(1)) of wall construction could be identified along the length of the middle section
of the garden and through into the east end. At the western end of Cutting 1 the wall-walk began with
a set of steps (Step 1.1), quite high up, starting just below the sea campion layer (F631(1)). A series of
seven steps ran steeply downwards towards the east; they measured between 0.58m and 0.72m
north/south and between 0.2m and 0.5m east/west.The top three steps were flat and had well-preserved
drystone risers, but the next four had moved and sloped downwards towards the east with the drystone
risers having slipped outwards, allowing the eastern end of the treads to fall (Ill. 2.110). East of these steps
was a run of six slabs that sloped more gently towards the east, and these measured 0.38–0.78m east/west
124
THE EXCAVATIONS
Q2
CUTTING 5E
wall face
wall
phase 2
CUTTING 5
Ill. 2.108—Plan
of F640 and
Cutting 3.
F640
N
CU
T
TIN
G3
wall
phase 3
wall face
0
5
metres
Q1
125
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Q1
Q2
step
3.1
step
3.2
Ill. 2.109—
Elevation of
F640 and
phasing.
west wall
step 1.2
phase 3
step
2.1
step 1.2
phase 2
step 1.3
phase 1
0
step 1.5
5
step 1.4
metres
Ill. 2.110—
Composite
photograph of
elevation from
the north (E.
Bourke).
and 0.5–0.62m north/south.
East of these was another set of three steps (Step 1.2), descending towards the east and sloping only
slightly downwards towards the east; they measured 0.22–0.82m east/west and 0.54–0.6m north/south.
The drystone risers were almost vertical and had held their position well. East of these was another flat
run of eight stones of the wall-walk, which sloped imperceptibly downwards towards the east; these
measured 0.36–0.78m east/west and 0.58–0.69m north/south.
The end stone of that run formed the top step of two steep steps (Step 1.3) descending to the east.
The middle step measured 0.43m east/west and 0.38m north/south; the drystone risers sloped slightly
but both treads were roughly flat. The next run of flat wall-walk consisted of four stones that measured
0.38–0.92m east/west and 0.38–0.4m north/south.
The end stone of that run formed the only step (Step 1.4) down to the next level of wall-walk.This
dropped down onto a flat slab that measured 0.84m east/west and 0.34m north/south. This slab did not
fill the whole depth of the wall-walk at this point; three smaller stones laid flat to the south of it filled
up the width of 0.48m.These stones were at a higher level and may represent a repair, possibly associated
with phase 2 of this wall, as they would allow someone using the wall-walk to avoid having to step down
into the area between Step 1.4 and Step 1.5.
The wall-walk rose again after this slab (Step 1.5), with a step up of only 0.12m onto the next gently
sloping run. This step also occurred at the point where Cutting 3 came to an end.
The next, slightly undulating run of seven flat stones of the wall-walk ran eastwards into the 1994
excavation of the east end of the garden. These measured 0.21–1.09m east/west and 0.37–0.43m
north/south.
126
THE EXCAVATIONS
Phase 2
A second phase of wall-walk (F640(2)) was built above the westernmost step of Step 1.3.This ran flat for
ten slabs measuring 0.11–0.9m north/south and 0.15–0.63m east/west.This phase appears to be a repair,
as the wall above it diverges gently back from the earlier wall-walk, with the initial part of this second
phase being so narrow as to be almost useless.The wall above the wall-walk is c. 0.78m wide at this point
and terminates where it collapsed over the last stone of the run described above. Beyond the end stone
of this run was a gap of 0.55m where the stones had been removed to the last stone which lay on the
surface of the wall; this stone, measuring 0.47m east/west and 0.41m north/south, formed the top stone
of a single step (Step 2.1) that dropped 0.25m to the east. Only one stone survived from the next run of
stones, and measured 0.5m north/south and 0.3m east/west.
Phase 3
A third phase of wall-walk (F640(3)) occurred at the western end of the site at the south-eastern end of
Cutting 1. Here a higher piece of wall-walk, at the same level as the third step of Step 1.1, was built into
the wall. The wall above appears to have been rebuilt and some damage had been caused by the
nineteenth-century wall above but no direct connection can be determined with Step 1.1, as the stones
which might have joined the two sets of wall-walk are now gone. The western end of this run of wallwalk is the top step of a pair of steps (Step 3.1) descending eastwards; these measured 0.41m by 0.32m
and 0.32m by 38m respectively.The bottom stone of this feature measured 0.7m by 0.38m and the feature
then stepped up again by 0.22m (Step 3.2). One stone, which measured 0.5m by 0.56m, remained at the
east end of this feature; no other stones of this phase survived. The top of the wall at this point was only
0.58m thick.
The narrowness of the second and third wall-walks and the lack of formality in the way that they
have to be approached from the phase 1 wall-walk imply that the second two phases are repairs, possibly
after the phase 1 wall-walk had become difficult to use, following the collapse of the wall of the inner
enclosure into the garden (see below).
The cuttings
In order to explain the stratigraphy in the cuttings, Ill. 2.111 shows the location of all the cuttings in the
middle section of the lower monks’ garden. A second plan (Ill. 1.112) shows the features revealed in each
cutting.
Cutting A was a test excavation dug in 1993 (Ill. 2.106) to establish the stratigraphy at the western
end of the middle section of the lower monks’ garden. It ran north-west/south-east across the garden
and was 5m long by 1.5m wide.
Cutting 1 ran north-west/south-east across the garden and was 9.5m long by 2m wide. This was
excavated to determine the stratigraphy at the western end of the mid-section of the garden.
Cutting 2 was a small test excavation carried out to determine the stability of the base of the wall
of the inner enclosure. It was located directly against the outer edge of that wall where it met a bedrock
ledge; it ran north-east/south-west and was 2.5m long by 2m wide.
Cutting 3, a trench excavated east of Cutting 1 and against the inner face of the outer enclosing
wall of the garden, ran north-east/south-west and measured 6.5m long by 3m wide.
Cutting 4 was excavated in order to build a retaining wall to consolidate the outer face of the inner
enclosure wall, after it was discovered that this had been rebuilt on unconsolidated rubble. It ran northeast/south-west and measured 13m by 1.5m. It was excavated in four sections, numbered 4.3, 4.4, 4.1
and 4.2 from west to east.
Cutting 5 was directly to the west of Cutting 1 and south of Cutting 4. It ran north-east/southwest across the site and measured 10m long by 2–2.6m wide.
127
N
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
M1
M2
Q2
CUTTING 5E
ING
CUTT
2
N2
N1
Ill. 2.111—Plan
of cuttings in
the middle
section of the
lower monks’
garden.
CUTTING 5
CUTTING 4.2
CUTTING 3
CUTTING 4.1
CUTTING 4.4
P1
CUTTING 4.3
P2
CUTTING 1
Q1
CUTTING 7
CUTTING 6
0
128
10
N
THE EXCAVATIONS
F637
F669
C.5E
F635
C.2
F636
F651
F635
C.4.2
C.5
Ill. 2.112—Plan
of features in
the middle
section of the
lower monks’
garden.
F657
C.4.1
C.3
C.4.4
C.4.3
C.1
F664
C.7
F661
F665
C.6
F662
0
10
129
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Cutting 5E was a roughly L-shaped extension southwards and eastwards of Cutting 5. Its total length
was 16m along the edge of Cutting 3; it was 2m wide at the west end, increasing to 5.6m at the east end.
Cutting 6 was directly to the west of Cutting 1 and ran north-east/south-west. It was 5.4m long
and 3.6m wide.
Cutting 7 was directly south of Cutting 6 and west of Cutting 1. It ran north-west/south-east and
measured 2m long by 3m wide.
The excavation
A second north–south cutting was opened up in 1995 (Cutting 1), which revealed the same basic situation
as in the 1993 cutting (Cutting A) further east. While there were some intervening features and layers, the
same pattern of collapse and redeposition could be shown to have taken place in both cuttings (Ill. 2.113).
P1
0
5
earlier
excavation
spoil F629
metres
F632
lighthouse
wall F630
monastery wall
P2
F645
F631
Ill. 2.113—
Section,
Cutting 1.
F646
pavement
walk F20
F632
F631 - campion layer (19th century pottery)
F635
F634
F646 - stone free dark soil (19th century pottery)
F645 - dark humic soil with small stones (19th century pottery)
original
wall
F640
F633
F635
F636
F632 - dark humic soil
F633 - stone layer with some dark humic soil
F634 - dark organic soil
F635 - yellow brown sandy soil
pavement walk F17
F636 - loose rubble/boulders
Layers that pre-date the medieval collapse
The earliest layers encountered in the lower monks’ garden consisted of features that underlay the collapse;
these features occurred in Cutting 4.3 and 4.4, along the north side of Cuttings 5 and 5E and Cutting 6.
Cutting 4.3 and 4.4. This cutting was excavated in order to put in a retaining wall to ensure that the
inner enclosure wall to the south (F628) did not collapse into the lower monks’ garden (see below).
Above the bedrock was a layer of natural, inorganic, red sandy clay (F653). This natural material (Ill.
2.114) has been recovered under the wall of the outer enclosing wall to the garden (F640) and under
the stones of the inner enclosing wall at the east entrance (F513).
Above this was a layer of dark brown to black soil with well-preserved organic material (F643). Its
structure was very fine and a bit gritty; it was 0.1–0.2m thick and survived because the shape of the
bedrock meant that the soil was completely waterlogged at this level in 4.3 and 4.4 (Ill. 2.115). It did not
occur in 4.2, as the wall foundation did not go down to bedrock, and was not encountered in 4.1. Owing
to the good anaerobic preservation, numerous samples were taken from this layer, which had produced
grassy material, but no finds were recovered and there was no material suitable for dating. Layer F643
directly underlay F636 in this cutting.
2
Where a number appears in brackets after the feature number for a layer, it refers to the cutting in which the feature was
found. Where a number appears in brackets after the feature number for a wall, it refers to the phase of the wall that is being
discussed.
130
THE EXCAVATIONS
Ill. 2.114—
Natural at base
of Cutting 4.
Ill. 2.115—F643,
organic material
from beneath
the collapse
(F636) in
Cutting 4 (A.
Halmschlag).
Cutting 5. This layer (F643(5))2 was also encountered at the north side of Cutting 5 at the level of
the base of the retaining wall. Only the edges of the feature were uncovered and it was less well preserved
than on the other side of the retaining wall. Excavation ceased at this level, as it was not regarded as safe,
or expedient, to dig beneath the foundations of the retaining wall. No finds were recovered but several
soil samples were taken.
Cutting 5E. Some of this layer of dark brown to black soil with well-preserved organic material (F643(5E))
occurred along the side of the new retaining wall in Cutting 5E, towards its eastern end; the state of
preservation of the layer was even poorer than in Cutting 5, but some samples were taken.
131
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Cuttings 6 and 7. The bedrock came close to the surface at the north end of Cutting 6 and stepped down
three times at its north-western corner. Above the bedrock was a layer of red sandy clay with bedrock chips
in it (F653(6)); this clay was equivalent to the layer found in Cutting 4 but contained patches of grey and
yellow. Above this was a small pocket of the dark organic material (F643(6)), 1m by 0.4m in extent, which
occurred next to the steps in the southern part of the cutting.The layer was very thin, less than 0.1m thick,
and its preservation was not as good as in the other cuttings. Also beneath F635(6)), further west in the
cutting, was a small layer of dark clayey soil (F660), 0.01–0.04m thick, which may represent the same layer
in an even worse state of preservation. Apart from one large boulder (1.55m by 1.33m by up to 0.3m),
F636 did not survive in this cutting, so the layers mentioned above directly underlay F635.
Early wall at east end of Cutting 5E
This wall (F637) ran northwards for a distance of 4.6m from the stone setting (F1030) exposed in 1994
towards the bedrock at the back of the cutting. Only its eastern side survived to any height. At the northeastern end it was c. 0.4m high, while at its southern end, where it abutted the stone setting, it was only
a couple of courses high. The stones were generally small, 0.2m by 0.1m by 0.6m. Its western side had
been badly disturbed, as most of the stones were completely out of place and a repair had been carried
out in the middle of the west face, where slightly larger, better-laid stones were added, probably after
some collapse had taken place. There were two finds from this wall, a sherd of modern glass and a sherd
of Ham Green A/B ware.
One stone of the collapse (F636) was found just west of the wall, and it is likely that this caused at
least part of the damage. There was a layer of possible pavement (2m east–west by 1.5m north–south)
west and north of this stone (F688), which appears to have been laid down on top of F636 after the
collapse. These features and layers were directly overlain by F633 in this area.
Medieval collapse of inner enclosure wall (F636)
Cutting 1. At the east end of Cutting 1 was a layer of large boulders lying against the wall (F636(1)).This
layer was never bottomed but a depth of up to 0.7m was exposed where it met the outer enclosing wall
(Ill. 2.116). Some of the looser rubble at the top of this layer of collapse lay on the stones of the phase 1
wall-walk of the enclosing wall to the south (F640(1)). The main stones of this collapse were of large
size, up to 0.8m by 0.4m by 0.12m in this cutting, though even larger stones from the same collapse
occurred at the east end of Cutting 3. There was a high concentration of small stones among the larger
stones and a number of chips of bedrock mixed between them.
Cutting 3. The layer of large boulders F636(3)) continued into Cutting 3, where the boulders were of
even greater size, some up to 0.9m by 0.7m by 0.38m; one measured over 1.3m by over 0.3m but its
depth could not be determined because of overlying large stones. The gaps between the boulders were
not filled with any other material; a steel reinforcing rod was inserted as a probe into this layer and showed
that it was at least another 1m in depth in the centre of the cutting. This layer produced a spatulateheaded pin, a stone cross, an iron object and two sherds of nineteenth-century pottery, and a fragment
of a stone plate was found between the wall and the boulders. These finds that came from the edge of
F636 and might be subject to contamination were given the feature number F638, to take into account
the fact that they came from areas close to the wall where organic material had percolated down into
the surface of F636.
Cutting 5. Three boulders of F636 protruded through Layer F635 at the west end of this cutting. None
of them was fully exposed, but the largest measured 0.9m by 0.7 by 0.1m, while another triangularshaped boulder measured at least 0.62m by 0.43m by 0.2m. Owing to the discovery of other features
132
THE EXCAVATIONS
Ill. 2.116—
Lower monks’
garden after
conservation,
showing stones
of boulder layer
left in situ (Con
Brogan,
DAHG).
associated with the layer above (F635), it was decided to conserve the middle section of the lower monks’
garden as it would originally have appeared, with the boulder layer exposed where it met the outer
enclosing wall.
Cutting 5E. The layer of boulders (F636(5E)) was only exposed along the line of the retaining wall next
to Cutting 5 and in a very small area just north of the flagstone path (F657).The stones were again quite
large, averaging 0.22m by 0.15 by 0.15m, being much blockier than the standard stones found in the
drystone walls elsewhere in the monastery. These boulders also appear to have done quite some damage
to the north–south wall (F637) that divided the east end of the excavation from the west.
Cutting 4. In Cutting 4.3 and 4.4 this boulder layer (F636) directly overlay F643 and was up to 1.4m
thick. There was little or no soil between the stones, as was noted in Cutting 3, and because of the size
of some of the stones the cutting was only bottomed at the western side. The one difference noticed in
this cutting was the angle of the stones, most of which had been deposited at a very steep angle. There
were very few small stones and chips of bedrock in this cutting, unlike the same deposit in Cutting 3.
There were a few animal bones at the surface of this layer, but there is a strong likelihood that they may
have washed into it from F635 above.
Cuttings 6 and 7. Apart from one very large stone (see above), this feature (F636) did not survive in
Cutting 6. It seems likely that this is due to the closeness of the bedrock to the surface and to the fact
that the lower courses of the inner enclosure wall (F628(1)) survive at this point. It may also be due to
the fact that the bedrock slopes away southward in steps at this point. A second large stone that underlay
F661 was not excavated, but it may also originally have been part of F636.
Redeposition in the garden following the collapse
Cuttings 6 and 7.The earliest feature connected with the backfilling is a structure (F661) at the northern
133
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
edge of Cutting 6, consisting of a platform of stones set on top of a large, flat boulder. This boulder,
which measured 1.26m by over 0.6m, was not excavated as it appeared to form the base of the stone
structure. The platform was 2.7m long and at least 0.7m wide. Its northern extent was not excavated, as
this might have undermined the south face of the inner enclosure wall. There were two courses of
masonry visible, a lower course made up of stones measuring on average 0.25m by 0.12m and an upper
course of seven stones measuring on average 0.5m by 0.3m; behind this upper course the stones were a
mixture of sizes (Ill. 2.117). This presented a flat platform against which F635 had been dumped. It is
tempting to see this as a work surface for the wall repair, as the bedrock was sloping southwards, with
Ill. 2.117—F661
from south (A.
Halmschlag).
natural steps in it, across Cutting 6.
Main backfill layer (F635)
A layer of dark yellowish sandy soil with some darker charcoal-stained patches appeared in all of the
cuttings except Cutting 2. It was over 1m deep in places, where bottomed.This layer was laid down over
the layer of boulders (F636) across the site and only petered out at the east end of Cutting 5E, where the
boulder collapse came to an end. It appears to have been laid out to form a garden once it was decided
that the collapse was too large to clear from the enclosure.
Cutting 1. In Cutting 1 this layer (F635(1))3 contained some pottery and a reasonable quantity of bird,
fish and animal bones. It was up to 1.02m deep at the south end of Cutting 1, where it came up to just
below the base of the phase 3 wall-walk of the outer enclosing wall (F640(3)).
The pathway down to the east end of the garden passed through Cutting 1 (see below). The three
stones in this cutting were planned, but the plan did not survive the fire. These stones are marked
approximately on plan as F648, while the rest of the pathway is fully recorded under the feature number
F657 (Ill. 2.118).
3
The finds from Cutting 1 were damaged in the fire in the archaeologists’ accommodation in 1995; the same is true for Cutting 2 and Cutting 4. Cutting 1 was extended to the east after the fire and the finds for that area are labelled F635(1E). All the
finds from the fire are registered under the feature number F1019.
134
THE EXCAVATIONS
North of these steps there were several
upright flagstones (F649) placed in an open
rectangle, perhaps to retain soil from north
of the pathway (F648). The relevant plans
and notes do not survive for this feature.
Cutting 3. This layer (F635(3)) extended
downward across Cutting 3 to where it
met the wall-walk of F640(3); it was up to
1m deep and ran across the whole cutting
in a southerly direction.
Cutting 5. This layer (F635(5)) extended
right across the cutting and was a little
over 1m deep at its southern end, where
the boulders had been exposed. The layer
is the same as in the other cuttings, if
perhaps a bit darker, with the same
selection of animal bone as elsewhere. It
also yielded two lumps of fused metal, one
of which was fused onto a stone, and a
whetstone of fine micaceous sandstone.
Because this was the layer and level at
which it was decided to conserve the
garden in this area, the layer was not
bottomed on its northern side.
Ill. 2.118—
Pathway F657
from west (A.
Halmschlag).
Cutting 5E. This layer (F635(5E)) covered
all of Cutting 5 and most of Cutting 5E.
It ran for c. 12m from the east baulk and only petered out against the pavement F1027, which surrounded
Cell G and ran out past the north/south wall (F1029/F637). One piece of slag was retrieved from this
layer. It became clear that the paving survived east of the collapse of F636, where it was partially preserved
as an entrance into the paved area at the east end of the garden.
Stone features preserved in F635(5) and (5E). An 8.5m-long curving wall (F669) ran east–west along the
cutting, west of F637 and north of the pathway F657. The junction between the two walls was badly
disturbed so it was not possible to say which was earlier, particularly as only the bottom two courses of
F637 survived and that wall had obviously been repaired on more than one occasion. It is likely that
F669 post-dates F637, as it survives to a height of c. 0.5m at the junction. The wall is of drystone
construction using small stones, but three large stones (F670) were added to the top of the wall midway
along its length.These measured 0.6m by 0.2m by 0.18, 0.48m by 0.18m by 0.12m and 0.38m by 0.28m
by 0.1m respectively. The wall had collapsed in the 0.7m gap between the most westerly of these stones
and the other two. It seems that these heavy stones may have been put up onto the original wall and
then destabilised it. This wall continues to the west, where it was only two courses high. The material
behind the wall, which functioned as a terrace, was not excavated, as this appeared to be a medieval
terrace wall associated with the conversion of the terrace into a garden.
South of this wall and level with its base was a pathway of stones (F657), set like stepping-stones
135
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.119—Wall
F669 from
south-west
during
conservation
(Con Brogan,
DAHG).
Ill. 2.120—Wall
F637 and
eastern end of
pathway F657
after
conservation,
from north
(Con Brogan,
DAHG).
from the end of F635 at the east and running upslope towards the steps leading down to the east entrance
of the inner enclosure. There were 34 of these steps, running east–west across Cutting 5. There was one
informal step up just south of the gap in F669 and a second informal step at the western baulk of the
cutting. Three stones of this pathway were noted in Cutting 1, but unfortunately the relevant plan was
destroyed in the fire. Their approximate location is marked by dotted lines on the plan (Ill. 2.112). Three
further stones of this pathway were uncovered in Cutting 7, where they again lay on the surface of F635
and stopped just short of the bend in the stairway down to the east entrance to the inner enclosure.
Cutting 4.This layer (F635(4)) was sandier at its upper (northern) end than at the other end of the cutting,
136
THE EXCAVATIONS
but it was still light brown to yellowish in colour and very uniform in its appearance. It was also thinner
here, being only 0.1–0.2m thick. Bird, fish and (to a lesser extent) mammal bones were present in this
layer, and a bulk sample was sieved for analysis. There was also some charcoal in the layer.
Cuttings 6 and 7.This layer (F635(6)) was yellower in these cuttings and became progressively wetter the
closer excavation came to bedrock.The bones retrieved, which were only from mammals, were in a very
bad state of preservation and crumbled when touched. The layer was c. 0.25m deep at the north end of
the cuttings, where it abutted F661, and its depth could not be measured at the south end, as the bedrock
was stepping downwards so steeply towards the south and there was a fear of undermining the steps
down to the east entrance of the outer enclosure. F635 was the lowest level excavated in Cutting 7, as
the western terminus of the pathway (F657) that had been found earlier in Cutting 5E and in Cutting
1 was laid on F635. This pathway seems to have led up to the corner of the steps leading down to the
east entrance of the outer enclosure.Three sherds of local medieval ware, one sherd of Orléans-type ware,
a cross of Valentia slate, a possible hinge strap and two pieces of slag were retrieved from this layer.
Dark organic soil (F634) redeposited over sandy layer (F635)
This layer of dark organic soil occurred in every cutting except Cuttings 6 and 7, Cuttings 4.1 and 4.2,
Cutting 2 and the east end of Cutting 5E. It contained a high proportion of bird and fish bones, with a
lesser quantity of animal bones.
Cutting 1. Layer F634(1) overlay F635 in this cutting. It was 0.1–0.3m thick and ran from 1m north of
the outer enclosing wall back to within 2m of the inner enclosing wall.
Cutting 5. F634(5) occurred on the south side of Cutting 5, thickening to c. 0.2m at its western end,
while further eastwards it became patchy and mixed with F633. These areas of mixing were given the
feature number F639 and yielded one piece of slag. There was still a large amount of bird and fish bone
in Layer F634(5)), together with a smaller number of mammal bones, and just in this area there was a
large quantity of badly decayed mollusc shells. The soil became darker and patchier than in Cutting 1
and there was still an amount of charcoal mixed into the soil. A charred seed, a thin-walled piece of iron,
possibly from a tin can, a roof slate, some stone and burnt mortar were retrieved from this layer.
Cutting 5E. The continuation of the dark organic layer (F634(5E)) petered out 7.5m from the west end
of Cutting 5 and it was only slightly over 0.1m in average thickness at the western baulk. An iron object
of uncertain function came from this layer, and a small bone gaming piece came from its base at the
point where it petered out at its eastern edge. This layer did not overlie the pavement F1027 at the east
end of the garden and in that area F633(5E) directly overlay F635(5E).
Cutting 4. The same layer appeared directly above F635 in Cutting 4. This layer (F634(4)) was dark
grey/black with a fine texture and contained both traces of charcoal and some lumps of charcoal together
with large quantities of fish and bird bone, but mammal bones were much less frequent. Bulk samples
were sieved from this layer. It directly underlay the later nineteenth-century collapsed material from the
outer wall of the inner enclosure that was repaired by the OPW (see below).
The nineteenth-century wall enclosing the lower monks’ garden
The nineteenth-century wall (F630) was assumed to have been erected in the 1820s (Ill. 121) and mostly
overlay the earlier monastic enclosing wall (F640) (Ill. 122). Cutting 3 was opened along the wall to
establish the exact connection between the two. The nineteenth-century wall (F630) was 1–1.2m wide
137
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.121—
Nineteenthcentury wall
F630 and its
position relative
to the early
medieval wall
F640 (Con
Brogan,
DAHG).
Ill. 2.122—Wall
F630 in the
1950s, from
north (DAHG
Archive).
138
THE EXCAVATIONS
Ill. 2.123 —
Photo showing
how F630 was
built on soil and
rubble above
F640 in places
(E. Bourke).
at its base, 0.6–0.7m wide on top and varied in height from 1m to 1.75m. The first thing that was done
was to strip a layer of sea campion with some brick and yellow sandstone from the lighthouse period
from the top of the wall.Then, as excavation of the upper layers of the backfilled soil continued, it became
clear that while layers F631 and F632 lay against the lighthouse wall, the latter oversailed the monastic
wall on the inside by between 0.06m and 0.4m along the interior face and lay directly on top of F635
and F636 as they rose and fell along the inner face of the wall.
Wall F630 consisted of two faces with a rubble core, and it was clear that it was not structurally
sound. Part of it had collapsed during the digging of the test-trench in 1993, as its rubble core was
unstable. It was decided to take down the lighthouse wall in order to reduce pressure on the monastic
wall beneath. During this operation a clay pipe bowl dating from 1800–30 was recovered from the rubble
core.
As the wall was being taken down, the reason for its instability became apparent. It seems that the
top of the monastic wall had not been cleaned or secured before the addition of the later wall, and in
places up to 0.4m of earth, dislodged stones, rubble and campion was sandwiched between the two walls
(F652) (Ill. 2.123).
Nineteenth-century layers and minor rebuilding work in the lower monks’ garden
Cutting 2 was excavated to explain how the wall was rebuilt at its eastern end. It established that the wall
was built on unconsolidated rubble, with a layer of dark brown organic material with stones and some
mortar inclusions (F633) left as the base layer after the wall was re-erected. The amount of mortar in
F633 decreased in the area not directly below St Michael’s Church and the layer did not occur at all in
Cuttings 6 and 7, which are to the west of the wall collapse. Above this a layer of dark organic material
(F632) had built up, equivalent to F656 in Cutting A. At the western end of the site, in Cuttings 6 and
7, there had been minor rebuilding work to the steps leading down to the east entrance in the lower
monks’ garden.
Cutting 2. This cutting was opened originally to establish the stability of the exterior wall of the inner
139
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.124—
Constructing
the revetment
wall in Cutting
4 (E. Bourke).
enclosure above F628. The layer of stone
collapse (F655) underlay the late
nineteenth-century repair at the base of
the cutting. This had been exposed in
Cutting A in 1993 and it was intended to
expand the area to see whether the wall
was as unstable as it had looked at that
time. The base of the reconstructed wall
(F628(3)) was on a series of north/southaligned slabs, upon which the rest of the
wall was built (Ill. 2.104). This raft of
stone lay directly on the rubble of the
earlier collapse; it was unstable and had
to be retained temporarily by grouting it
with concrete. The layer of collapse
contained two stone settings that looked
as if they were temporary settings to
allow entrance onto the collapse from
either side.
The first setting (F1050)4 was at the
east end of the cutting, propped against
bedrock. It was about 1.2m high and
between 0.5m and 0.8m wide. Its top was
approximately at the same level as a step
in the bedrock to the east of the cutting.
The entire structure proved to be
Ill. 2.125—
North–south
wall (F637)
west of Cell G
during
excavation (A.
Halmschlag).
4
While the section for this area survives from 1993, the plan and detailed sections for Cutting 2, as well as the finds and notebooks, were destroyed in the fire in August 1995.
140
THE EXCAVATIONS
completely unstable and it partly collapsed shortly after it was exposed. Several pieces of nineteenthcentury pottery were found in the soil around it.
On the west side of the cutting a second, smaller stone setting (F1051) was uncovered, c. 0.4–0.5m
wide and c. 0.6m high. It was built on top of a very steep layer of collapsed stones and its structure was
very irregular and unstable.
Above the collapse was a thick layer of dark soil containing lumps of mortar (F633(2)). A stone cross
was banked up against the rubble within this layer and two sherds of nineteenth-century pottery were
also found. This layer was not bottomed in Cutting 2.
Above the collapse, and also above F633, a layer of dark humic soil had been deposited (F656). This
layer sloped southwards into Cutting 5E to about midway towards the outer enclosure wall, where it
tapered off to nothing. It seems to be equivalent to F632.
Cutting 4. In Cutting 4 a layer of collapsed building stone overlay F634(4).This layer (F655) is associated
with the later nineteenth-century collapse illustrated by Dunraven (Ill. 2.126) which damaged St Michael’s
Church, and a drawing was prepared by the OPW architect showing how the conservation work was to
be carried out (Ill. 2.105).This was supposed to have involved excavation to a certain depth and rebuilding
on firm foundations. This did not in fact happen, and at the east end of the wall in Cutting 2 it was
apparent that a raft of stones was laid out on the original collapse and that the new wall was built up
from that. This collapse (F655) was apparent in both Cuttings 2 and 4 and measured, where bottomed,
up to 1m in depth.This layer also contained some mortar, and many of the stones were suitable building
stones, either flat or roughly squared. Above this was a layer of dark brown humus and small stones (F656),
similar to F632, which had been thrown over the stones of F655 to disguise the shallowness of the
foundations. This layer was up to 0.5m thick and directly underlay the campion, F631(4).
Ill. 2.126—
Photograph
from Dunraven
1875–7,
showing
collapse of wall
of inner
enclosure.
141
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Cutting 1. Above F634 in Cutting 1 was a layer of dark humic soil with some stones (F633(1)). This
contained practically no mortar inclusions, as it is west of St Michael’s Church; it appears that F633 only
had mortar inclusions in the areas directly beneath the church, with little or none being incorporated
into the layer further west.
Above F633 another layer of humic soil containing few stones (F632(1)) ran the whole way across
the cutting. It was up to 0.4m deep at the north wall and up to 0.7m deep where it met the outer
enclosure wall.This cutting was extended after the fire and produced seven sherds of nineteenth-century
pottery and a rubber/hammer stone. The surface of this layer had been disturbed in the nineteenth
century, firstly by a layer of dark humic soil with some stones (F645), which only occurred in the centre
of the cutting and did not reach far beyond the baulk to either side. This layer was up to 0.4m deep at
its centre and measured 2.3m north–south and about 2m east–west, occupying an area from the centre
of the cutting northwards. It was cut by another small layer of stone-free dark soil (F646(1)), which cut
the southern end of F645 and was also cut into the surface of F632.
Cutting 5. Above F634 was a layer of loosely compacted, very dark brown soil (F633(5)). This contained
few bird bones, none of which was burnt; it also contained several large building stones, one of which
had a straight chiselled edge. This layer can be interpreted as soil and some stones from the nineteenthcentury collapse of the inner enclosure wall. It continued below the new retaining wall on the north
side of the cutting, particularly at the east end. It also contained a good number of large lumps of mortar.
Two sherds of Ham Green A/B transitional ware came from this context. This layer lay on F634(5) at its
western end and directly overlay F635(5) at its eastern end. Mostly the transition was clear, but in some
pockets there was a mixture of F635(5) and F633(5); a new feature number (F650) was given to the
single sherd of unglazed pottery from the mixed layer.
Above this layer throughout Cutting 5 was a layer of dark brown loose humic soil (F632(5))
containing some animal bones and five sherds of nineteenth-century pottery. Between the two layers
was a transitional layer with both layers mixed together (F641(5)); it is not certain how this layer was
formed but it may have been the result of spade cultivation.
Cutting 5E. This fine, loose, dark brown soil (F633(5E)) was up to 1.12m thick at its deepest around the
middle of the cutting. It still contained lumps of white mortar, and the only change noted in it was the
occurrence of decayed marine molluscs in its lowest levels. A small hammered bronze wire was found in
this layer near the retaining wall.
A group of six large boulders (F659) were found within F633(5E). Four of them were almost vertical
but had subsided at an angle southwards; one was lying flat on the slope inside them and one only just
peeped up through F635(5E). The recumbent boulder measured 0.95m by 0.64m by 0.18m, while the
uprights averaged between 0.45m and 0.85m in length and between 0.08m and 0.18m in depth. They
were up to 0.4m high and sloping southwards. It is difficult to say what this feature represents. Its base
was in F635(5E) but it stood up well into F633(5E); it may have been the edge of a terrace or it could
have been a platform, given the one surviving recumbent stone.
Above the collapse, and also above F633, a layer of dark humic soil (F656) had been deposited. It
sloped southwards into Cutting 5E to about midway towards the outer enclosure wall, where it tapered
off to nothing. This layer clearly seems to be equivalent to F632.
Cuttings 6 and 7. F633 did not occur in Cutting 6. A dark humic layer lay above F635 in both cuttings.
This material had been dug through on several occasions for the insertion of the small walls F662, F664
and F665, but distinctions between the digging and backfilling associated with this work could only be
142
THE EXCAVATIONS
Ill. 2.127—Wall
F664 from
north (A.
Halmschlag).
made where the trenches for the walls had been dug into F635.
The layer of loose dark soil (F632(6)) directly overlay F635 in Cutting 6, and during the time that
it was accumulating there was a remodelling of the steps leading down to the east entrance. Two walls
were built at this stage. The first (F662) was built against the north face of the steps as they exited into
the upper monks’ garden. A trench (F663) was dug into F635 and the wall was constructed within it.
The trench was 0.42–0.87m wide originally and its backfill was a mixture of F635 and F632. This wall
(F662) was 3.2m long and about 0.4m wide, except at its east end, where it widened to c. 0.6m. It was
built against the outside of the steps and its base partly oversailed the steps themselves.The wall, given its
thickness, was composed of quite small stones, with only a few spanning the full width from the front of
the wall to the back.
Next to this wall was a second wall, only the lower courses of which survived (Ills 2.127 & 2.128).
This wall (F664) curved eastwards from the end of F662 and was located in an extension of the
foundation trench dug for both walls. It was c. 1.8m long and only consisted of a front face, roughly
constructed against the slope of the garden. At its eastern end it was two stones thick, with the stones
aligned north/south at right angles to the direction of the wall. The wall appeared to be petering out as
it reached the eastern baulk.There was one incised cross-slab in its front face and there was a brick, with
mortar still attached to it, at its eastern end. One piece of yellow sandstone—‘Portland Stone’—from the
lighthouse period was also built into the wall.
South of wall F662 was another (F665) that abutted F662. This wall, which consisted of a front face
only, was laid out with the short axis of the stones facing the edge of the steps. Its stones ranged from
very small (0.12m by 0.06m) to medium size (0.52m by 0.22m).The wall was dug into F635 and survived
to a height of c. 0.6m. It ran straight north-east/south-west and ignored the curve of the steps. The area
east of the wall was backfilled with loose dark soil with few mortar inclusions (F666).This layer was finer
in consistency than F633 and there were no finds from it. On the western side of F665 the gap between
the wall and the steps was backfilled with loose rubble (F667).
All this material beneath the campion and above F635 was registered under F632(6) and F632(7). A
clay pipe, eleven sherds of nineteenth-century pottery, a sandstone piece of ‘Portland Stone’ and a waterrolled pebble were found in Cutting 6. A clay pipe, a hone stone and three sherds of nineteenth-century
143
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.128—Wall
F664 from
south (A.
Halmschlag).
pottery were found in Cutting 7. Cuttings 6 and 7 are to the west of the nineteenth-century collapse of
the outer wall of the inner enclosure and thus none of the debris associated with that collapse would be
expected in these cuttings.
Campion layer
Cutting 1. Above F632 was the campion layer, which survived across the cutting from north to south.
This layer (F631(1)) measured up to 0.34m in thickness. Two modern glass tumbler fragments, which
survived the fire, were found in this layer. Above this was a layer of earlier excavation spoil (F629), which
had formed the base of a stone stack to contain stones from the 1993 excavation.
Cutting 3. This layer (F631(3)) averaged 0.3m in depth in this cutting. Three sherds of glass bottle, a
turquoise glass bead, a sherd of an iron cauldron and twelve sherds of nineteenth-century pottery were
recovered from it.
Cutting 4.This layer (F631(4)) averaged 0.15m in thickness in this cutting. All finds were destroyed in the
fire.
Cutting 5.This layer (F631(5) and F631(5E)) averaged 0.2m in thickness.The finds from it were destroyed
in the fire.
Cuttings 6 and 7.This layer (F631(6&7)) ranged in depth from 0.07m at its northern edge to 0.2m at the
south. Two sherds of modern glass and seven sherds of nineteenth-century pottery were recovered from
it.
Dating
A series of radiocarbon dates was obtained from the deposits in the lower monks’ garden. The first of
these is a date of AD 1024–1154 from F633, the layer of nineteenth-century collapse. Although some
144
THE EXCAVATIONS
Ham Green B ware was also recovered from this layer, there is no doubt that it is actually part of the
debris from the collapse of the south wall of the inner enclosure. The fact that there was a great deal of
mortar in this layer from beneath St Michael’s Church and the fact that it ran down from the base of the
rubble in Cutting 2 indicate this clearly.
There were three dates from the earlier redeposited soil (F634), AD 726–887, AD 773–946 and AD
689–870. There was no surviving pottery from this layer but there were two fragments of stone and
mortar, a bone gaming piece and an iron rod of uncertain function. These dates show that the soil that
made up this material must have been redeposited from elsewhere in the monastery, as there are secure
pottery dates of the twelfth to fourteenth centuries from the layers below.
F635 produced a date of AD 1037–1167, but there are sherds of local pottery and a sherd of Orléanstype pottery securely from this context. Again this argues for the soil having been redeposited in the
garden.
There is a date of AD 723–943 from F642, which is a transitional layer between F635 and F636.
Again, there is no doubt that the material has been introduced from elsewhere, as the layer underlying it
produced a later radiocarbon date.
A date of AD 1021–1153 was obtained from charcoal in F636. This layer produced a fish-bone pin
and bead. The layer of collapsed stone was not normally filled with soil; where found, there was air
between the stones. F638—a feature number given to finds recovered from close to the surface of this
layer to take account of the fact that they may be residual (see above)—produced an iron object, a
fragment of a stone plate and two sherds of nineteenth-century pottery.
Thus we can see that we must go for the latest uncontaminated date, and that is for the deposition
of the medieval pottery. So the only thing we can say for certain is that the initial collapse took place at
Ill. 2.129—General view of monastery after conservation (Con Brogan, DAHG).
145
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
some time prior to the mid-fourteenth century (Ills 2.129–30).
Interpretation
It was clear from the excavations at the east
end of the garden (see above) that there were
a number of phases pre-dating Cell G that
post-dated the outer enclosure wall (F640).
Only hints of this were revealed in the
excavation of the middle part of the garden
in Cutting 4, parts of Cutting 5 and in
Cutting 6, where a layer of dark organic
material overlay the bedrock at the north side
Ill. 2.130—
General view,
of the excavation, just beneath the wall of the
after
inner enclosure and underneath the collapse
conservation,
layer (F636).
from north-east
(Con Brogan,
The next stage was the collapse of part
DAHG).
of the wall of the inner enclosure (F628(1)),
which left a debris field of large stones
sloping southwards across the cuttings. This
appears to have been too difficult to move
and a layer of redeposited yellow sandy soil
was laid out across this portion of the site. A
smaller layer of organic medieval material
(F634) was deposited on the surface of this
layer. There was an informal pathway down
to the east end of the garden, which still
housed a standing building (Cell G), and
there were stone terraces on the north side of this pathway, perhaps to create a small garden. The fact
that there is not more medieval activity may not be surprising, as there are indications that the site was
semi-abandoned from the thirteenth century onward. The partial rebuilding of the wall-walks of the
outer wall (F640(2 and 3)) appears to date from about this time.
At some stage in the 1820s the outer wall of the outer enclosure of the monastery was raised, probably
associated with the use of Cell G by the lighthouse-builders and the need to provide shelter on the way
down to it. This wall was poorly built and was not fully founded on the earlier wall.
The second collapse of the outer wall of the inner enclosure occurred in the late nineteenth century.
This wall was reinstated by the OPW but not on secure foundations.The debris from this collapse (F633)
ran down to the nineteenth-century wall, and a subsequent development of soil raised the area inside
the wall by between 0.5m and 0.8m. There was little subsequent activity, except for the development of
a layer of campion that covered the whole garden (Ill. 2.131).
The fact that material with early medieval dates was redeposited from elsewhere on the site means
that, although we cannot talk in specific terms about the individual deposits, we can speak in general
terms about the occurrence of insects, plants and wood on the island in an early medieval milieu (see
general discussion on environmental issues).
The pottery from the lower monks’ garden is more exotic than expected, with non-local pottery
predominating, and the finds, though few in number, are intriguing. The history of collapse in this part
of the monastery is complex and shows that the evolution of the monastery continued even after its
primary function was over.
146
Lower monks' garden ( middle section)
Features at east end
(selected features only shown)
Modern:
F631
Late 19th century:
F661
F665
F664
F662
F646
F666
F645
F656
F632
F641
F633
Early 19th century:
F630
F628(3)
F655
F1051
F1050
F650
F659
F652
Medieval collapse and redeposition:
F634
F649
F648
F657
F635
F640 (3)
F636
F638
F640 (2)
Monastic phase 3:
F1027
F637
Cell G
Monastic phase 2:
F1022
F1021
Monastic p hase 1:
F628
F640 (1)
F660
F653
Key:
F1028
147
Modern:
F680
Medieval collapse and redeposition:
Late 19th century:
F687
Late alterations to wall:
Early 19th c entury:
F695
Monastic phase 3:
F635
Monastic phase 2:
F1022
F640 (2)
Monastic phase 1:
F643
Natural:
F653
F1027
Ill. 2.131—Matrix for lower monk’s garden (middle section) with selected layers from the east end added to show all phases.
THE EXCAVATIONS
Natural:
Bedrock:
F643
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
THE UPPER MONKS’ GARDEN (90E54)
Claire Cotter
Introduction
Repair work to the southern retaining wall of the terrace known as the upper monks’ garden (i.e. the
outer enclosing wall) necessitated the removal of the deposits immediately inside it. This report refers to
the westernmost section of the wall (Ill. 2.132). An area measuring 8m east–west by 5m north–south was
opened and stepped down to a maximum depth of 2.8m adjacent to the retaining wall (Ill. 2.133). The
excavation revealed that the present retaining wall was a secondary feature constructed largely on the
foundations of a more substantial earlier wall. The only finds recovered were from the secondary levels
and date from the nineteenth century (Ills 2.135–6).
Ill. 2.132—
Cutting during
excavation
from southwest (Con
Brogan,
DAHG).
Ill. 2.133—
Cutting during
excavation
from north-east
(C. Cotter).
148
THE EXCAVATIONS
Entry to the upper monks’
garden is via an opening through a
north–south step in the south wall
roughly midway along its length (Ill.
2.134). A pathway, demarcated by
low uprights and some horizontal
slabs, runs east from this opening
towards the lintelled entrance to the
inner enclosure of the monastery
(south entrance 2, inner enclosure)
(Ill. 2.139). The pathway (F707)
measures 1–1.3m in width and was
Ill. 2.134—The
marked for most of its length by
nineteenthuprights that varied in height
century
entrance in the
between 0.14m and 0.47m. Six large
1950s (DAHG
slabs formed the southern side of the
Archive).
path towards its western end. The
slabs rested on uprights, thus
forming a box-like construction that
prevented subsidence.The surface of
the path consisted of compacted
rubble and clay, about 0.1m in depth
on
average, which
overlay
nineteenth-century deposits. This
pathway was not excavated but
remained in situ adjacent to the
cutting.
The present retaining wall (F700) along the south side of the upper monks’ garden is 1.5m in height
at the entrance and batters inwards from a width of 0.71m (just below present ground level) to 0.42m
at the top. The easternmost section was rebuilt in 1998 and the walling between this and the entrance
curves markedly inwards. There is a vertical drop of c. 5m on the outer face of the wall and the ground
slopes away fairly steeply beyond this.
The projecting foundations of an earlier wall (F701) can be seen along the outer face of the terrace
(Ill. 2.138). The surviving upper edge of this earlier wall varies from 0.45m below present ground level
at the western end of the excavated area (Ill. 2.137) to 2m below present ground level at the eastern end.
These foundations were laid on large, natural boulders resting on the bedrock. Over time the surviving
upper courses of the wall, particularly at the western end, have been pushed outwards, rendering the
whole structure unstable.
Prior to excavation, it was apparent that the present retaining wall represented a rebuilding, as it curved
back inside the line of the original structure. As part of the repair work involved the insertion of concrete
shuttering behind the wall, an area 8m by 5m was excavated and stepped down to facilitate this work.
The excavation
Monastic period
The earliest layer on the site was a layer of large boulders (F706) which ran across the base of the cutting
and could be seen protruding from underneath the front of the original east wall (F701).
Towards the centre of the cutting bedrock outcropped c. 1m below present ground level and then
149
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
N
stepped
down
F707
path
F706
boulders
F701
remains
of primary
wall
F700
ck
bedro
Ill. 2.135—Plan.
L3
L4
probable
line of
primary
wall
present
retaining
wall
surviving
face of
primary wall
L1
L2
0
3
metres
dropped vertically to a group of large boulders. The boulders measured up to 1.6m in length and 0.9m
in height, and in places continued under the foundations of the original retaining wall.
The internal face of the earlier wall (F701) was not visible before the excavation began. Removal of
the terrace topsoil (F702) revealed that the inner face survived only at the western end of the excavated
area. The extant section runs for a distance of 0.9m from the west baulk and curves outwards west–east.
Three courses of faced walling were evident at a maximum depth of 0.75m below present ground level.
Below this level the wall consisted only of unfaced rubble. The original wall was therefore about 1.2–
1.3m wide (allowing 0.2–0.3m for the displacement of the present upper courses). Over the remainder
of the site, the bulk of the original wall had collapsed down to just above the boulder layer (F706). At
150
THE EXCAVATIONS
PROJECTED SECTION
L1
L2
path
surface
F707
F700
present
retaining
wall
old sod
red ash
and charcoal
F702
tip
lin
e
top of
primary wall
from section
C-D
SECTION A-B
F704
F709 rubble
F701 primary wall
F705
L3
Ill. 2.136—
Section
L4
F702- campion and topsoil
rubble
0
0.5
1
F706
SECTION C-D
F704- dark brown loose clay with
charcoal lenses
F705- stone chippings and clay
F706- boulders
2
metres
Ill. 2.137—Boulder layer inside cutting (C. Cotter).
Ill. 2.138—Boulder layer at exterior face of wall (C. Cotter).
151
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.139—
Pathway (C.
Cotter).
roughly midway along the cutting, remains of the rubble core were evident, thus providing an
approximate line for the wall. The wall had been butted back against the large natural boulders (F706)
that formed the original ground surface prior to the construction of the terrace itself.
Above the boulder layer and behind F701 was a layer of stone chippings in a matrix of mid-brown
silty clay (F705).The layer was fairly homogeneous throughout, with localised variations occurring mainly
as a result of water-sorting or settling. The chippings consisted in the main of thin-sectioned spalls, 0.1–
0.15m long, which had been tipped in from above. This layer was up to 2.2m deep in the western half of
the excavated area but shallower elsewhere, indicating the possibility that a quantity of this material may
have been carried away when the wall (F701) collapsed. A large flagstone was uncovered within this layer
in the south-west quadrant. No finds were recovered and only an occasional charcoal fleck was observed.
Nineteenth-century material.
The secondary wall (F700) was constructed largely on the foundations of the original wall but was much
narrower. At the eastern and western ends of the excavated area it was set back c. 0.5m from the original
external face. Between these points, however, the secondary wall was deliberately curved inwards, with
the result that, at roughly the mid-point of the cutting, the external face of the later wall lies more or less
directly over the internal face of the original wall. At the western edge of the excavation the internal
face of the original wall projects 0.17–0.25m inside the later wall.
The greatest area of collapse of F701 was in the eastern half of the cutting. At the eastern edge of
the cutting, the secondary wall (F700) extended down for 2m below present ground level, with a
substantial batter on its outer face. At the western edge of the excavated area, the foundations of the
secondary wall lay 0.3m below present ground level and rested on a layer of rubble (F709) overlying the
original wall.
A mixed deposit of dark brown loose clay, with spreads of red ash and some charcoal (F704), abutted
the internal face of the secondary wall F700. The layer varied in depth from 0.5m at the western end of
the excavation to 1.3m at the eastern end, where a more substantial portion of the earlier wall had fallen.
In places this layer lay directly on top of the silty clay F705, but in other places a fragmentary layer of old
sod lay on the surface of F705. Fragments of coal, about a dozen pieces of iron slag/bloom and a sherd
152
THE EXCAVATIONS
Pathway:
F707
Topsoil:
F702
Lighthouse period:
F700
F703
F704
Ill. 2.140—
Matrix for
upper monk’s
garden.
F709
Monastic Period:
Boulder layer:
F701
F705
Rubble
F706
Bedrock:
Key:
Lighthouse and modern:
F1025
Monastic period:
F1028
of nineteenth-century pottery were recovered from F704. A clay pipe bowl decorated with a thistle
motif was found c. 0.35m below the present ground level. This type dates from the last quarter of the
nineteenth century.
Above F704 at the eastern end of the cutting was a layer of peaty sod (F703), up to 0.12m thick.This
sod lay 0.18–0.3m below present ground level and ended abruptly at the south side of the pathway. It
probably represents the ground level in the terrace prior to the spreading of soil there in 1988 (Ill. 2.140).
Modern material.
The upper layer encountered consisted of topsoil (F702), which varied from 0.18m to 0.3m in depth.
Some of this material was spread here in 1988/9, when this part of the terrace was cultivated
experimentally.
The finds5
Pottery
Five sherds of undiagnostic earthenware and creamware were recovered from the wall (F700) and from
the nineteenth-century levels (F704).
Clay pipe
An almost complete clay pipe with a thistle motif was found in the upper levels of nineteenth-century
material. The pipe bears all or part of the maker’s stamp—ACY and mould number 203. These pipes
were manufactured mainly in the period 1870–1900, but later examples of the type are also known. A
portion of an unmarked clay pipe stem was also recovered from the same material.
5
The current location of these finds is not known, thus they are not included in the finds catalogue (Appendix I).
153
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Interpretation
The original retaining wall at the south side of the upper monks’ garden was approximately 1.2–1.3m
wide and appears to have curved slightly outwards. The inner edge of the wall is faced to a depth of
0.75m below present ground level, indicating the probable original ground level. No finds were recovered
from the wall or the deposits that abutted it. The layer of stony chippings F705, which formed the bulk
of the fill over the western half of the site, appears to be contemporary with the original wall and may
be the detritus from quarrying or similar activity.This material is very similar to the foundation material
from the garden terrace on the South Peak.
At some stage in the history of the site a portion of the retaining wall collapsed, probably carrying
some of the terrace soil, backfill layer (F705) and other material with it. The greatest collapse appears to
have occurred in the eastern half of the excavated area, where the later wall extends down to 2m below
present ground level, as opposed to 0.3m at the western baulk. No hiatus was noted between the sterile
layer of rock chippings and the nineteenth-century material. It is likely that some excavation was carried
out here to facilitate the reconstruction in the nineteenth century.
The clay pipe was the only closely datable object recovered from the nineteenth-century material.
Lying as it did near the surface of this horizon, it can only be considered as a terminus ante quem for the
construction of the secondary wall.The pipe was manufactured around the last quarter of the nineteenth
century, by which time the monastery was no longer occupied by the lighthouse-builders. Some structural
work was carried out by the Board of Works at the monastery in the last quarter of the nineteenth
century, and it is possible that the retaining wall was rebuilt at this time.The inclusions in the nineteenthcentury levels and in the wall itself (coal, ash and slag) suggest, however, that the reconstruction was
undertaken during the occupation of the site by lighthouse personnel.
2.2.9 STRUCTURE AT
Edward Bourke
BASE OF EAST STEPS
(93E195)
Introduction
In 2002, following the monitoring of the repairs to the east steps, a structure was identified at the base of
the steps (Ill. 2.141) which warranted investigation and excavation. At the current base of the east steps
there was a slight platform and a small area of sloping ground with some rock-cut steps at the edge of the
cliff (Ill. 2.142). This cliff was formed when the lighthouse-builders dynamited out the landing below and
commenced building the modern road to the lighthouses. Inspection of the base of this platform showed
that the lowermost steps were much narrower than the others and turned at almost a right angle to them,
in order to avoid an area of bedrock and also to respect a structure built into the side of the platform. It was
clear that there was the north–south wall of a building with a return and a jamb in it. It appeared that some
form of long, narrow structure had originally stood on the area of sloping ground.
Excavation was intended to provide information as to the structural soundness of the base of the
plinth, and to see whether any trace of the south wall of the structure survived. It was also intended to
examine the stratigraphy on the inside of the structure in the hope of determining its purpose and, if
possible, its date.
Excavation
Excavation commenced towards the end of the season in 2002 and revealed that there were indeed some
further structural remains in addition to those visible in the initial inspection. A threshold was revealed,
as well as some layers of ash and paving.The site was covered up and left for the winter, and was excavated
during the 2003 season (Ill. 2.143).
154
THE EXCAVATIONS
Features pre-dating structure
The basal layer was a dark grey friable
soil that underlay most of the building
(F690). This layer was left in situ when
it became clear that it ran under the
orthostats at the base of the west wall
of the structure, and therefore its depth
was not ascertained.Three large samples
were taken for analysis in relation to
insects, seeds and charcoal, but no useful
data could be derived from them. A
stone cross fragment was recovered
from the surface of this layer and
directly underlay the threshold of the
structure.
Ill. 2.141—
General view
before
excavation
(Con Brogan,
DAHG).
Description of the structure
The western wall of the structure is the
only one that survives (F696). It is 5.5m
long with a jamb, 0.4m by 0.85m, at the
north end. In the area of the jamb the
wall is of drystone construction, using
smallish stones. Further south, the
construction consists of an orthostatic
base surmounted by similar drystone
masonry (Ill. 2.144).The middle part of
the wall is built directly on bedrock, but
the northern end is built upon layer
Ill. 2.142—End
of rock-cut
steps above the
landing (E.
Bourke).
155
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
ing
Y1
nd
Y2
la
at
bo
F680
wall
F683
puffin
F686 burrow
F682
F684
F690
east steps
SITE
F680- campion layer
F684- peaty brown earth
0
F683- ash & charcoal
3
metres
F686- burnt floor
F682- yellow flooring
N
)
ried
uar
(q
rock
bed
to rock-cut steps
step
location of
cross-slab
missing
step
threshold
Ill. 2.143—Plan
and section.
surviv
F684
F693
wall top
section
baulk
Y2
Y1
bedrock
F692
limit o
f exca
vation
F684
156
le line o
ing w
F689
probab
95
all F6
s
tep
st s
r ea
p fo
ram
f wall
F687
THE EXCAVATIONS
Ill. 2.144—Wall
showing
orthostatic
base (E.
Bourke).
Ill. 2.145—Wall
and entrance
jamb (E.
Bourke).
F690.The wall displays evidence of corbelling, particularly at its north-west corner close to the jamb (Ill.
2.145). A threshold stone (F694), aligned east–west and 1.19m wide, appears to give a minimum width
for the entrance of the structure. It was a long, flat stone, 1.19m by 0.38m, with four smaller stones on
its the inner side.
The structure is built onto the end of the east steps, which respect it as they rise up in front of the
structure, heading in a westerly direction for three steps and then turning south along the back wall (Ill.
2.146). In front of the structure there is an area of flat ground, at the end of which a set of rock-cut steps
would have descended to the original landing. Only five of these steps survived the dynamiting out of
the modern landing place.
157
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Occupation layers
Overlying the threshold there was a layer of ash (F687), which measured 1.32m east–west, 1.26m north–
south and up to 0.15m in thickness. Abutting this layer to the south was a layer of compacted small stones
(F688), which petered out at the break in slope. A similar layer of compacted stones (F689), measuring
0.46m by 0.32m by about 0.05m in thickness and containing a mixture of gritty dark brown natural, lay
at the same level against the paved area. Bedrock was exposed along parts of the western edge of the
interior of the structure. A layer of paving stones (F694) had been placed against the bedrock to level out
this area. The paving stones measured up to 0.55m by 0.58m and were up to 0.09m thick.
On top of all the features described above there was a layer of yellow stones in a gritty matrix (F684),
which seems to have formed a floor level across the structure and measured up to 0.18m in depth. The
western part of the layer was level, but it seems to have subsided downslope to the south. This feature
was overlain by a layer of brown gritty soil with small stones and occasional tiny lenses of charcoal (F686),
which was up to 0.14m thick and occurred at the edge of the break in slope. Above this was a layer of
ash with some charcoal, which covered the eastern end of the structure (F683); it measured 2.45m east–
west, 2.45m north–south and up to 0.06m in maximum depth. At its western end it had slumped into
an area of erosion between the west wall and the underlying floor, where its thickness increased to 0.2m.
A smaller spread of ash and charcoal (F691) at the western end of the structure measured 0.6m east–
west, 0.41m north–south and up to 0.07m in thickness; this also overlay F684.
Features post-dating the structure
Overlying these layers was a layer of
brown peaty earth with stones (F682).
This was up to 0.95m thick against the
west wall of the structure and petered
out to the east at the break in slope.
The stones were mixed up through the
layer, lying in every direction. A large
number of roof slates were found in
this layer, which seems to represent the
destruction of the structure.
Above this was a layer of campion
roots and peaty soil (F681), 0.2–0.3m
thick, which followed the slope
downwards towards the east. A second
layer of campion, 0.1–0.2m deep, had
developed on top of this (F680) and
also covered the entire area excavated.
Ill. 2.146—
Entrance jamb
and steps (E.
Bourke).
Interpretation
This structure at the base of the east
steps has posed almost as many
questions as it has answered. Its wall
had the same orthostatic base as the
plinth under the large oratory, the large
oratory itself and the inner wall of the
oratory on the South Peak. The fact
that it is built onto a set of steps is
158
THE EXCAVATIONS
Ill. 2.147—The
site after
conservation
(Con Brogan,
DAHG).
159
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
reminiscent of the oratory on the South Peak, where the steps run up the side of the building (Ill. 2.146).
The top three layers can definitely be assigned to the lighthouse period and later. There was one
piece of iron from the top of the wall of the structure (F696) and no finds in the campion layers. A large
quantity of slate was mixed through F692 and this appears to be from around the time of the dynamiting
and the destruction of the south wall of the structure. There are no finds associated with any of the
stratigraphy below this, except for the cross fragment from under the threshold, which pre-dates the
construction of this building.
From an architectural point of view, the structure is unusual in that it uses bedrock for its rear wall,
an orthostatic base with an upper level of flat-laid stones for its west wall and horizontally laid stones for
the interior of the jamb. There are traces of corbelling along the west wall. Extrapolating to make the
building roughly symmetrical, the internal dimensions would have been 2.3m by 5.1m with a wide,
north-facing door, given that the threshold stone is c. 1.2m long. It is unlikely that the building was
completely corbelled, as such a long, narrow building would be difficult to roof in that way. In fact, it is
difficult to see how it might have been roofed. Given the large quantity of slates recovered from F682, it
is possible that the lighthouse-builders reroofed the building, before accidentally destroying it.
The function of the building is also unclear. It might have been a storehouse placed close to the east
landing so that material could be moved there prior to being carried up to the monastery. It may perhaps
have been a place high enough above the landing to drag a boat out of the way of even summer storms
(Ills 2.147–8). This makes a certain amount of sense, but it would require the route from the landing to
have been of a reasonably gentle gradient and considerably less precipitous than the route above the
structure— something we can never know, as the lighthouse-builders dynamited away all traces of this
route in building their landing.
Ill. 2.148—The site during excavation, showing relationship between wall and steps (C. Brogan, DAHG).
160
Lighthouse period:
F680
F681
F682
Possibly monastic:
F683
F691
F686
F692
F684
F687
Monastic Period:
F695
F693
F690
Bedrock:
Unexcavated:
Lighthouse and modern:
F680
Possibly monastic:
F687
Monastic:
F695
THE EXCAVATIONS
161
Ill. 2.149—Matrix for structure at base of east steps.
F689
F694
Basal layer:
Key:
F688
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
2.3 THE SOUTH PEAK
2.3.1 INTRODUCTION
Alan R. Hayden
Eighteenth- and nineteenth-century maps and accounts of pilgrimages to the South Peak and a few
brief twentieth-century second-hand accounts mention the existence of man-made structures on the
218m-high South Peak of the island. It was only in the late 1980s, however, when Walter Horn, Jenny
White Marshall and Grellan D. Rourke explored and surveyed the South Peak that the true extent and
significance of the remains that survived there were revealed. They described and discussed their
discoveries in The forgotten hermitage of Skellig Michael (Horn et al. 1990),6 concluding that the dramatically
situated drystone terraces and rock-cut features constituted an early medieval hermitage and pilgrimage
site, perhaps unique in the physical challenge it provided.
Ill. 2.150—
LiDAR image of
the South Peak
looming behind
the monastery
on Skellig
Michael
(DAHG).
Ill. 2.151 (left)—
The South Peak
from the east
(A.R. Hayden).
Ill. 2.152
(right)—The
South Peak
from the southeast (A.R.
Hayden).
6
The second edition (published 2002) is also referred to below, as it contains a number of different photographs.
162
THE EXCAVATIONS
After the general completion of conservation works in the monastery on the island in 2003, the
OPW focused the attention of its conservation works on the South Peak, where monitoring had
demonstrated a visible and ongoing deterioration of the structures that survived there.
Seven seasons of archaeological excavation, survey, exploration and conservation works were
undertaken on the South Peak between 2004 and 2010.
Ill. 2.153—The South Peak from the south-west (A.R.
Hayden).
Ill. 2.154—The South Peak from the west (A.R. Hayden).
Ill. 2.155—The South Peak from the north-east (A.R.
Hayden).
Ill. 2.156—The South Peak: looking up at the east end
from the north steps (A.R. Hayden).
163
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The South Peak is a harsh, erosive and unforgiving environment for both built structures and human
beings. Its height, exposed location, steep slopes and confined spaces presented a special challenge to the
undertaking of archaeological excavations and conservation works.
The works undertaken on the South Peak vividly revealed the highly collapsed and vulnerable state
of the surviving remains and demonstrated that much original structure had unfortunately been lost over
many centuries. Indeed, given the conditions that prevailed on the Peak it was often remarkable that
anything at all survived in some parts, and the remnants spoke volumes about the flexibility of drystone
architecture and the skill and daring of the original builders. The works also dramatically illustrated and
recorded that in many places erosion and collapse were ongoing and that, if not halted, would have
resulted in increased—and, in places, total—loss of the surviving structures and thus all evidence of their
former existence.
There were two temporally distinct routes leading to the top of the South Peak. Both initially
followed the same path upwards from Christ’s Saddle before diverging at the base of the cliff beneath
the infamous Needle’s Eye. The final route is that described by Horn et al. (1990), while the primary
route was only discovered during the recent work.
??
N
218m
200m
Ill. 2.157—The
routeways on
the South Peak.
?
150m
150m
0
5
10
20
metres
150m
Common route
Primary route
Final route
Route to outer terrace
Quarrying route
Lighthouse builders’ route
The routes to the peak were naturally broken into separate terraces, traverses and climbs by the
local physical conditions, and generally each of these small sections was excavated, recorded and conserved
individually for practical reasons. For ease of description (and following the usage of Horn et al. 1990),
each section was allocated a name that reflected its character or position but which attempted to avoid
the ascription of a subjective interpretation of function. The titles given by Horn et al. (1990) to the
various terraces and areas on the Peak, which have, by this stage, become firmly entrenched in the
language of Skellig, were retained for continuity and simplicity.
164
THE EXCAVATIONS
Ill. 2.158—
LiDAR image of
the South Peak,
with the main
routes marked
(DAHG and
A.R. Hayden).
Lower part of route
Final route
Primary route
Route to outer terrace
The recent works also clearly demonstrated the degree to which the monks altered the South Peak;
in fact, there is almost nowhere on either of the routes to the top that the climber actually stands on a
natural surface.
While the works on the South Peak were generally conducted from the top down, for ease of
description this report will commence at the base of the climbs and follow them upwards. The lower
part of the climb, which is common to both routes, is described first. The upper part of the primary
route to the top of the Peak is next described, followed by the upper part of the final route.
Archaeological excavations on the South Peak
The works on the South Peak were undertaken under the overall on-site direction of Grellan Rourke
(Senior Conservation Architect, Office of Public Works [OPW]) and in consultation with Dr Ann Lynch
(Senior Archaeologist, National Monuments Service of the Department of Arts, Heritage and the
Gaeltacht [NMS, DAHG]). Edward Bourke (Archaeologist, NMS, DAHG) directed the first season of
archaeological excavation in 2004, and Alan Hayden (Archaeologist and Director, Archaeological Projects
Ltd) directed the seasons from 2005 to 2010.
Because the island is only reliably accessible from late May to late September, each season of work
was limited to a period of less than five months. Excavations were generally undertaken in three or four
two-week sessions each year. Only one archaeologist was employed on the site at any time.The weather,
the isolated nature of the site and in particular the steepness and ruggedness of the South Peak also had
a major influence on the schedule and progress of the works undertaken.The severe constraints of access
and space, and those imposed by the presence of protected wildlife and the very important issue of safety,
also rendered working conditions somewhat challenging.
All these factors, along with the fragility and vulnerability of some of the remains uncovered, meant
that excavation and conservation work had to be closely integrated at all times.
A schedule and timetable of the proposed works was drawn up prior to each season’s work.
165
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ministerial consent and appropriate licences were applied for and received before each season of work
was undertaken. Safe rope access was first established to the areas to be examined, and scaffolding
constructed where it could be and would be required. The structures to be conserved were assessed
before any works were undertaken, and a final detailed programme of work was then agreed upon before
the works were begun.
The individual features in each area were first photographically recorded before any works were
carried out. Then (where required) they were excavated, recorded, photographed and planned at a scale
of 1:20 during or after excavation. Excavated spoil was surveyed by metal-detector under licence. In
general, conservation was undertaken immediately after the completion of excavation and recording, but
some structures had to be conserved or secured during the excavations owing to their fragility and
structurally unsound state.
The conservation works were also fully recorded, described, photographed and filmed, and the
conserved features were all subsequently fully surveyed, drawn and photographed after the completion
of works. Photographs were also taken of the conserved features each year to record vegetation growth,
wind damage etc.
The works on the South Peak were filmed on a regular basis by professional film crews
commissioned by the OPW. Aerial photographs were also taken periodically before, during and after
the work was undertaken.
Reports detailing each season’s work were submitted annually to the National Monuments Service
and summaries were published in the annual Excavations Bulletin series.
Site conditions
No occupation deposits survived anywhere on the South Peak and excavation generally consisted of
little more than the removal of sea campion and other growth, followed by the removal of collapsed
stonework to reveal the surviving drystone remains. The only deposits encountered consisted of the
quarry waste used to infill the terraces and limited areas of naturally accumulated silt. In many areas
excavation simply involved the clearing of plant growth from bedrock. As a result, sieving of excavated
material was not possible, nor required.
Excavation of the individual terraces often had to be undertaken in a piecemeal fashion owing to
the restrictive terrain, the requirement to retain all spoil and stones, and the lack of space in which to
store them.
Apart from plant growth, all the spoil from the excavations was sorted and retained and reused in
the conservation works. The well-drained and acidic nature of the site and the lack of any occupation
deposits meant that few finds were uncovered.Those uncovered consisted of roughly carved cross-shaped
slabs and a few water-rolled pebbles.
[Left] Ill.
2.159—
Excavation on
the oratory
terrace (A.R.
Hayden).
[Right] Ill.
2.160—
Conservation
works on the
north-west
passage (A.R.
Hayden).
166
THE EXCAVATIONS
Progress
Excavation of the lower traverse, the two platforms above the Needle’s Eye, the small enclosure above
the Needle’s Eye, the climb from the Needle’s Eye to the ‘garden’ terrace and a test-trench across the
‘garden’ terrace was undertaken under the supervision of Edward Bourke in 2004.
The works in 2005–7 concentrated on the upper terraces and structures above the Needle’s Eye—
the ‘garden’ terrace, the oratory terrace, the upper traverse and the outer terrace. From 2008 onwards the
structures and features of the newly discovered primary route to the South Peak and the beginning of
the climb to the South Peak were explored, surveyed and excavated.
The excavation and conservation works were completed in 2010, but there are several areas and
features on the Peak that will be recorded and surveyed in the coming years.
Schedule of works undertaken on the South Peak
2004
Excavation: the lower traverse, the upper and lower platforms, the small enclosure above the Needle’s Eye
and a test-trench on the ‘garden’ terrace.
Conservation: the lower traverse, the upper and lower platforms and the small enclosure above the Needle’s
Eye.
2005
Excavation: the ‘garden’ terrace and part of the oratory terrace.
Conservation: the ‘garden’ terrace and part of the oratory terrace.
2006
Excavation: completion of the oratory terrace (including the interior of the leacht), the upper traverse,
sondage in the outer terrace, the ledges beneath the oratory terrace, and the possible mini-terraces
in the upper gully.
Conservation: completion of the oratory terrace, the upper traverse and the possible mini-terraces in the
upper gully.
Post-conservation survey: the ‘garden’ terrace.
[Left] Ill. 2.161
—Aerial
photography in
progress (A.R.
Hayden).
[Right] Ill.
2.162—Filming
on the northwest passage
(A.R. Hayden).
2007
167
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Excavation: the outer terrace, ledges below the outer terrace, the southern traverses of the north-west
passage, and the platform and steps at the base of the cliff beneath the Needle’s Eye.
Conservation: the outer terrace, started on the southern traverse of the north-west passage, and the platform
and steps at the base of the cliff beneath the Needle’s Eye.
2008
Excavation: the northern traverse and the stairs of the north-west passage, the northern cliff ledges, the
lower gully, and the second narrow ledge from the blind corner to the cliffs below the Needle’s
Eye.
Conservation: completion of the southern traverses of the north-west passage, the northern traverse and
stairs of the north-west passage, and the second narrow ledge from the blind corner to the cliff
below the Needle’s Eye; repairs to the outer terrace, the upper traverse, the oratory terrace, the
platforms above the Needle’s Eye and the north steps.
Post-conservation survey: the outer terrace, the upper traverse, the oratory terrace, and the upper and lower
platforms above the Needle’s Eye.
2009
Survey: parts of the route from the start of the climb to the blind corner, the Needle’s Eye, the steps in
the gully, photographic survey of all steps and a three-day EDM survey.
Post-conservation survey: the north-west passage, and the path from the blind corner to the cliffs beneath
the Needle’s Eye.
2010
Excavation: the ledge with fallen stones and the broad ledge.
Conservation: the ledge with fallen stones and the broad ledge.
Post-conservation survey: the ledge with fallen stones and the broad ledge.
Ill. 2.163—The
South Peak
crew 2008 in
front of the
UNESCO
World Heritage
Site plaque on
Skellig Michael.
Left to right:
Micheál
O’Sullivan, Tom
Kerrisk, Mike
Connors, Colin
McGorlick, Alan
Hayden, Grellan
Rourke, Pat
O’Shea, and
Johnny Lyne
(A.R. Hayden).
168
THE EXCAVATIONS
A note on the plans and photographs
The plans included in this report are but a preliminary working of the field plans and surveys undertaken
on the site. The plans of the lower part of the route and the primary route have been colour-shaded, as
their intricacy meant that man-made features might not be readily recognisable amongst all the
background detail. The plans of the uppermost terraces on the Peak have not yet been so treated. The
detail in the shaded plans remains to be checked on site.
The detailed natural setting—rock ledges, cliffs, slopes, gullies etc.—forms an important part of the
site and is vital to its interpretation. The legible illustration of the complex nature of these features has
required a lot of thought and work.They have to be shown in enough detail to be understandable while
not obscuring the man-made features, which were cut into and built on them. Graduated shading (with
the darkest shade indicating the base of a slope and the darkness of the shading representing the steepness
of the slope) has been utilised and combined with lines indicating the edges and shape of the slope to
attempt to convey the natural features.The different character (direction of grain, smoothness, slope etc.)
of the bedrock bases of the ledges had also to be illustrated. This was generally done using lines running
with the grain of the rock. Little of the ‘flat’ bedrock at the base of the terraces was in fact flat and level.
Illustrating these slopes proved difficult if they were not to be confused with steeper slopes and obscure
the illustration of the character of the rock. These slopes have generally been shown flat, as it was felt
that the accompanying photographs would adequately illustrate the slope of individual sections. Areas
where vegetation survived and was not removed during the excavations are shown with a green shade,
and areas where soil and stones survived and were not removed are shaded a light yellow-brown.
On top of all this, the man-made structures had also to be shown. Again a graduated shade, but
without parallel lines, is used to show stones lying at an angle. A dark grey-brown is used to show vertical
stones, and a light grey shade indicates flat stones. The sloping faces of walls are illustrated with a darker
grey shade. The top and bottom edges of walls and structures and rock-cut features are outlined in red,
and the possible/probable lines of now-vanished structures are illustrated with dashed red lines.
The excavation field plans were drawn by A.R. Hayden at a scale of 1:20. Plane-table surveys of the
conserved sites were drawn and surveyed by G.D. Rourke and A.R. Hayden at a scale of 1:50. A threeday EDM survey of parts of the lower part of the route, the north-west passage and the area above the
Needle’s Eye by Seán Obida (Malachy, Walsh and Partners) represents the only such survey undertaken
in the seven years of work on the Peak. The final plans and the shading scheme used were devised and
drawn by C. McHale (Archaeological Projects Ltd), with some amendments by A.R. Hayden.
Photography, both on the ground and aerial, was also used extensively to record the works and only
a small part of the archive has been used for illustration in this report. In many areas it is difficult to
adequately convey the site and its features without photographs.The location of steps and features marked
on some of the photographs remain to be fully checked on site.
Ground photography had its limitations, as the cramped spaces and partly obscured views sometimes
restricted the ability to show what was intended. Aerial photographs taken by Con Brogan (NMS, DAHG)
from various helicopters proved very useful.They illustrate much that was not accessible or visible, and also
allow more panoramic and more contextually inclusive views than are possible from the ground.
It is also hoped in the future to use the LiDAR survey of the island to illustrate long vertical sections
through the Peak. This will perhaps best illustrate the verticality of the site, which is impossible to
adequately convey in plans. Although a work in progress, the LiDAR survey unfortunately does not
appear to include sufficient detail to replace ground survey, and in the steeper locations it shows little or
no detail of the structures and features that survive; indeed, in places it does not show substantial structures
at all. It appears to have worked best on the flatter and more gently sloping terrain of the island, most of
which lies elsewhere than on the South Peak! It is, however, very useful for generating more general
views of areas, as the images can be turned and viewed from almost any angle.
169
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
2.3.2 THE
Ill. 2.164—The
south face of
the South Peak,
showing the
main geological
fractures: red =
bedding planes;
blue = cleavage
planes; yellow =
fracture lines
(A.R. Hayden).
INFLUENCE OF THE GEOLOGY OF THE
SOUTH PEAK
The underlying geology7 had a fundamental influence on the structures built on the South Peak.
The Peak was formed of sandstone laid in roughly horizontal bedding planes, which were deformed
by folding and now lie at slight to moderate angles.Where these bedding planes are exposed, they erode
to form rough ledges that slope usually relatively gently in one or two directions.The routes up the Peak
generally follow these ledges and it was on these ledges that the drystone terraces were built.
The vertical cleavage planes in the bedrock run generally on north-east/south-west lines through
the South Peak. They split the rock into thin vertical sheets, rather like a deck of cards set on edge. They
also had a strong influence on the
layout, construction and survival of the
terraces.Where the cleavage planes run
parallel to the cliff face, the cliff face is
vertical, smooth and straight (as it
follows the cleavage planes), and hence
any walls built on these cliffs (generally
on the south and north sides of the
Peak) were straight but had a poor
resistance to sliding and collapse.
Where the cleavage planes run
transversely to the cliff, the cliff face is
deeply fissured and uneven as a result
of erosion along the exposed ends of
the cleavage lines. The walls running
along these cliff faces (generally at the
west side of the Peak) tend to be
curved or sinuous. Because the walls
cross the eroded cleavage lines, their
bases rise and fall and consequently
have a greater ‘grip’ on the rock,
rendering them more stable.
Where the cleavage planes run parallel to the cliffs, they also allowed large blocks of rock both to
detach naturally and to be removed by quarrying. Large stones detach and fall naturally from the cliffs
on the north and south sides of the Peak, which run parallel to the cleavage planes. These falling stones
naturally gather on the roughly horizontal bedding plane ledges.The accumulated stone on these ledges,
while it provided a source of building material, had to be cleared to make the ledges accessible and usable.
There are a number of ledges that were not built upon but that lack naturally fallen stone, which the
monks appear to have stripped for building work elsewhere.
To complicate matters further, a series of vertical faults run through the South Peak, generally on
north-east/south-west and north-west/south-east lines, dividing the rock into large blocks, which are
often twisted at an angle or raised or lowered to a different height than adjacent ones. The fault lines are
often eroded out to form deep vertical gullies, which provided vertical access routes.The faults also break
the horizontal flat bedding plane ledges into blocks divided by deep gullies. Hence the fault lines often
determine the size or length of terraces.
7
A detailed report on the geology of the South Peak is being prepared by the Project Geologist, Michael O’Sullivan. In this report I have liberally used information provided by Michael while we were working on the South Peak.
170
THE EXCAVATIONS
The flat ledges
eroded out
along the
bedrock
bedding planes
naturally gather
fallen stone.
Left: Ill. 2.165—
A ledge on the
cliffs below the
north side of
the monastery.
Right: Ill. 2.166
—A ledge on
the west side of
the rock tor
west of the
scree slope
(A.R. Hayden).
2.3.3 THE
LOWER PART OF THE ROUTE
Introduction
The lower part of the route to the top of the South Peak, which was common to both the primary and
final routes, was only briefly examined and described by Horn et al. (1990, 23–9) and had not been
previously surveyed, explored in detail or excavated. As the conservation and excavation works on the
South Peak were undertaken generally from the top down, it was only after the completion of the works
on the upper terraces of the primary and final routes in 2008 that work was undertaken on the lower
part of the route.
Much of the lower part of the route consisted of rock-cut features, which were fully exposed and
only required cleaning, recording and surveying. Perhaps, because of the large amount of human traffic
on this part of the route and its vulnerability to stone falling from the cliff above, all the areas where
masonry had originally been constructed were highly eroded and only the last remnants of the original
structures survived. Little of any of the surviving masonry structures was visible or recognisable before
excavation.
The lower part of the route as far as the blind corner also sees the most human traffic on the South
Peak. The narrowness of the path and the high and vertical cliff at the blind corner deter many people
from proceeding further up the climb.The surviving remains up as far as the blind corner were therefore
highly vulnerable and, as they represented but the last vestiges of formerly substantial structures, there
was a real possibility that they could have been lost and all evidence of their former existence could have
disappeared.
The scree slope
The route to the South Peak begins at Christ’s Saddle (elevation 130m OD), the valley between the
lower (elevation 185m OD) eastern peak of the island (where the monastery lies) and the 218m-high
South Peak.
The route initially extends up the steep scree slope that lies at the western side of Christ’s Saddle.
171
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.167—The
ledges on the
lower part of
the route.
I
Ill. 2.168—Composite
photograph of the south
face of the South Peak. The
route follows the
prominent bedrock
bedding plane ledge that
curves upwards from the
lower right (A.R. Hayden).
172
THE EXCAVATIONS
Ill. 2.169—
Aerial view of
the lower part
of the route;
the top of the
scree slope is at
the lower right
(Con Brogan,
DAHG).
Ill. 2.170—The ledges at the beginning of the lower part of the route, viewed from the south (A.R. Hayden).
173
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.171—
LiDAR image of
the lower part
of the route
(DAHG).
Ill. 2.172—The
highly eroded
paths on the
scree slope
when it was
open to public
access (A.R.
Hayden).
174
THE EXCAVATIONS
This slope is covered by thick campion growth but several well-worn modern tracks run up it.The scree
slope also slopes steeply down to the north-east. A high and easily climbed rock tor delimits its west side.
A substantial number of large rocks that have tumbled off this high tor lie on the scree slope.
The lighthouse men cleared the lower part of the scree slope to create a small field, which is partly
defined by a drystone wall composed of large stones robbed from the top end of the north steps, a short
section of the south steps and probably from the masonry stairway that led up the scree slope to the start
of the climb up the South Peak (see below).
The steepness of the scree slope and the loose nature of the stones and soil with which it is covered
render it very vulnerable to erosion.There was a noticeably high degree of erosion on the slope when it
was open to the public. Erosion of some of the pathways has continued, however, even though it has
been closed to public access and the works crew were supposed to be the only human traffic. If the scree
slope is to be reopened to the public, recording, surveying and possibly excavation and conservation
works will probably be required, as early features have been noted on it.
No archaeological excavations, conservation work or surveys have been undertaken on the scree slope.
Cross
There is a small, cross-shaped slab set erect at the
base of the west side of the scree slope. It is very
similar in form to a smaller cross that was found
fallen from, and lying on a ledge below, the
oratory terrace. There was formerly a second
cross here but it is no longer extant (G. Rourke,
pers. comm.).The surviving cross lies in the area
cleared by the lighthouse men to create a field
and so may not be in its original position. A cross
is shown here on the first-edition OS map of
1841, however, and so one or both of these
crosses could mark the start of the route to the
South Peak. There appear to be crosses and
leachta along the route up the Peak, and so it
would be expected that one would have been set
up to mark the start of the route.
Ill. 2.173—
Cross-shaped
slab in the
lighthouse field,
possibly
marking the
start of the
route to the
South Peak
(A.R. Hayden).
The ‘cave’
Two small chambers were created by partly
blocking off the areas beneath large fallen rocks
on a ledge to the west of the Saddle. A clear
drystone-walled terrace leads to them. There are two cross-shaped slabs in and beside these chambers.
No excavation or survey work was undertaken in this area.
The stairway
It had long been suspected that there must have been some form of formal pathway or stairway leading
up the scree slope to the start of the climb to the South Peak, as the slope is slippery and awkward to
cross and all the rest of the route is provided with terracing and stairways. No definite features were,
however, previously noted on the slope.
The warm and dry weather of the spring and early summer of 2010 somewhat restricted campion
growth on the slope and resulted in an early die-back of the plants. As a result, several man-made features
175
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
became clearly evident there for the first time. An opportunity arose to examine the scree slope in some
detail, and several features were discovered that clearly prove the former existence of a masonry stairway
leading up the slope.
Ill. 2.174—The scree slope,
showing the location of the
visible early features (A.R.
Hayden).
Clearance
Large rocks naturally detach from the high rock tor on the west side of the scree slope and tumble down
onto it. It is noticeable that the mass of fallen rocks abruptly ends in unnatural straight lines. In places
along the edge of the fallen rocks stones are piled on top of each other, two or three in height. To the
east of the edge of the fallen stones there is a flat area about 3m wide extending up the slope. This area
176
THE EXCAVATIONS
is free of large fallen stones and interrupts the otherwise uniform west–east slope of the area. East of the
flat area fallen stones again are present. It is therefore probable that the flat area was cleared of naturally
fallen stone.
Ill. 2.175— Looking down the scree slope.
Stone fallen from the high rock tor lies to
the right of the ranging rod; the area
cleared of stone lies to its left (A.R.
Hayden).
Stair risers/cross-walls
Within the narrow, flat area, six definite
lines of drystone masonry were
discovered extending transversely
across the slope. There are up to half a
dozen other possible lines of masonry
also visible, but too little of each is
evident to confirm that they are walls.
All these lines of masonry are
potentially either remnants of drystone
risers of the stairway or the remains of
structural walling within the stairway.
Ill. 2.176—One
of the visible
lengths of
drystone walling
on the scree
slope (A.R.
Hayden).
177
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Stair treads
On the upper half of the flat area there are
numerous thin, flat and subrectangular
stone slabs, which could have been the
treads of the stairway. A few may still be
roughly in situ but more have slipped out
of place. Several thin flat slabs also survive
at the top of the scree slope; it is probable
that these are not fallen rocks and they
may be further remnants of treads of the
stairway.
Ill. 2.177—Slipped stair treads on the upper
half of the scree slope (A.R. Hayden).
The stairs in the lighthouse field
No stone treads appear to survive on the lower half of the scree slope, within the lighthouse field. The
lighthouse men clearly removed the stone treads of the three stairways on the Saddle to build their field
walls. Nevertheless, the survival of at least one drystone riser/wall of the stairway to the South Peak
within the area of the field and the form of the surviving field walls suggest that the lighthouse men
removed only the larger stones from the stairways.This means that some of the original risers/cross-walls
of both the steps to the South Peak and the top end of the north steps could survive within the area of
the field. The field also lies at the base of the scree slope and hence should have a thicker soil cover
because of the natural downslope migration of soil and stones owing to erosion, which could have covered
and protected the early remains.
The top of the scree slope
A 2m-high, east/west-aligned, rising bedrock cliff delimits the top of the scree slope. At the base of the
cliff there are two large, flat stones, which could be the last remnants of masonry steps facilitating access
to the climb up the low cliff to the ledge with fallen stones above.
Possible leacht on rock tor to west of the scree slope
A group of loose quartz blocks and a small loose stone with a possible carved notch on one of its sides
were noted on a flat ledge near the base of the tor on the west side of the scree slope, about threequarters of the way up the slope. The small flat ledge measures about 1.5m across and is accessible from
the south up a narrow gully in which there are several probably natural ledges, which make convenient
steps.There is a lot of stone at the base of this gully and it is possible that some may be collapsed drystone
masonry or steps.
178
THE EXCAVATIONS
[Left] Ill. 2.178—Possible masonry (left arrow) at base of climb to possible leacht (right arrow) on the rock tor at the
west side of the scree slope (A.R. Hayden).
[Right] Ill. 2.179—The possible masonry or steps at the base of the climb to the possible leacht (A.R. Hayden).
The presence of the quartz blocks on
the platform suggests that it may have been
some form of leacht or marker, now hidden
beneath campion cover. This area has not
been excavated. The ledge lies in a
prominent position on the side of the tor
on the west side of the scree slope, and any
structure that may have stood on it would
have been clearly visible when climbing
the scree slope or when looking down
over it from the beginning of the climb to
the South Peak. Other small leachta,
platforms, crosses and gatherings of quartz
blocks were noted elsewhere along the
route to the Peak. They may have acted as
stations or markers along the route.
Stone was also apparently quarried
from the rock tor for use on the South
Peak (Michael O’Sullivan, pers. comm.),
and it is easily climbed.
Ill. 2.180—The
climb up the
cleft to the
possible leacht
(A.R. Hayden).
The ledge with fallen stones
Introduction
Above the low cliff at the top of the scree slope there is a long, east/west-aligned broad ledge (the ledge
with fallen stones), which follows a bedrock bedding plane. The ledge slopes quite steeply down from
west to east and also from north to south. A north/south-aligned bedrock fault divides the western third
of the ledge from its eastern two thirds. Steep rising cliffs mark the inner (north) side of the ledge.
Before any works were undertaken here, bedrock was exposed over most of the western third and
the southern 2m of the ledge. A thin, campion-anchored soil cover existed on the inner side of the
179
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
eastern two thirds of the ledge. Several stones fallen from the cliffs to the north lay on the ledge. A very
large flat rock lay near the eastern end of the ledge, several smaller ones lay at its centre and two large
stones lay slightly overlapping at its west end.
On the western third of the ledge the bedrock floor of its inner (northern) 1–1.5m was extremely
rough and difficult to walk on. The bedrock on the outer side of the ledge was smooth and weathered,
however. There was a clear, straight, east–west linear boundary between the two areas. The difficulty of
crossing the rougher bedrock in wet weather (when it is one of the most awkward and slippery parts of
the climb) and the differential erosion suggested that formerly there could have been terracing covering
the rougher area of bedrock.
A number of small rock-cut features were also evident on the exposed bedrock at the centre of the
terrace. In order to record these features, the sparse vegetation growing in cracks in the bedrock on the
western half of the ledge was cleaned away and the bedrock brushed down in 2009. The soil-covered
eastern half of the ledge was excavated in 2010 to determine whether any remains of a putative terrace
survived there.
Early features—quarry and tripod holes
In the centre of the 2m-high cliff at the top of the scree slope that marks the southern side of the ledge
with fallen stones there is a small, subrectangular rock quarry, measuring 1.8–2m east–west by 2.4–2.6m
north–south. Its west side is straight, as it is defined by the fault line that divides the ledge with fallen
stones, while its east side is ragged. The north side is also straight, as the quarrymen utilised the straight,
east/west-aligned bedrock cleavage planes to remove large, straight-sided slabs.The quarry pre-dates the
layout of the lower part of the route to the South Peak, as the first nine steps of the climb run through
it (see below).
The quarrying appears to be contemporary with an interesting and possibly unique group of
features. A south-east/north-west line of four small bowl-shaped hollows (A–D) was carved into the
sloping bedrock of the ledge east of the quarry. The hollows measured 180–220mm across and were up
to 120mm deep. There could originally have been a fifth (E, the eroded notch at the north-east corner
of the quarry) in this line.
Ill. 2.181—The tripod holes
(yellow rings) surrounding
the small quarry on the
ledge with fallen stones
(A.R. Hayden).
180
THE EXCAVATIONS
Another four rock-cut hollows (F, G, K and
L), measuring 160–200mm across, survive to the
north of the quarry. One of these (G) was partly
cut away when the final slab was removed from the
quarry. West of the quarry there are three further
similar hollows (H–J) on a north-east/south-west
line (lying at right angles to the first-mentioned
series). One of this line lies almost on the cliff edge
(which also appears to have been substantially
quarried back over much of its length) and is much
larger than the others. It measures 400mm across
and 200mm in depth.
The holes all appear too small to be steps; they
are certainly not typical of steps elsewhere on the
South Peak, as they are bowl-shaped rather than flatbased and are too small to accommodate a foot.
Michael O’Sullivan (Project Geologist) interpreted
these features as holes to secure a tripod that would
have been used to raise large stones up the low cliff
at the start of the climb. They are perhaps a unique
survival.
Ill. 2.182—The
eastern line (A–
E) of tripod
holes (A.R.
Hayden).
[Left] Ill.
2.183—Three
(F, G and K) of
the northern
tripod holes; G
was partly cut
away by the
quarry at right
(A.R. Hayden).
[Right] Ill.
2.184—The
western line
(H–J) of tripod
holes (A.R.
Hayden).
It is probable that the tripod holes and the quarrying were contemporary, the holes being used to
secure a tripod to lift the quarried stone. This would explain why the west and north legs of the tripod
seem to have been periodically moved further to the north, allowing the tripod to be repositioned over
the stones as the quarrying progressed inwards (northwards). One of the holes at the north side of the
quarry has also clearly been partly removed by the quarrying.
Michael O’Sullivan noted that some of the stone used on the South Peak was quarried from the
181
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.185—The
cliff at the top
of the scree
slope leading
up to the ledge
with fallen
stones, showing
the rock-cut
access steps,
quarry and
cross base,
before
excavation
(A.R. Hayden).
[Left] Ill.
2.186—The
base of the
lowest rock-cut
steps (A.R.
Hayden).
[Right] Ill.
2.187—The
upper steps
leading up the
low cliff at the
top of the
scree slope to
the ledge with
fallen stones
(A.R. Hayden).
rock tor at the west side of the scree slope, and stone quarried there for use further up the Peak could
also have been hauled up the low cliff using the tripod.
Large stones later either fell into or were deposited in the quarry, partly filling its base and east side.
The impact of the falling stones, or the removal of slabs during quarrying, broke and twisted southwards
the protruding bedrock spines at the east side of the quarry. The uppermost of the stones filling the
quarry is a very large, thin slab that appears to have two notches cut from its southern side.These do not
182
THE EXCAVATIONS
appear to be natural and are not of use as handholds for the climb.While it is possible that this stone may
have been the last removed from the quarry but not utilised (its length matches the width of the quarry),
the notches do not correspond to any features that survive around the edges of the quarry.
The climb to the ledge with fallen stones
In the low rock face below the quarry there are two clear rock-cut steps (1 and 2) in the 0.8m-high
western end of the rising cliff. They lead upwards to the east to a short and narrow flat ledge, where
there is a third large and shallow rock-cut step (3).The fourth and fifth steps should also lie on this ledge
but its surface is heavily eroded and no definite worked areas survive.The route continues up the eastern
side of the quarry. The bare stump of a sixth step (for the left foot) and just the innermost end of the
seventh (for the right foot) survive close to the edge of the rising cliff slightly higher up. The eighth and
ninth steps survive to a greater extent and lead up to the outer edge of the ledge with fallen stones.
Cross base
A drystone structure was built in the disused quarry on top of the fallen stones immediately to the left
of the rock-cut steps described above. Three short lengths of walling were constructed, each set back
[Above left] Ill. 2.188—The cross setting in the disused quarry viewed
from the west (A.R. Hayden).
[Above right] Ill. 2.189— The cross setting in the disused quarry
viewed from the south-east (A.R. Hayden).
[Right] Ill. 2.190—Drystone walling under the south side of the cross
setting in the disused quarry (A.R. Hayden).
183
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
from the underlying one, between the western edge of the quarry and the fallen stone choking its eastern
side. They were all composed of small stones and all survived poorly.
At the top of the quarry several large stones were laid down flat. The topmost and largest of these
was deliberately and deeply notched to create a socket, into which a number of flat slabs were set upright.
These may have secured the base of an upright stone cross or pillar, which would have marked the side
of the route to the South Peak.
The ledge with fallen stones (centre and east)
It was previously
unclear where the
route initially ran
across this sloping
ledge, as there are no
traces of any rock-cut
steps
here.
The
topography suggested,
however, that the
access
route
was
unlikely
to
have
extended south from
the top of the rock-cut
steps owing to the
steepness
of
the
bedrock slope here. It
appeared most likely
that it would have led
first eastwards along the outer edge of the
ledge before turning south for a short
distance and then turning again to the west
to run along the inner side of the ledge.
The impact of the huge fallen stone
now lying on the central part of the eastern
end of the ledge could have removed much
of any possible surviving terracing but also
prevented much of the area of the ledge from
being excavated.
A broad and deep gully extended
eastwards along the outer edge of the ledge.
It is possible that this gully could have been
used to anchor the outer wall of the putative
terrace.There were a number of stones within
the gully but too few survived to determine
whether they had been deliberately laid. One
much larger stone bridged the top of the
gully towards the eastern end of the excavated
area. It is possible that it could have stood near
the south-east corner of the terracing, but it
Ill. 2.191—The
ledge with
fallen stones
after
excavation,
from the east
(A.R. Hayden).
Ill. 2.192—The
possible eastern
end of the
masonry
terrace at the
east end of the
ledge with
fallen stones
(A.R. Hayden).
184
THE EXCAVATIONS
Original edges of terracing?
cliff edge
Rock-cut steps
185
Ill. 2.193—Plan of features uncovered on the ledge with fallen stones.
cliff edge
0
metres
Quarry
5
Cross
setting
Slipped step
risers
Tripod holes
Cross
N
Possible leacht
may also have slipped into
this present position.
East of the large fallen
stone a number of stones
survived on the sloping
bedrock across the full width
of the ledge. These were in
the main relatively large and
flat and appeared to have slid
downslope into their present
positions. There were also
several large flat stones and
smaller stones trapped
beneath the large fallen
stone. One group of smaller
stones at the centre of the
end of the ledge appeared to
be intact; the setting had a
relatively straight eastern side
and it is possible that this
may be the last remnant of
the eastern end of the
terracing.
A small quantity of
loose stone lay on the
northern (inner) side of the
centre and eastern end of the
ledge. Beneath it were several
larger stones, which appeared
to have been deliberately and
neatly laid flat in natural
gullies in the bedrock. The
surviving structure consists
of a 0.7m-wide, east/westaligned band of largely
unstructured stones that
lacked a clear defining wall.
These stones did not
naturally accumulate in this
location (M. O’Sullivan, pers.
comm.), however, and
represent the last and very
disturbed remains of a
possible terrace. A wide natural break in the east/west-aligned bedrock ridges could have marked the
eastern end of the terrace, as it would have formed a natural setting in which to secure the eastern end
wall. It lined up with the straight edge of the masonry noted towards the eastern end of the ledge.
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.194—
Remains of the
drystone
terrace on the
centre of the
ledge with
fallen stones,
looking east
(A.R. Hayden).
The southern edge of the stone spread appears to follow a shallow cleft in the bedrock and it is
probable that it was used to anchor the original retaining wall, none of which survived. This cleft lines
up with the line between the differently eroded sections of bedrock further west, which also seemed to
mark the outer edge of the putative terrace (see below).The western end of the stone spread was defined
by the fault line where the ledge suddenly narrows and rises.
Two larger stones, which appeared to have originally lain with their east sides in a line, remained
amongst the stonework. They could have been the base of a stair riser or cross-wall. One of the stones
had slipped slightly out of place.
Ill. 2.195—Detail of slipped
step riser, visible at the left of
the preceding photograph
(A.R. Hayden).
186
THE EXCAVATIONS
Cross
A small stone, cross-shaped slab was also found amongst the loose rubble at the centre of the terrace.The
cross appears to have been too small to stand in the stone setting built in the disused quarry on the
southern side of the ledge. Examination of the rock ledges higher up and north of the cross failed to
reveal any structures into which this cross could have been set. It could have been associated with a
possible leacht (see below). It is, however, not impossible that it could have fallen from the oratory terrace
and tumbled down the cliff to land where it was uncovered; a cross fallen from the oratory terrace was
found on the ledges below the oratory terrace, which lie immediately above this point (see below).
Ill. 2.196—The cross-shaped slab
found on the ledge with fallen
stones (A.R. Hayden).
Possible leacht
A group of flat stones survived on a small ledge in the rising bedrock cliff about 1m above the centre of
the ledge. Two fist-sized pieces of quartz lay on the stones and against the cliff face. These were the only
quartz stones found anywhere on this ledge. The bedrock here and in the cliffs at the inner side of the
ledge lacks any quartz veining, which suggests that the two pieces of quartz were deliberately deposited
by a human hand. It is possible, therefore, that the stonework beneath the quartz could be some sort of
small leacht-like structure.The stonework was examined and excavated but little survived and no definite
structure was detected.
Ill. 2.197—The possible leacht at
the back of the ledge with fallen
stones; note the quartz (A.R.
Hayden).
187
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The ledge with fallen stones (west)
Immediately west of the fault line, which marks the western side of the early quarry, the ledge narrows
on its northern (inner) side and begins to rise more steeply.This is the area where the differential erosion
of the bedrock may evidence the previous existence of now-vanished masonry. The innermost metre or
so of the ledge here has a very rough surface, contrasting with the smoothly eroded rock surface of the
remainder.The junction between the two areas, if extended eastwards, lines up perfectly with the groove
in the bedrock, which possibly held the outer edge of the stonework uncovered at the centre of the
ledge.There is a clear worked and artificially
flattened area (no. 10) measuring 0.8m east–
west by 0.5m north–south in the rougher
area. It could have been cut to help secure
the base of stonework. It certainly does not
Ill. 2.198—The
work as a step in its own right.
west end of the
Possible masonry visibly survived
ledge with
fallen stones,
beneath and between the two large fallen
showing the
stones that lie near the western end of the
large fallen
ledge. It was possible to excavate only a
rocks and
differential
small area, as the masonry mainly lay under
erosion of the
the stones, which were too massive to move.
bedrock (11);
A thin layer of small stone fragments and
10 marks a
flattened area,
clay covered the masonry.What appeared to
and L is the
be regularly laid masonry lay beneath the
northernmost
larger, western and uppermost of the fallen
of the tripod
holes (A.R.
stones. It appeared to have a roughly straight
Hayden).
southern edge that could have been its
original face, but it was not possible to see
this with any clarity. When the larger stone
fell, it landed partly on the smaller eastern
fallen stone and partly on bedrock about 2m
[Left] Ill. 2.199—The possible drystone masonry surviving beneath and between the two large fallen stones at the western end of the
ledge with fallen stones (A.R. Hayden).
[Right] Ill. 200—The short broad ledge. The westernmost of the two large fallen stones at the western end of the ledge with fallen stones
is visible on the right on the east side of the fault line that marks the start of the short broad ledge (A.R. Hayden).
188
THE EXCAVATIONS
to its west, and so it may not have directly hit this masonry. The remains of the underlying structure
could therefore be relatively intact.
The smaller fallen stone, however, had clearly landed on the possible stonework, as there was a clear
impact crater around it, marked by crushed stones and disturbed stones lying nearly vertically. Such impact
craters were also noticed around stones fallen on terracing elsewhere on the South Peak (see below).
Immediately west of the fallen stones, the ledge ends on a straight, roughly north/south-aligned
fault line.
The short broad ledge
After the fault line, the ledge
broadens on its north-western (inner)
side, but the cliff edge on its outer
side also abruptly steps back
northward, so the connection
between the two ledges is narrow.
Dynamited Area
The slope of the bedrock plane also
reduces, so the surface of the short
Flattened
broad ledge only slightly rises to the
Cliff cut
west. No excavation work was
Area
back
required on this ledge as all the
surviving features here were rock7
cut.
6
There are a number of clear,
5?
shallow, rock-cut steps leading across
4
this ledge. The first (1) is a small
3
triangle on the southern edge of the
narrow gully that marks the fault line.
The step originally would have been
2
wider; the erosion of the side of the
cliff has removed much of it. The
second and third steps (2 and 3) are
much larger but very shallow. Only
1
the northern side of the fourth (4)
survives on the side of a natural gully,
where it is cut into the side of a low
spur of bedrock. No clear worked area survives where the fifth step would have lain.The sixth and seventh
steps (6 and 7) lie slightly higher up on a narrow ledge.
The western end of the ledge turns slightly to the north-west, where there is a worked and flattened
area that would have been the next step. The cliff at the corner here was also clearly cut back to widen
the passage. A wide fault marks the western end of this ledge.
The ragged ledge
After the wide fault line the bedrock is fractured by a number of fissures eroded along the cleavage planes
of the bedrock, and several deep faults cross the ledge on north–south lines. This has resulted in the
division of this 7m-long area into uneven narrow ledges of irregular width and uneven slopes. The
western 2m of the ledge, which is less fissured, turns to the north-west around the inner bedrock cliff. A
number of rock-cut steps survive on this ledge.
189
Ill. 2.201—
Features on the
short broad
ledge, looking
west (A.R.
Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.202—The
ragged ledge.
The wide fault
separating it
from the short
broad ledge is
at the right. A
deep fault also
runs across the
centre of the
ragged ledge
(centre of
photograph).
The lower end
of the broad
ledge lies at the
corner of the
cliff to left of
centre (A.R.
Hayden).
No excavation work was required here, as all the surviving features are rock-cut.
On the east side of the fault line that marks the end of the short broad ledge there is a roughly
horizontal drilled blasting hole, 40mm in diameter and 380mm long, low down on the west-facing rock
face on the north side of the ledge. The north-western side of the hole has largely been blasted away.
The lighthouse-builders created the hole. There are at least two sizes of drilled holes (40mm and 65mm
in diameter) visible along the road to the lighthouses, and the smaller of these corresponds to the size of
the hole on the South Peak. The route where the drilled hole lies probably originally ran straight
westwards from the end of the short broad ledge toward the cliff edge and skirted around a formerly
protruding part of the cliff (which was blasted away), before turning at a right angle to the north to join
the first surviving section of steps on the ragged ledge from the south. The first three steps (1–3) on the
ragged ledge run from south to north.
The junction between the ragged and broad ledges is marked by two closely spaced fault lines, the
easternmost of which is particularly wide, and so it is not surprising that a section of rock close to the
cliff edge could have fallen away, thereby severing access. The blasting removed part of the cliff to reestablish access to the next section of the route.The fact that the lighthouse-builders went to the trouble
of reopening this route suggests that it was of importance to them. It is possible that the lower part of
the route combined with the southern part of the north-west passage provided an access route and short
cut between the monastery (where the lighthouse-builders lived) and the upper lighthouse (see below).
On the west side of the wide fault line there is a narrow spine low down on the left that has clearly
been flattened and forms the first step (1), which may originally have been wider but has been reduced
by erosion and perhaps by the blasting.The step clearly links now-absent steps to its south-west (beyond
the modern cliff edge) with the second and third steps to its east. The second and third surviving steps
are large and finely cut. The route then turns through a right angle back to the west, where there is a
wide flattened area (4) and then a step formed by flattening the top of a narrow ledge (5) on the rock
spur that rises at the inner side of the ledge.
190
THE EXCAVATIONS
Ill. 2.203—The
original and
modern routes
and the area of
the cliff blown
away at the
junction of the
short broad
and ragged
ledges (A.R.
Hayden).
[Left] Ill. 2.204—Features at the junction of the short broad ledge and the ragged ledge, looking east (A.R. Hayden).
[Right] Ill. 2.205—Close-up of the drilled blasting hole (A.R. Hayden).
191
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The route then crosses another
deep fault line. It is probable that
there could have been masonry here
spanning this fault, as it could have
been easily built and securely footed
in the fault. No stonework survived
here, however, which is not
surprising, as the blasting of the rock
at the lower end of this section of
the ledge could have removed any
masonry that might have remained
at that time.
The next step (6) is on a
narrow ledge on the rising rock on
the right. Then there are two fine
rock-cut steps side by side (7 and 8),
followed by a series of five (9–13)
very finely preserved and large rockcut steps that lead up to a slight turn
to the right, where the cliff on the
inner side of the ledge was clearly
cut back. Around the slight corner
there is a worked and flattened area
(13) on the surface of the ledge, and
then, where the ledge rises and turns
to the north-west, there are four
Ill. 2.206—
Features on the
ragged ledge,
looking west
(A.R. Hayden).
[Left] Ill. 2.207—
Masonry built by
the OPW crew
on the ragged
ledge to aid the
transport of
stone up to the
broad ledge for
conservation.
There could well
have been
masonry here
originally also
(A.R. Hayden).
[Right] Ill.
2.208—The
western end of
the ragged ledge
as it turns and
opens onto the
broad ledge
(A.R. Hayden).
192
THE EXCAVATIONS
clear rock-cut steps (nos 14–17). The first and the last two are poorly preserved. There is a handhold on
the rock to the right above the junction between steps 16 and 17.
The broad ledge
After this corner the route opens onto a 4–4.5m-wide ledge that rises evenly and steeply upwards from
south-east to north-west for a length of 12m. This is the broad ledge, one of the naturally widest parts
of the climb to the South Peak.
Ill. 2.209—The
broad ledge
from the west
before
excavation
(Con Brogan,
DAHG).
[Left] Ill.
2.210—The
broad ledge
before
excavation,
from the south
(A.R. Hayden).
[Right] Ill.
2.211— The
broad ledge
before
excavation,
from the northwest (A.R.
Hayden).
193
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
High and ragged cliffs, where the ends of the bedrock cleavage planes are exposed and unevenly
eroded, mark the inner side of the ledge. The outer edge of the ledge is also ragged for the same reason.
About two-thirds of the way up the broad ledge there is a near-vertical gully in the cliff forming
the inner side of the ledge. There are rock-cut steps in the upper half of this gully, and it provided access
to the area beneath the oratory terrace where stone was won or quarried and raised to build the oratory
and ‘garden’ terraces (see below). There are many large, loose rocks in the gully and it is clear that many
have tumbled down onto the broad ledge—and, indeed, more are likely to do so in the future.
Ill. 2.212—
Continuing
erosion evident
on the broad
ledge. The large
stone visible
near the top of
the ledge in the
top left
photograph
had
disappeared by
2004. (Top left,
Horn et al.
2002, fig. 15;
remainder, A.R.
Hayden).
194
THE EXCAVATIONS
One such large stone lay at the lower end of the broad ledge, close to the cliff on the outer side of
the ledge. A second huge stone, shown in 1980s photographs on the upper end of the ledge, fell over the
cliff before 2004. In general, however, this ledge was devoid of fallen stones and so must have been cleared
by the monks, as the width of the ledge means that a large quantity of stone would naturally have
accumulated on it.
Before excavation, bedrock was exposed at the upper end of the ledge and along its outermost
1.5m or so.There are a number of what appear to be man-made hollows on the exposed bedrock at the
outer side of the ledge. These were partly obscured beneath scree and campion.
The area along the base of the rising cliff on the back (north-east) side of the ledge contained
vertical or near-vertical stones set in a mineralised silt matrix.This is typical of an area where much water
is flowing down a cliff face and eroding the ground (like a drip gully); a similar feature was observed on
the inner side of the upper traverse. The water in the drip gully flowed downslope and then across the
ledge close to its lower end, where it had gouged out a significant hollow. The flowing water also
deposited a large debris fan on the outer side of the ledge where it flowed over the cliff.
Before excavation a few large stones were visible protruding through the thin cover of soil, stones
and campion on the ledge.Two stones lay on top of each other about two-thirds of the way up the ledge
and appeared to be step treads. They were roughly in situ but both appeared to have slipped downhill
somewhat and there was a likelihood of further slippage, as the ground beneath them was heavily eroded
and undercut. The remains of the base of
a possible drystone riser were also visible
towards the lower end of the ledge. These
stones also lay at a steep downhill angle;
they had slipped considerably and were
clearly likely to move further downhill.
These survivors seemed to represent the
last remnants of a very poorly preserved
masonry stairway, which originally could
have extended up the full length of the
ledge.
The thin soil cover that protected the
last remnants of the stairway was also being
eroded at an increasing rate.This was very
noticeable over the three wet and windy
summers of 2007–9. The amount of soil
and vegetation on the terrace was
noticeably less than in the 1980s (compare
to Horn et al. 2002, fig. 15, which shows
the ledge almost completely covered in
growth). Falling stone also posed a danger
to the surviving remains. The large stone
shown towards the north-western end of
the terrace in the 1980s was probably
knocked off the ledge by another falling
stone.
The increase in human traffic if the
area was to be reopened to the public
would only add to the erosion. The climb
195
Ill. 2.213—The
broad ledge
after initial
clearance and
excavation
(A.R. Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
N
Gully
Step treads
Step treads
Step risers
rock-cut features
0
5
original edges of stair
metres
scale 1:50
Ill. 2.214—Plan of the broad ledge after initial clearance and excavation.
up this part of the route to the Peak is also relatively easy until the blind corner, just beyond this ledge,
is reached. It is at that point that many people turn back. Therefore it was anticipated that there would
be a relatively large amount of human traffic, and hence disturbance, on this ledge when the site was
reopened to the public.
The broad ledge was excavated in 2010.
Rock-cut features
A number of small, shallow depressions, cut into the smoother surface of the bedrock, lay just southwest of the fault line about midway up the ledge. These were fully revealed by the excavation. In this
area there are also a number of small bedrock ledges, which looked like they had been levelled and
flattened. All these features were very heavily eroded. It is not clear what they represent.The cut features
appeared to be too shallow and small to have been rock-cut steps and they did not appear to conform
to any recognisable pattern that would have aided a climber. In view of the survival of the tripod holes
on the ledge with fallen stones, it is possible that the features on the broad ledge could have functioned
in a similar way and provided anchorage for timbers used in moving large stones.
The stairway
Excavation revealed the last remnants of a number of parts of a formerly extensive stairway beneath the
very thin covering of shillet, clay and campion.
The possible step riser noted before excavation towards the lower end of the ledge appeared to be
196
THE EXCAVATIONS
just a coincidental alignment of stones, several of which had been moved to their present position owing
to the impact and damage caused by falling stones. A definite step riser was discovered further downslope,
however, near the lower end of the ledge. It was a drystone structure and composed of quite small stones.
Its eastern edge may have been intact, but some of it could have been disturbed and removed by erosion
down the drip-gully channel. Its western end was partly demolished when the huge fallen stone slid
over it towards the cliff edge and pushed it slightly downhill. There was a clear impact crater in the
surviving stonework on the eastern side of the large slipped stone.
The very base of a second drystone riser was revealed a few metres upslope. This, however, was
greatly disturbed and only one or two courses of stonework survived.
[Left] Ill. 2.215—Detail of the lowest step riser (A.R. Hayden).
[Right] Ill. 2.216—The second step riser (above the ranging rod). The large fallen stone that damaged the first riser is visible at the left
(A.R. Hayden).
Ill. 2.217—Step risers and
treads on the lower twothirds of the ledge (A.R.
Hayden).
197
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Further up the ledge three
stair treads survived but all had
slipped out of position to varying
degrees. The uppermost two were
those visible before excavation.The
lowest and newly revealed stone had
slipped downwards considerably
and had cracked in half. It was
removed after recording, but only
disturbed stonework survived
beneath it.The two uppermost large
Ill. 2.218—The
stones were still roughly in position
three surviving
stair treads
but had rotated clockwise and
midway up the
slipped slightly downhill.
ledge (A.R.
A straight-edged fault runs
Hayden).
along the length of the ledge about
1.5m from its outer edge.The slope
of the ledge follows the bedding
plane of the rock south-west of the
fault and is smooth. North-east of
the fault the bedding planes of the
rock are generally horizontal
(resulting in a series of flat-topped
natural steps) except against its
immediate north side, where it is
twisted, broken and uneven and
protrudes in places above the lower rock to its south-west. The edge of the fault line was utilised by the
monks to retain the outer edge of the masonry staircase; the rougher rock surface north-east of the fault
line also provided a more secure base on which to build.
Southwest
Northeast
slipped & broken tread
modern
topsoil silt
mineralised
material
Ill. 2.219—Top:
southwest/north-east
section across
the lowest
surviving step
tread and the
drip gully.
Bottom:
longitudinal
section up
along the ledge.
soft grey silt
West
Northwest
0
5
metres
Southeast
198
THE EXCAVATIONS
The surviving remains of the stairway show that it ran roughly straight up the lower two-thirds of
the ledge before turning to the west near its top.The stairway appears to have originally measured about
1.5m in width.
The bare inner side of a number of step treads and the walling supporting their inner edge survived
at the beginning of the upper third of the ledge, where the bedrock cliff on its inner side turns to the
west. The stairway there also turned at about 45º to the west. Only its neatly aligned and straight inner
(north-eastern) edge survived. Much of the remaining stonework in this area was badly crushed; indeed,
the stones that survived were only partial remnants of the originals, the remainder having been crushed
and broken away. This destruction was caused by the impact of large stones, which fell down the steep
gully at the inner side of the ledge.The bedrock in the gully is unstable and fractured and there are several
large stones there, which will likely fall at some time in the future. Stone falling from this gully was
probably also responsible for most of the destruction evident on the lower part of the stairway.
After the turn, the individual steps would have been fewer and more elongated, as the slope the
stairway covered was much more gentle. The stairway is likely to have continued right to the western
end of the ledge, where it would have merged into the rising bedrock. Its north-western end led on to
the rock-cut steps on the next ledge.
Ill. 2.220—The surviving
inner side of the stairway at
the northern end of the
ledge, from the east (A.R.
Hayden).
199
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.221—The
upper end of
the stairway
from the west;
note the
crushed and
fractured
stones (A.R.
Hayden).
N
Gully
Step treads
Ill. 2.222—Plan
of the broad
ledge after the
completion of
excavation.
Step treads
Step risers
rock-cut features
5
0
metres
scale 1:50
200
original lines of stair
THE EXCAVATIONS
Drip gully
The base of an original drip gully, naturally formed by water trickling
down the steep cliff face at the inner side of the ledge, underlay both the
modern drip gully and the surviving stonework of the stairway and was
filled with soft, grey, silty clay. The inner side of the stairway was set away
from the cliff on the inner side of the ledge, sensibly maintaining a drain
along the base of the cliff. Deposits built up in the drain/drip gully after
both the construction and the destruction of the stairway. These deposits
consisted of heavily mineralised stones and clay and were so hard that they
could only be removed by chiselling and pickaxing.
The silting up of the drip gully at the eastern end of the northern
cliff restricted drainage and caused the formation of a pool of water
between the cliff and the inner side of the steps at the northern end of
the ledge. The fill of the pool was only partly excavated, as it was
unnecessary to remove it all. Its fills were again heavily mineralised.
Ill. 2.223—The
gully that gives
access to the
ledges beneath
the oratory
terrace and
down which
stones fell,
damaging the
stairway on the
broad ledge;
note the
crushed stones
at the base of
the gully (A.R.
Hayden).
Ill. 2.224—The drip gully on the
inner side of the stairway (A.R.
Hayden).
201
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The first narrow ledge and the blind corner
The broad ledge narrows abruptly at its upper end, where bedrock was again fully exposed before
excavation.The ledge then turns to the west and, narrowing further, extends upwards to the blind corner.
There are two well-preserved runs of rock-cut steps on this part of the climb. The ledge in this area has
been almost entirely cut from bedrock.
No excavation was required in this area, as all the surviving features were rock-cut.
Ill. 2.225—The
first narrow
ledge, leading
from the broad
ledge (right) to
the blind
corner (left).
The rock-cut
steps and lower
ledge outside
them are
clearly visible
(A.R. Hayden).
Above the top of the masonry stairway there is a low spur on the cliff at the northern end of the
broad ledge. The top of the eastern end of this spur was worked and flattened. There is a clear rock-cut
step (1) on the centre of the spur, followed by two more rock-cut steps (2 and 3) further to the west. A
fourth rock-cut step lies to the north of the last and brings the route around a slight turn to the north,
where the cliff has clearly been cut back. There is then another clearly worked and flattened area (5).
There is a further worked and flattened area in the angle formed by the rising cliff on the right.This was
not a step, as it lies off the line of the route.
The ledge then narrows considerably and is clearly totally man-made, having been dramatically cut
from the steep cliff face. There is a fine run of nine shallow rock-cut steps (6–14) on the inner side of
the quarried ledge.There are also two rock-cut handholds (about 1m above steps 10 and 13) on the cliff
on the inner side of the route before it rounds the blind corner, and a third one and a fourth, less definite
one opposite step 14 immediately after the blind corner.
There is a long, narrow ledge, partly natural and partly carved following a natural bedding plane,
outside and 200–300mm below the level of the rock-cut steps on the first narrow ledge. A photograph
taken in the 1980s (Horn et al. 1990, fig. 17) clearly shows drystone masonry capped by a large flat slab
lying on the lower end of this ledge. This masonry does not survive today but its former presence shows
that the pathway around this vertiginous corner was augmented by masonry built on the lower ledge on
202
THE EXCAVATIONS
[Left] Ill.
2.226—The
junction of the
broad ledge
and the first
narrow ledge
(A.R. Hayden).
[Right] Ill.
2.227—The
upper part of
the first narrow
ledge (A.R.
Hayden).
N
Cliff
cut back
Ill. 2.228—Plan
of the top of
the broad ledge
and the first
narrow ledge.
Rock-cut
steps
Rock ledge
Blind
Corner
Broad
Ledge
0
5
metres
scale 1:50
its outer side, and so the passage here would have been wider and less daunting than it appears today.
Masonry with an identical function survives on a similar ledge just around the blind corner, and also on
the new route discovered in 2010 leading to the early entrance to the monastery on the other peak of
the island.
It is also possible that the eight rock-cut steps before the blind corner may have been intended to
secure stonework rather than being actual steps on which to walk.To securely anchor the masonry on the
narrower ledge it probably would have been necessary to key it into stonework laid on the rock-cut steps.
203
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
It is also possible, of course, that the masonry on the lower ledge could be a later augmentation, and
it is therefore unclear whether the features here represent a single phase or several phases of activity.
The passage here is of sufficient width to have allowed the construction of a narrow parapet wall
along the cliff edge, but no evidence of such a structure survived.
[Left] Ill.
2.229—
Masonry that
survived in the
1980s on the
narrow ledge,
evidencing the
former
existence of a
wider pathway
or parapet wall
(Horn et al.
1990, fig. 17).
[Right] Ill.
2.230—The
same view
today (A.R.
Hayden).
Ill. 2.231—The third handhold at the blind corner (A.R.
Hayden).
204
Ill. 2.232—The blind corner viewed from the west (A.R.
Hayden).
THE EXCAVATIONS
The second narrow ledge
The route turns to the north-west as it
rounds the blind corner, and here the
pathway is again completely cut from the
rock cliff. It runs flat and straight, as its
base was quarried down to a naturally flat
bedding plane in the bedrock. The
pathway then turns back to the west,
where there are two large stones fallen
from the cliff above. It then broadens and
extends a short distance to the northwest to worked and enhanced rock
ledges (the stepped ledge) that lie at the
base of the cliff beneath the Needle’s Eye.
The last remnants of some very
poorly preserved stonework were visible
before excavation on a ledge just below
the middle section of the second narrow
ledge. The masonry appeared to have
fulfilled the same function as that which
had been lost since the 1970s on the first
narrow ledge and therefore could have
been the only surviving masonry that
evidenced the former existence of such
walling on either of these ledges.The masonry was loose and very collapsed and hence highly vulnerable.
The full length of the path from the blind corner to the base of the cliff below the Needle’s Eye was
excavated in 2008 to reveal any surviving masonry and allow it to be recorded and conserved.
Ill. 2.233—View
south-east from
the stepped
ledge along the
second narrow
ledge towards
the blind
corner, before
excavation.
Note the large,
naturally fallen
stones (A.R.
Hayden).
[Left] Ill.
2.234—The
quarried
pathway of the
second narrow
ledge and
remnants of
walling on the
ledge below it,
looking northwest (A.R.
Hayden).
[Right] Ill.
2.235—The
quarried
pathway of the
second narrow
ledge and
remnants of
walling on the
ledge below it,
looking southeast (A.R.
Hayden).
205
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Stepped Ledge
N
fallen flag
Ill. 2.236—Plan
of the second
narrow ledge
from the blind
corner to the
stepped ledge.
wall
rock-cut
steps
fallen
stones
wall
cliff cut back
rock ledges
0
5
metres
Blind
Corner
Ill. 2.237—The second narrow ledge: rock-cut steps leading
up from the western end of the narrow quarried ledge
(A.R. Hayden).
206
Ill. 2.238—Walling surviving west of the large fallen
stones along the cliff edge where the second narrow
ledge widens and turns to the west (A.R. Hayden).
THE EXCAVATIONS
The path beyond the blind corner initially is completely cut from the almost vertical rock cliff and
measures about 400mm in width. Its outer side is ragged and uneven. There are narrow ledges outside
and approximately 300mm and 600mm below the main ledge, which retained some poorly preserved
stonework at their western ends.This masonry clearly demonstrates that the pathway here, like that before
the blind corner, was wider than at present, as it had masonry on its outer side. An approximate height
of 600mm of drystone walling survived over a length of 2m south-east of the first large fallen stone. Its
base was constructed on a narrow rock ledge that lay about 600mm below the rock-cut base of the
pathway. The face of the wall was poorly preserved but it appears to have been battered. There was a
second, 200mm-wide ledge at a higher level (about 200–300mm below the path level) at the eastern
end of the surviving wall. It extended east to the blind corner and it too could originally have held
masonry. These ledges were clearly worked and widened.
These ledges indicate that the passage here was formerly of sufficient width to have allowed the
construction of a narrow parapet wall along the cliff edge, but again no evidence of such a structure
survived.
The pathway steps upwards towards the western end of the length of surviving walling, just before
it reaches the first large fallen stone. A deep rock-cut ledge survived on the bedrock here. The quarried
flat top of the bedrock immediately above this ledge provided a second step up. These steps may have
held masonry and were not necessarily intended as steps for the feet of the climber. Originally there
must have been at least two masonry steps here on top of the rock-cut ledges, which raised the pathway
upwards to its evident level beyond the large fallen stone.
North-west of the large fallen stone the pathway broadened and turned to the west. A maximum
of two courses of drystone walling, 1.2m in length, survived of the original wall that defined the outer
side of this part of the pathway, which was approximately 1m wide. The walling ran along the cliff edge,
Ill. 2.239—The
stepped ledge
viewed from
above (north).
The rock-cut
slot visible in
the bottom
right corner of
the planning
frame drains
water from the
inner side of
one of the
quarried
surfaces (A.R.
Hayden).
207
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
which sloped upwards to the west. The impact of the large fallen stone, which still lay on the walling,
formed a readily recognisable crater where the underlying walling was heavily distorted; many of the
original stones in the walling were smashed or twisted to lie nearly vertically. This feature is paralleled
on the first section of terracing of the north-west passage, which was also hit and partly demolished by
a large stone that fell from above (see below). This side of the Peak is particularly prone to damage from
stones falling from the cliff above owing to local geology.
The stepped ledge
The broad pathway at the western end of the second narrow ledge leads up to a wide and open area
where there are a series of flat rock ledges, which were quarried from the rising bedrock following natural
bedding planes.There is a clearly evident drain cut into the outermost of these ledges. It prevented water
pooling on this broad step, the surface of which sloped slightly downwards to its inner side. A number
of huge fallen stones lie on the gently sloping ground inside and to the north-east of the exposed bedrock
of the stepped ledge.
No excavation was undertaken here as the surviving features were rock-cut and fully exposed.
2.3.4 THE
PRIMARY ROUTE TO THE
SOUTH PEAK
Introduction
From this point the two routes to the top of the South Peak diverge. The final route climbs up the cliff
to the north-east.The primary route runs west and around the north-west corner of the Peak (the northwest passage). It crosses four sections of terracing (the southern traverses) as it runs to the north-west
corner of the Peak. Turning the corner to the north and east, it descends down a masonry stairway and
across another masonry traverse into a steep gully. It then ascends this gully directly to the top of the
South Peak. This route was only discovered during the recently undertaken works.
The north-west passage
The north-west passage was composed of a number of elements. Four sections of terracing—the southern
traverse—led west to the north-west corner of the Peak. Rounding the corner, the route descended to
the north-east down drystone stairs—the northern stairway—to a short length of terracing—the northern
traverse—which ended at a steep gully.
Before excavation an obvious route led west from the stepped ledge along a prominent ledge to
the north-west corner of the Peak. No structures were visible along most of this route, but just before
the corner at the western end of this part of the route there were a number of rock-cut steps leading
down to a well-preserved, although partly collapsed, section of drystone terracing (southern traverse,
section 4). Around the corner there was a possible routeway following a narrow ledge hugging the vertical
cliff. It sloped downwards at first but then levelled out and ran into a long, steep gully that extended up
from the sea (where it is called Glengarrif on the six-inch OS maps, sheet 104, from 1841 onwards) right
to the top of the Peak. A few flat stones protruded through the soil cover on these latter parts of the
route.
The terracing on section 4 was cleared of plant growth in 2007, revealing a substantial section of
surviving terracing. The survival of this masonry suggested that other masonry structures could have
been erected on the other parts of the southern traverse, and they were subsequently excavated in that
year.The discovery of the very poorly preserved terracing on these sections suggested that the route they
represented was important and potentially of early date. The ledges around the corner on the northern
side of the Peak (the northern stairway and northern traverse) were excavated in 2008.
208
THE EXCAVATIONS
N
218m
Upper gully
200m
Lower gully
Ill. 2.240—The
upper part of
the primary
route to the
Peak.
Northern stairway
& traverse
North-west
Passage
Southern traverses
0
5
10
20
metres
150m
Ill. 2.241—The
southern
traverse (red
line) of the
north-west
passage, viewed
from the lower
lighthouse. The
fourth section
of the traverse
lies behind the
rock tor at the
left. The blue
line indicates
the final
sections of the
lower part of
the route (A.R.
Hayden).
209
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill. 2.242—Looking west along the southern traverse of the north-west passage before excavation (A.R. Hayden).
[Right] Ill. 2.243—Looking down on the fourth section of the southern traverse; the collapsed terracing is clearly visible
(A.R. Hayden).
Ill. 2.244—LiDAR
image of the
southern traverse
of the north-west
passage and the
north-west corner
of the Peak
(DAHG).
210
THE EXCAVATIONS
Ill. 2.245—
Aerial view of
the southern
traverse of the
north-west
passage during
excavation
(Con Brogan,
DAHG).
N
Second Section
Section
Second
Section
Third Section
Third
Corner
Corner
0
Fourth Section
Section
Fourth
First
First
Section
Section
5
metres
Ill. 2.246—Plan of the southern traverse and north-west corner of the north-west passage. The red lines indicate the
edges of masonry steps and walls.
The southern traverse
First section
The easternmost section of the southern traverse consisted of a very wide and irregular-shaped terrace.
Its curving outer wall was built following the cliff edge.Very limited and poorly preserved walling, only
one to two courses in height and laid along the curving edge of the cliff, survived along the outer edge
of the eastern end of the traverse. At the south-west end of the traverse none of the outer face of the
walling survived, but a tell-tale vertical edge containing some small backing stones survived fronting the
infilled stone and quarry waste lying just back from the cliff edge.This edge marks the former line of the
inner face of the terrace wall, which was typically composed of stones smaller than those used in the
outer face.
211
Stepped
Stepped
Ledge
Ledge
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
N
Ill. 2.247—Plan
of the first
section of the
southern
traverse, with
the stepped
ledge at the
right. The wall
designed to
stop debris
from washing
down onto the
terrace lies at
the base of the
slope at the
back of the
terrace (top).
in
dra
ff
cli
ge
ed
0
5
metres
scale 1:50
Ill. 2.248—Looking
down on the first
section during
excavation.
Collapsed paving
slabs can be seen
in the foreground,
and part of the
surviving traverse
wall at the back
(A.R. Hayden).
212
THE EXCAVATIONS
The west end of this first section of terracing is marked by a deep gully, which severely constricts
the width of the access route. To allow transit across it, a wall was built in the gully, based on a narrow
ledge at a lower level than the main wall of the traverse. Only a small part of this drystone wall survived
and it was in very poor condition.The cliff on the inner side of the traverse just before (east of) the gully
also appears to have been partly quarried back to facilitate transit; the end of a protruding rock spur has
clearly been quarried away down to just below the finished level of the terrace—it was left untouched
at a lower level.
Ill. 2.249—Looking east
at the outer side of the
first section. Traces of
outer walling survive at
the right, and the wall
crossing the gully is
visible at the left (A.R.
Hayden).
There was a low drystone retaining wall,
where a steep-sided fan of debris cascaded
down from higher gullies, on the inner side of
part of this traverse. The wall was laid in a
straight-sided cut made through the lower end
of the debris fan and was obviously intended to
retain material washing down the cliff and stop
it from spreading onto the traverse. After the
traverse went out of use and the wall collapsed,
a considerable amount of debris again
accumulated here, as owing to the surrounding
topography this area is a natural collection point
for material falling from the cliffs above.
Originally this traverse was paved with
Ill. 2.250—Drystone retaining wall at the back of the
first section, looking east (A.R. Hayden).
213
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
some very large, flat stones that later slipped southwards and downwards after the collapse of the outer
wall. The collapse of the terrace wall may have been, at least partly, caused by large stones falling from
the cliffs above. One huge rock, weighing several tonnes, which fell from the cliff landed on the terrace
and still lay embedded in its south side. There was a clear impact crater around the stone where the
underlying walling had been pulverised, pushed downwards and partly trapped beneath the stone. Stones
turned to lie vertically marked the outer edge of the crater.
Ill. 2.251—First section: fallen stone and crater
around it. Paving and drystone retaining wall at the
inner side of the traverse visible at right (A.R.
Hayden).
Ill. 2.252—Worked
stone found on the
first section of the
southern traverse
(A.R. Hayden).
214
THE EXCAVATIONS
Second section
The next section of the southern traverse consisted of a broad bedrock ledge that slopes slightly upwards
to the west.There were traces of an outer wall along the straight cliff edge, but again the walling survived
very poorly and to only one or two courses in maximum height. This section of the traverse was also
paved with very large stones, although only a few survived and those that did had also slipped downwards
and southwards after the collapse of the outer wall. The terrace-builders left some of the naturally
accumulated rock in situ at the western end of this section, as it lay beneath the intended finished level
of the traverse. Its presence highlights the absence of other, similar naturally accumulated stone elsewhere
on the north-west passage and clearly demonstrates that the ledges were cleared of naturally fallen stone
[Left] Ill.
(which provided a convenient source of building stone) before the terraces were built.
2.253—Looking
west along the
second section.
Traces of the
original outer
wall and paving
visible. Wall
crossing gully
from first
section at front
left (A.R.
Hayden).
[Right] Ill.
2.254—Looking
back eastward
along the
second section.
Naturally fallen
stones left on
bedrock ledge
in foreground,
as they lay
beneath the
finished terrace
level (A.R.
Hayden).
Ill. 2.255—Plan of the second and third sections of the
southern traverse.
N
f
clif
cli
ff e
dg
e
ge
led
ion
p
cam
ge
ed
0
5
metres
215
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Third section
A deep gully crossing the ledge marks the start of the next section of the southern traverse. In the gully
there were again traces of the base of external walling built on a lower ledge to facilitate transit.The cliff
on its inner side was also partly quarried back to widen the passage and facilitate the crossing of the
gully.The surface of the main ledge here slopes quite steeply upwards to the west.The base of the cleared
level of the ledge lies at the same level as the finished top of the first and second sections to the east and
shows that originally the tops of these three sections of the traverse lay at the same level. The finished
level of these sections lay well above any of the surviving stones on the rock ledge of the third section,
except at its western and higher end, where the bedrock would have lain close to the original finished
level.The western end of this traverse was largely devoid of any stone, and the bare, sloping bedrock was
exposed before excavation.
[Left] Ill. 2.256—Looking east from the third section; walling at its east end visible in foreground (A.R. Hayden).
[Right] Ill. 2.257—The walling at the east end of the third section viewed from the east (A.R. Hayden).
Fourth section
The top of the fourth section of the southern traverse lay more than 2m below that of the preceding
one and was separated from it by another fault line, this time not marked by a gully but by a 4m drop in
the level of the underlying bedrock bedding plane.
A small piece of bedrock stands proud of the rest of the cliff leading down to the fourth section of
the traverse. It is possible that this was once carved into a cross, but it was very highly eroded and no
evidence of working remained.
A narrow, east/west-aligned cleft close to the cliff on the inner side of the western end of the third
section of the traverse contained three shallow and narrow rock-cut steps that led down towards the top
of the cliff down to the fourth section of the traverse.The stumps of three further rock-cut steps survive
at the base of this low cliff. These latter steps are very poorly preserved and there clearly must have been
more steps here originally. The cliff here is highly eroded. There are two clear handholds on the top of
the rock spine on the northern side of the climb down. There may also have been another on this side
216
THE EXCAVATIONS
Ill. 2.258—The cliff leading down
from the third to the fourth
section. Note the small piece of
protruding bedrock at right,
possibly once a way-marker
cross (A.R. Hayden).
and one on the southern side, but the rock
is very fractured and no clear remains
survived.
The fourth section of terracing was
the best-preserved structure on the whole
north-west passage, surviving to almost 2m
in height. Nevertheless, even this terracing
had severely collapsed inwards, causing
some of the basal stones to burst outwards,
and there were substantial voids in the base
of the walling. The stones of the upper part
of the wall had moved inwards considerably
and slumped downwards at their inner ends.
The uppermost courses of walling, however,
were horizontal and level and had clearly
been rebuilt at some time, possibly by the
lighthouse-builders.There is a possible route
that may have been used by the lighthousebuilders from the upper lighthouse up the
gully at the north-west corner just beyond
this terrace (see below).
The eastern end of the terrace would
originally have measured almost 2.5m in width, but owing to collapse its top survived to little over 1m
wide. The terrace originally narrowed to only 1m in width over its western half. The inner side of the
terrace must have consisted of loose material deposited inside its outer walling. The presence of a high
cliff close to the rear face of the terrace wall meant that water gathered and flowed down though the
back of the terrace and removed much of the finer material behind the wall, causing it to collapse inwards.
The stones used in the wall of this traverse were noticeably large, perhaps because of the easy
availability of large rocks, which fell from the broken ends of the adjacent cliffs. Many huge, naturally
fallen rock slabs litter this area. Most pre-date the construction of the traverse, as they are incorporated
217
Ill. 2.259—The
rock-cut steps
(white) and
handholds
(yellow) in the
cliff leading
down from the
third section to
the fourth
section of the
southern
traverse (A.R.
Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.260—Plan
of the fourth
section of the
southern
traverse.
[Left] Ill. 2.261—The gaps at the base of the wall of the fourth section (A.R. Hayden).
[Right] Ill. 2.262—The gaps in the base of the wall were built up with stone to support the wall as the excavations
progressed (A.R. Hayden).
[Left] Ill.
2.263—Looking
down on the
fourth section
terracing from
the east. The
flat stones on
the lower ledge
are visible on
the left (A.R.
Hayden).
[Right] Ill.
2.264—Looking
down on the
ledge below
the fourth
section from
the west (A.R.
Hayden).
218
THE EXCAVATIONS
into it, but some also clearly fell at a later time.
Excavation in this section consisted simply of removing the sea campion growing between the
joints in the outer face and top of the terrace wall and cleaning over the ledge at the base of the terrace.
The voids in the base of the terrace wall had to be infilled with masonry to secure the wall before the
lower ledge was excavated.
Several large, flat stones were visible on the broad, roughly flat ledge below and outside the base of
the terrace wall. This ledge was cleared of plant growth but not further excavated. The large stones may
have been deliberately laid either to facilitate construction of the terrace or to provide a usable space on
this wide ledge.There are a number of other, lower-lying broad and flat ledges on the cliffs below, which
are easily accessible from this ledge. None of these were excavated or examined in detail, but there is
possibly a route here leading down the cliff face.
The north-west corner
Westwards from the fourth section of the southern traverse, the route follows a natural ledge around a
tight corner at the north-west end of the South Peak.There is a deep and initially narrow gully extending
downwards from this corner. Several large and naturally fallen stones were lodged in the head of the
gully and provided a convenient transit across it.
Possible lighthouse-era route to upper lighthouse
This gully descending from the north-west corner looks like it could be easily climbed, and it is possible
that there was once a routeway here leading down to the (lighthouse-era?) terraces that lie above the
back of the upper lighthouse. This route has not been climbed or explored in any way as yet.
The blasting hole at the junction of the short broad ledge and the ragged ledge and the repaired
top of the terracing of the fourth section of the southern traverse suggest that the lighthouse-builders
may have used this route as a short cut between the monastery (where they lived) and the upper
lighthouse.
West corner of
north-west passage
Terrace behind
upper lighthouse
Terraces (?)
219
Ill. 2.265—
LiDAR image
showing the
possible route
from the northwest corner of
the north-west
passage down
the gully to
drystone
terraces and on
to the long
terrace at the
back of the
upper
lighthouse
(DAHG and
A.R. Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.266—The
possible route
from the back
of the upper
lighthouse
(arrowed left)
up the gully to
the corner at
the western
end of the
southern
traverse
(arrowed right)
(A.R. Hayden).
Ill. 2.267—The
northern side
of the northwest passage
before
excavation
(A.R. Hayden).
Rounding the north-west corner
The bedrock ledge leading from the head of
the gully around the corner is heavily eroded
owing to its crossing the exposed western ends
of the cleavage planes of the bedrock. It is not
clear whether this ledge was man-made, partly
enhanced or totally natural. No excavation was
undertaken here, as there were no suggestions
of man-made structures surviving in the area.
The northern side of the north-west
passage
After the corner the route turns back to the
north-east, where there are a few worked
natural bedrock ledges leading downwards to
two more sections of masonry lying on narrow
ledges on the vertical to overhanging cliff.
These two sections were discovered, excavated
and conserved in 2008.
The first section of masonry consists of a
flight of masonry steps (the northern stairway)
leading down along natural sloping bedrock
ledges, past a small rectangular platform on the
cliff edge to a long, narrow terrace (the northern traverse).The terrace led into a prominent natural gully
that runs up from the sea right to the top of the South Peak.
220
THE EXCAVATIONS
Ill. 2.268—The northern stairway and
traverse after excavation (A.R.
Hayden).
metres
NORTHERN
TRAVERSE
0
cliff edge
5
N
GULLY
Ill. 2.269—Plan of the
north side of the northwest passage.
cliff ed
ge
RECTANGULAR
STRUCTURE
NORTHERN
STAIRWAY
NORTHWEST CORNER
221
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The northern stairway
Stepping down off the bedrock after the corner there is evident masonry. The last step down onto a
bedrock ledge is high and there are traces here of the base of a drystone step, which, originally being
higher, would have rendered the step down easier. The base of another masonry step survives below it.
This is the uppermost of eight former steps that led downslope for a length of 3.8m to the east.
Ill. 2.270—The northern
stairway after excavation,
from the north-east (A.R.
Hayden).
The masonry of all but one of this flight of steps survived. The existence of the missing step is
evident, however, as the steps were all of roughly equal width and the missing section of stonework was
of this width.The steps survived as little more than one and occasionally part of a second course of stone
that were just the bare bases of the original steps. Tell-tale stones had slipped to the vertical against the
lower ends of some of the steps, clearly showing the former existence of more masonry on the surviving
stonework. Sloping bedrock immediately underlay the surviving stones.
Traces of a narrow wall that had almost totally collapsed outwards survived on the outer side of the
steps. The steps would then originally have had a narrow parapet wall along their outer side that would
have incorporated their outer ends within it. Including the wall, the structure here measured 1.3m in
width. It is not clear how far down the steps the parapet wall stretched, as no trace of it survived over
their lower half.
The stairway turned at a right angle to the north-west at its lower end, where the bases of two steps
survived at the lower end of the bedrock slope.These steps led down to a large and shallow notch cut into
the next sloping bedrock ledge below. This notch was too far down to have been the surface of a step in
itself and so was probably cut to anchor a masonry step on the sloping rock. From this point the route turns
back again through 90º to the north-east. Originally, there must have been further steps here owing to the
evident height difference but only the very base of the lowest step survived. A line of small stones that had
slipped to the vertical marked a joint in the masonry, showing the position of this lowest step.
Rectangular platform
Just after the second turn of the steps the bedrock cliff turns abruptly inwards (south) along a fault line.
222
THE EXCAVATIONS
[Left] Ill. 2.271—The junction of the northern stairway and the
northern traverse, looking east (A.R. Hayden).
[Right] Ill. 2.272—Detail of the junction, showing vertical stones,
looking west (A.R. Hayden).
The cleavage planes here run east–west and the edge of the cliff consists of a narrow bedrock spine with
a steep-sided cleft inside it. A large stone was jammed upright across the eastern end of the rock cleft.
This was deliberately done, as it was secured in place by several smaller, tightly wedged stones.The vertical
stones retained the eastern side of a short length of drystone walling (of which both faces survived) that
extended northwards for the short distance to the cliff edge. A notch was cut in the end of a protruding
bedrock spine to retain the western face of this wall at one point.The wall appeared to turn to the southwest, following the cliff edge. This suggests that there was a small rectangular terrace or platform here. It
is not clear where its south-western end lay; it may have extended up the cliff in line with a clear narrow
step in the bedrock that would have provided a secure footing for its end wall. If this were the case, the
structure would have measured about 3m in length by 1.4–1.8m in width, but the bedrock there slopes
so steeply upwards to the south-west that it is difficult to see how any usable enclosure or structure could
have been constructed. The form and function of this putative structure remain unclear. It would,
however, have faced in
roughly the same direction
as the small enclosure
above the Needle’s Eye and
that at the north-western
end of the ‘garden’ terrace
and so could have fulfilled
a similar function, i.e. as
some form of small
contemplation or prayer
station. Alternatively, the
structure could have been
built to house a leacht, cross
or some other form of
marker alongside the route.
223
Ill. 2.273—The
rectangular
structure at the
turn of the
northern
stairway. Its
north-eastern
wall lies to the
left of centre at
the bottom of
the
photograph; its
south-western
wall follows the
cliff edge (A.R.
Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ills 274 and 275—Details of the north-eastern
wall of the rectangular structure, showing
upright stones jammed into a bedrock fissure as
support (A.R. Hayden).
The northern traverse
The northern stairway led down to a
narrow, drystone-walled terrace—the
northern traverse. The terrace had an outer
retaining wall set on a bedrock ledge, which
sloped down to the north-east. There was a
clear masonry face on the outer side of the
south-western end of the terrace wall, where
bedrock was at its highest. Along the
remainder the wall of the terrace has largely
collapsed, but the original line of its face was
clear, as it followed the cliff edges of the
bedrock ledges down to the north-east.The
outer face of the terrace originally ran on a
straight line but because of the irregular line
of the cliff on its inner side the terrace varied
from 1m to 1.4m in width.
The easternmost 2m of the terrace was
very poorly preserved, heavily disturbed and
covered by material (stones and soil) that
Ill. 2.276—
Looking down
on the
northern
stairway and
the northern
traverse from
the upper
platform above
the Needle’s
Eye (A.R.
Hayden).
224
THE EXCAVATIONS
cascaded down the steep
gully that lies at its northeastern end.
The surviving part of
the terrace measures 3.9m
in length from the line of
the step at its western end
to the first rise in bedrock
at its eastern end.
Originally it may have
measured 5.6m in length if,
as seems likely, it covered
the bedrock ledges beyond
its north-eastern end. Its
surface was probably level
and appears to have been
paved, as several large, flat
stones survived among the
general collapsed rubble on its surface.
The steps described above indicate that the finished surface level of the terrace originally lay at the
same level as the base of the final rock ledge north-east of the terrace. Therefore the terrace would
originally have measured just over 1m in height at the most.
Including the masonry steps, a total of 12m of linear structure survives on this side of the northwest passage. Combined with the 27.5m of terracing uncovered on the southern side of the north-west
passage, this gives a total length of nearly 40m of new stonework discovered on this route.
The northern cliffs
Several natural bedrock ledges and large
lumps of slipped bedrock lead upwards
from the eastern end of the northern
traverse into a large gully. The gully
extends upwards from the sea, passing
under the uppermost of the two
platforms above the Needle’s Eye and
extending higher up, beneath the
northern end of the upper traverse, before
petering out just below and to the north
of the top of the Peak.
It was initially thought that the
north-west passage crossed this gully and
led on to roughly flat ledges beyond it on
the west-facing cliff and around the
corner to further ledges on the north side
of the Peak. These ledges were excavated
in 2008 but no definite man-made
structures were noted.
Ill. 2.277—The
northern
traverse after
excavation,
from the north
(A.R. Hayden).
Ill. 2.278—The
northern part
of the northwest passage
(right arrow)
runs down and
into the gully
(left arrow) up
which the
primary route
then runs to
the top of the
Peak. Roughly
flat ledges
extend around
the cliff to the
left from the
point of this
arrow (A. R.
Hayden).
225
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
First cliff ledge beyond the northern traverse
A clear ledge led north-westwards from the other side of the gully beyond the end of the northern
traverse to the south-western corner of the cliff. The rock at the level of the ledge contained several
horizontal layers of softer material, which eroded out to form small holes and flat ledges. There were a
number of areas of this softer rock that had been heavily worn by feet and which may also have been
carved to create footholds on the south-eastern half of the ledge, where the passage was restricted and
awkward.
Ill. 2.279—The ledge on the
cliff north-east of the end of
the northern traverse, before
excavation (A.R. Hayden)
The north-western end of this ledge was broad and relatively flat, and it was initially thought that
the clearly worn areas might have led to terracing at this end of this ledge.The broad area was excavated
but no evident structure survived and all the stones on the ledge appear to have accumulated naturally.
The stone consisted of a single layer lying on the uneven original rough surface of the ledge, which is
composed of fallen rock and in situ bedrock. A small ledge slightly higher up, which lay beneath a low
overhang in the cliff to the north-east, was also excavated. It proved to have been covered with only a
single layer of very small fallen stones.
Ills 2.280 and
2.281—Worn
and carved
footholds on
the ledge (A.R.
Hayden).
226
THE EXCAVATIONS
The second cliff ledge
The broader ledge led to a slightly
higher and easily climbed flat-topped
spur of rock, which marks a turn in the
cliff around to the north-east.The turn
leads to a broad (1.5–2m in width),
northward-sloping ledge that extends
eastwards for a length of about 10m.
This ledge was also excavated in 2008.
There was no man-made structure
evident on this ledge and only a very
small quantity of fallen stone was
present.
Ill. 2.282—The
area excavated
at the end of
the ledge, from
the south-west
(A.R. Hayden).
[Left] Ill. 2.283—The area excavated on the ledge, from the north (A.R.
Hayden).
[Below] Ill. 2.284—LiDAR image showing the location of the excavated
northern cliff ledges (DAHG and A.R. Hayden).
227
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill.
2.285—The
ledge beyond
the corner on
the northern
side of the Peak
before
excavation,
looking west
(A.R. Hayden).
[Right] Ill.
2.286— The
ledge beyond
the corner on
the northern
side of the Peak
after
excavation,
looking east
(A.R. Hayden).
The cliff face north of the gully
As no evidence of structures was found on the ledges north of the gully, the cliff face to the north of the
gully was also carefully examined to determine whether it contained any possible climbing routes, as it
initially appeared to
present an easier climb
than the gully. The
climb up this cliff face
is awkward, however, as
the
rock
often
protrudes out above
and beyond lower
ledges, and as a result
they are difficult to
climb
onto
and
Ill. 2.287—The
impossible to follow.
cliff face north
There were many
of the gully
(Con Brogan,
places where the climb
DAHG).
was difficult, and at
none
of
these
locations—or, indeed,
anywhere on the cliff
face—was there any
evidence of man-made
features.
228
THE EXCAVATIONS
The gully
Introduction
The absence of man-made structures elsewhere suggested, unlikely as it seemed at the time, that the
route evidenced by the north-west passage must have continued up the gully at the north-eastern end
of the northern traverse.
Upper traverse
Outer terrace
Upper platform
above needle’s eye
Ledge beneath
outer terrace
Gully
Ill. 2.288—
LiDAR image of
the southwestern side of
the South Peak
(DAHG and
A.R. Hayden).
Northern staircase
& Ttaverse
Excavated cliff ledges
Rope access was established down
the gully from the upper platform above
the Needle’s Eye, and the lower quarter of
the gully was cleared of campion and silt
in 2008 and cleaned down. This revealed
no features. Subsequent examination of
the upper part of the lower half of the gully
downwards from the upper platform
beneath the Needle’s Eye revealed,
however, that quite a number of rock-cut
steps and handholds survived there. These
clearly showed that the primary route did
in fact extend up this almost vertical climb.
Con Brogan of the DAHG took
photographs, looking horizontally into the
gully from a helicopter, in July 2009. In
August 2009 the features in the gully were
examined and recorded in poor weather
conditions, and their locations were
roughly marked on enlarged prints of these
photographs. Close-up photographs and
Ill. 2.289—The
north-west
passage and the
route up the
gully to the top
of the Peak
(Con Brogan,
DAHG).
229
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
video of the individual steps and handholds were also taken during the examination of the gully. The
steps and other features in the gully have yet to be surveyed in detail.
The gully—lower section
The gully at the level of the end of the northern traverse was choked for a height of 8m or so with rock
and silt that had fallen or had been washed down it, all covered by thick sea campion growth. The fallen
rock lies at an angle of about 45º and impedes transit across the gully.There are no evident foot- or handholds
[Left] Ill.
2.290—Looking
up the gully
from the end of
the northern
traverse at the
start of
excavations
(A.R. Hayden).
[Right] Ill.
2.291—Looking
up the gully
from the end of
the northern
traverse after
excavation;
note the fan of
debris at the
base of the
gully (A.R.
Hayden).
Ills 292 and
293—Views of
the gully,
showing the
smoothed and
highly eroded
bare rock in its
lowest 20m or
so (A.R.
Hayden).
230
THE EXCAVATIONS
in the fallen bedrock and no
remains of walling, so crossing
the gully was awkward, but not
impossible.This fan of fallen rock
and debris was not removed; the
campion and plant growth was
simply cleaned from its surface.
The bedrock in the lower
20m or so of the gully
immediately above the end of
the northern traverse was highly
smoothed by erosion, and there
were several eroded ‘bruises’
(shallow,
bowl-shaped
indentations) on the rock face
caused by the impact of falling
stones.
The first man-made
features (nos 1–4) were noted
between about 21m and 21.5m
up the gully above the level of
the top of the northern
traverse. The position of these
features is significant, as they lay
below the ledge on the cliff
face, which could have been the
top of an alternative climb up
the cliff face to the north that
attempted to avoid using the
lower, steepest part of the gully.
These steps suggest that the
lower part of the climb must
originally
have
ascended
straight up the gully.The lack of
any man-made features in the
lower part of the gully is clearly
the result of erosion; those that
survive above were all heavily
eroded. Silt and stones washing
down and stone falling down the gully have clearly eroded the bedrock. More severe erosion could
have occurred here, however, as the gully is located on the intersection of two major faults. One was
aligned east–west and is reflected in the line of the gully itself. The second is aligned north–south and
is reflected in the line of the cliff to the north of the gully. This intersection meant that the bedrock in
the gully was very unstable and liable to fracturing and breaking away in large blocks, and hence any
man-made features would not survive. The instability of the rock here also suggests, perhaps, that the
lower part of the gully in the past may not have been as steep as it is now and may once have been
more easily climbed.
231
Ill. 2.294—
Steps (red) and
handholds
(yellow)
recorded in
the lower half
of the gully
(Con Brogan,
DAHG; survey:
A.R. Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill.
2.295—Rockcut step (1) on
a narrow ledge
(A.R. Hayden).
[Right] Ill.
2.296—Flat top
of rock spine
(2) that may
have been
worked (A.R.
Hayden).
Ill. 2.297—The
finely preserved
rock-cut step
(3) (A.R.
Hayden).
Ill. 2.298—The
left-hand side
of the ledge
crossing the
gully, with a
possible step
(5) in the deep
notch on the
left and another
(6) on the flattopped rock
spine at the
centre (photo:
A.R. Hayden).
Features recorded in the lower gully
Some of the rock-cut steps in the gully were clearly evident and contain definite evidence of human
working—these are described below as definite steps. Climbing the gully revealed several locations where
there clearly should be steps. In some of these locations there are notches or hollows in the rock face
where either too little survived or no clear working was evident that would allow them to be definitely
identified as steps. These are described below as possible steps.
No evidence of any man-made
features survived in the lowest 20m of
the gully, where the rock was most
eroded. About 20m up the gully, a
narrow ledge extended across its full
width and here the first rock-cut steps
and features survived.There was a very
definite rock-cut step (1) 200mm
from the south rock wall of the gully
on this ledge.To the right of this there
was a broad rock spine, attached to the
south wall of the gully. It had a flat top
(2) that appeared to be worn and its
surface could also have been worked.
Half a metre above the northern
end of the ledge there was a large and
very definite inner end of an
exceptionally well-preserved rock-cut
step (3) that measured 200mm in
width and 150mm in depth. About
400mm above this step were the bare
remains of the triangular inner end of
a possible rock-cut step (4).
These steps show that the climb
here ascended the left-hand side of
the gully and led up to a pronounced
ledge that crossed the gully.This is the
ledge that could represent the top of
a possible alternative climb up the
232
THE EXCAVATIONS
cliff face to the left (north). It should originally have held at least one step but no definite worked areas
survived. Possible steps may be indicted by a deep, square notch (5) at the left side of the ledge, and the
flat and smooth top of a rock spine (6) that stands proud of the centre of the ledge. No definite evidence
of working survived on either of these features, however.
Above this ledge the route may have moved southward towards the right-hand side of the gully.
No definite steps survived here, but two small notches that measured 100mm in width and depth in the
steep bedrock (7 and 8) could be the last remnants of steps. They retained no evident traces of working.
The route then seems to have run back towards the northern side of the gully, where it reached
another narrow ledge. There were
two possible rock-cut steps (9 and
10) at the left side of the gully, one
just below and one on this ledge.
No. 9 measured 220mm in width
and 180mm in depth. No. 10 was
270mm wide and 100mm deep,
and its possible tread was largely
broken away.
The next group of steps lay a
short distance above this ledge,
roughly at the centre of the gully.
They were the best-preserved run
of steps in the lower part of the
gully and seven consecutive steps are evidenced,
although only the very innermost ends of the lower
four steps survived. The first four (11–14) lay in a
vertical line spaced 300–400mm apart, and were cut
into the end of a narrow vertical spine of rock
protruding slightly from the gully. Three further
steps (18–20) lay above and slightly to the left of
this line and led up and onto another ledge. Some
of the right-hand side of this ledge (21) may also
have been worked flat.
Ill. 2.299—The
ledge crossing
the gully, with
possible step 9
lying at the left
side of the
photograph
and possible
step 10 at the
centre (photo:
A.R. Hayden).
Ill. 2.300—The
inner ends of
steps 11–14 on
the narrow
rock spine at
the centre of
the gully (A.R.
Hayden).
Ill. 2.301—Steps 18–20 run from the bottom left corner of
the photograph diagonally up and onto a narrow ledge, the
right-hand side (21) of which also appears to have been
worked (A.R. Hayden).
233
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
There are two handholds (15 and
16) carved one above the other to the
right of the lower four steps close to
the right side of the gully. A small
notch (17), which could have been
another handhold or could have been
cut by a rope, lay on the rock wall of
the right side of the gully.
The next surviving step (22) lay
about 1.6m above the ledge, towards
the left side of the gully. It was
relatively well preserved and measured
130mm in width and 120mm in
depth. The route from this point
upwards hugged the left (northern)
side of the gully.
A short distance above the last step there was
a large, rectangular, natural or enhanced hole (23)
that measured 400mm in width and 500mm in
depth. There were three or four rough courses of
small, horizontally laid stones in its base. These
could have been deliberately placed but were in
too bad a condition to determine whether or not
they represent definite structure. The masonry
certainly helps the climb to the step above.
There was a fine handhold (24) on the rock
to the left of the hole (23).
Ill. 2.302—
Handholds 15
and 16 lie one
above the
other on the
sloping rock on
the left, while
the notch (17)
lies on the right
(A.R. Hayden).
Ill. 2.303—Step
22 (A.R.
Hayden).
Ill. 2.304—Rectangular hole with possible masonry (23)
(A.R. Hayden).
234
THE EXCAVATIONS
Ill. 2.305—Handhold 24 at the left side of the gully (A.R.
Hayden).
There were three very fine knobbed and notched handholds (25–27) for the right hand on a rock
spine slightly to the right just above this hole. The topmost lay immediately below the next surviving
step (28). Step 28 was 170mm wide and 300mm deep, and its tread was very worn.
[Left] Ill. 2.306—The three handholds (25–27) running diagonally up the crest of the rock at the centre of the gully. Step
28 lies at the top left corner of the photograph.
[Right] Ill. 2.307—Step 28, with the two uppermost handholds (26 and 27) visible below it (A.R. Hayden).
235
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Out on the left side of the gully
there was a possible step (29). It
appeared to be worked but lay out of
line with the other steps in the area. If
it is a step, it may indicate that the
route up this part of the gully was
realigned at some stage.
Ill. 2.308—Possible step (29) at the left
side of the gully (A.R. Hayden).
Next there were two fine, rock-cut steps (30 and 31) at the centre of the gully. No. 30 was a wellpreserved rectangular step that measured 250mm in width and 300mm in depth, while step 31 was less
well preserved and narrower, measuring 150mm in width and 280mm in depth.
Ill. 2.309—Steps 30 (left) and 31 (right)
(A.R. Hayden).
There were two fine, rock-cut handholds (32 and
33) on the rock at the left side of the gully just above
the level of step 31.
Above step 31 the route seems to have turned to
run across the gully to the right and up its centre or right
side. There are several narrow natural ledges in the
bedrock here just below the base of the wall of the upper
platform above the Needle’s Eye, which crosses and
blocks the route. The ledges provided natural steps and
were unworked.
Ill. 2.310—Looking down the gully from the upper platform
above the Needle’s Eye. The northern stairway and traverse are
visible below on the left; the large terrace behind the upper
lighthouse is visible at the centre right (A.R. Hayden).
236
THE EXCAVATIONS
Ill. 2.311—The
steps and
handholds
recorded in the
upper half of
the gully to
below the
upper traverse
(Con Brogan,
DAHG; survey:
A.R. Hayden)
The gully—upper section
The upper section of the gully was examined in 2006 and a number of rock-cut steps were noted in it.
Three small lengths of very poorly preserved possible terracing (the mini-terraces), built on narrow ledges
on the north side of the gully, were also discovered and excavated in 2006. The rock-cut steps in the
gully were recorded in 2009 but have not yet been fully surveyed.
237
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill.
2.312—The
primary route
where it
reappears
above the
upper platform
(A.R. Hayden).
The primary route in the upper gully. The route re-emerges above the upper platform outside the righthand (southern) side of the gully and runs diagonally to the right (south) up a vertical, rough and very
eroded rock face to a roughly flat rock ledge. As many as ten steps could survive here. The uppermost
steps (39–43) on the flatter surface survived the most clearly. Those (34–38) below on the vertical rock
face were very poorly preserved and it was difficult to determine which were actual steps.The steps may
have been partly removed by quarrying associated with the construction of the upper platform above
the Needle’s Eye. The quarrying left the rock face
stepped and broken. Nevertheless, a left–right
sequence does seem to be indicated by the
surviving steps. A single rough handhold (44)
survived to the left of these steps.
The route then turns back to the north and
runs up a narrow, stepped ledge back northward
towards the gully. The steps (45–51) on this ledge
[Right] Ill.
2.313—Looking
down on the
first steps
above the
upper platform
(A.R. Hayden).
[Left] Ill.
2.314—The
route above
the upper
platform to
where it turns
back into the
gully (A.R.
Hayden).
[Right] Ill.
2.315—The
steps up the
narrow ledge
where the
route turns
back into the
gully (A.R.
Hayden).
238
THE EXCAVATIONS
were generally very well preserved and clearly visible. At its northern end this line turns to the northeast around a protruding rock spur and re-enters the gully. There were two carved handholds above step
49 and three (54–56) above step 51 on the corner of the rock spur. The lowest handhold (54) was very
finely carved and well preserved.
Back in the gully the rock is very fractured and eroded. There is a natural rock ledge on the lefthand side, on which the lowest of the
mini-terraces was built (see below). Only the
bare inner side of a single step (57) appeared to
survive in this part of the gully. It lay to the right
of the base of the second of the three miniterraces.
Three more bare stumps of steps (58–60)
appeared to survive to the right of the uppermost
of the three mini-terraces.
The gully above this point cut through a
thick stratum of much harder and darkercoloured rock. Here two very fine and large steps
(63 and 64) survived above three less wellpreserved ones (60–62).
Above an eroded bedding plane the gully
continues vertically upwards for about 2m, where
the rock was very eroded and no features
appeared to survive.
The route then reappears and initially hugs
the left-hand side of the gully, where there was a
run of six poorly preserved steps and possible
steps (65–70) carved into the fractured rock. The gully turns to the right slightly above these steps and
after a short gap the route reappears, running upwards along the southern side of an upward-sloping
rock ridge, the top of which was highly fractured. One possible step (71) and two very definite broad
and well-preserved rock-cut steps (72 and 73) lead to the base of the wall of the upper traverse (see
below). The wall of this terrace was actually built on top of step 73 and completely blocks the original
route.
The possible mini-terraces. There were three short rock ledges on the north side of the gully about midway
up the climb of its upper part. They lay in a vertical line, each about 2m above that below. The ledges all
sloped slightly down to the south across their narrow axes. The remains of possible stonework survived
on all three. The stonework was very loose and extremely poorly preserved. Some of the uppermost
stones on each ledge could have fallen from above and lodged there. The lower stone on the ledges
appeared to be structural, and each length had a rough but straight outer face. So little structure survived,
however, that it was difficult to be sure whether these actually were lengths of very disturbed walling.
If these stones were walls, they were of no use in the climb up the gully; if they had been higher
originally, they may in fact have hindered it somewhat. It is difficult to be sure what they represent.They
lie under the almost vertical cliff beneath the south side of the outer terrace and would appear most
likely to have been connected with the building of that terrace. They may simply have been storage
spaces for stone that was to be lifted up to the terrace for construction or repairs.
239
Ill. 2.316—The
rock-cut
handholds
where the
route turns
back into the
gully (A.R.
Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.317b—Steps 65–73 in the gully below the upper traverse (Con
Brogan, DAHG).
Ill. 2.317a—Steps 60–64 in a harder band of rock (A.R.
Hayden).
[Left] Ill.
2.318—The
two uppermost
steps (72 and
73) beneath
the upper
traverse. The
wall of the
upper traverse,
which blocks
the primary
route, was also
clearly built on
step 73 (A.R.
Hayden).
[Right] Ill.
2.319—The
upper section
of the gully
before
excavation
(A.R. Hayden).
240
THE EXCAVATIONS
[Left] Ill.
2.320—The
three possible
mini-terraces in
the upper gully,
from the south
(A.R. Hayden).
[Right] Ill.
2.321—The
three possible
mini-terraces in
the upper gully,
from the southeast (A.R.
Hayden).
Ill. 2.322—The three possible mini-terraces in the upper gully from above and east (A.R. Hayden).
241
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The top of the gully
Re-emerging above from behind the upper traverse, the gully extends upwards to a shoulder about 3.5m
below the northern side of the top of the Peak. This short but awkward climb contains a clear run of at
least seventeen rock-cut steps. No excavation was required here as the steps were clearly evident.
Ill. 2.323—The
top of the Peak
from the west
(Con Brogan,
DAHG).
Ill. 2.324—The steps in the top of the gully
above the upper traverse (Con Brogan,
DAHG; survey: A.R. Hayden).
242
THE EXCAVATIONS
The last part of the climb and the top of the Peak
From the shoulder at the top of the gully the climb turns to the south and runs up a 2m-high sloping
and stepped rock slope to the base of a 1.5m-high vertical cliff, which marks the last part of the climb.
There are several small rock-cut steps on the north-western and western sides of this cliff, which lead up
to a broad rock ledge that slopes down to the east. The top of the Peak lies on a narrow, 1m-high rock
ridge bordering the southern side of this ledge. From the top of the Peak there are fine views east over
the monastery and the Little Skellig.
Ill. 2.325—The awkward climb up from the upper traverse
(A.R. Hayden).
[Left] Ill.
2.326—The
top of the Peak
on the left, the
rock ledge
below its
northern side
on the right
(A.R. Hayden).
[Right] Ill.
2.327—
Lighthouse
weather-vane
before its top
broke off in
2006/7 (A.R.
Hayden).
243
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The partial remains of an iron
weather-vane of lighthouse date survive on
the top of the Peak. Its pointed top broke
away recently and presently lies on the
ledge above and south-east of the upper
traverse.
Eighteenth-century accounts of
pilgrimages to the Peak mention that the
final act of penitence was a terrifying crawl
out along the vertiginous and narrow spit
(a 200mm-wide ridge that slopes markedly
down to the east) to a tall, slender stone set
upright in a rock cleft at its end. Apparently
the stone had a small cross roughly
scratched on it. It is not known when this
stone was erected but it fell and was lost in
the 1970s.
Ill. 2.328—The
stone that
formerly stood
at the end of
the spit (photo:
Horn et al.
1990, fig. 47).
Ill. 2.329—The
view eastwards
along the spit
from the top of
the South Peak,
with the
monastery,
Little Skellig and
the Kerry
coastline from
Valentia to
Waterville
visible in the
background
(A.R. Hayden).
244
THE EXCAVATIONS
2.3.5 THE
SOUTH PEAK
FINAL ROUTE TO THE
Introduction
The final route to the top of the South Peak is that discovered and published by Horn et al. (1990). The
lower part of this route is shared with that of the primary route. Its upper part begins at the stepped
ledge at the base of the cliff beneath the Needle’s Eye, where the north-west passage also begins. The
final route leads up the cliff to the north via rock-cut steps to masonry steps and a small traverse (the
lower traverse) at the base of the Needle’s Eye (a widened, almost vertical rock cleft). Exiting the top of
the Needle’s Eye, there are two masonry platforms that give access to the base of a long climb up rockcut and masonry steps to another widened rock cleft, which leads up to the ‘garden’ terrace. From this
terrace there are two routes leading upwards to the oratory terrace and the shrine terrace, one via a small
traverse (the outer passage) on the southern side of the Peak, the other via rock-cut steps and ledges
higher up. There are likewise two routes from the oratory terrace up to the next terrace (the upper
traverse): one doubles back on the second route to the oratory terrace from the ‘garden’ terrace mentioned
above, while the second leads up the cliff at the north side of the oratory terrace. The final route rejoins
the uppermost part of the primary route to the top of the Peak at the northern end of the upper traverse.
??
N
8
218m
7
200m
6
4
2
3
1
150m
0
?
5
5
10
150m
20
metres
150m
Final route
Lower part of routes
Route to outer terrace
Ill. 2.330—The final route to the Peak: (1) the lower traverse, (2) the Needle’s Eye, (3) small enclosure, (4) platforms
above the Needle’s Eye, (5) the ‘garden’ terrace, (6) the oratory and shrine terraces, (7) the upper traverse and (8) the
outer terrace.
245
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Steps and platform below cliff under the Needle’s Eye
Some traces of masonry survived
beneath a small amount of stone
(collapsed from the lower traverse)
and sea campion on the sloping
ground at the base of the vertical
rock face that gives access to the
Needle’s Eye. This area was
excavated in 2007.
The last remnants of a number
Ill. 2.331—
of heavily disturbed drystone steps
Remnants of
masonry steps
were revealed below campion cover
between the
and loose stones and silt that had
stepped ledge
washed down from the cliffs above.
and the base of
the cliff beneath
Very little definite structural
the Needle’s
masonry survived, however. All that
Eye (A.R.
remained were small and much
Hayden).
disturbed parts of the core of a
masonry stairway. Part of the
possible face of one of the steps (the
second up) may be evidenced by an
alignment of three stones. These
appeared to have slipped downslope
somewhat and were probably not in
their original position. Judging by
the height and gradient of the
slope, there would have been four or five masonry steps here originally.
A small, low and poorly preserved rectangular drystone platform stood above the masonry steps
against, and giving access to, the base of the cliff. The platform had a small extension leading west to the
base of natural steps in the cliff, which in turn led to a long, narrow ledge leading west up and across the
cliff face. This ledge was free of fallen stone and appears to have been cleared and possibly widened and
levelled by the monks. It was cleaned of growth in 2008 but no man-made features were revealed on it.
The ledge is easily traversed but does not lead anywhere.
Ill. 2.332—The small platform at the base of the cliff
beneath the climb to the Needle’s Eye before excavation
(A.R. Hayden).
246
Ill. 2.333—The small platform at the base of the cliff beneath
the climb to the Needle’s Eye after excavation (A.R.
Hayden).
THE EXCAVATIONS
The climb to the lower traverse
From the top of the small masonry platform seven rock-cut steps lead up and slightly to the west across
the bare rock of the rising cliff. There is a fine handhold in the rock on the right. Five rock-cut steps
then lead back and up to the east along a narrow natural ledge. There are two very definite and two
more possible handholds on the top of the rock above these steps. All these features were clearly visible
and no excavation was undertaken here.
[Left] Ill.
2.334—Aerial
view of the
climb up to the
lower traverse
and the
Needle’s Eye
(Con Brogan,
DAHG).
[Right] Ill.
2.335—The
climb up to the
lower traverse
and the
Needle’s Eye
from below
(A.R. Hayden).
The climb crosses a cleft and then
there are six rock-cut steps leading
straight up along the end of a
protruding rock spur.The topmost step
is little more than a notch, but the
others are fine examples. There is a
clear handhold on the west side of the
topmost step. The cliff on the east side
of these steps has also been very clearly
cut back to facilitate passage.
No excavation was undertaken in
this area, as all the features were rockcut.
Ill. 2.336—The rock-cut steps and handholds
(yellow) on the climb to the upper traverse,
from below (A.R. Hayden).
247
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.337—The rock-cut
steps and handholds
(yellow) on the climb to
the upper traverse, from
above (A.R. Hayden).
[Left] Ill.
2.338—
Handholds on
the left side of
the climb to the
upper traverse
(A.R. Hayden).
[Right] Ill.
2.339—Closeup of steps
14–18 and
handholds on
the climb up to
the lower
traverse (A.R.
Hayden).
Masonry steps and the lower traverse
The rock-cut steps give access to the base of a very well-preserved run of eight masonry steps, which
lead upwards to the south-east end of the lower traverse, a terrace constructed to allow easy passage to
the base of the Needle’s Eye, which lies near its western end.
The masonry steps were cleared of campion in the 1980s, but this had grown back by 2004 and
largely covered the stairs. In 2004 a sondage was opened up through the campion cover, which revealed
that it measured 200–350mm in depth. The campion was subsequently cleared from the steps in the
same year. All the steps lay in their original positions and were intact.
248
THE EXCAVATIONS
Ill. 2.340—Looking down on the masonry steps at the
east side of the lower traverse (A.R. Hayden).
The steps were built between the vertical
cliff and the south-eastern end of the lower
traverse. The lower traverse was c. 4m long and
1.3m wide. It was cleared of campion growth in
2004. This revealed that the base and top of its
eastern end wall, which lined the western side of
the masonry steps, was loose and partly missing.
Nevertheless, this end of the traverse survived to
almost its original height. Its western end, which
lay directly below the entrance to the Needle’s
Eye, was less well preserved. Perhaps a height of
about 400mm of its top had collapsed. The
collapse was probably largely due to the impact of
visitors jumping down onto this end of the
traverse from the lowest surviving step, which lay
about 1m above, in the base of the Needle’s Eye,
the lowest original steps having broken away long ago.
The Needle’s Eye
The route next ascends almost vertically through an artificially widened fault line—the Needle’s Eye—
that lay in the cliff at the back (north) of the lower traverse, close to its western end.The side walls of the
cleft were clearly quarried back and it appears that originally the cleft must have been only a narrow
crack, which was subsequently extensively widened to facilitate access up through it. A huge fallen stone
lies vertically lodged against the upper half of the outer side of the cleft, effectively transforming this part
of the cleft into an enclosed chimney. A long, narrow piece of stone lies horizontally trapped between
the larger stone and the outer side of the rock to either side of the cleft.
There was no need for excavation here, as the rock-cut steps were all clearly visible and free of
plant growth.
A fine run of rock-cut steps
survived in the cleft. The steps at the
very base of the Needle’s Eye were,
however, poorly preserved, but there
appear to have been at least two
narrow masonry steps in the narrow
base of the cleft, followed by another
two or three rock-cut steps at the
centre and right side. It is difficult to
be sure where these latter steps lay
originally, as there are several possible
narrow ledges in the eroded bedrock,
none of which retain evidence of
working, which could be the stumps
249
Ill. 2.341—The
lower traverse
and the
Needle’s Eye
before
excavation,
from below
(Con Brogan,
DAHG).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
of the original steps.The rest of the steps up through the Needle’s Eye were much better preserved. First,
there were two clear steps on the eastern side of the cleft, followed by one on the west and another on
the east. The steps then run up the western side of the cleft, where there is a clear line of seven steps,
with an extra and smaller step further west between the fourth and fifth steps of this run. This step is of
no help in the climb and may relate to the period when the cleft was initially quarried and worked.
There are a number of narrow ledges in the deeper and narrower southern side of the cleft. Several of
these have flat tops, which could have been deliberately worked.These, while not necessary to the climb,
we have found useful when transporting equipment through the Needle’s Eye.
[Left] Ill.
2.342—The
lower steps in
the Needle’s
Eye (A.R.
Hayden).
[Right] Ill.
2.343—Rebuilt
masonry steps
at the base of
the Needle’s
Eye (A.R.
Hayden).
[Left] Ill.
2.344—The
middle and
upper steps in
the Needle’s
Eye (A.R.
Hayden).
[Right] Ill.
2.345—Looking
down the
Needle’s Eye.
Note the
narrow stone
trapped behind
the larger fallen
one (A.R.
Hayden).
250
THE EXCAVATIONS
The cleft then narrows and there are three
more steps that occupy its full width. The final step
leads onto a drystone traverse—the lower platform
above the Needle’s Eye.
Ill. 2.346—Looking down at the uppermost steps in the
Needle’s Eye (A.R. Hayden).
Ill. 2.347—The exit from the top of the Needle’s Eye onto the lower platform. The upper platform lies at the front left,
the lower platform is to its right and the small enclosure is at the back right (A.R. Hayden).
251
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.348—Section of the Needle’s Eye (A.R. Hayden).
252
THE EXCAVATIONS
The small enclosure
A small, drystone-walled enclosure occupied a broad rock spur above and west of the top of the Needle’s
Eye. Edward Bourke excavated this structure in 2004. The small enclosure is accessed from above the
western side of the exit from the Needle’s Eye.The initial part of this climb out to it is a slightly awkward
one, and is not aided by handholds or rock-cut steps.The small enclosure was built on a flat, subrectangular
rock ledge, which was quarried or cleared down to a 40mm-thick layer of quartz, which formed its
finished internal surface. The ledge was surrounded by drystone walling on three sides; a high rising
block of bedrock delimited the eastern side. Excavation revealed that a 350mm-thick layer of campion
and small stones covered the interior. The walling survived in good condition, apart from some loss of
stone at the base of the outside of the wall at the south-west corner of the enclosure.Three water-rolled
pebbles, which must have been brought onto the island from the mainland, were found in the interior
of the structure. A stone cross-shaped slab and the stump of a candle were found in the small enclosure
in the 1980s (Horn et al. 1990, 32).
[Left] Ill.
2.349—The
small enclosure
before
excavation
(Horn et al.
1990, fig. 21).
[Right Ill.
2.350—The
small enclosure
immediately
after excavation
(E. Bourke).
Ill. 2.351—Plan of the small enclosure before
excavation (Horn et al. 1990, fig. 22).
253
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The east/west-aligned long axis of the enclosure leads the viewer to look west, where there is
nothing to see but the ocean. Indeed, on a misty day there is nothing at all to be seen from here. This
place, then, could have been a prayer or contemplation station.
The lower and upper platforms above the Needle’s Eye
On exiting the top of the Needle’s Eye, there are
two drystone-walled terraces (the lower and
upper platforms), one above the other, linked by
a short run of one masonry step and three poorly
preserved rock-cut steps. These platforms were
excavated and conserved in 2004.
The surface of the lower platform sloped
steeply down to the cliff to the north-west and
was covered by a thick layer of campion growth.
Ill. 2.352—The poorly preserved steps linking the lower
and upper platforms above the Needle’s Eye (A.R.
Hayden).
Ill. 2.353—The
upper platform
viewed from
above (A.R.
Hayden).
254
THE EXCAVATIONS
During repairs in 2009, necessitated by wind and water damage, traces of distorted paving, composed
of fine, large, flat slabs, were noted on the upper platform a few centimetres beneath its surviving surface.
A worked stone, possibly part of a cross, was also uncovered. The upper platform cuts across and blocks
the line of the primary route up the gully to the top of the Peak. The two platforms were not strictly
necessary on the climb up the final route and so could have been built either to provide storage space
for materials or as a gathering place for prayer or ritual associated with the pilgrimage to the Peak.
Ill. 2.354—Distorted paving at the back of the upper
platform (A.R. Hayden).
Ill. 2.355—
Possible cross
fragment from
the small area
excavated at
the back of the
upper platform
(A.R. Hayden).
Ill. 2.356—Looking down from the outer terrace on the
climb from the upper platform (bottom right) to the
‘garden’ terrace (upper left). The heavily quarried surface of
the slope behind is clearly visible (A.R. Hayden).
255
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The climb to the ‘garden’ terrace
From the uppermost of the two platforms above the Needle’s Eye the final route runs upwards to the
‘garden’ terrace. Initially the route climbs north-eastwards through an area that was heavily and unevenly
quarried but where there are many clear rock-cut steps leading upwards to the point where the primary
route turns back onto the gully. From this point the final route follows a steep and stepped natural cleft
or gully upwards to the south-east. There are many well-preserved rock-cut steps, some masonry steps
and rock-cut handholds on this climb. At the top of the climb, the rock cleft narrowed just before it
reached the ‘garden’ terrace. There are clear tool marks visible on its rock sides here, showing that it was
substantially widened to facilitate access.
This run of steps and the quarried rock surfaces were cleaned down in 2004.The lower section was
surveyed by EDM in 2009, but the upper part of this climb remains to be surveyed.
[Left] Ill.
2.357—The
start of the
climb from the
upper platform
to the ‘garden’
terrace (A.R.
Hayden).
[Right] Ill.
2.358—The
middle and
upper sections
of the climb;
note the
masonry steps
(A.R. Hayden).
[Above] Ill. 2.359—Looking down on the top of the
climb (the ‘garden’ terrace lies to the left); note the rockcut steps and the cut-back face of the rock cleft (A.R.
Hayden).
[Right] Ill. 2.360—Possible tool marks still visible on the
side of the widened cleft at the top of the climb (A.R.
Hayden).
256
THE EXCAVATIONS
The ‘garden’ terrace
This is the only terrace/traverse on the South Peak whose given name subjectively allocates a function.
Even though it is by no means certain that this structure was either a ‘garden’ or a ‘terrace’ (as opposed
to a traverse), the name is utilised here for continuity.8
This sinuous structure is aligned with its long axis on a north-west/south-east line. The bedding
plane of the bedrock in the area dips downwards to the south-east, which means that if the terrace had
a level top the walling at its southern end would have been the highest built here. The vertical cleavage
planes in the bedrock run east–west, roughly parallel to the short axis of the terrace. This means that the
cliff along the outer edge of the south-western side of the terrace is deeply fissured and uneven, so the
terrace wall is of sinuous shape. The base of the wall rises and falls as it crosses the gullies that mark the
ends of the unevenly eroded cleavage lines. It is only at the north-western and south-eastern ends of the
terrace that the wall was aligned parallel with the cleavage planes. There it was built on straight, sloping,
bedrock bedding planes that ran parallel to its line.
The outer face of the retaining wall of the northern half of the terrace appeared to be in good
condition and little changed from photographs taken in the 1980s. It did, however, contain some gaps and
plant growth. The wall at the southern end of the terrace was much more poorly preserved. It had partly
[Left] Ill.
2.361—The
‘garden’ terrace
in the 1980s
(Horn et al.
1990, fig. 25).
[Right] Ill.
2.362—Plan of
the ‘garden’
terrace in the
1980s (Horn et
al. 1990, fig.
26).
collapsed and much was lost by the 1980s,
when it was first recorded in detail.
The surviving top of the terrace
echoed the slope and level of the
surviving top of the outer wall. It was
roughly level in the north-eastern quarter
of the terrace, but the southern threequarters slopes downwards to the
south-east, gently at first but with
increasing steepness towards its southeastern end.The steep slope was the result
of the collapse of the outer wall at the
lower, southern end of the terrace and the
subsequent erosion of the material that
filled its interior. Examination of this end
Ill. 2.363—The
steep slope of
the southeastern end of
the terrace,
before
excavation
(A.R. Hayden).
8
As informally defined by Horn et al. (1990), a ‘traverse’ is a section of terracing built to provide access to another area, while a
‘terrace’ is a built self-contained unit, a destination in itself.
257
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
of the terrace in 2004 revealed that the erosion was continuing and that some of the outer walling that
survived in the 1980s had collapsed over the cliff edge. More of the finer material that constituted the
infill behind the wall had also been lost. The south-eastern end of the terrace existed as a steep and
unstable slope on which there were many loose stones lying at a steep angle on loose soil, and hence it
was in a very vulnerable and unstable state.
The north-eastern end of the terrace wraps around two sides of a high spur of rock that forms a
small extension, possibly at a higher level than, and almost separate from, the rest of the terrace.This may
form a small enclosure similar to that above the Needle’s Eye, which also faces in the same direction.
[Left] Ill.
2.364—The
wall at the
south-western
side of the
terrace,
gradually
lessening in
height as it
progresses
southwards
(A.R. Hayden).
[Right] Ill.
2.365—Only
the basal stones
of the wall
survived at the
south-eastern
end of the
terrace. Note
the loose and
slipped stones
ready to fall
(A.R. Hayden).
[Above] Ill. 2.366—The small outshot at the northeastern end of the terrace in the 1980s (Horn et al.
1990, fig. 24).
[Right] Ill. 2.367—The small outshot after conservation,
showing that its shape, its alignment and possibly its
purpose echo that of the small enclosure below (A.R.
Hayden).
258
THE EXCAVATIONS
The high rock outcrop was quarried down to a layer of white quartz (as was the base of the small
enclosure) that formed its finished surface.This is surely no coincidence.The importance and significance
of white quartz is well known from this and other early ecclesiastical sites. It is probable that the finished
top of the terrace here lay level with the top of the rock outcrop. It is also possible that a wall formerly
stood on top of the northern side of the rock outcrop, enclosing the outshot on three sides.
A shallow and roughly north/south-aligned test-trench was opened midway along the terrace in
2004 to examine the condition of the internal face of the terrace wall. This revealed what appeared to
be a second roughly constructed drystone wall inside the outer wall. The short length of the internal
face of the terrace wall exposed was also cleared and conserved in 2004. The terrace was excavated in
2005 to allow for the conservation and stabilisation of its southern end, to determine what the apparent
inner wall represented and to allow the inner face of the main wall to be conserved as it was in poor
condition. Scaffolding was first bolted to the cliff outside the southern end of the terrace. The lower
(southern) end of the terrace was then excavated to bedrock, as there the terrace wall appeared to have
almost totally collapsed. Much of the material filling the terrace had also been lost at this end, and the
steepness of the slope of the surviving deposits meant that loss was continuing. Subsequently the
remainder of the terrace was excavated but generally to a much shallower depth (less than 0.5m), as there
was no requirement to fully remove all the material from its interior.The interior of the narrow northern
end of the terrace, beside the high rock outcrop that marks its northern end, was also excavated to
bedrock.
The southern end of the terrace was excavated northwards until intact and stable layers of the
original infill were revealed.
The wall at the southern end of the terrace had largely collapsed but its base and more of its
easternmost end survived. The higher eastern end of the section of the wall was partly keyed into an
almost vertical natural bedrock fissure in the cliff. This crack rose at a slight inward slope to well above
the probable finished level of the terrace. It would have held and secured the full height of this end of
the outer face of the terrace wall. This keying of the wall into a fracture or behind a bedrock ridge is
repeated elsewhere on the South Peak.
The short length of wall surviving at this end was also the only place on the South Peak where the
outer face of a terrace wall was sealed behind collapsed material, as it was only here that the wall was
built set back from the cliff edge. The facing stones were still tightly jointed and their external surfaces
were well preserved and
unbroken, showing how fine the
masonry originally looked. Its
appearance contrasted sharply
with that of the rest of the
terrace wall, where the stones
were often cracked, their edges
eroded and the joints between
them larger and more open.
A 1.5m length of the wall
at the south-western corner of
the terrace had collapsed
completely and no traces of it
remained.
The area excavated here
was the only place on the Peak
where such a depth of infill
259
Ill. 2.368—The
surviving outer
walling at the
south-eastern
corner of the
terrace, looking
south, with
collapsed stone
inside it. The
wall at the
south-western
corner (right)
had gone
completely
(A.R. Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
survived on a terrace and where it was possible to see the full construction sequence in section. Any
fallen stone first appears to have been removed from the natural bedrock ledge before building work
began; the ledge seemed to have been cleared right down to bedrock. The underlying bedrock rose to
the north-east in a series of irregular steps, following natural shelves and joints. The outer face of the
terrace wall (which consisted of large, flat stones) was first built to a height of two or three rough courses,
a total of 300–400mm in height. The interior of the terrace was then infilled with dumps of large and
small stone fragments in a matrix of smaller stone chips and clay, to the level of the top of the retaining
wall. This infilled material is clearly the waste from the quarrying and breaking up or shaping of stones.
The base of this material had a thick manganese mineral pan within it. The terrace wall was next built
higher but in a wider form, now with an inner face resting on the infilled quarrying waste. Further
dumps of quarry waste were subsequently deposited behind the wall as it was built, raising the level of
the terrace to its full height.
The wall at this end of the terrace may
originally have stood to a considerable height, as the
bedrock ledge on which it was built sloped down
to this end of the terrace (see below). The
underlying bedrock slope also meant that much of
Ill. 2.369—The
the drainage on this terrace would have been
surviving
southern end
through the wall at this end, which would have
of the intact
resulted in the leaching out of the infilled material
wall at the
behind the wall and the degradation of the
western side of
the terrace.
stonework by fluctuating moisture levels. The
Note that the
compaction over time of the infill behind the wall
inner face rests
would also have destabilised it. The poor quality of
on infilled
material while
the inner face and its founding at least partly on soil
the outer face
rather than bedrock may also have contributed to
rests on
its weakness. In addition, the outer face of the wall
bedrock (A.R.
Hayden).
was footed on a smooth bedrock ledge that sloped
downwards towards the outer side of the wall and
hence provided limited grip for its base. It is
noticeable that the section that survived best was
that keyed into the cleft in the bedrock at the southeastern corner of the terrace.
It is not clear here whether the wall initially collapsed outwards or, as evidenced elsewhere, slumped
inwards to a point where it lost its stability, collapsed and was lost over the cliff.
The inner face of the wall over the rest of the terrace was also in general composed of much smaller
stones than those used to form its outer face, probably an indication of the shortage of large stones
available. It was also in generally poor condition. This was most apparent at its northern end. There the
standing wall was also wider than elsewhere, except at the southernmost end of the terrace. The wall
here had partly slumped inwards but initially appeared to be in good condition. When the stones were
examined in detail, however, it became clear that they were all cracked and fractured, many were little
more than dust and the wall retained little structural integrity. A combination of the wetting and drying
of the adjacent material that filled the terrace, the exposure of much of the outer face of the wall to
constant sunshine during the daytime and movement of the wall caused by the leaching out of the finer
deposits filling the terrace may have been responsible for its deterioration.
260
THE EXCAVATIONS
[Left] Ill. 2.370—The inward slumping of the wall at the
northern end of the terrace, with the reinforcing wall at its
base (A.R. Hayden).
[Above] Ill. 2.371—The degraded state of the stonework of
the inner face of the wall at the northern end of the terrace
(A.R. Hayden).
Up to Upper Traverse
& Oratory Terrace
Up from
Needle’s Eye
A
N
B
black
layer
topsoil
quarry
debris
Enclosure(?)
topsoil
Reinforcing
wall
wall
drystone rubble
clay &
rubble
Section
Test
Trench
B
To Oratory Terrace
via Outer Passage
Sloping
wall face
Top of
terrace wall
A
The excavation of the remainder of the terrace revealed that the inner wall noted in 2004 was built
to reinforce the main wall of the terrace and evidenced an attempt to prevent it from slumping inwards.
A trench, varying from 700mm to 850mm in width, was first dug inside the line of the terrace wall.
Flat stones were next set upright and flat against the inner face of the terrace wall. Then a drystone wall,
up to 450mm wide, was built in the outer side of the trench, tight up against the flat stones and the inner
face of the main wall. This wall was largely composed of small stones but the occasional larger one was
261
Ill. 2.372—Plan
and section of
the ‘garden’
terrace.
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.373—The
reinforcing wall
viewed from
the south-east
(A.R. Hayden).
Ill. 2.374—The
reinforcing wall
viewed from
above (A.R.
Hayden).
also used. It had a reasonably good face on its inner side but was much rougher on the side that lay
against the inside of the terrace wall. Larger and thicker and more block-like stones were then tightly
rammed down between the drystone wall and the inner edge of the trench. A large number of these
stones consisted of roughly rectangular quartzite blocks. The use of quartz blocks may be of no more
significance than that they were the only available type of stone that came in thick blocks—the sandstone
that makes up most of the peak generally breaks into thinner slabs. Thin flat stones were also hammered
vertically down between the stonework to further tighten the structure.
In the narrow turn at the northern end of the terrace the reinforcing wall filled the whole interior
of the terrace between the outer wall and the high bedrock outcrop to its north-west. The reinforcing
wall survived to its greatest height at the northern end of the terrace and gradually became lower and
more ruinous as it extended southwards. Over the southern third of the terrace the feature survived only
as a trench filled with loose soil and stones, and it did not survive at all at the southern end of the terrace.
This form of buttressing finds a parallel in the nearby early monastic site on Church Island in Valentia
Harbour, excavated by O’Kelly in the 1950s. There the inward slumping of part of the inner face of the
cashel wall was buttressed by
hammering flat stones vertically
downwards, with their broadest
faces against the inner face of the
wall.
None of the original
finished surface level of the
terrace survived. Rock-cut steps
survive in the bedrock at the
inner side of the southern end of
the terrace leading up and over
to the oratory terrace via the
outer passage (see below). The
lowest of these steps lies well
above the present top of the
southern end of the terrace but
is roughly level with its surviving
northern end.This could suggest
that originally the terrace had a
level top (which is to be
expected) and therefore the
surviving southern end now lies
well below its finished level.This
would also suggest that the wall
at the southern end of the
terrace would originally have
been about 3m in height.
It is possible, however, that
the whole terrace could
originally have been about 1m
higher. It is likely that the
surface of the out-turn of the
northern end of the terrace
262
THE EXCAVATIONS
would originally have been level with the flat-topped bedrock outcrop there, and so was originally much
higher than evidenced today. It is not clear whether this higher area was originally elevated above the
rest of the terrace or whether the whole terrace was built to this level.
The rock-cut steps at the southern end of the terrace could, however, pre-date the construction of
the ‘garden’ terrace, and hence the height of the lowest step may be of irrelevance.The possible burial of
the lowest part of this route would have been of no consequence as it would still have been usable
subsequently.
The digging of the wide foundation trench for the reinforcing wall meant that when it was being
built the usable width of the terrace varied from only 0.65m at its narrowest to 2.15m at its widest
(reduced from the 1.5–3m width defined by the original wall). Horn et al. (1990, 67–9) speculated that
a building could once have stood on the wider southern end of the terrace, but the erection of the
reinforcing wall suggests that this was highly unlikely.
The southern end of the ‘garden’ terrace gives access to the route that runs via the outer passage to
the oratory terrace, and hence the terrace may in fact have been a traverse.The presence of this routeway
(if it was contemporary with the use of the ‘garden’ terrace) and the reinforcing wall also means that it
was unlikely to have been a garden, at least in the later stages of its use.
The two routes to the oratory terrace
Introduction
There were two distinct routes linking the ‘garden’ terrace with the next terrace above—the oratory
terrace. One route led from the south-eastern end of the ‘garden’ terrace up and around the cliff at the
southern side of the Peak via a short length of terracing—the outer passage—to the south-west corner
of the oratory terrace. The second route ran from the north-western end of the ‘garden’ terrace up the
cliff and then turned to the south and east around a corner, following partly quarried and partly natural
ledges to the north-west corner of the oratory terrace.
The route via the outer passage
A number of poorly preserved rock-cut steps
survive on the 2m-high cliff at the inner side of
the ‘garden’ terrace, close to its south-eastern end.
They mark the start of the first route to the
oratory terrace. There is also a handhold on the
higher rock to the left.
Ill. 2.375—The short climb up to the outer passage from
the south-eastern end of the ‘garden’ terrace (A.R.
Hayden).
263
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.376—Looking down on
the rock-cut steps leading up
to the outer passage (A.R.
Hayden).
[Left] Ill.
2.377—The
location of the
rock-cut
handhold on
the climb to the
outer passage
(A.R. Hayden).
[Right] Ill.
2.378—Closeup of the
handhold (A.R.
Hayden).
The outer passage
The steps lead up to a short length of terracing—the outer passage—on the top of the cliff on the south
side of the Peak. The outer passage consists of walling erected on a narrow rock ledge. A large stone was
laid flat on top of drystone walling to form the surface of the pathway. The infill behind the walling had
washed out and as a result the inner side of the large stone slumped downwards.
There is a quarried ledge that retains clear tool marks outside and below the base of the wall of the
outer passage. It would have provided a foothold on the outer side of the wall when it was being built.
A worked ledge, also retaining tool marks, survives in a similar location outside the southern wall of the
oratory terrace (see below).
A number of rock-cut steps lead down from the eastern end of the outer passage to the top of the
264
THE EXCAVATIONS
outer wall of the oratory terrace. There must originally have been one or two masonry steps (which did
not survive) here on the top of the western end of the wall of the oratory terrace. This suggests that this
route to the oratory terrace was in use after the oratory terrace was built and therefore was not just a
route used during the construction phase.
No excavation or survey was undertaken on the outer passage, as conservation involved simply
rebuilding the areas of missing walling to support the large stone.
[Left] Ill.
2.379—The
outer passage
from the west
(A.R. Hayden).
[Right] Ill.
2.380—The
outer passage
from the east.
Note the
worked ledge
on the cliff
below the
masonry (A.R.
Hayden).
[Left] Ill. 2.381—The junction of the outer passage and the oratory terrace from the east before conservation (A.R.
Hayden).
[Right] Ill. 2.382—The junction of the outer passage and the oratory terrace from the east after conservation (A.R.
Hayden).
265
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The second route to the oratory terrace
The second route to the oratory terrace starts from the opposite end of the ‘garden’ terrace, just above
the point where the route up from the Needle’s Eye enters the terrace. The ‘garden’ terrace was not
necessary to this route, as this climb does not involve actually setting foot on the terrace.The route leads
up a steep rock slope where there are a large number of well-preserved rock-cut steps and handholds.
One of the lower steps has largely broken away and makes the start of the climb somewhat awkward.
At the top of the cliff there was a broad, flat rock ledge extending to the left and the right. Turning
left brings the climber towards the upper traverse. Turning right and following partly natural and partly
[Left] Ill.
2.383—The
rock-cut steps
and handhold
on the climb up
from the west
end of the
‘garden’ terrace
(A.R. Hayden).
[Right] Ill.
2.384—Closeup of handhold
(A.R. Hayden).
[Above] Ill. 2.385—The wide sloping rock ledge west of
the oratory terrace; the route follows its inner edge
(A.R. Hayden).
[Right] Ill. 2.386—The ledge and steps leading down to
the northern side of the oratory terrace (A.R. Hayden).
266
THE EXCAVATIONS
quarried rock ledges leads to a wide rock ledge, which slopes down to the east. This ledge was at least
cleared of fallen stones and may have been partly quarried. It has a layer of quartz on its surface.The cliff at
the lower end of this ledge lies on a north/south-aligned fault, which continues into the cliff on the left
(north) as a gully. East of the gully there is a narrow rock ledge on which several steps were cut. The steps
lead east and down along the 1.5m-high cliff at the northern side of the oratory terrace to just beyond the
north-western corner of the oratory.There they turn to the west, where there are masonry steps formed in
the wall of the north-western corner of the oratory leading down to the surface of the terrace.
No excavation or survey was undertaken on this route. All the features were rock-cut and free of
plant growth.
The oratory terrace
Introduction
The oratory terrace occupies a broad rock
ledge on the steep cliffs on the southern
side of the Peak. Major north/southextending faults mark the eastern and
western ends of the terrace.The rising cliff
at the western end is vertical and even,
while the descending cliff at the eastern end
of the terrace is deeply fissured and drops
away steeply. A high cliff that extends up to
the top of the Peak delimits the northern
side of the terrace. A minor fault runs on
the line of the western wall of the oratory
and influenced the shape and line of the
southern side of the terrace. The bedding
plane of the bedrock on this terrace is
roughly flat on the north–south axis but
dips downwards to the east, while the
vertical cleavage planes in the rock run on
east–west lines. These two sets of planes
formed narrow flat ledges on the cliff on
the southern side of the terrace.
The walls defining the
southern side of the terrace
survived poorly, and only two
short sections of their outer
faces were visible since at least
the 1980s. They were both
partly collapsed and the
stonework was gapped and
loose. The outer edge of the
terrace sloped down to the cliff
edge and there had evidently
been much loss of walling and
infilled material along all but
the westernmost end of the
Ill. 2.387—The
oratory terrace
before
excavation,
viewed from
the west. The
route from the
western end of
the ‘garden’
terrace runs
down the rock
at the left side
of the picture
(A. R. Hayden).
Ill. 2.388—The
oratory in the
1980s (Horn et
al. 2002, fig.
31).
267
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
terrace. Loose stones lay on the
slope and it was clear that
collapse here was ongoing. No
traces of the eastern wall of the
terrace were visible before
excavation.
The remains of the north
and west walls of a small
oratory built at the eastern end
of the terrace survived and
were clearly visible before
excavation. The north wall of
the oratory survived to about
1.5m in height, but less than a
0.5m height of the west wall
survived. The doorway in the
west wall was clearly evident
and a large slab, probably its
lintel, lay inside it. No traces of
the east or south walls of the
building were visible. The area
of the oratory was partly
cleared of campion growth in
the 1980s but it had regrown
over and covered it again by
2005. A 300mm-thick layer of
campion growth also covered
the rest of the terrace.
West of the oratory,
some paving composed of
large stones was visible in the 1980s but was again
covered by campion in 2005.The top of a rectangular
drystone leacht was partly cleared of campion growth
and revealed west of the oratory in the 1980s and was
visible standing proud of the ground in 2005. Part of
a fine slate slab inscribed with a ringed Latin cross was
found between the leacht and the oratory in the 1980s
and was removed to the OPW depot in Killarney.
Two rectangular inter-linked water cisterns were
cut into the rock ledge on the inner side of the terrace
just west of the north-western corner of the oratory.
They have previously been described in detail (Horn
et al. 1990, 45–6). The cisterns are fed by water
trickling down a series of small grooves carved in the
cliff face to their north. Some of these grooves run
behind the north wall of the oratory, and so it is clear
that the water collection system was created before
Ill. 2.389—The
oratory from
the west
before
excavation.
Note the large
flat stone in
the doorway,
possibly the
lintel. Water
basins visible
at left (A.R.
Hayden).
Ill. 2.390—
The rock-cut
water
cisterns; note
the channels
carved on the
rock face (A.
R. Hayden).
Ill. 2.391—The
rock-cut water
cisterns, from
above (A. R.
Hayden).
268
metres
the oratory was erected. It is likely that the cisterns had some form of cover or lid, as they lie on the
lower end of the route down to the terrace and have often proved to be a water trap for unwary feet!
The grooves on the rock face were clearly carved before the oratory was erected, as they run behind and
beneath its north wall.
The instability of some of the structures surviving on the oratory terrace required that they be
269
Ill. 2.392—Plan of the features on the oratory terrace and the shrine terrace.
Hauling platform
Worked ledges
from ‘garden terrace’
via outer passage
West wall
from ‘garden terrace’
Leacht
rb
Wate
asins
Cross slab socket
Oratory
Altar
to upper traverse
N
Carved crosses
0
Slab Shrine
Bridging slab
Cuts for anchoring
wall base
5
THE EXCAVATIONS
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.393—The
western end of
the oratory
terrace. Outer
wall visible at
top, leacht and
paving in
foreground. The
two paving
slabs on the
right are
modern (A.R.
Hayden).
Ill. 2.394—The
western end of
the eastern
section of outer
walling, with the
western section
above in the
background.
The small
hauling
platform lies at
the left (A.R.
Hayden).
conserved as the excavation progressed, and so the area was excavated and conserved on a phased basis,
progressing from west to east.
The instability and the loose nature of the outer wall of the terrace required excavation to facilitate
its conservation. Scaffolding was first bolted to the cliff just below the base of the walling at the southern
side of the terrace and the area of the wall was excavated in 2005. This revealed that the upper part of
the wall only survived at the westernmost end of the terrace. Over the remainder of its line the top of
the wall had fallen over the cliff and the material infilled behind it had also washed away. This severely
undermined the base of the southern side of the leacht.The surface of the terrace west of the oratory was
excavated to allow the leacht to be conserved in 2005. After underpinning and consolidation, the leacht
was enclosed in a timber frame composed of scaffold planks for protection.
The oratory was excavated later in 2005. This revealed its southern and eastern walls and a small
altar, a vertical cross-slab and paving
inside the east end of the building.
These parts of the oratory were,
however, very poorly preserved.The
outer face of the south wall of the
oratory had completely collapsed,
while the east end of the structure
slumped alarmingly downwards to
the east. The slumping and
movement proved to have been
caused by the slippage of a large
stone, which underpinned the east
end of the oratory. The stone had
been placed across the top of a deep
gully but one end had slipped down
into the gully, causing the collapse
of the oratory. The large stone lay at a precarious
angle and had minimum purchase on a quarterinch-wide ledge on the northern side of the gully.
In 2006, after the completion of the major
conservation works on the oratory terrace, the
interior of the leacht was excavated.
Outer walling
The walls defining the southern side of the oratory
terrace survived to varying degrees. At the southwestern corner of the terrace the wall stood to
almost full height, but its base had partly collapsed
where it bridged a deep fissure that marks the major
fault line at the western end of the terrace.This part
of the terrace wall runs on a straight line as far as
the minor fault just west of the oratory.The eastern
end of this section of walling had totally collapsed.
The wall was footed on a narrow bedrock ledge
following a bedding lane, which sloped down to the
east.
270
THE EXCAVATIONS
East of the minor fault,
the rock is twisted slightly
anticlockwise, and so the
cleavage planes and the cliff
edge run slightly more
northwards; hence the wall
defining this half of the
terrace runs also at a slightly
different angle. Here the cliff
also extends further outwards.
The outer walling here was
footed on a lower ledge than
that to the west and followed
a sloping bedding plane, and
so the base of the wall lies
further to the south and at a
Ill. 2.395—The
eastern section
of outer walling,
with the base
of the southeast corner wall
visible in the
background
(A.R. Hayden).
lower level than the western section. The more
easterly wall only survived in small lengths, but a
clear corner was uncovered at its eastern end,
marking the south-eastern corner of the terrace. A
notch was cut into the bedrock on which the wall
was built, to counteract the tendency for the wall to
move downslope (east).There was also a larger notch
cut into the bedrock higher up and well inside the
line of the western end of this section of the wall.
This may have functioned in a similar fashion to the
other mentioned but no walling survived here.
Ill. 2.397—The
hauling
platform. Note
hollow cut in
rock ledge to
secure its base
(A.R. Hayden).
Ill. 2.396—Hollow cut on rock ledge to secure
foundations at east end of eastern section of outer
walling (A.R. Hayden).
The junction between the two lengths of
walling did not survive. As both the sections of wall
ran on different lines and the bases lay at different
heights, it is not clear how this junction was
handled.
This eastern length of walling was not
necessary to support the south wall of the oratory,
as it is clear that its south wall rested on a bedrock
ridge at a higher level (see below). Coupled with
271
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
the evidence of access around the east end of
the oratory afforded by a large bridging stone
(see below), it is clear that this wall was built
purely to allow access around the south side of
the oratory. Circumambulation of the oratory
therefore appears to have been an important
consideration for the builders of the terrace.The
narrowness, slope and smoothness of the ledges
on which the terrace walls were built did not
provide much of a grip for the base of the walls,
and so they would have been vulnerable to
slippage and collapse.
Apart from the southern corner, none of
the eastern wall of the terrace survived.This was
due to the catastrophic collapse caused by the
movement of a very large stone under the
oratory, on which the wall would have been
footed. There could have been a narrow wall
here resting on the outer side of the then
collapsed stone, leaving enough room for a
narrow passage around the back of the oratory.
Ill. 2.398—The
worked ledges
on the cliff
below the
southern wall
of the oratory
terrace (A.R.
Hayden).
Hauling platform
A small, isolated, rectangular
block of drystone masonry
stood on a ledge at a lower
level below the western end
of the eastern section of
walling. It was not
connected with the terrace
walls and may have been a
constructional
feature,
possibly used to haul rock
from
the
extensively
quarried ledges beneath the
oratory terrace (see below).
Ill. 2.399—Detail of the
tool marks still visible on
the worked ledges
below the southern wall
of the oratory terrace
(A.R. Hayden).
Worked ledges below the southern wall
A number of narrow sloping ledges below the outer wall of the terrace retained evident tool marks.These
ledges (like that on the outer passage) would have provided a foothold during construction.
Terrace infill
Before excavation the west end of the terrace was covered by a thickness of 350–400mm of sea campion
and its rooted soil, while the paving at the centre of the terrace was devoid of any cover and was fully
exposed.
The rough bedrock surface of the interior of the terrace was levelled with quarrying debris,
composed of small stone chippings and dusty clay, before the paving and leacht were constructed. Bedrock
272
THE EXCAVATIONS
lay directly beneath the surface of the most of the terrace, but along its southern half, where the rock
sloped away steeply, a greater degree of filling was necessary to provide a level surface.The infilled material
here would originally have measured up to about 400mm in thickness but no more than a depth of
about 200mm survived.
Two small patches of oxidised clay were encountered in the infilled material at the western end of
the terrace; similar material was also present on the rock surface beneath the centre of the leacht. The
oxidised clay was sampled but on analysis did not contain any charcoal or organic material and so was
undatable. It is likely to represent small cooking fires used by the monks during the construction of the
structures. The burning was neither intensive enough nor widespread enough to suggest that fire was
used to aid the removal of rock from the terrace.
A low, flat rock ledge was left after the clearance of the terrace along its northern side at the base
of the rising cliff. This extended almost the full length of the terrace and also ran beneath the oratory
(see below). The rock-cut cisterns were cut into this rock ledge.
The westernmost end of the ledge in the north-west corner of the terrace was either cut away or
was naturally cut out, forming a depression 220mm deep. Apart from the quarry waste used to infill the
lower-lying areas of the terrace, no stratigraphy survived apart from a layer of modern topsoil and campion
roots.
[Above] Ill. 2.400—The gully under the eastern end of the
oratory terrace viewed from the east (A.R. Hayden).
[Right] Ill. 2.401—The collapsed upper bridging stone over the
gully viewed from the north-east before excavation (A.R.
Hayden).
The bridging stones beneath the oratory
The eastern end of the ledge on which the oratory was built was ragged and deeply fissured along the
lines of the exposed cleavage planes in the bedrock. The triangular inner end of a large and deep gully
ran under almost the full length (east–west) of the centre of the oratory. The top of this gully had to be
bridged to allow building here. Several stones were placed across the top of the gully, resting on spines
in the centre of the gully and on the rock on its northern and southern sides. Water washing down the
gully in the rising cliff north of the stone appears to have eroded and ‘rotted’ both the bedrock and the
northern end of the uppermost and largest bridging stone. This caused its northern end to fracture off,
and this end of the stone slipped downwards into the gully, lodging on a very narrow ledge, less than
25mm in width, on the northern face of the gully.The movement of this stone would have been sudden
and would have caused the catastrophic collapse of most of the oratory.
273
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill. 2.402—The bridging stones beneath the oratory viewed from the north (A.R. Hayden).
[Right] Ill. 2.403—The bridging stones beneath the oratory viewed from the west (A.R. Hayden).
The larger and upper bridging stone lay much further to the east than was necessary to support the
east wall of the oratory. It therefore appears to have been laid down to allow for access around the east
side of the oratory.The width allowed here would have made it possible to build a parapet wall along the
eastern edge of the stone.
There was a small hollow at the centre of the upper surface of the bridging stone. A narrow drain
was carved from it leading to the eastern edge of the stone, where there was a deeper notch carved on
the edge of the stone to allow water to drain away efficiently.
The movement of this large stone caused the collapse of both the oratory and the western end of
the walling of the shrine terrace. The collapse of the wall of the shrine terrace in turn undermined the
paving on it and the wall around the shrine itself, causing both to slump and collapse (see below).
Ills 2.404 and 2.405—The upper bridging stone replaced, showing that there was ample room for access around the
east end of the oratory, even though the stone was replaced further to the west than its original position (A.R.
Hayden).
274
THE EXCAVATIONS
The oratory
The northern wall and the northern two-thirds of the west wall, which includes the doorway of the
oratory, were visible before excavation. The interior of the oratory appeared to have been cleared out
previously, as the 400mm thickness of campion-rooted soil in its interior contained very little stone: there
should have been a much greater quantity from the collapse of the structure. The realignment of the
paving in front of the altar in the east of the oratory after the altar moved significantly (see below) also
suggested that the oratory had been cleared out for reuse after its partial collapse. Removal of modern
topsoil and the small amount of collapsed stone revealed the base of the southern and eastern walls of
the oratory, allowing its full plan to be reconstructed.The building was almost square in plan, measuring
2.32m in length and 1.9–2.28m in width.
Ill. 2.406—The oratory terrace
after the completion of the
excavation and conservation of its
western end and before the
excavation of the oratory (A.R.
Hayden).
Ill. 2.407—The interior of the
oratory after excavation, from the
south-west (A.R. Hayden).
The corners and the outer face of the southern wall of the structure had collapsed entirely and did
not survive. A bedrock ledge, on which the southern wall would have been footed, was uncovered at a
lower level than the interior of the oratory, however, and clearly marks the original line of the outer face
of the wall.Three courses of stonework survived of the wall’s inner face.While they were roughly in situ,
275
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.408—The
slope and
loose stone
beneath the
line of the
southern side
of the south
wall of the
oratory (A.R.
Hayden).
the individual stones had all
slumped dramatically down to
the south and lay at an
approximate 30º slope. There
was a clear channel inside the
face of the wall from where
the stones had slipped,
marking the original line of
the inner face.
Ill. 2.409—The surviving northern
face of the south wall of the
oratory. Note how stones have
slumped outwards. Altar in left
foreground (A.R. Hayden).
Ill. 2.410—The southern end of
the inner face of the east wall.
Note how it has twisted and
moved (A.R. Hayden).
276
THE EXCAVATIONS
Ill. 2.411—The east wall of the
oratory, showing its partly intact
inner face and possible remnants
of its slipped outer face (A.R.
Hayden).
The full inner face of the east wall survived, but its southern end was deformed and had moved
eastwards. A possible short length of the original outer face of the northern end of the wall survived but
it was not in situ, having slipped eastwards.The east wall had split and moved substantially outwards at its
centre and downwards into the underlying void.
The base of the north wall of the oratory consisted at its western end of a ridge of bedrock that
stopped abruptly about midway along the wall, where it was replaced by large flat slabs set upright on
edge, giving the impression of almost cyclopean masonry.There were a number of grooves carved in the
top of the bedrock supporting the wall. They seem to have been designed to channel water down the
rock face to a shallow and narrow drain below. The drain ran along the inside of the north and west
walls, exiting through the south wall. The drain was in reality little more than a shallow scooped gully, a
few centimetres deep, covered with a few small flat stones.
[Left] Ill. 2.412—The drain and rock ledge inside the north
wall of the oratory (A.R. Hayden).
[Below] Ill. 2.413—Base of the north wall of the oratory.
The upright rock on the left is bedrock; two upright stones
support the wall on the right. Note the grooves in the top
of the bedrock, which channel water down to the drain
(A.R. Hayden).
277
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The aforementioned bedrock ledge left along the northern edge of the terrace lay just below the floor
of the north side of the oratory, while the remainder of the interior was covered with infilled clay, stones
and gravel. The only surviving original flooring consisted of the two large threshold slabs in the western
doorway; over the remainder the surviving ground level lay beneath the original floor of the building.
Altar
The collapsed remains of a small altar survived against the inner face of the east wall of the oratory. The
altar was literally sucked downwards into the void created by the collapse of the underlying footing
stones. It survived to a maximum of three courses in height and to a trapezoidal shape, as its eastern
corners had been pulled down and inwards. It had also rotated clockwise, its north-west corner lying
some 100mm east of its original position. Originally the altar appears to have measured approximately
900mm east–west by 960mm north–south. Its front was vertical but the base of the sides sloped outwards
noticeably. The sloping sides would originally have been partly hidden beneath floor level.
The centre of the altar had been completely sucked downwards and only a hole choked with
collapsed stone survived. Indeed, so great had been the collapse that one of the southern side stones lay
turned through 90º in the horizontal and 90º in the vertical, so that it was pointing directly downwards.
Ill. 2.414—The altar, stump of
upright cross-slab and kneeling
stone, from the west; collapsed
inner face of east wall of oratory
visible behind (A.R. Hayden).
Ill. 2.415—The altar from the
north-east, showing the dramatic
slumpage (A.R. Hayden).
278
THE EXCAVATIONS
Kneeler, paving and cross-slab
The base of a vertically set stone, which is likely to have been a cross-shaped or cross-inscribed slab,
survived in front of the altar, anchored in place by smaller stones hammered vertically around it.
There was a small area of re-laid paving in front of the altar. A large, flat, slate kneeling slab lay in
front of the original position of the altar. This stone was not in its original position, however, as it had a
man-made notch at the centre of its west side, which clearly corresponded to the surviving base of the
small slab set vertically in front of the centre of the altar. The stone had clearly been turned around and
re-laid at some time.There were eighteen very finely scratched simple Latin crosses on the upper face of
the paving slab. The fineness of the scratches suggests that they were made with a metal blade, and their
survival on the exposed upper face suggests that they could be of relatively recent origin. A number of
small slabs were inserted in the gap between the kneeling stone and the altar, presumably after the kneeling
slab had been realigned but also after the altar had moved eastwards by about 150mm. This suggests that
these slabs were laid down after the collapse of the oratory and, coupled with the small amount of
collapsed stone in the interior of the oratory, indicates that the structure was cleared out after it had
collapsed so that the altar could be reused. It is not clear when this happened but it could have been in
post-medieval times, when regular pilgrimage to the site was still popular.
Ill. 2.416—The finely incised
crosses on the kneeling
stone marked by chalk
(A.R. Hayden).
Main terrace paving
The terrace west of the oratory was originally paved with flat slabs. The paving immediately west of the
oratory was visible before excavation, while the western end of the paving underlay a 300–400mm
thickness of sea campion. The surface of this area of the terrace originally consisted of three broad steps
leading upward to the west, with the leacht and the curving west wall (see below) located on the
uppermost level. Paving survived outside the western end of the oratory at the same level as the threshold
slabs in the doorway. Bedrock ridges lay unevenly directly beneath these slabs, with the hollows infilled
with quarry waste.The paving was stepped upwards twice as it extended westwards.The edge of the first
step did not survive, but the edge of the uppermost step did and it lay on the line of the eastern end of
the leacht.
Two paving slabs overlay topsoil that in turn lay on the paving slabs at the north-western side of
the leacht. These two slabs were therefore placed here relatively recently. They may have been laid down
to give a firm surface to the damp north-western corner of the terrace, as water and silt gathered here
after being washed down the gully at this corner of the terrace. It is possible to jump down to this point
from the higher rock above, where the route crosses the gully that marks the beginning of the oratory
terrace. This was the most obvious way to get down to the oratory terrace in recent times.
279
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.417—The western
end of the oratory terrace
during excavation. The two
rearmost paving stones to
the right of the leacht are
modern. Note steps up in
the original paving (A.R.
Hayden).
Ill. 2.418—The western
end of the oratory terrace
after the completion of
excavation (A.R. Hayden).
Wall bench
A low, narrow drystone wall was revealed after the removal of the sea campion cover at the western end
of the terrace. The wall ran on a gentle curve, reflecting the line of the higher cliff that delimited the
western end of the terrace. Its centre was disturbed probably in relatively recent times, as it contained
only loose topsoil and campion. It is not clear whether the wall extended right into the north-western
corner of the terrace, as both natural erosion (caused by water running down the fault line) and human
erosion had greatly disturbed this area. The wall may not have stood to much more than its uncovered
height (320mm) and would most likely have been used either as a bench on which weary pilgrims could
rest or as a shelf on which to display devotional objects etc.
280
THE EXCAVATIONS
[Left] Ill. 2.419—The wall bench and leacht at the western
end of the oratory terrace, from the north-east (A.R.
Hayden).
[Below] Ill. 2.420—The wall bench and leacht at the
western end of the oratory terrace, from above and west
(A.R. Hayden).
The leacht
The leacht was located with its eastern end on a line with the step down off the highest level of the
paving at the western end of the terrace.
The drystone-walled leacht, which measured 1.56m east–west by 1.14m north–south and stood to
just over 400mm in height, was built in a 100–150mm-deep trench cut through the infilled material of
the terrace down to bedrock. The construction trench was only revealed at the western end and the
western end of the northern side of the leacht owing to its erosion on the southern side and the presence
of paving which was not removed on its northern and eastern sides.
[Left] Ill.
2.421—The
leacht viewed
from the northeast (A.R.
Hayden).
[Right] Ill.
2.422—The
south face of
the leacht (A.R.
Hayden).
The southern side of the leacht was the most poorly preserved and the wall at the centre of this side
had partly collapsed. The two ends of this side had also moved outwards. The undermining of this side
of the leacht resulted from the outwash of the terrace infill beneath it after the outer wall at the southern
side of the terrace had collapsed. The northern and western sides of the leacht were in good condition
but many of the facing stones of its north-western quarter were missing, and this corner of the structure
only survived to about 200mm in height. There were many fist-sized and larger lumps of quartz on the
surface of the leacht and on the ground around it.
281
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
After the conservation of its outer walls, the quartz on top of the leacht was removed and an
irregularly shaped area (measuring a maximum of 1.2m by 1m) of the interior was excavated in 2006.
The facing stones of the leacht were left undisturbed. The interior, which contained loose soil and only
a few stones (including several pieces of quartz), proved to have been totally disturbed by two animal or
bird burrows. A number of bird bones and parts of a dead puffin were found in the base of one of the
burrows. A small area of burnt clay was revealed on the bedrock under the leacht; analysis found that it
did not contain organic material, however, and so it was undatable.
A few small, roughly carved stone crosses were found in the collapsed stone and topsoil around the
leacht.
[Left] Ill.
2.423—The
leacht after
conservation
and before
excavation of
its interior, from
the west (A.R.
Hayden).
[Right] Ill.
2.424—The
interior of the
leacht after
excavation
(A.R. Hayden).
Cross-slab and setting
A flat stone was driven vertically into and across
one of the natural crevices in the bedrock just to
the east of the leacht.This setting is similar in form
to that used to hold the small vertical slab erected
in front of the altar in the oratory. It is therefore
possible, and indeed probable from its location,
that this is the setting for the fine ninth-century
cross-inscribed slab that was previously found
lying prone just to the north-east of this point
(Horn et al. 1990, 45).
Ill. 2.425—Stone jammed upright into and across a
natural bedrock furrow (to left of uppermost end of
top red segment of ranging rod), probably part of
setting to hold the upright cross-slab just east of the
leacht (A.R. Hayden).
282
THE EXCAVATIONS
[Above] Ill. 2.426—The cross-inscribed slab found on
the oratory terrace in the 1980s (Horn et al. 1990, fig.
36).
[Right] Ill. 2.427—Reconstruction drawing of the original
slab (Horn et al. 1990, fig. 59).
Carved crosses
Three crosses carved into the bedrock cliff on the northern side of the oratory were discovered and
photographed in 2008. One, which is a finely pocked simple Greek cross, lies only a few inches above
the pathway where it reaches the north-western corner of the oratory. The two others, which lie higher
up on the same rock face close to its end above the eastern end of the oratory, are larger incised crosses.
They are both very faint and hard to see. There could well be more crosses carved on the cliffs here, as
the rock faces have not yet been examined in detail.
[Left] Ill. 2.428—The location of the small pecked cross on the cliff behind the northern side of the oratory (A.R.
Hayden).
[Right] Ill. 2.429—Detail of the small pecked cross (A.R. Hayden).
283
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.430 (above)—The location of the two larger incised
crosses on the cliff behind the north-eastern corner of
the oratory (A.R. Hayden).
Ills 2.431(above right) and 2.432 (right)—Details of the
two larger incised crosses on the cliff behind the northeastern corner of the oratory (A.R. Hayden).
Ill. 2.433—The view from the oratory terrace on a foggy
day.
284
THE EXCAVATIONS
The shrine terrace
Immediately north-east of the major fault that marks the eastern end of the oratory terrace there was an
almost separate long, narrow, triangular terrace—the shrine terrace. After the fault line the rock is twisted
in an anticlockwise direction, so the cleavage lines run north-east/south-west; the outline of the shrine
terrace follows this alignment and hence is at a slight angle to the line of the oratory terrace.
Traces of parts of its outer wall were visible in the 1980s but had either fallen away or were hidden
beneath campion growth by 2005. Three large, flat stones set upright on the back of this terrace were
interpreted as the remains of a shrine. The surface of this ledge, particularly at its western end, sloped
down very steeply to the cliff edge and there were loose stones visible there, which were in clear danger
of sliding off.
The shrine terrace was excavated in 2006.
The terrace
Access to this terrace was gained in two ways.The first route ran straight on eastwards from the northern
entranceway to the oratory terrace, along a stone-flagged path above and to the north of the oratory.
The second approach circumnavigated the southern side of the oratory, crossing over the large bridging
stone.
[Above left] Ill. 2.434—The western end of the shrine terrace
before excavation, looking west. Collapsed bridging stone
beneath oratory visible at left (A.R. Hayden).
[Left] Ill.
2.436—Looking
down on the
shrine terrace
during
excavation,
from the northeast (A.R.
Hayden).
[Above right] Ill. 2.435—Looking east along the shrine terrace
before excavation (A.R. Hayden).
[Right] Ill.
2.437—The
scar left by the
collapse of the
outer wall of
the shrine
terrace, from
the east. The
shrine is visible
at the upper
right (A.R.
Hayden).
285
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Owing to the collapse caused by the movement of the large stones beneath the oratory, no
stonework survived where the shrine and oratory terraces joined. Indeed, the movement of these large
stones had probably also caused the collapse of much of this end of the shrine terrace itself.Water flowing
down the deep fissure marking the major fault at the south-western end of the terrace probably
destabilised the large stone beneath the oratory and also removed more material once the terrace wall
had collapsed.
The outer face of the wall defining the edge of the shrine terrace had completely disappeared. Its
inner face, however, which consisted of a straight line of small stones, survived to just a single course in
height.The wall was clearly built along the top edge of the cliff, which dipped downwards to the southwest. The collapse of the terrace wall in turn caused the collapse and severe slumping of the infilled
material behind it and the structures that originally stood on its surface.
Ill. 2.438—The
western end of
the shrine
terrace during
excavation,
from the east.
Note the steep
angle of the
surviving
paving. The
ranging rod lies
at the outer
edge of the
wall retaining
the shrine
(A.R. Hayden).
After removal of the campion, the surviving surface of the terrace infill was revealed. It lay at a very
steep angle and it is clear that much material had been lost. The remains of paving, consisting of flat
stones lying at an angle of 45º, survived on the western half of the terrace. All the stones were loose and
it is clear that they represent only the very last remnants of the surface of the terrace, the remainder
having slid over the cliff. The surviving stones had to be removed after recording and planning, as they
were in danger of sliding over the cliff.
The shrine
Three large, flat slabs set upright, two parallel to each other and one set transversely, were visible before
excavation at the back of the centre of the terrace.These are possibly the last remnants of a partly collapsed
slab shrine. Traces of a collapsed wall consisting of a line of pitched small, flat stones, a single course in
height, survived to the south of the upright slabs. This wall appears to have defined the platform on
which the shrine stood. There were also traces of a small length of walling at the eastern end of the slab
shrine, which may have been its original eastern end.
286
THE EXCAVATIONS
[Left] Ill. 2.439—Looking down on
the shrine after excavation. The
ranging rod lies on the line of the
collapsed retaining wall (A.R.
Hayden).
[Below left] Ill. 2.440—The
collapsed retaining wall of the
shrine, from the south-east (A.R.
Hayden).
[Below right] Ill. 2.441—The stones
of the shrine. Note that they were
held in place only by the campion,
with little solid material supporting
them (A.R. Hayden).
The surviving ground level around the upright slabs lay beneath their bases and they were only
held in place by the campion and collapsed material. Hence it was not possible to demonstrate, as seems
likely, that originally there were more upright slabs here. It is therefore difficult, if not impossible at this
stage, to reconstruct the original form of the putative shrine. Its placement on a raised platform and its
close proximity to the oratory are, however, typical features of the slab shrines found in other early
medieval ecclesiastical sites on the Iveragh Peninsula (Marshall and Walsh 2005, 55–8, 166–71; Sheehan
2009, 198–200; Hayden 2011).
The two routes to the upper traverse
There were two routes leading from the oratory terrace up to the next terrace—the upper traverse—on
the final route to the Peak.The western route is the one most used today and, while possibly less exposed
and apparently safer, is in fact the more difficult.
The southern route
From the path from the ‘garden’ terrace on the low cliff above the north-western side of the oratory
terrace a narrow rock ledge runs diagonally eastwards up the steep cliff. There are several rock-cut steps
on this ledge and the ledge itself also appears to have been worked. A rock-cut step above the eastern
end of the ledge leads to the end of another sloping ledge, which runs diagonally back and upwards to
the west. Again there are several rock-cut steps on this ledge and the ledge is also worked. At the western
end of this ledge the climb proceeds straight up the cliff to the north, where there is a series of large
287
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
rock-cut steps leading straight up to a spacious flat ledge overlooking the oratory terrace. Following this
ledge west leads to a point about 1.5m above the southern end of the eastern side of the upper traverse.
Two small rock-cut steps lead down the rock slope to a large stone on the upper traverse (see below).
This stone was placed here after the construction of the upper traverse, so it is clear that this route
represents one that continued in use after the building of the upper traverse.
This route is very easily climbed in an upward direction but is less easy going down.
Ill. 2.442—The ledges and rock-cut
steps of the lower part of the
climb from the northern side of
the oratory terrace to the ledge
above the upper traverse (A.R.
Hayden).
Ill. 2.443—The rock-cut steps of
the upper part of the climb from
the northern side of the oratory
terrace to the ledge above the
upper traverse, from the east (left)
and from above (right) (A.R.
Hayden).
288
THE EXCAVATIONS
The western route
The second route to the upper traverse requires doubling back along the ledge to the top of the climb
up from the western end of the ‘garden’ terrace. Passing the top of the climb up from the ‘garden’ terrace,
the route follows a narrow ledge for the short distance to the southern end of the upper traverse. Much
of this ledge has been carved from the cliff and there are several rock-cut steps here. There are at least
half a dozen carved handholds on the top of the rock on the right-hand side.Today this short part of the
climb is quite awkward to cross, as the rock on the right protrudes at shoulder height.There is, however,
[Left] Ill.
2.444—The
path to the
upper traverse,
looking north:
red = rock-cut
steps; dashed
yellow = cutback and
worked areas;
solid yellow =
handholds; blue
= possible
worked ledges
that could have
held masonry
(A.R. Hayden).
[Right] Ill.
2.445—The
path to the
upper traverse,
looking south
(A.R. Hayden).
a narrow ledge on the cliff just below the path here,
and it is possible that there was once masonry
anchored on it that would have provided a wider
passage, as on the narrow ledges on either side of
the blind corner (see above). The lower rock ledge
is very narrow and uneven and there were no traces
of any masonry surviving here. Perhaps the number
of handholds on the rock to the right might suggest
that this part of the route was always awkward to
cross.
Ill. 2.446—Close-up of the carved handholds at the end
of the route to the upper traverse (A.R. Hayden).
289
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The upper traverse
In the area of this terrace the vertical cleavage planes of the bedrock run east–west, and as a result the
cliffs here are fissured and uneven owing to the erosion of the weaker cleavage planes. The long axis of
the terrace, which runs roughly north–south, lay at right angles to the cleavage planes. The terrace was
roughly straight, but at its southern end it turned outwards and skirted around a protruding rock outcrop.
The high cliff behind the terrace was roughly vertical, but it overhung in places and was also very
fissured and uneven. An eroded major fault line (the gully that the primary route follows up the Peak)
marks the northern end of the terrace, and a minor one marks the opposite end.
[Left] Ill.
2.447—The
upper traverse
from below
before
excavation
(A.R. Hayden).
[Right] Ill.
2.448—The
remains of the
poorly
preserved
external wall at
the southern
end of the
upper traverse
before
excavation
(A.R. Hayden).
[Left] Ill.
2.449—The
exterior face of
the wall at the
northern end
of the upper
traverse before
excavation
[Right] Ill.
2.450—The
exterior face of
the wall at the
northern end
of the upper
traverse after
excavation. The
large stone at
the base of the
corner of the
wall sits on a
rock-cut step
that was part of
the original
route (A.R.
Hayden).
290
THE EXCAVATIONS
Before excavation, standing on the upper
traverse, it appeared almost as a natural flat ledge, as
virtually no walling was visible on its surface and its
top was roughly level and covered with plant growth.
The outer wall was clearly visible from below,
however. The outer face of the walling (although
somewhat obscured by plant growth) appeared to be
largely in good condition, except for its northern and
southern ends. At the southern end only the very
basal stones of the wall appeared to survive. The rest
of the outer face there appeared as a tumble of stones
covered with plant growth and it was obviously
unstable. At its northern end more of the outer face
was visible, but its base was hidden below plantcovered fallen debris and the top of the wall was
evidently missing and marked by a steep, plantcovered slope.
The outer face of the terrace wall was cleaned
of plant growth and the terrace was excavated in 2006
to determine how much of the northern and southern ends of the enclosing wall survived and to examine
its condition.
Initial clearance of the thick campion cover from the face of the outer wall and from the surface of
the terrace revealed the outline of only part of the inner face of the terrace wall. A short length of a
partly collapsed inner facing survived along the southern central section of the terrace. This walling,
however, sat on top of a loose fill of stones and clay, which in turn lay on top of the original terrace wall.
It clearly evidenced a secondary partial rebuild of the inner face of the wall. This secondary wall stood
to a maximum of three courses (300mm) in height and measured less than 400mm in width. It lacked a
Ill. 2.451—The
base of the
external face of
the terrace wall
rose and fell
considerably, as
it was erected
crossing deep
fissures, which
gave the wall a
secure
foundation. The
ranging rod lies
in the same
position as that
in the
preceding
photograph
(A.R. Hayden).
Ills 2.452 and
2.453—The
collapsed
secondary
facing of the
inner face of
the wall of the
upper traverse
(A.R. Hayden).
291
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
built outer face and may simply have been erected as a revetment to loose material piled against its outer
side. The stones in the wall were very loose; it had collapsed inwards significantly and was very unstable.
After recording, this walling was removed to allow access to the underlying original terrace walling.
The inner face of the original terrace wall survived to a maximum of 600mm in height above the
finished interior surface of the terrace. It was composed of much smaller stones than were used to
construct the outer face and had moved inwards and partly separated from the wall itself.The inner face
survived over all but the southern end of the terrace, where it had completely collapsed. Only the
collapsed core and a few of the stones of the very basal course of the outer face of the wall survived at
this end of the terrace.
Owing to its instability, the top of the inner face of the wall over its southern half had to be removed
to allow excavation of the interior of the terrace.
The interior of the terrace had been infilled with quarry waste, raising its level to the top of the
surviving original wall. The secondary wall facing was built on this infill.
[Left] Ill. 2.454—The slumped and partly detached original inner face of the terrace wall (A.R. Hayden).
[Right] Ill. 2.455—The paving on the interior of the terrace. The large stone bridging the drip gully is visible at the lower
right (A.R. Hayden).
The interior of the terrace was originally paved with medium-sized flat slabs.The paving slumped
downwards, however, and lay at a 30º angle, sloping down towards the outer wall. Several of the paving
slabs also extended partly beneath the original inner face of the wall. The leaching out of the finer
material infilling the base of the terrace under the paving caused the slumpage. This erosion and the
movement of the paving (which partly underlay the inner face) partly undermined and destabilised the
inner face of the wall. The movement of the inner face of the wall and the paving explains why the
interior of the terrace was infilled and raised, for it would no longer have been possible to cross it
comfortably or safely.
292
THE EXCAVATIONS
up to top of South Peak
N
Drip
Gully
up from
Oratory
Terrace
up from Garden &
Oratory terraces
0
5
metres
Ill. 2.456—Plan of the excavated features on the upper traverse.
293
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill.
2.457—The
paving at the
southern end
of the terrace.
The large stone
bridging the
drip gully is
visible at the
back left (A.R.
Hayden).
[Right] Ill.
2.458—The
slumped and
wide northern
end wall of the
terrace (A.R.
Hayden).
[Left] Ill.
2.459—The
inner face of
the wall at the
northern end
of the terrace
(A.R. Hayden).
[Right] Ill.
2.460—The
top of the
masonry steps
at the northern
end of the
upper traverse
leading to the
climb to the
top of the Peak
(A.R. Hayden).
At its northern end the top of the outer face of the terrace wall proved to have collapsed extensively
outwards where it turned back to meet the cliff and gully. The base of the outer face survived, however,
partly hidden beneath plant growth and fallen debris. The base of the outer face of the terrace here sat
on top of a clear rock-cut step that was part of the original route to the Peak (see below).The inner face
of the wall at the northern end had also moved to the north and lay at an angle. The wall here was of
much greater width than elsewhere, probably because its inner side included steps leading from the
traverse up to the rock-cut steps in the gully. The top of these masonry steps survived against the cliff
face. A large slab lying against the inner face of the wall on the interior of the terrace may be a remnant
of the first of these masonry steps.
The outer face of the wall along the length of the terrace rose and fell as it crossed the fissured
bedrock. It was generally in good condition, as the uneven bedrock provided a good key for its
construction. Its southern end, however, was constructed on a steep outward slope that lacked any deep
fissures, and this may explain why it collapsed so extensively. The inner face of the wall at the northern
294
THE EXCAVATIONS
end of the terrace was also prone to wind damage. Easterly winds are funnelled along the southern side
of the cliff at the back of the rock ledge above and east of this end of the terrace, and descend down to
the northern end of the terrace with considerable force. After the inner face was conserved and raised,
its top edge was disturbed by wind over the winter of 2008/9. The top course of relatively large stones
on its inner face was uprooted and flipped over by wind, and the stones lay upside down on the centre
of the top of the wall.
[Left] Ill.
2.461—The
external face of
the wall at the
southern end
of the upper
traverse after
excavation
(A.R. Hayden).
[Right] Ill.
2.462—Wind
damage to the
conserved
inner face of
south end of
the terrace wall
A deep drip gully extended along the rear of the terrace against the base of the high cliff face. It (A.R. Hayden).
extended downwards through the secondary fill and the primary paving. It contained flat stones pitched
into it at an angle, suggesting that it was gradually opened over time. It was formed by the leaching of
the terrace infill pulling loose material away from the cliff face combined with the erosion caused by
water flowing down the cliff. A large boulder and several smaller stones were placed into the top of the
drip gully at the southern end of the terrace.These appear to have been intended to bridge the gully and
provided a step up to two rough and small rock-cut steps found on the bedrock slope above.These steps,
which lead to the broad ledge that overlooks the northern side of the oratory terrace, mark the uppermost
end of the route up the cliff from the northern side of the oratory terrace.The placement of the boulder
Ills 2.463 and
2.464—Crossshaped slab
found on the
upper traverse
(A.R. Hayden).
295
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
over the drip gully shows that it was laid down well after the terrace had been built. This suggests that
the route up the cliff from the southern side of the oratory terrace continued to be used after the
construction of the upper traverse and was not simply a route used during construction.
As little of the bedrock beneath the upper traverse was revealed by excavation, we do not know the
extent to which this area was quarried to facilitate the construction of the traverse. The base of the
traverse lies on a bedrock bedding plane whose original width and regularity is not known. Therefore
we do not know whether or not it may have been possible to cross this area before the construction of
the traverse.
2.3.6 THE
OUTER TERRACE
The outer terrace is reached from the head of the gully below the very top of the Peak and so could
connect with either or both of the routes up the Peak. There is no evidence to show when the outer
terrace was built relative to the other terraces, and so it is not clear on present evidence when it was
built.
The route to the outer terrace
The route to the outer terrace begins at the head of the gully, just below the start of the final climb to
the Peak. From this point there were a number of possible ways to reach the outer terrace. Unlike the
other routes on the Peak, no rock-cut steps or handholds survived here to aid the monk or pilgrim.
The easier but more roundabout route runs north-east from the top of the gully along a level ledge
and down either one of two easily climbed 45º slopes to a broad and flat ledge. This ledge runs back
horizontally to the south-west.The descents of the slopes on this route are much easier and less exposed
than those on the probable original route (see below). These easier routes appear, however, not to have
been used. After these slopes were cleared of growth, Michael O’Sullivan (pers. comm.) suggested that
they were never used, as several narrow and fragile rock spines were still present; these would have been
broken off by human traffic.
[Left] Ill.
2.465—The
route to the
outer terrace:
the flat ledge to
which the
easier but
apparently
unused climbs
lead down
(A.R. Hayden).
[Right] Ill.
2.466—The
apparently used
route extends
down this
awkward slope
to the western
(near) end of
the ledge in the
previous
photograph
(A.R. Hayden).
296
THE EXCAVATIONS
The more direct but more difficult route leads by a different way down to the western end of this
flat ledge.This route runs south-west from the top of the gully, along the crest of the rock ridge and then
down its north-western face.This involves a vertiginous and difficult clamber down a steep face adjacent
to the cliff. This rock face contains several small, rough but natural ledges that aid the climber. None
were improved but several appeared worn, and so could indicate that this was the route used by the
monks. Some of this wear could, however, be modern. A short length of very poorly preserved drystone
masonry lies at the western end of the flat ledge at the base of this rock face. These have not fallen here,
as they are geologically dissimilar to the surrounding rock, and therefore appear to have been deliberately
placed (Michael O’Sullivan, pers. comm.). No recognisable structure survived, but they may have formed
a small platform providing the climber with a more secure footing at the base of the steeper part of the
climb. An area of bedrock just above the masonry has also been worked flat.
[Above] Ill. 2.467—Masonry at the western end of the flat ledge down
to which the first part of the climb leads. Worked area visible to its left
(A.R. Hayden).
[Right] Ill. 2.468—Detail of worked area at western end of flat ledge
(A.R. Hayden).
From this point the route is clear, as it follows the flattened top of the narrowing spine ridge stepping
down to the south-eastern corner of the terrace.The western half of the spine may originally have been
covered by stonework—the top of the adjacent terrace wall—as again there are delicate pieces of rock
surviving on the flat top of the spine,
which would not have survived if the area
had been walked on extensively (Michael
O’Sullivan, pers. comm.).
There is a cache of at least twelve
pieces of quartz laid on a ledge on the
northern side and just below the start of
this final descent to the terrace. They do
not form any recognisable structure.
Ill. 2.469—Cache of quartz blocks on the
northern side of the climb down to the outer
terrace (A.R. Hayden).
297
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Today the final descent onto the outer terrace takes one down the face of a rock ridge to the right,
where there are a number of small rock-cut notches and handholds. Originally, however, the climb down
would have been on rock-cut steps and masonry steps constructed on bedrock ledges and finally on the
top of the inner side of the southern wall of the terrace itself, which did not survive here. The final run
of steps was on stones protruding from the inner face of the wall. Only the lowest of these steps survived.
[Left] Ill.
2.470—The
final part of the
climb down
originally lay
partly on the
wall top, which
no longer
survives (A.R.
Hayden).
[Right] Ill.
2.471—The
one surviving
protruding
stone on the
inner face of
the wall that
provided the
last step down
to the terrace
(A.R. Hayden).
The outer terrace
Of all the terraces on the South Peak, the outer
terrace is the most exposed (being the only one
built on the northern side of the Peak), remote and
difficult to access. It does not form part of either
of the two routes to the Peak, but is accessible from
both of them. In addition, it is the least well
preserved of the upper terraces; indeed, it is not
clear whether it was ever completed. Nevertheless,
the interpretation of its temporal position and its
function could be fundamental to the
understanding of the development and meaning of
the structures on the South Peak (see discussion
below).
Ill. 2.472—The South Peak from the north-east. The outer
terrace is visible just below the top right side of the Peak.
The prominent ledge below it may have provided the
stone for its construction (Con Brogan, DAHG).
298
THE EXCAVATIONS
The outer terrace occupies the south-eastern end of a high and narrow spine of, in places,
overhanging bedrock protruding from the south-western corner of the Peak.The vertical cleavage planes
of the bedrock run north-east/south-west in this area, so there are vertical cliffs that are prone to the
scaling off of large slabs on the northern and southern sides of the terrace, while the cliffs to the west are
fissured and uneven owing to the erosion of the exposed ends of the cleavage planes.
Ill. 2.473—The south side of the outer
terrace, viewed from the upper
traverse. The climb down follows the
sloping wall on the right (A.R.
Hayden).
The walls of the precipitous outer terrace hug the edges of the curving cliff that defines the western
end of the Peak.The uppermost part of the terrace looks down to the south over the top of the gully up
which the primary route climbed.The terrace curves around to the west and descends, and at its northern
side peers vertically down over the steep and exposed cliffs on the north side of the island. From its
eastern side there is a clear view of the north landing and the north steps, the monastery behind them
and Little Skellig in the distance.The terrace is exposed and even on the warmest days it was always cold,
often bitingly so, and windy. Indeed, this terrace provided, by a long way, the toughest environment in
which work was undertaken on the Peak.
The enclosed area of the terrace is composed of four east/west-aligned rock ledges of varying width,
which step steeply down to the north like a flight of broad, uneven and 1m-high giant steps. There is a
difference of 7m in height between the southern and northern sides of the enclosed area.
Before any works were undertaken, the walls of the outer terrace were examined in 2006. This
revealed that although the walls at the southern and south-western corners of the terrace were stable
they were badly cracked and slumped in and out in places. The walling along the western side of the
terrace was in atrocious condition, however. It leaned inwards hugely and was close to total failure and
loss.
299
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill.
2.474—
Examination of
the walls of the
outer terrace
(A.R. Hayden).
[Right] Ill.
2.475—The
highly collapsed
and unstable
centre of the
western wall of
the terrace
(A.R. Hayden).
Ills 2.476 and
2.477—The
lower surviving
section of the
western wall
(A.R. Hayden).
Ill. 2.478 and
2.479—The
sondage
opened in the
south-western
corner of the
outer terrace
(A.R. Hayden).
300
THE EXCAVATIONS
A small sondage (measuring less than 1m by 1m across) was opened in the curved west-facing corner
of the terrace in 2006 at a point that was partly cleared in the 1980s, revealing possible paving (Horn et
al. 1990, fig. 51). The test-trench revealed a surface containing a few flat stones that could indeed have
been very poor and gapped paving at the level of the base of the internal face of the enclosing wall.
Scaffolding was erected around the western side of the terrace in 2006 (an engineering feat in itself)
and the terrace was excavated in 2006–7. Excavation was first undertaken on the western side of the
terrace to facilitate the conservation of the highly vulnerable walling there, and then moved down to
the northern side of the terrace to determine whether anything of its northern wall survived. The
northern wall was of key structural importance, as it lay at the lowest part of the terrace and would have
retained all the terrace infill. Rather like the situation on the ‘garden’ terrace, the collapse of this wall had
the potential to destabilise all the other walls and whatever structures were built on the terrace itself.The
interior of the terrace was next excavated to see whether any features survived there, and finally the
better-preserved walling and the access route down to the terrace at its southern and upper side were
excavated.
Ill. 2.480 (Left and below)—The spectacularly sited scaffolding erected at the western side of the outer terrace,
which was an engineering feat in itself (A.R. Hayden).
Ill. 2.481—
Looking up
(east) along the
access to the
terrace. Steps B
and C are
visible above
the bedrock
ledge (D) near
the top of the
picture. Wall E
is visible below
at the centre
right, with the
internal wall
visible below
(A.R. Hayden).
Access walling and features
The access route runs down into the eastern end of
the southern side of the terrace. A rock-cut step (no.
1) at the end of the highest part of the bedrock
spine leads down to two or three steps in the eastern
section of the southern wall of the terrace. The
upper step may have been on the outer part of the
parapet wall itself, the second was in the wall (B)
and the final one was a separately built masonry step
(C) erected on the flat bedrock ledge.The masonry
of this step was not keyed into the outer wall.
301
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ills 2.482 and 2.483—(Left) Detail of steps B (top) and C
(bottom), leading down to (below) a flat bedrock ledge (D)
(A.R. Hayden).
The step led down to a flat bedrock ledge (D). A few large stones lay on the ledge and it could
originally have been completely covered in masonry. At its south-western end there was another rockcut step (no. 2), leading down to the top of a short length of walling (E) built against the vertical northern
side of the bedrock ridge. The latter wall was composed of notably larger stones than those used in the
adjacent terrace wall and appears to have stood as a completely separate section of walling, at least in its
surviving (lower) part. Its upper part could have been keyed into the terrace wall where it had originally
stood above the level of the bedrock ridge.
[Above] Ill. 2.484—Looking west along the bedrock ledge
(D), part of step C in foreground. Stepping down off the
far end of the ledge led to the top of wall E (A.R. Hayden).
[Right] Ill. 2.485—Looking west along wall E; note the
projecting stone on its inner face (A.R. Hayden).
302
THE EXCAVATIONS
There should have been three steps in this wall originally—only the lowest step survived—to bring
one down to the level of the highest bedrock ledge at the south-eastern side of the terrace.The top steps
would have lain in the wall itself. The bottom step, which survived, was a flat stone projecting from the
face of the wall.
A number of clear rockcut notches and a handhold
survive in the rock inside the
main terrace wall. They indicate
that this was an alternative route,
which could have been used at
either an earlier or a later time
than that described above,
leading down to the terrace
from the end of the high rock
spine. This alternative route is
the natural modern one, as the
collapse and erosion of the
terrace wall and wall E
prevented their use as a route
down to the terrace, and could therefore be of more recent origin. It is, however, possible that this was
the route used by the terrace-builders themselves, before they had constructed the wall at this side of the
terrace. The modern use of this route may also have contributed to the damage to the possible leacht
described below, as it lay where one lands after jumping down onto the terrace from these rock-cut
notches.
Both these routes led down to the highest-level rock ledge in the terrace, where a small leacht
survived (see below). The finished terrace level, however, seems to have lain about 1m lower, on the next
rock ledge down.
Two poorly preserved rock-cut steps (nos 5 and 6) led down the rock slope at the north-western
corner of this ledge to the main level of the terrace.
[Left] Ill. 2.487—Looking down the access route at the southern side of the terrace, after conservation (A.R. Hayden).
[Right] Ill. 2.488—Wall E with reconstructed steps on its top, after conservation (A.R. Hayden).
303
Ill. 2.486—The
north face of
wall E, with the
projecting
stone step
visible at centre
(A.R. Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The northern, western and south-western walls
The defining drystone walling of the terrace curves sinuously around the south-western end of the steep
rock cliff. Its base rises nearly 7m from the north to the west side of the terrace. The wall is described
below, running upwards from the northern side to its western and then southern sides.
A vertical rib of bedrock on the cliff edge, lying parallel to the wall line, marks the eastern end of
the surviving walling on the northern side of the terrace. The original walling here had almost totally
fallen away over the steep cliff and only a single course of stones survived, showing that the wall ran at
least as far east as the western side of this bedrock rib. The bedrock rib actually leans slightly outwards
and so the wall base lies outside it, but as the wall rose it must have become tucked behind it, which is a
typical construction feature found on the Peak.
Ills 2.489 and
2.490—Looking
down on the
north-eastern
corner of the
terrace during
excavation. The
few surviving
stones of the
northern wall
lie on the right
(southern) side
of the bedrock
rib at the cliff
edge. No trace
of an eastern
wall survived.
The conserved
lower end of
the western
wall is visible at
the left (A.R.
Hayden).
It was impossible to determine exactly where the north-eastern corner of the terrace lay, or indeed
whether the wall here had actually been finished, as no stonework at all survives north-east of the bedrock
rib. The northern wall could only have continued on for a maximum of 2m or so, as beyond this point
there is nowhere to anchor the base of a wall because the cliff there is vertical to overhanging and lacks
any ledges. There were no traces of stonework or notches cut in the bedrock to show where the wall
would have turned south-eastwards to define the eastern end of the terrace. There would have to have
been a wall here somewhere if the terrace was completed, as the fine material infilled behind the northern
and western walls of the terrace had also to be contained on this side.
The wall at the northern side of the terrace, if completed, would originally have stood to at least
3m in height if it were to retain material to the level of the second rock ledge down, which appears to
have formed the finished internal level of all but the southern side of the terrace, which lay at a higher
level (see below).
Two small notches were noted on
a protruding bedrock rib on this steep
slope. They were probably worn by
ropes used by the terrace-builders to
haul stone up from the lower ledge on
the cliffs on the northern side of the
island (see below).
Ill. 2.491—Rope-worn notches on the
bedrock at the northern side of the terrace
(A.R. Hayden).
304
THE EXCAVATIONS
At the north-western corner of the terrace the wall survived to less than 0.5m in height, but it rose
steadily as it extended south along the western side of the terrace, where the bedrock rose in a series of
broad vertical steps. From the north-western corner of the terrace the wall turned to the south, where
it met a transverse and vertical upward step in the bedrock that marked the transition to the second
bedrock ledge.The base of the wall abutted against this vertical rock face but higher up it extended over
it, linking it to the next section of walling to the south.The southern end of the base of this latter section
of walling also ran into and abutted another transverse vertical rise in the bedrock, which marked the
edge of the next rock ledge up. The top of this ledge appears to have been the finished level of the
interior of the terrace. Again the upper part of the wall here continued over the bedrock rise, where it
turned abruptly to the south. The presence of the bedrock rise accounts for the abrupt change in the
line of the wall here. This middle section of the wall of the western side of the terrace was the most
visibly deformed, as it both survived to the greatest height and had collapsed inwards the most
dramatically owing to the severe erosion of the fills behind it. The stones from the top of the front face
of the wall, which had collapsed inwards, stood vertically.
This central section of the western wall was the only part where it was necessary to remove
stonework, and only the poorly preserved and highly collapsed stonework of its upper part was removed.
The stone forming the lower part of the wall, although severely slumped, lay at a lesser angle and was
jacked up and secured from the inside.
[Left] Ill.
2.492—Looking
up from the
north-western
corner of the
terrace after
the removal of
campion and
before
excavation
(A.R. Hayden).
[Right] Ill.
2.493—The
western wall
after excavation
and removal of
the loose
stone. Note
collapsed inner
facing stones
lying vertically
at right (A.R.
Hayden).
305
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
There was no clear inner wall face surviving here owing to the high degree of collapse. If there had
been a face, as seems likely, it has detached and totally collapsed. The stones in the outer face of the wall
were large and flat and notably bigger than those used elsewhere, especially in the walls at the southwestern corner and eastern side of the terrace.This is probably due to the much greater height to which
the wall here would have stood originally. The stones in the surviving part of the western wall were all
very badly decayed and fractured.Where the wall had collapsed, most of the stones were reduced to small
fragments and often little more than dust, while the surviving stones in the upper part of the standing
walling were all heavily cracked and fractured. The degree of deformation of the wall dramatically
demonstrates the substantial quantity of infill that was washed out from behind the wall. The collapse of
Ills 2.494, 2.495,
2.496 and
2.497—Stages
in the removal
of the central
section of the
western wall of
the outer
terrace (A.R.
Hayden).
306
THE EXCAVATIONS
the northern wall of the terrace allowed virtually all of this infilled material to be washed over the cliff
at the northern side of the terrace. The closeness of the steps in the bedrock to the back of the wall also
contributed to this erosion, as water would naturally have drained down their vertical faces, removing
the finer material packed in behind the wall.
Some of the original infill
survived behind the wall beneath
the base of the collapsed stone. It
consisted of the usual mix of small
stone fragments and clay, which
was the waste from the quarrying
and shaping of the stone used to
form the walls. The remains of a
fire (red to orange ash with no
apparent charcoal flecking) were
clearly evident in this material
directly behind the wall. Traces of
a small fire or fires were uncovered
at a similar level on the oratory
terrace and probably represent no
more than cooking episodes
during the construction of the terrace.These fires were all lit on the infilled material, so were clearly not
related to quarrying. The burnt material was removed for examination but consisted only of oxidised
clay; it did not contain any charcoal and hence was unsuitable for radiocarbon dating.
Ills 2.499 and 2.500—Burning on bedrock beneath the
surviving terrace infill at the centre west side of the terrace
(A.R. Hayden).
The wall around the curving south-western corner of the terrace survived to a much greater extent.
There, as noted above, it was composed of much smaller stones. There were, however, several major
vertical cracks in the outer face of the wall, and a large hole had opened at its base at one point.The base
of the outer and inner faces of the wall had also slumped inwards in places and the stones in the wall
there were loose. The inner face of the wall in the tight corner was roughly corbelled inwards at its top.
307
Ill. 2.498—The
condition of the
stones in the
centre section
of the western
wall of the
outer terrace
(A.R. Hayden).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
While the oversailing was uneven, it appears to be original.
The unevenness of the face and the small size of the stones
used suggest that it could never have formed an actual
corbelled roof over the inner side of the corner.
None of the stones of this area of walling had to be
removed, and it proved possible to stabilise the wall by
inserting stones in its faces to fill the holes and cracks and
to tighten and secure the structure.
Ill. 2.501—The
slumped outer
face of the wall
north of the
south-western
corner of the
terrace (A.R.
Hayden).
Ill. 2.502—Large hole in the base of the wall at the
south-western corner of the terrace (A.R.
Hayden).
[Left] Ill.
2.503—Vertical
crack in the
wall south of
the southwestern corner
(A.R. Hayden).
[Right] Ill.
2.504—Eastern
end of the
walling at the
south-western
corner tucked
behind a
bedrock rib
(A.R. Hayden).
The base of the wall at the south-eastern side of the tight southern corner was butted inside the
end of a high and narrow ridge of bedrock that ran almost parallel to the line of the wall.The upper part
of the wall did not survive here but would have run partly over this ridge.
The bedrock rib mentioned above separated the walling at the south-eastern corner from that
delimiting the eastern end of the southern side of the terrace. This last section of wall survived well but
to only a low height. It was in reality largely just a thin skin of walling lying against the outer face of the
308
THE EXCAVATIONS
rising bedrock rib. As the bedrock rib
narrowed as it rose, the wall would
originally have widened as it extended
upwards. The inner face of the wall
below the highest bedrock ridge was
curved and notably battered. This
appears not to be original. The inner
face is composed of very small stones
(the outer face is composed of larger
ones) and has slipped and moved
considerably, as the stones were fractured
and lay at an odd angle. The surviving
outer face here, however, was vertical.
Ill. 2.505—The
south-western
corner of the
terrace during
excavation. The
leacht is at the
lower right, the
internal wall is
at the centre
and the
oversailing
south-western
corner of the
outer wall is at
the top left of
the picture
(A.R. Hayden).
Internal features
The excavations revealed a number of new features on the southern side of the terrace—a leacht, a short
length of internal drystone walling and paving.
N
Edges of
rock ledges
inner face
of wall
collapsed
here
Ill. 2.506—
Preliminary plan
of the
excavated
features on the
outer terrace.
Leacht(?)
Paving
Line of
access route
Wall top
edges
Rock-cut
steps
Step B
Step C
Internal
wall
Ledge D
Protruding
step in wall E
0
5
Quartz
metres
309
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The leacht
The remains of a possible leacht survived on the highest bedrock ledge on the southern side of the terrace,
flanking one side of the final part of the route (steps 5 and 6) down to the terrace. There was a limited
and sharply defined pile of two to three layers of very fine-grained and flat slate stones (quite different
in nature to the stones forming the terrace and other walls in the immediate area). The corner of the
bedrock ledge would have defined the north-western and south-western sides of the possible structure,
but there the stone heap was most disturbed. The south-eastern side of the stone pile was delimited by
a low and straight rise in the bedrock. The north-eastern side of the heap had a possible disturbed face.
Although the stone here was only one to two layers in thickness, the pile appeared to have a roughly
straight edge, which appears to have moved to the north-east somewhat. The stone heap also ended
abruptly along this line. The most convincing evidence of its original face here was the presence of a
number of pieces of quartz, which ran on a line outside the face as if they had fallen off the structure.
A number of quartz blocks, which had probably fallen off the structure, landed on the paving on
the ledge below and north-west of the leacht. Amongst the quartz were a number of flat stones pitched
on edge, which had clearly fallen from the higher ledge.
The small size and seemingly rectangular shape of the stone pile, coupled with the presence of the
quartz, suggest that the structure may have been a small leacht. A small stone cross which was previously
found on the lower, northern side of the terrace could conceivably also have fallen relatively recently
from the possible structure, as its find-spot lay at the base of the natural slope to the north of it.
[Above] Ill. 2.507—The possible leacht viewed from the south-east;
note the quartz blocks rolled off its top on the right (A.R. Hayden).
[Right] Ill. 2.508—Quartz blocks fallen from the leacht lying on the
paving on the ledge below and north-west of the possible leacht
(A.R. Hayden).
Internal wall
A short length of walling survived lining the steeply sloping western end of the highest bedrock ledge,
on which steps 5 and 6 were cut, at the south-western side of the terrace.The wall survived very poorly,
as its base had been constructed on a narrow (100–150mm wide), irregular flat ledge on the slope. The
base of the wall may have been footed on stones hammered vertically down into the south-west/northeast-aligned clefts in the bedrock, as there was nothing else to support it. Only the top two courses of
the wall survived in any way intact. The base had collapsed outwards and down to the south-west, and
its northern end had slumped down to the north. The wall could conceivably also have been used to
310
THE EXCAVATIONS
support some form of cover that roofed the narrow south-western corner of the terrace, where there
may have been a rough, narrow shelter.
Ills 2.509 and
2.510—The
short length of
internal walling
lining the
western end of
the highest
ledge at the
south-western
corner of the
terrace (A.R.
Hayden).
Paving
The interior of the south-western corner of the terrace was roughly paved with flat stones. There was
no paving, or it did not survive, where the bedrock ledge ran back to the north. This was the finished
level of the terrace here and logically would seem to have been the finished level of the northern end of
the terrace also, if it was to have been a usable space.
The largest of the paving stones extended into the delimiting wall of the terrace and were the only
[Left] Ill.
2.511—The
possible paved
shelter at the
south-western
corner of the
terrace, looking
east; internal
wall at right (A.
R. Hayden).
[Right] Ill.
2.512—The
possible paved
shelter at the
south-western
corner of the
terrace, looking
south-west;
internal wall at
left. Note paving
at lower right
extending into
the face of the
outer walling
(A. R. Hayden).
311
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
parts of the paving to do so. In this and in its general nature the paving is strongly reminiscent of that
revealed on the traverse.
The paved area was covered by a thick layer of stone that had collapsed off the surrounding walls,
mixed with heavily rooted topsoil. No occupation debris or recognisable deposits survived on the paving.
Shelter
The top of the curving inner face of the wall at the south-western corner of the terrace oversailed
significantly. The wall was in excellent condition here and this appears to have been a deliberately built
feature as opposed to later slumpage. The terrace wall here also stood about 1.5m above the finished
level of the terrace and so is, by a large margin, the highest surviving walling relative to finished terrace
level anywhere on the South Peak.The presence of the small internal wall lining the rock slope just back
from this corner and the fact that the base of this corner was paved could suggest that this area might
have been some form of small shelter.There is sufficient room in the small space created for an individual
to squat or lie down, well sheltered from the prevailing winds. It is not clear whether this area could
have been roofed over originally. The poor quality of the internal wall may suggest that, if so, it was not
with a stone corbelled vault.
Cross
An upright cross-shaped slab was found at the northern corner of the terrace in the 1908s (Horn et al.
1990, 63).
Beyond the outer terrace
To the east of the lower part of the outer terrace there is another broad, flat, natural ledge, which is
accessible down a 2m-high and easily climbed cliff. The fault line that marks the western end of the
ledge continues down the external cliff face below as a narrow gully. A short length of possible drystone
masonry survived in and near the top of this gully at the cliff edge. It may have been erected to aid the
climb up or down the top of the gully, which leads down to the ledges below the outer terrace (see
below).
No excavation, survey or any other work has been undertaken on this ledge.
[Above] Ill. 2.513—Location of ledge and gully east of the outer
terrace (Con Brogan, DAHG).
[Right] Ill. 2.514—Possible masonry in gully below ledge east of the
outer terrace (A.R. Hayden).
312
THE EXCAVATIONS
2.3.7 AREAS WHERE
STONE WAS QUARRIED AND WON
Two areas where stone was quarried or won and hauled up the cliff to build structures above were
excavated and illustrate some of the methods used by the monks to acquire building materials.The ledges
beneath the oratory terrace were the first of these areas excavated. Later a parallel situation was observed
on the ledges below the outer terrace and so they were also excavated.
The area beneath the oratory terrace
There are a number of sloping ledges on the steep cliffs directly beneath the southern side of the oratory
terrace. Small flat stones lying on the ledges could be seen protruding from the grass cover. Some of
these had clearly fallen from the oratory terrace when its walls collapsed and had lodged on the ledges.
In other parts, stones protruded from the campion cover at the edges of the ledges.To determine whether
the latter represented structures, and in order to reuse the fallen stones for conservation work on the
oratory terrace, these ledges were explored and selected areas were excavated in 2006 and 2007. The
small drystone platform noted below the southern side of the oratory terrace probably aided the lifting
of stone from these ledges up to this terrace.
Ill. 2.515—The
south face of
the South Peak,
showing the
features
beneath the
oratory terrace:
the access
route up the
gully from the
broad ledge
(white dashed
line), the south
platform (lower
red arrow), the
north platform
(upper red
arrow) and the
hauling
platform
(yellow arrow).
The ledges
excavated lay
below the latter
(A.R. Hayden).
Two stone platforms (the south platform and the north platform), composed of flat stones bridging
the tops of gullies, were visible to the west and below these ledges. They were probably also used in the
movement of stone up this part of the Peak. They have not been surveyed, recorded or excavated in any
way.
These ledges were accessed from below up the vertical gully at the north-eastern corner of the
broad ledge. This route remains to be fully recorded and surveyed.
313
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.516—LiDAR
image of the area
below the oratory
and ‘garden’
terraces, with the
locations of the
north platform
(upper arrow) and
the south platform
(lower arrow)
marked. These
substantial
structures do not
appear to be visible
on the LiDAR
(DAHG).
The access route
The access route from below to this area of the Peak extended up the gully at the north-eastern corner
of the broad ledge.The rock in the lower part of this gully was very fractured and covered in much plant
growth, which was not removed. A number of small rock-cut steps survived on a spine of rock protruding
from the centre of the gully close to its top. There were additional steps and handholds on the left-hand
side of the gully. These have not yet been recorded or surveyed. The climb up this gully is an awkward
one, especially where it exits at the top.
Ill. 2.517—
Aerial view of
the features
beneath the
oratory terrace
after excavation
of the rock
ledges: the
south platform
(lower red
arrow), the
north platform
(upper red
arrow) and the
hauling
platform below
the oratory
terrace (yellow
arrow) (Con
Brogan,
DAHG).
314
THE EXCAVATIONS
[Left] Ill. 2.518—The gully at the north-eastern corner of the broad ledge, which
provided access from below to the ledges (A.R. Hayden).
[Above Ill. 2.519—Some of the rock-cut steps near the top of the gully. This area has
not yet been fully recorded, and there are more steps and handholds here than are
shown in the photograph (A.R. Hayden).
The south platform
A huge, long stone lies on the ledge at the top of the gully and protrudes out over its centre. Three large,
flat stones were deliberately laid down as bridging between this stone and the northern side of the top of
the gully. They formed a flat platform (the south platform) protruding out over the gully. It was probably
used as a place from which to raise or lower stone between the broad ledge and the rock ledges above.
[Left] Ill.
2.520a—The
south platform
from below,
looking up the
gully (A.R.
Hayden).
[Right] Ill.
2.520b—The
south platform
from above,
looking east
(A.R. Hayden).
315
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The north platform
A second platform (the north platform), constructed in a similar way, lay at the head of a gully on the
cliff edge higher up. It was much larger and was constructed of several layers of stones laid bridging the
top of the gully. A stone stands upright, trapped between a huge fallen rock and the cliff edge at the
western side of the platform. It is not clear whether human hands placed this here or whether it was a
natural feature. If it was deliberately set erect, it could have aided in the raising of stone to this point, or
alternatively may have been a marker on the route to the Peak; it is clearly visible from many points on
the route.The gully beneath this platform has not yet been climbed or investigated to determine whether
it contained any features.
[Left] Ill.
2.521—The
north platform
from below and
south-west
(A.R. Hayden).
[Right] Ill.
2.522—The
north platform
from above
(A.R. Hayden).
[Left] Ill. 2.523—The upright stone beside the north platform viewed from the south steps to the monastery (A.R.
Hayden).
[Right] Ill. 2.524—The upright stone beside the north platform viewed from the southern traverse of the north-west
passage (A.R. Hayden).
316
THE EXCAVATIONS
The ledges beneath the oratory terrace
The ledges directly under the southern side of the oratory terrace were first examined to see whether
any of the stonework was structural.The first and only find uncovered on these ledges lay on the surface
of the grass directly beneath the oratory terrace and so had only recently fallen. This was a small and
very fine cross-shaped slab, which the first person down discovered immediately on reaching the ledge.
The cross is very similar in form to that which possibly marked the very start of the route on Christ’s
Saddle, and is also very like one in the monks’ graveyard in the monastery.
Beneath the grass cover the ledges were devoid of large, naturally fallen stones. Only small stones
survived on the ledges and there was no evidence of any man-made structures. These small stones had
evidently fallen here from the oratory terrace when its southern wall and the oratory collapsed. The
stones were hauled back up again to the oratory terrace for use in the conservation works.The scaffolding
used to haul the stones up to the oratory terrace was footed on the same flat ledge on which the small
drystone platform noted beneath the oratory terrace was built.This small structure could, then, also have
functioned to aid the transport of stone up the almost vertical cliff.
[Left] Ill.
2.525—Mike
Shea, the first
man down to
the ledges
beneath the
oratory terrace,
finds a fine
cross-shaped
slab lying on
the grass (A.R.
Hayden).
[Right] Ill.
2.526—The
cross found is
similar to the
left-hand one
shown here in
the monks’
graveyard in the
monastery
(A.R. Hayden).
[Left] Ill.
2.527—The
ledges below
the oratory
terrace before
excavation,
from below
(A.R. Hayden).
[Right] Ill.
2.528—The
ledges below
the oratory
terrace before
excavation,
from above
(A.R. Hayden).
317
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill.
2.529—Looking
down the cliff
from the
oratory terrace.
The hauling
platform is at
the right, the
excavated
ledges below,
and the north
platform is
visible in the
background
(A.R. Hayden).
[Right] Ill.
2.530—
Clearing and
raising stone
from the ledges
(A.R. Hayden).
Ill. 2.531—Even when the winch broke
down, the OPW crew still managed to haul,
by hand, a substantial quantity of stone up to
the oratory terrace from the ledges below
(A.R. Hayden).
Gully
With modern equipment it is only a simple abseil from the
oratory terrace down to the ledges below, although the climb
back up is much harder! The fault line that defines the western
side of the oratory terrace descends down the cliff beneath it
as a deep and narrow gully. In two places near the top of the
gully there were stones jammed into it. It is possible that these
may have been deliberately placed there to aid the climb up
or down this gully, which formed the most obvious route
between the oratory terrace and the ledges below.
Ill. 2.532—The gully descending the cliff beneath the oratory terrace
(A.R. Hayden).
318
THE EXCAVATIONS
Ills 2.533 and
2.534—Stones
possibly
deliberately
jammed into
the gully may
have aided the
climb up or
down it (A.R.
Hayden).
The ledges below the outer terrace
As noted above, the northern wall of the outer terrace was the least well preserved, and almost all of it
had collapsed and fallen over the cliff. If the terrace had been finished, a substantial quantity of stone
must have fallen, as the original wall would have been around 3m in height.
Some clearly fallen small stones were visible on a number of ledges on the steep cliff about 30–
50m below the northern side of the outer terrace. Two of these ledges were excavated in 2007 to
determine whether the monks had cleared them of stone and to retrieve fallen stone for the conservation
of the outer terrace walls.
[Left] Ill. 2.535—The South Peak from the northeast, showing the excavated ledges (arrowed)
and the rough line of the route up from the cliff
ledges beyond the north-west passage (Con
Brogan, DAHG).
[Below] Ill. 2.536—The steep climb up to the
ledges (A.R. Hayden).
319
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 2.537—The uppermost of the ledges below the outer terrace before excavation (A.R. Hayden).
Ill. 2.538—The uppermost of the ledges below the outer terrace after excavation (A.R. Hayden
320
THE EXCAVATIONS
Ill. 2.539—Hauling stone up to the outer terrace from
the ledges below. The cliff is vertical to overhanging
here, which made this activity relatively easy (A.R.
Hayden).
The ledges can be reached by abseiling down the cliff below the outer terrace or down the gully
east of the outer terrace. Alternatively, there was probably a route ascending the steep slope east and north
from the excavated cliff ledges across the gully from the end of the northern traverse of the north-west
passage. The use of this route to access the ledges below the outer terrace could explain the presence of
the rock-cut features noted on the first of the ledges excavated on the northern cliff. This route is
awkward, as the exposed cleavage planes of the bedrock mean that the rock is very fractured and easily
broken. There was no immediate routeway across this area, but it has not been explored in any detail. If
there are further rock-cut features here, they will probably be poorly preserved and hard to find owing
to the eroded state of the bedrock. In any case, if this was a route used only for quarrying there may
never have been many such features there.
The largest and uppermost ledge below the outer terrace was stripped of soil cover and trowelled
down. This revealed a considerable quantity of small stones fallen from above.Virtually no larger stones
survived beneath the small stones, apart from a few along the outer edge of the ledge.The ledge is broad
and flat and should have caught a substantial quantity of large stones fracturing off the cliffs above, which
over time would have accumulated to several metres in thickness. This clearly suggests that these ledges
were stripped of the naturally fallen stone to provide material for the building of the outer terrace.
Stone for conservation works was also retrieved from a ledge about 30m further down and further
to the east.This ledge was stripped of soil cover and then trowelled down. Like the upper ledge, it retained
little fallen bedrock and also appears to have been stripped by the monks.
The removal of stone for building from ledges far below this terrace is paralleled by the findings
from the oratory terrace and the ledges below it. There was no need on the outer terrace for a platform
to aid in raising the stone, as was built on the oratory terrace, as the cliffs below the northern side of the
outer terrace were vertical to overhanging. The lowest rock ledge forming the base of the northern side
of the outer terrace also provided a natural platform for hauling stone up the cliff. The two rope-cut
notches noted on the outer terrace close to its northern side (see above) might be evidence of this activity.
321
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
2.4 NEWLY DISCOVERED STEPS AND ACCESS ROUTES
Alan R. Hayden
2.4.1 NEW
FEATURES ON THE EAST STEPS
The cleaning of campion growth from the east steps (which is part of the ongoing maintenance
programme) appeared to reveal some previously unnoticed features. What appear to be previously
unrecorded rock-cut steps were uncovered on the western side of the lower section of the masonry steps.
Along with the run of rock-cut steps previously recorded, they form part of two rock-cut stairways that
pre-date the masonry east steps.
Ill. 2.540 and 2.541—Newly discovered early rock-cut
steps on the western side of the east steps (A.R. Hayden).
Ill. 2.542—Newly discovered early rock-cut steps on the
western side of the east steps (A.R. Hayden).
The second-lowest surviving length of rock-cut steps (which lie to the east of the structure described
in Section 2.2.9) was also cleaned of campion growth.The clearance of the vegetation revealed that there
were actually two superimposed sets of steps there, both of which pre-dated the masonry steps; they
actually extend under the masonry stairs.The westernmost of the two lines appears to be the earliest and
was the most eroded. Several of these steps have disappeared owing to the breaking away of chunks of
322
THE EXCAVATIONS
Ill. 2.543—Looking down on the two
flights of rock-cut steps to the east of
the platform at the base of the east
steps (A.R. Hayden).
bedrock. The steps in the eastern
line were both larger and better
preserved and appear to be partly
cut through the remains of the
other set.
These steps have not yet
been surveyed in detail.
Rock-cut cross
Part of a large cross, still attached
to the bedrock, was also noted on
the eastern side of the steps close
to their base. A large, freestanding, upright and thin slab of
rock has a clear notch cut in its
side. The notch is too far away
from the masonry steps to have
been cut to accommodate them
and is also carefully carved
around three faces of the stone,
which would have been
unnecessary if it had been simply
cut for allowing access. The
opposite edge of the stone is
heavily weathered and no
working survives on it. It is
possible that the working signifies
that an attempt was made to form a cross from the bedrock.
Ill. 2.544—
Cross cut from
bedrock near
base of east
steps (A.R.
Hayden).
323
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Terrace at base of east steps
Some drystone masonry was also noted on the cliff edge at the surviving base of the east steps. It would
have formed part of the outer wall of the small terrace on which the ‘boathouse’ structure was erected.
The masonry lies in a precarious position on the cliff edge, which is much fractured owing to having
been partly blasted away by the construction of the lighthouse road. The masonry that is visible appears
to be very loose, is in very poor condition and is highly vulnerable.There could be quite a lot of additional
masonry surviving, as there is abundant vegetation covering this area.
Ill. 2.545—
Remains of
masonry
terrace on cliff
edge (bottom
left), with
masonry base
of east steps
above, viewed
from the east
landing (A.R.
Hayden).
2.4.2 THE
STEPS TO THE SOUTH-EAST LANDING
The opportunity was taken to re-examine the south-east steps in September 2010, as they had not been
visited and photographed for quite a while. Most of this area was heavily blasted and radically altered by
the lighthouse men on at least two occasions.
There are three sets of steps in this area; two are clearly of lighthouse date while the third is likely
to be of monastic origin.
The later lighthouse steps
The westernmost line of steps leading to the south-east landing is the latest in date. This route starts a
short distance down the south steps below the lighthouse road. A line of worked squares and other
features extend along a heavily worked and altered rock ledge running east from the south steps. This
ledge was severed by a partial collapse of the bedrock or by more recent blasting associated with the
construction of the helipad, and the steps further along it are no longer accessible from it. Earlier blast
holes that were drilled to make this ledge accessible are also clearly visible.
At the eastern end of the ledge the route runs downslope, where there is a very finely carved rockcut staircase. There are also several blast holes here, which visibly pre-date the steps. The lowest section
of the steps is very finely wrought and displays the fine chisel-pecked dressing that is typical of lighthouse
work (it is also clearly visible on the earlier set of lighthouse steps to the south-east landing and on the
lighthouse steps to the south landing). Of course, this would not survive if the steps were of much earlier
date because of the aggressive nature of the environment in which they are located. At the base of the
324
THE EXCAVATIONS
Ill. 2.546—The steps (indicated by lines) to the
south-east landing, from a photograph taken in
2007: yellow = monastic steps, red = earlier
lighthouse steps, and blue = later lighthouse
steps. The blue shading indicates the area
blown away when the latest steps were
constructed. The flat storage area created at
the base of the steps is indicated by a blue
arrow (A.R. Hayden).
[Below left] Ill. 2.547—The worked ledge
leading east from the south steps to the latest
lighthouse steps to the south-east landing
(A.R. Hayden).
[Below right] Ill. 2.548—The base of the latest
lighthouse steps to the south-east landing
(A.R. Hayden).
325
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
steps here is also a clearly
worked linear area extending
eastwards, showing that the
route led to the cliff at the
west side of Cross Cove.
Ill. 2.549—
When these steps were
Detail of
constructed, the earlier line
tooling on
of lighthouse steps (which lie
latest lighthouse
steps (A.R.
on top of a ridge to the east)
Hayden).
were partly blown away, and
there are many blast holes
visible that clearly evidence
this work. This blasting was
undertaken to create a large,
flat area at the base of the
new steps. This is similar to
several flat rock ledges at the base of the south steps, which were enclosed with iron railings. These flat
areas were clearly intended as temporary storage areas for material offloaded from, or for loading onto,
ships.
The earlier lighthouse steps
The earlier (easternmost) set of lighthouse steps take a different route but run to the same destination.
The topmost section was removed, probably during the construction of the helipad. The three or four
uppermost surviving steps survive to the east of and a few metres below the helipad. They run down to
the south for a short distance to the cliff at the eastern edge of the area blasted away when the large, flatbased storage area was created.The steps then turned to the east, but the corner of their line is completely
gone.They reappear heading east a few metres lower down.These steps are quite rough and almost square
[Left] Ill.
2.550—The
middle section
of the earlier
lighthouse
steps. The top
of this line was
later blown
away (A.R.
Hayden).
[Right] Ill.
2.551—The
finely carved
lower end of
the earlier
lighthouse
steps. These
steps retain
tooling identical
to that on the
later lighthouse
steps (photo:
A.R. Hayden).
326
THE EXCAVATIONS
in shape, and they extend down the south-western side of the narrow and sloping rock spine that forms
the western side of Cross Cove.The steps then turn to the south once again, and the final south-running
flight is wider and of much better quality. They are carved as a single staircase, as opposed to individual
steps. They also display the chisel-pecking that is typical of lighthouse construction.
The monastic steps
The construction of the middle section of the surviving part of the earlier line of lighthouse steps cut
through an earlier flight of steps, of which only a short length lie at the top edge of the rock ridge. The
stumps of the inner side of the six or seven uppermost steps and the full width of the lowest step are all
that remain of the early line, and they run downwards to the east along the south-western side of the
rock ridge close to its top.The narrow width of the lowest surviving step and the high degree of erosion
evident on all these steps suggests that they are of early and probably monastic origin. They clearly led
down to the same point as the later steps, however.
[Left] Ill.
2.552—The
middle section
of the earlier
lighthouse steps
lie at the
bottom left of
the
photograph,
with the
monastic steps
above them to
the right (A.R.
Hayden).
[Right] Ill.
2.553—Closeup of the
surviving
monastic steps
(A.R. Hayden).
The south-east landing
The cliff at the base of these three
flights of steps frames the western side
of Cross Cove.The cove is one of the
most sheltered anchorages on the
island, and today the Skellig ferries
often tie up there when waiting for
their clients to return. At low tide a
large, flat rock ledge is exposed at the
base of the cliff on the western side
of the cove, immediately beneath the
area to which the steps lead. It seems
possible that materials could have
Ill. 2.554—The
western side of
Cross Cove
viewed from
the east at midtide. Note the
large, flat rock
near the end of
the peninsula.
The south-east
steps lead
down to the
wide notch in
the cliff above
the rock (A.R.
Hayden).
327
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
been unloaded from boats onto this ledge, or at high tide could have been unloaded directly off boats,
and hauled up the cliff.The cargo could have been temporarily stored in the large flat area created when
the latest set of steps were made. It was probably the need for such a temporary storage area that prompted
the lighthouse men to realign their steps to this landing place, as the earlier flight did not appear to have
provided access to such a facility.
The construction of the helipad also necessitated the removal and blasting of rock from the upper
part of this area. Clear blast holes are visible all around the base of the helipad.
2.4.3 A
NEW ROUTE TO THE MONASTERY
In June 2010, when checking above the east landing for loose rocks, the Irish Rope Access supervisor,
Colin McGorlick, discovered some previously unknown steps on the south-eastern side of the island.
The features discovered appear to constitute part of a new route from the sea up to the monastery. Judging
by the line the route is taking (it appears to lead towards one of the early entrances in the eastern wall
of the monastery), it is potentially the earliest set of steps built on the island and is therefore of considerable
importance.
Time constraints have meant that the entire route has not yet been examined; the lower part, where
it descends to the sea, and the upper part, where it runs to the monastery, remain to be explored.
Ill. 2.555—
LiDAR image
showing the
lowest
explored part
of the new
route to the
monastery (red
line) at the
eastern side of
the island. The
yellow lines
indicate the
possible lines of
the as-yetunexplored
lower part of
the route
(DAHG and
A.R. Hayden).
The new route has been traced down to the head of a gully, which descends almost vertically to a
small cove. It seems unlikely that the steps descend into this cove owing to the steepness of the cliffs
here. It would appear more probable that the lower part of the route ran to the south-east and met the
sea somewhere at the south-eastern corner of the island, where there is low-lying and relatively flat rock
where the island meets the sea. A prominent natural ledge, visible from the sea, extends along the slope
there. It descends gradually down towards the sea and may be part of the lower line of the route.
328
THE EXCAVATIONS
Ill. 2.556—
Photograph
taken from sea
level to the
east, showing
the features of
the lowest
explored part
of the new
route to the
monastery
(A.R. Hayden).
Ill. 2.557—
Details of the
lowest
explored part
of the new
route to the
monastery
(A.R. Hayden).
329
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
At the lowest traced point of the route there appears to be a short length of poorly preserved
terracing (1) extending on a north–south line. The outer side of the possible terrace has a clear straight
edge. From this point the route ascends a steep and wet gully (2), which is choked with debris and
covered with thick vegetation. Several clear lengths of drystone walling extend across the gully and there
are many slipped flat slabs that could have been stair treads.
[Left] Ill.
2.558—The
lowest point of
the route
reached, the
head of a steep
gully (A.R.
Hayden).
[Right] Ill.
2.559—The
possible terrace
(1) at the
lowest part of
the route
reached (A.R.
Hayden).
Ill. 2.560 and 2.561—The first gully (2); note the displaced flat stones, which may be slipped stair treads (A.R. Hayden).
330
THE EXCAVATIONS
At the head of the gully, where bedrock is exposed, there is a short flight of extremely well-preserved
and well-built masonry steps leading up to a broad, sloping ledge (3), which overlooks the east landing.
A worked stone, possibly part of a small cross, was found lying on these steps. Several large stones lying
on the ledge above the steps appear to have fallen there in relatively recent times.While there is no clear
walling visible on the outer side of the ledge, there are many smaller stones there and hence a wall could
originally have defined the outer side of the terrace.
[Left] Ill. 2.562—The well-preserved masonry steps at the
top of the gully (2), leading to the ledge (3) above the east
landing (A.R. Hayden).
[Below] Ill. 2.563—The ledge (3) overlooking the east
landing, with possible masonry visible at cliff edge (A.R.
Hayden).
[Left] Ill.
2.564—Looking
down on the
east landing
from the ledge
(3) (A.R.
Hayden).
[Right] Ill.
2.565—Crossshaped slab on
steps at top of
gully (2) (A.R.
Hayden).
From the base of the well-preserved steps the route turns at a right angle to the north and follows
natural rock ledges which may have been worked and where there are some possible rock-cut steps (4)
which lead upwards for a short distance to a large terrace (5). Here a natural rock ledge appears to have
been cleared of fallen rock and augmented with walling.
From the northern end of this terrace the route ascends another steep, wet and vegetation-filled
gully (6). Here again there are several short lengths of walling and many flat rock slabs, which could be
slipped stair treads. At the top of the vegetated part of the gully the route turns to the north again, and
here there is a clearly evident drystone-walled terrace (7).
331
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Above] Ill. 2.566—Terrace (5) from the north (A.R.
Hayden).
[Right] Ill. 2.567—Slipped stair treads in gully (6), with
terrace (7) above (A.R. Hayden).
[Left] Ill.
2.568—Looking
up towards
terrace (7) and
the top of the
eastern side of
the route (8)
(A.R. Hayden).
[Right] Ill.
2.569—Rockcut step and
terrace leading
up to the
masonry steps
at the top of
the eastern side
of the route (8)
(A.R. Hayden).
From the northern end of this terrace the route ascends straight along a very narrow rock ledge at
the side of the cliff (8). There are many rock-cut steps on the ledge. On a slightly lower ledge outside it
there is clear drystone masonry, which extends up and partly covers the inner ledge in places. Several
masonry steps survive and clearly there was once a long flight of drystone steps here. It is not clear
whether the rock-cut steps represent the first access here, which was later improved by the construction
of the masonry steps, or whether they were simply cut to secure the inner edge of the masonry steps.
332
THE EXCAVATIONS
The construction here, utilising a lower ledge to found the base of the stairway/path, has very close
similarities with some of the work on the South Peak.
There is a fine handhold on the rock cliff on the inner side of these steps about two-thirds of the
way up.
[Left] Ill. 2.570—Close-up of rock-cut steps (8) above terrace (7) (A.R. Hayden).
[Right] Ill. 2.571—Masonry on ledge outside rock-cut steps (8) (A.R. Hayden).
[Left] Ill.
2.572—
Surviving
masonry on
top of rock-cut
steps (8) (A.R.
Hayden).
[Right] Ill.
2.573—Carved
handhold on
cliff near top of
steps (8) (A.R.
Hayden).
333
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
At the top of this stairway there are a number of exceptionally well-preserved masonry steps, which
stand on top of fine drystone walling where the route enters a narrow cleft in the cliff.
From this point the route turns west at a right angle and ascends a roughly 2m-high cliff (9). Here
there are clear but very heavily eroded rock-cut steps and possible handholds. The steps lead up onto an
artificial ledge, which has been cut from the cliff face.
[Left] Ill.
2.574—
Masonry steps
above terrace
and steps (8)
(A.R. Hayden).
[Right] Ill.
2.575—Closeup of masonry
steps at top of
climb (8) (A.R.
Hayden).
[Left] Ill.
2.576—Looking
down along the
rock-cut steps
and masonry
(8) to terrace
(7) and gully (6)
(A.R. Hayden).
[Right] Ill.
2.577—Colin
standing on
rock-cut ledge
(9), with
masonry steps
of top of (8)
visible below
(A.R. Hayden).
334
THE EXCAVATIONS
From this point the route is not clear for a short distance. There are a number of possible lines that
it could have taken to extend over the rounded top of a rock ridge. The going here is easy, and so very
little alteration of the natural landscape may have been necessary to facilitate transit.
The route shortly reappears and there are three flat, vegetation-covered ledges (10–12), which look
as if they have been cleared of fallen rocks and which have possible drystone walling defining their outer
Ill. 2.578—The
features of the
upper explored
part of the
route and
possible routes
to the
monastery
above (A.R.
Hayden).
[Above] Ill. 2.579—Looking north towards the cut-back
rock (13) (at left) and terraces (10–12) (A.R. Hayden).
[Right] Ill. 2.580—Terrace (12) leading to the cut-back
rock (13) from the east (A.R. Hayden).
335
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
edges. Towards the western end of the third terrace there is a large protruding spine of bedrock, which
originally must have blocked the route. Its lower end has been clearly cut back to allow passage (13).
Beyond the rock there is a small possible length of terracing (14) crossing a vegetation-covered gully.
The route from this point onwards has not been examined, but from here it appears to have ascended
towards the earliest entrance into the monastery, which lies at its eastern end. It is very unlikely that the
route ran to meet the east steps, as owing to the topography it would have to descend several times before
again rising to meet the east steps. None of the other routes to the monastery do this.
There appear to be several possible relatively easy approaches from the explored end of the route
up to the large ledge beneath the eastern end of the monastery.
[Left] Ill.
2.581—Closeup of the
western side of
the cut-back
rock (13) (A.R.
Hayden).
[Right] Ill.
2.582—
Terraces (12
and 14) and the
cut-back rock
(13) from the
east (A.R.
Hayden).
Discussion
The newly discovered route is clearly of early date, as not only are the steps very ruinous and have clearly
been abandoned for a considerable period but also they appear to lead to the earliest entrance into the
monastery. They could, then, be the earliest route built by the monks on Skellig Michael.
The route leads up from the sea somewhere at the north-eastern corner of the island. There the
bedrock slopes down to the sea at a relatively gentle angle and, perhaps unlike any other part of the
island, there are no high cliffs where the land meets the sea. This, then, was probably the easiest place to
climb onto the island from a boat. This location provided little shelter for boats, however, and hence
would have been accessible only at limited times. The other landing places on the island (the east, south,
south-east and north landings) were much more sheltered but all required considerable effort and
expertise to render them usable, as they were only accessible down very steep cliffs.
When these other routes were opened to the more sheltered harbours on the island, there would
have been little point in retaining the newly discovered route, and hence it may have been abandoned
quite early in the history of the monastery.
It is possible that the east steps also originally led up to the earliest entrance in the monastery.There
is a very prominent right-angled corner at the top of the wide gully that they ascend about three-quarters
of the way up their route. The angled corner is different from any of the other corners on the east steps,
336
THE EXCAVATIONS
which are all more sinuous. It is possible that the east steps may originally have run straight on upwards
instead of turning this corner. From this point they could also have led up to the earliest entrance into
the monastery.
This could further suggest that the east steps, which led to a sheltered anchorage, replaced the newly
discovered route. The lowest part of the east steps represents one of the greatest engineering and
construction achievements of the monks, as they were built up an almost vertical cliff face. They do,
however, lead down to one of the most sheltered landing places on the island, which is still in use today.
If the east steps did originally run to the early eastern entrance in the monastery, the realignment
of their upper part would have happened after the entrance was moved to the southern side of the
monastery. These southern entrances are not accessible from the newly discovered route and this would
provide another reason for its abandonment.
337
3. THE FINDS
3.1 INTRODUCTION
The finds assemblage from all phases of excavation on Skellig Michael is comprised of approximately
380 artefacts, including the fragments of stone crosses. With the exception of seven stone crosses found
on the South Peak, all artefacts were recovered from the monastery and associated access steps.
Approximately 30% of the artefacts belong to the early medieval or monastic period, with the remainder
dating primarily from the nineteenth/twentieth centuries, the period of occupation by the lighthousebuilders. During the supervision of works by Paddy O’Leary in 1988 in the lower monks’ garden and in
1990 by Claire Cotter in the upper monks’ garden a number of nineteenth/twentieth-century artefacts
were found. These comprised pottery, clay pipe fragments and pieces of iron, and are described in the
site records. As these particular objects cannot be located at present, they are not included in the finds
catalogue (Appendix I) but are described in the relevant section of the excavation report.
The artefacts recovered during the 1986–7 excavations were numbered using the National Museum
of Ireland prefix E338.The registration number changed, however, in 1990, when the excavation licence
number and finds registration numbers were combined, with the result that artefacts recovered in
subsequent years were registered using the prefix 93E195.
Individual specialists have analysed specific groups of artefacts, e.g. medieval pottery, post-medieval
pottery etc., and Julie Franklin of Headland Archaeology Ltd compiled the general section on small finds,
which was subsequently edited by Ann Lynch. Analysis was carried out using a hand-lens and
stereomicroscope (magnification up to x 40) where necessary. Relevant parameters have been recorded
on a find-by-find basis. The most interesting finds have been presented as an in-text catalogue while
others are described as a group. A complete catalogue of all the finds is included in Appendix I. For items
of stone and skeletal material the identification of the materials used is noted after the description, and,
where appropriate, the likely source. The geological identifications are by Michael O’Sullivan, Skellig
Michael project geologist. Measurements have been recorded and these are consistently in millimetres
(mm), with the following abbreviations used: length (L), height (H), width (W), thickness (T), diameter
(Dia). Context notes are given at the end of each catalogue entry.The report has been ordered under the
following headings: medieval pottery; post-medieval pottery; clay pipes; crucifixes; coins; copper alloy;
lead alloy; iron; ironworking waste; bone and antler objects; stone; glass; leather and miscellaneous.Within
certain groups the finds are divided by period: monastic period (seventh–thirteenth centuries); lighthouse
period and later (1820–present).
3.2 THE MEDIEVAL POTTERY (Ills 3.1, 3.2)
Clare McCutcheon
Introduction
A total of 21 sherds of medieval pottery was recovered during the excavations. Following identification
and some reassembly this was reduced to nineteen sherds.
Methodology
The pottery was identified visually and the information has been entered onto an Access database table,
as per the requirements of the National Museum of Ireland.The pottery identification presented in Table
3 shows the quantity of sherds in each fabric type and the minimum number of vessels (MNV), an
objective number based on the presence of rim/handle sherds in the assemblage. The more subjective
339
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
minimum number of vessels represented (MVR) is also listed and is based on the numbers of diagnostic
pieces, such as differently shaped rims, quantity of handles etc. The most likely form of the vessels
represented by the sherds and the known date of distribution of the fabric type are included in the table.
Table 3—Medieval pottery identification, Skellig Michael (93E195).
Fabric type
Sherds
MNV
MVR
Form
Date
Ham Green A ware
Ham Green A/B ware
Ham Green B ware
Orléans-type ware
Saintonge green-glazed
Saintonge sgraffito
Local medieval ware
Total medieval
1
4
2
2
3
1
6
19
–
–
–
–
–
–
–
–
1
1
1
1
1
1
2
8
Jug
Jug
Jug
Jug
Jug
Jug
Jug
1120–1160
1160–1180
L12th–M13th
L12th–E13th
13th–14th
L13th–E14th
13th–14th
Ham Green wares (Ill. 3.1)
The kiln at Ham Green outside Bristol, and the ware produced there, has been extensively described
(Barton 1963; Ponsford 1991). Two hand-built glazed wares were produced, i.e. Ham Green A, dating
from c. 1120–1160, and Ham Green B, dating from c. 1175–1250, as well as a cooking ware that appears
to have been contemporary with both glazed wares (Ponsford 1991, 98).
Ham Green A ware
A single body sherd (93E195:197) has horizontal ridging with the addition of diagonal slashes in a
herringbone pattern. Damaged in hut fire, probably from lower monks’ garden, F1019.
Ham Green A/B transitional ware
Three body sherds (93E195:1, 93E195:2, 93E195:18) are decorated with a combination of vertical and
horizontal lines, lightly incised. The fourth sherd (93E195:33) is a large, well-glazed sherd with three
horizontal ridges at the neck and a pattern of horizontal combing around the body, interspersed by at
least four bands of vertical lines. Two sherds were found in deposit F633(5) and one (:33) was found in
wall F637 in the lower monks’ garden, while one sherd (:18) was recovered from the fill (F615) of Cistern
3, which is a mixed/modern deposit.
Ham Green B ware
Two sherds, consisting of a body (93E195:110) and a base (93E195:303). The clay matrix of the body
can be differentiated from the other Ham Green bodies although the body itself is undecorated. The
base has the thumbing typical of Ham Green B ware. One sherd (:110) was found in a spoil heap derived
from the lower monks’ garden and the other (:303) was found in the lighthouse period wall (F500) above
the east entrance.
‘Orléans-type’ (Ill. 3.2)
This ware has been recovered consistently in Cork (Gahan et al. 1997, 122–3, fig. 51.6; McCutcheon
2003, 210), Waterford (Gahan and McCutcheon 1997, 306, fig. 11.10.10, 11) and Dublin (McCutcheon
2006, 99–101, pl. xxii, fig. 42), although to date no exact provenance has been found in France.The name
‘Orléans-type’ was attributed following the recovery of a jug at Exeter that was compared to similar jugs
340
THE FINDS
Ill. 3.1—Ham
Green wares.
341
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 3.2—
‘Orleans’,
Saintonge and
local wares.
342
THE FINDS
found in Orléans by Chapelot (Allan 1984, 108). A similar jug, undatable, has been found at Tours (Husi
2003, fig. 6, pichet 9), but there are no recent finds from Orléans and thus the exact source is still
unknown.
The jugs are small and flat-based with a rounded belly, straight neck and no lip or spout.The hollow
rod handle is attached by being plugged into the body of the vase, with the typical French rolled-over
edge on the handle.The fabric is a distinctive white clay with deep grooving and stabbing on the exterior,
primarily on the rounded belly of the jug.The bright green glaze covers the belly and handle, extending
up to the flat-topped rim but not down to the base.
Two sherds were recovered at Skellig, a rim and a body sherd. The rim (93E195:200) is typical of
the vessel type, with a flat, thickened top. The body sherd (93E195:32) is decorated with two parallel
incised lines, very diagnostic of the vessel type, although the clay is not the most typical bright white
clay. One sherd (:32) was found unstratified in the garden soil of the lower monks’ garden and the other
(:200) was damaged in the hut fire but probably came from the lower monks’ garden, F1019.
Saintonge green-glazed (Ill. 3.2)
The term ‘Saintonge’ has been used as something of a catch-all in Irish ceramic studies. It is becoming
increasingly apparent, however, that a number of production centres in the wider Bordeaux area shared
similar clay and forms, thus making it difficult to distinguish on the basis of chemical analysis. It may be
that certain vessels will be recognised in the future by their individual decorative motifs as coming from
a particular area.
The mainstay of the Saintonge wares are tall standard jugs with minimal decoration, flat splayed
bases, strap handles and applied spouts. As with the majority of the French wares, the clay is off-white,
micaceous and containing quartz and haematite.The application of a lead glaze containing copper filings
leads to the mottled effect that constitutes the primary decoration. Vertically applied thumbed strips
(usually only three) or lightly incised horizontal lines are the only other decoration.
Sherd 93E195:107, from the left side of a spout, has an incised mark on the interior. This consists of
a quartered circle that was cut after firing. While such marks have been called makers’ marks (Dunning
1968; Brown 2002, 27), there is stronger evidence for calling them owners’ marks (McCutcheon 2006,
114–18). Since the purpose of an owner’s mark is to distinguish one jug from a number of similar vessels,
in the case of the assemblage at Skellig Michael this would seem to make no sense. As the glazed jugs
were most probably gifts rather than purchases, however, it is likely that this jug had already been in the
possession of one careful owner on the mainland.
A body sherd (93E195:196+:202) is very typical of Saintonge ware, with no decoration other than
the glaze. The final sherd (93E195:3) is an unglazed portion of a base, not splayed.
Two of the Saintonge sherds (93E195:196, 93E195:202) were damaged in the hut fire and probably
came from the lower monks’ garden, F1019. Sherd 93E195:107 was found during cleaning of the east
steps, and sherd 93E195:3 was retrieved from redeposited soils (F650(5)) in the lower monks’ garden.
Saintonge sgraffito (Ill. 3.2)
A portion of spout (93E195:345) from a sgraffito decorated jug was recovered. These vessels date from
the later thirteenth century to the early fourteenth century. This sherd most closely resembles several
sherds from the Guernsey wreck, where ‘an overall red slip, undefined by carving at the edges, has been
scratched through in a variety of linear motifs including chevrons and cross-hatching’ (Thomson and
Brown 1991, 68, fig. 11). A similar sherd was recovered at Wood Quay, Dublin, and examples have also
been recovered in Waterford, Cork and Galway (McCutcheon 2006, 124–5, fig. 50.9).
This sherd was found in a layer of collapse and stones (F638(1E)) in the lower monks’ garden.
343
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Local medieval ware (Ill. 3.2)
No specific locally made medieval pottery type has to date been recognised from Kerry. It is possible
that the Skellig assemblage was sourced in the Limerick area but there is nothing very diagnostic about
the sherds recovered.
Four sherds (93E195:6, :34, :35, :105+:106), all found in the lower monks’ garden, appear to be from
the same vessel. Sherd 93E195:6 was recovered from the top of steps in Cutting 6, and sherds 93E195:34,
:35 and :105+:106 came from redeposited soils F635. They have a pink clay with a beige interior. Some
organic nodules were burnt out, leaving grey void surrounds on both surfaces, with shell, stone and iron
visible in the clay. One sherd (93E195:34) has a green glaze, possibly just by the shoulder of the jug, with
a narrow, double incised line decoration.
The other two sherds (93E195:7, 93E195:199) have been included with local medieval ware by
default as they are neither French nor English. Sherd 93E195:7 was a surface find from the east steps and
sherd 93E195:199 was damaged in the hut fire but probably came from the lower monks’ garden, F1019.
Discussion
The assemblage is remarkable, given the nature of the site—an isolated island community of monks living
an austere life in beehive huts. Given the wide date range of the material, from the mid-twelfth to the
early fourteenth century, this suggests a long-term if occasional import of jugs into the community. If
the single sherd of Saintonge sgraffito is excluded, then the date range of the assemblage could be reduced
to a possible 50-year period from the later twelfth to the early thirteenth century, still representing several
generations. It may be that visitors on pilgrimage to the island brought the new jugs as gifts to the
community. Did the material come as a group directly from the most likely urban centres, such as Cork
or Limerick, with other supplies, or were these special items of value to individuals and so offered up to
the holy men on the island? In spite of the relative inaccessibility of the island of Skellig, there appears
to have been no bar to the use of these items in such a community. Similar contemporary material was
recovered at the Celtic hermitage on St Helens, on the Isles of Scilly (Dunning 1964).
3.3 THE POST-MEDIEVAL POTTERY (not illustrated)
Rosanne Meenan
An assemblage of 178 sherds of post-medieval pottery from Skellig Michael was examined.The minimum
vessel number (MVN) is estimated on the number of rims present in the assemblage unless otherwise
stated. The sherds were recovered primarily from the nineteenth/twentieth-century levels in the lower
monks’ garden, including in and around Cell G, and from the surface campion layer. Additional sherds
were recovered from the east steps, south entrance 2 (inner enclosure) and from the floor surface within
the large oratory.Three sherds were recovered within the peat adjacent to the ‘latrine’ on the small oratory
terrace. The assemblage has been treated as a single unit for discussion purposes.
Late seventeenth-century pottery
Two sherds of North Devon gravel-free ware represent the earliest material in the post-medieval
assemblage. They came from a thin-walled vessel such as a jug, which was presumably brought onto the
island by visitors/pilgrims at the time when it was most commonly in use, i.e. in the late seventeenth/early
eighteenth century.
Nineteenth-century pottery
The assemblage is predominantly nineteenth-century in date. A small number of coarseware sherds are
344
THE FINDS
Table 4—Post-medieval pottery identifications.
Ware
19th-century stoneware
Banded ware
Black-glazed ware
Earthenware (Staffordshire)
Glazed red earthenware
Hand-painted fine ware
Mocha ware
North Devon gravel-free
Pearlware
Porcelain ?
Shell-edge ware
Transfer-printed ware
Unidentified
White earthenware
TOTAL
Sherd
count
10
12
3
3
5
4
1
2
3
1
30
25
4
75
178
MVN
Forms
2
4
2 bottles
Bowls (identified by base)
1
1
1
1
Small food storage jar
Bowl/jar
Bowl
Bowl
3
3 saucers (identified by base)
11
8
11 plates
2 saucer bowls, 2 cups, 1 bowl, 2 saucers, 1 plate
17
50
12 plates, small storage jar, saucer, 2 bowls(?), jug
present but the assemblage is predominantly made up of fine tablewares, i.e. cups, saucers, jugs and plates,
in wares that were probably made in Staffordshire potteries.
Stoneware
Two bottles were recognised in the stoneware assemblage by their rim/necks (93E195:20, 93E195:55).
These are unusual vessels in that they resemble in form squat wine bottles of the mid-eighteenth century.
One of them (:20) features a cordon below the rim, while the other (:55) features a cordon around the
edge of the rim. Neither was glazed on the interior and the fabrics appear to be quite different.
Body sherds of storage vessels were also present.
Black-glazed ware
One (93E195:103) of the three sherds of black-glazed ware probably derives from a storage vessel.These
sherds, with matt glazes, do not resemble the black-glazed wares that were imported from west England
and from Wales and may have been manufactured more locally in Ireland.
Glazed earthenware
One sherd (93E195:48) of glazed red earthenware is the rolled-over rim from a possible bowl or jar;
another piece (93E195:204) appears to have been worked into a gaming piece. These were probably of
Irish manufacture. There was also a rim (93E195:15) from a small food storage jar, which featured a
groove under the rim into which a piece of string to tie the lid would have fitted.This piece was factoryproduced, possibly in Britain.
Fine earthenwares
White earthenwares. The largest single group comprised white earthenwares. These were tablewares, with
plates being the commonest form. Bowls, saucers, jugs and a small food jar were also present. A small
number were decorated with raised ridges or moulded patterns and there were also a small number of
sherds decorated with very sparse amounts of paint, e.g. single lines of colour around the rim. It should
be pointed out that some of the undecorated sherds may have come from vessels that were decorated on
other parts of the vessel.
345
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Shell-edge ware.There is a large group of shell-edge ware sherds. These were all from plates with green or
blue paint applied over the moulded decoration on the rims. Two of the rims (93E195:64, 93E195:65)
can be dated to the first three decades of the nineteenth century, while others (93E195:63, E338:1)
featured embossed decoration along the rim under blue paint and have been dated to the years 1820–
30 in America, where Orser has noted that shell-edge wares were the cheapest forms of decorated
tableware by the 1820s (Diagnostic Artifacts in Maryland, www.jefpat.org/diagnostic). Rim sherds from
another plate (E388:2) may date from as late as the 1860s.
Transfer-printed wares. These wares are also well represented.The vessel forms are mixed but again comprise
tablewares. The patterns are mixed and blue is the predominant colour. Three sherds from one saucer
bowl were examined by Audrey Whitty (National Museum of Ireland), who identified them as dating
from approximately 1820.While she could not identify the pattern exactly, the closest parallel was Broseley
Pattern A, dating from approximately 1820–30 and manufactured by the Don pottery of Swinton,
Yorkshire (A. Whitty, pers. comm.). This is a variation of the Willow pattern and was also known as the
‘Two Temples pattern’. It was also used on a porcelain body made by Spode.
The patterns used on the other sherds were not identifiable.
Other finewares. The remaining finewares are much less numerous. Banded wares were factory-made, using
different colours of slip on the exterior, which were worked by various methods to produce different
decorative motifs. One such variation was mocha ware.These wares were produced from the later decades
of the eighteenth century through the nineteenth century and again filled the lower end of the market
with mass-produced table goods.They were in full circulation in the first decades of the nineteenth century.
Pearlware. Three sherds of pearlware represented three different saucers and were hand-painted in
combinations of blue, brown and yellow. The sherds of hand-painted fineware were also painted in this
palette. One sherd was identified as being possibly porcelain, albeit very coarse.
Discussion
Two sherds of North Devon ware indicate some activity on the island in the late seventeenth/early
eighteenth century, but otherwise the assemblage dates predominantly from the first decades of the
nineteenth century. Those sherds that could be dated by type of ware or by the decorative techniques
slotted into the years 1800–40, with just one plate dating from c. 1860.
The pottery was recovered from the excavations within the monastery. It is known that the builders
of the lighthouses used some of the buildings within the monastery as dwellings during construction of
the lighthouses, which took place c. 1820–60. It is probable, therefore, that the pottery assemblage was
associated with the construction workers.
The wares that are present in the assemblage are relatively fine and therefore fragile and perhaps
would not have been expected to be used by a group of construction workers in such difficult conditions,
particularly as breakages would not have been easily replaced. It is, however, becoming clear from
archaeological evidence in Britain and further afield that mass-produced plain white earthenware, transferprinted wares and shell-edge wares were coming into common usage even in the most remote locations.
They have been found on South Uist in the Hebrides (Parker Pearson et al. 2004, 181) in early
nineteenth-century contexts.They have also been found on St Kilda, in Tasmania and the Falkland Islands
in contexts dating from the 1820s and onwards (Alasdair Brooks, pers. comm.). Shell-edge wares in
particular were inexpensive and breakages would not have been costly to replace.
The construction workers might have brought with them wooden or enamel vessels that would
have been more durable than ceramic examples. When describing island households on Great Blasket
346
THE FINDS
Island during his childhood in the 1850s and 1860s, Tomás Ó Criomhthain (Ó Crohan 1979, 31) wrote
that they used bowls and plates and wooden mugs. Unfortunately he does not record what the bowls
and plates were made of, but it is interesting to note the use of wooden mugs at this late date. It is
tempting to suggest that wooden vessels may have been used by some of the lighthouse-builders, but
mugs and plates in wood would not have survived well in the archaeological record and therefore it is
not possible to determine whether they were brought onto the island at the time in question.
Ó Criomhthain (ibid., 32) also wrote that tea was not known on nearby Great Blasket when he was
growing up in the 1850s–1870s. This begs the question of how much tea was drunk on Skellig Michael
in the 1820s, although it is clear that the construction workers would have come from the mainland,
where tea had probably appeared at an earlier date than on Great Blasket. There is evidence for some
cups in the assemblage, although the other table forms are proportionately more numerous.
The question of provisioning the construction workers is interesting, as all their foodstuffs had to be
imported from the mainland; possibly wooden barrels containing preserved meat etc. were brought onto
the island. The low proportion of black-glazed ware suggests that there was little use of ceramics for
food-processing or storage. The black-glazed ware and the stonewares represent the only storage vessels
present in the assemblage.
3.4 THE CLAY PIPES (not illustrated)
Joseph Norton
The number of clay pipe items recovered totalled eight stems and four bowl fragments, a surprisingly
small assemblage given that the lighthouse-builders resided in the monastery in the nineteenth century.
The stems comprised eight plain pieces and four mouthpiece fragments. All the stem material dated from
the nineteenth century. The stem pieces are all very short lengths and have no diagnostic features. One
mouthpiece (93E195:124) has traces of a red substance on it, possibly the remnant of a protective slip,
sometimes applied to this part of the stem to prevent burning when smoking.
There are four incomplete bowl fragments, two of which (93E195:37b and c) join to make a partial
bowl of early nineteenth-century date, i.e. c. 1800–30.
There is one almost complete spurred bowl (93E195:90), which has a very small spur and dates from
the first half of the nineteenth century, probably c. 1800–30. There is another incomplete bowl
(93E195:340) with a slightly larger spur; it too dates from c. 1800–30.
The clay pipe fragments were recovered primarily from the upper levels in the lower monks’ garden
(particularly from around Cell G) and from a nineteenth-century layer in the south entrance 2, inner
enclosure, cutting.
3.5 CRUCIFIXES
Two crucifixes were recovered during the excavations and both date from the nineteenth/twentieth
centuries.
E338:38 (Ill. 3.3)
A wooden crucifix on a stepped base (H 420mm).The figure is unusually corpulent and crudely carved,
with the head slightly inclined to Christ’s right and arms outstretched. The body is painted in a skin
tone, the hair is dark brown (no crown of thorns evident), the loincloth is white, and splashes of red paint
along Christ’s left side represent blood from the wound under the right breast. The cross and base are
347
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill. 3.3—
Wooden
crucifix
E338:38. (Con
Brogan,
DAHG).
[Right] Ill. 3.4—
The altar in the
large oratory in
the 1950s. The
lintel of the east
window is just
visible
in the
background
(DAHG
archive).
painted black, and the individual pieces are nailed
together with small iron nails or tacks.The big and badly
modelled loincloth (flowing to Christ’s right, which is
unusual) is typical of figures as early as the seventeenth
century but the suppedaneum and superscription are
more probably eighteenth/nineteenth-century in date
(C. Hourihane, pers. comm.). An altar or devotional use
is indicated by the base.
An example of folk art, possibly carved by one of the
lighthouse personnel, found in a fragmentary state within
the make-up of the altar inside the large oratory. It is
similar to (but not identical to) a cross visible on the altar
in a photograph dating from the mid-twentieth century
(Ill. 3.4).
Ill. 3.5—Wood
and bronze
crucifix
93E195:206
(Con Brogan,
DAHG).
93E195:206 (Ill. 3.5)
A wood and bronze crucifix (H 205mm). The nimbus
and superscription and the style of the figure indicate a
date of c. 1900–50 (C. Hourihane, pers. comm.). Found
in sea campion layer in the lower monks’ garden (F631),
directly below the large oratory, from which it probably
emanated.
348
THE FINDS
3.6 COINS (not illustrated)
Michael Kenny
E338:23. Silver long-cross penny
English, Henry III (1216–72)
The long-cross penny was introduced in 1247 and was initially struck at numerous mints throughout
England. After 1250 production was limited to London, Canterbury, Durham and Bury St Edmonds.
This coin is extremely worn and pitted but there are some features discernible, including the outline of
Henry’s facing bust. Some letters are still readable, particularly on the reverse, which suggest that the coin
was struck at Bristol. The full legend is ELIS ON BRVSTO, ‘Elis (the moneyer) of Bristol’. Since the
obverse is so worn, it is difficult to work out the subgroup or class to which this coin belongs. According
to the numismatic publications and essays on this period, the provincial mints, such as Bristol, did not
work after 1250.This, if the basic identification is correct, gives a fairly precise minting date of 1247–50.
Large oratory, south-east quadrant, F302, disturbed floor surface.
E338:22K. Penny
Irish, 1971
Large oratory, south-east quadrant, F302, topsoil.
E338:22L. Penny
English, 1971?
Worn and corroded. Large oratory, south-east quadrant, F302, topsoil.
E338:22M. Twopence
Irish?
Very worn and corroded. Large oratory, south-east quadrant, F302, topsoil.
3.7 THE SMALL FINDS
Julie Franklin
3.7.1 COPPER ALLOY
Monastic period (seventh–thirteenth centuries)
Pins
The two copper-alloy pins recovered are arguably the most decorative items in the assemblage.
Significantly, the head of each is decorated with crosses on both sides.The pins would have secured outer
clothing such as a cloak and would have belonged to individuals of some status, possibly abbots or wealthy
visitors.
Crutch-headed pins such as 93E195:213 are almost exclusively found with stirrup rings (Fanning
1994, 11). Stirrup rings are easily parted from their heads and are often found separately, as could easily
have happened here (ibid., 9). They are a common type; a quarter of the Dublin pins in Fanning’s study
were of this type (ibid., 10).The moulded cross decoration is unusual. Ring-and-dot is the more common
technique used on this type of pin. In Dublin this type of pin was found in deposits dating from the
eleventh and early twelfth centuries, with a floruit between c. 1000 and c. 1070. Later examples may in
fact be residual or heirlooms (ibid., 41–6). Finds from elsewhere in Ireland are more thinly scattered but
349
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 3.6—Metal
artefacts.
350
THE FINDS
support the Dublin dating. A similar pin was found at nearby Church Island (O’Kelly 1959, 105, fig.
11.2).
The rounded spatulate head of stick-pin 93E195:205 places it in O’Rahilly’s class 13 (O’Rahilly
1998, 29).The evidence from Dublin dates this type of pin to the late twelfth to mid-thirteenth century.
In Waterford, however, they appeared to have a slightly earlier range, and were found in deposits dating
from between c. 1130 and c. 1200. The pin, then, is most likely to date from between the mid-twelfth
and mid-thirteenth centuries, although of course it may already have been of some age when deposited.
It is of a relatively unusual form; only three of the 232 stick-pins found in Waterford were of rounded
spatulate form. Stick-pin decoration takes many forms but is generally quite simple. Crosses often appear
on the tops of rounded pin heads, but the spatulate form here allows decoration on the side which would
have been more easily visible when the pin was in use.
93E195:213. Ringed pin (Ill. 3.6). Stirrup-ringed, crutch-headed pin. Tapering ring ends narrowing to
points, held within perforation running length of head. Ring with oval section, now a little loose, possibly
through distortion or corrosion, and can be removed from head. Panelled decoration, featuring moulded
cross motifs. On top of the head this fills the available rectangle. On the front and back of the head the
horizontal arm of the cross fills the rectangular field, while the shaft extends from the top of the head
down the upper part of the pin shaft, ending at a double incised line around the shaft. Ring divided into
four panels by three sets of double incised lines. Shaft undecorated, straight-sided, tapering towards tip,
which is slightly bent. Pin L 101, ring Dia 12. Found in void in the west jamb of south entrance 1 (inner
enclosure), prior to excavation.
93E195:205. Stick-pin (Ill. 3.6). Rounded spatulate head. Curved, pointed projections on either side at
join with shaft, with central leaf-shaped moulding extending down shaft on one side. Incised decoration
on head. On one side, cross formed from dots: central dot and three dots for each arm. On reverse, cross
formed from two bands filled with cross-hatching, crossing in middle (though partially obscured by large
corrosion blister). Lozenge-sectioned shaft towards head, bulging and changing to rounder section in
middle before narrowing to completely round tip.Very slight S-bend towards tip, probably from use. L
117, shaft max.W 4.3, head L 7, head W 8 by 2. Found in wall collapse (F636(3)) in lower monks’ garden.
Lighthouse period and later (1820–present)
Wire (not illustrated)
The wire is undiagnostic and of uncertain function. It is, however, in very good condition and was found
associated with bottle glass, and thus is more likely to date from the lighthouse period or later.
93E195:208.Wire. Length of wire bent into hairpin shape with L-bend at one end. L (straightened) 126,
T 1.3. Redeposited soils (F633(5E)) in lower monks’ garden.
3.7.2 LEAD
Monastic period (seventh–thirteenth centuries)
Writing lead
The single lead object (93E195:207) was probably a writing implement of some type. Styli were used
351
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
for writing on wax tablets, while writing leads were for parchment or paper. Styli were often of iron or
copper and could be well formed and decorated. One end was pointed, the other spatulate for erasing
and smoothing down the wax.Writing leads were generally more simply made. On paper they produced
a light grey impermanent mark, similar to modern graphite pencils. Early writing leads, such as the
ninth- to eleventh-century examples found at Winchester, are simple rods, tapering to a point at one
end. Later examples often had a flattened end for ruling lines (Biddle and Brown 1990; Egan 1998,
270–1). The Skellig artefact was blunted but was clearly tapering to a point at one end. The lack of a
spatulate head implies that it was intended for parchment rather than wax, but it is possible that the top
end is broken. A number of similar leads, though typically with flattened ends, were found in thirteenthcentury contexts at Tintern Abbey, Co. Wexford (Lynch 2010, 167), and Jerpoint Abbey, Co. Kilkenny
(Scully 2007, 391).
93E195:207. Writing lead (Ill. 3.6). Round-sectioned thick rod, bent in middle. One end tapering to a
rounded point, probably broken at tip. Other end blunt, possibly broken. L 49, W 6. Leacht area, topsoil,
F561.
3.7.3 IRON
Monastic period (seventh–thirteenth centuries)
Knife
An iron knife with the remains of a wooden handle (E338:30) was found in a clay deposit possibly
associated with the construction of the large oratory, or at the very least with the early phases of its use.
Unfortunately it is broken and very worn and provides little useful dating information. It is a whittletang knife, the form in use during the period of the monastery’s occupation. Though scale-tang knives
were introduced in the late medieval period and became the more common type by the fifteenth century
(Cowgill et al. 1987, 25), whittle-tang knives continue to be made to this day. The relatively short tang
indicates that it was not a particularly large knife. The dishing visible on the back of the blade may be
the result of using it as a strike-a-light.
E338:30. Knife (Ill. 3.6).Whittle-tang iron knife blade, single-sided, broken, very worn and blunted. Back
of blade also slightly concave, possibly through wear. Partially mineralised remains of wooden handle
survive along length of tang. Short pointed tang visible in X-ray; top level with back of blade, bottom
stepped in slightly from blade edge, tapers straight to point. L 63, tang L 36, W 20. Large oratory, southeast quadrant, in a pocket of F306, in top of deposit (F315) relating to construction of oratory.
Building material?
The only other iron finds of potential medieval date are building-related. These include a possible door
stud (E338:22J). Door studs were nails with large, domed, lozenge-shaped heads serving a dual function:
they secured two thicknesses of wood together in the construction of the door, and the domed head
provided a certain amount of protection from edged weapons during an attack.Though from a disturbed
deposit associated with modern finds, this may once have been part of the door to the large oratory.
Finally, a piece of strap (93E195:223) may be part of a large strap hinge. These finds are potentially of
medieval date.
E338:22J. Door stud? Nail with large, domed, lozenge-shaped head, broken at one end. Short stump of
352
THE FINDS
shaft. Head L 35+ (estimated 38), head W 28, shaft L 17. Large oratory, south-east quadrant, F302,
disturbed floor surface.
93E195:211. Nail. Nail with oval head and short stump of shaft. Head L 29, head W 24, shaft L 17. Leacht
area, F577, possibly contemporary with Burials 1 and 2.
93E195:223. Hinge strap? Fragment of thick strap. Fused to stone.W c. 32. Redeposited soil (F635(7)) in
lower monks’ garden.
Lighthouse period and later (1820–present)
93E195:349. Bayonet (Ill. 3.6)
Handle and part of blade of bayonet. Single-edged fullered blade, broken at end. Cross-piece with sweptforward quillon ending in a rounded finial, muzzle ring missing. Handle of thick cast iron, cast in one
piece with blade. Thick cast-iron pommel, thinning towards blade to allow for attachment of grips of
black, hard rubber-type substance held by two iron rivets, with a long leaf spring attached by means of
the rivet closest to the blade. Overall L 243+. Unstratified.
This bayonet is something of a mystery. The distinctive lobed shape of the end of the grip as well as
the long leaf spring and fullered blade identify it as an Italian weapon, a Sciabola-Baionetta M70, model
no. BE7017. These were adopted by the Italian army in 1870, remaining unchanged until 1887. Despite
various improvements, some of these bayonets were still in service in World War I, when the blades were
shortened from the original 518mm to around 235mm (Carter 1974, 69–71, illus. 139, 142). The blade
of this example only survives to 116mm and thus it is unclear whether it has been cut down or not.
How a late nineteenth-century Italian weapon came to be on Skellig Michael is a matter of
conjecture. The fact that it is specifically an anti-personnel weapon—as opposed to, for example, a rifle,
which could be used to shoot wildlife—makes it even stranger. The shaped grip does make it a useful
hand-held weapon when detached from a rifle, but the blade, even in its cut-down version, would have
been a little unwieldy for hunting purposes.There was a large wave of emigration in the late nineteenth
century from Italy to more prosperous countries, though it is difficult to see how such a journey would
end at the Skelligs. It may have been in the possession of one of the lighthouse-keepers, possibly an exnavy or military man who may have brought it back from his travels or who may have fought on the
side of the Allies during World War I.
93E195:336. Cauldron (not illustrated)
Cast-iron body sherd.Thick-walled sherd, curving in two planes. Large bump on exterior, possibly blister,
decorative feature or repair. L 55, W 47, T 6. Surface campion layer (F631(3)) in lower monks’ garden.
As the cauldron sherd is of cast iron it is unlikely to pre-date the eighteenth century (Butler et al.
2009, 1). It was presumably in use during the occupation of the site in the 1820s by the lighthousebuilders.
Miscellaneous
Other iron finds are less easy to identify. There appear to be some fragments of thin-walled round
containers, probably food cans (93E195:219, 93E195:224). Canning was developed at the beginning of
the nineteenth century and by the 1820s canned food was beginning to be used on long sea voyages. It
is possible that the lighthouse-builders would have made use of this innovation during their stay. A Tshaped piece of iron (E338:19) is probably part of a strap hinge. A large rod (93E195:209) is of uncertain
function. Other finds are too fragmentary for identification. Some of these finds are clearly of modern
353
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
date, and some are assumed to be so by association with modern bottle glass and other finds.
Ironworking waste (monastic period)
(with identifications by Kath Crooks)
There were five small lumps of ironworking waste (totalling 255g). Four of the pieces are of slag
(93E195:214–217), three of which were recovered from redeposited soils in the lower monks’ garden
while the fourth (:215) came from the surface campion layer (F1011) at the east end of the same area.
One piece, also recovered from redeposited soils (F635(6)), with possible hammerscale embedded in the
surface (93E195:221) is a clear indication of smithing. The other piece is a fragment of ceramic furnace
lining (93E195:311) found in a mixed deposit (F1020) adjacent to Cell G. These all appear to suggest
medieval smithing taking place within the monastery. There is no evidence, however, of the kind of
metalworking workshop that was in evidence at Illaunloughan (Marshall and Walsh 2005, 202–7). Indeed,
if there was a forge in use on the island for any period of time, more waste would be expected. It may
be that this waste is from a temporary workshop to provide fixtures and fittings for monastic building
works.
3.7.4 BONE AND ANTLER
(identification of skeletal materials by Claudia Tommasino Suárez)
Monastic period (seventh–thirteenth centuries)
Comb
A single comb sideplate (E338:16) was recovered as a surface find and is described according to the
terminology outlined in Dunlevy’s (1988) typology. It is from a single-sided composite comb of type F
or G. The distinction between the two types comes chiefly from the shape of the spine, arched for type
F, straight-backed for type G.The fragment has a slight curve but the piece is too small to place it in type
F on this basis alone. Instead, the relatively flat section of the plate and the fineness of the teeth evident
from the tooth-cutting marks on the plate are more characteristic of type G combs.Tooth-cutting marks
on the sideplates are also a common feature of type G combs. Decorated examples (termed type G
ornate) also commonly feature well-decorated spines, an otherwise quite unusual feature. Two examples
from Dublin feature very similar decoration (E71:2538, E81:5685; Dunlevy 1988, 406). Type G ornate
combs are dated to between the eleventh and thirteenth centuries. Finds of combs from twelfth- and
thirteenth-century contexts in Cork and Waterford show the predominance of type G combs in the
south of Ireland in the twelfth and, particularly, thirteenth centuries (Hurley 1997a, 654–6; 1997b). The
change to straighter combs is thought to be due to the decreased supply of antler and hence the increased
use of bone, a material which lends itself better to long straight shapes (Hurley 1997a, 654).
E338:16. Comb (Ills 3.7, 3.8). Sideplate from single-sided composite comb, probably of Dunlevy’s type
G ornate. High, flat section. Broken at both ends at rivet holes. Remains of iron rivets 32mm apart, one
present, one represented by staining. Slight curve to back of plate, teeth-cutting marks visible on opposing
side, fine teeth (ten per cm). Incised decoration of row of dot and double circle along front of plate, with
criss-cross along spine. Post-cranial long bone, probably a tibia from a large animal, probably cattle, horse
or red deer. L 46, W 17, T 4. St Michael’s Church, surface find.
93E195:348. Part of handle? (Ill. 3.8)
Fragment of worked bone (possible rib) with the remains of two iron rivets still in place. Slightly curved,
354
THE FINDS
Ill. 3.7—Bone
comb fragment
E338:16 (Con
Brogan, DAHG).
Ill. 3.8—Bone and antler artefacts.
355
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
with saw marks visible on the flat (back) surface. This could be the sideplate of a handle or similar
composite artefact. L 45, W 19 (max.), T 1–2. Lower monks’ garden, lower collapse (F638(1E)).
Gaming piece
A carved antler gaming piece (93E195:228) was recovered from the lower monks’ garden. It is of unusual
form compared to typical gaming piece assemblages of the period (cf. Hurley 1997a, 666–9). It appears
to have been flat-bottomed but the split base may hide the remains of a peg hole or possibly an integral
peg (cf. MacGregor 1985, fig. 71t–u;Turner and Orton 1979). Indeed, the forming of such a feature may
have caused the split. Arguably, without a peg the piece is a little too light and unstable to make an
effective gaming piece. Pegged pieces were for use with perforated gaming boards such as that found at
Ballinderry, Co. Meath (Murray 1951, 59), and would have been particularly useful while travelling.
It is most likely to be either a hnefatafl piece or a stylised chess piece. The former was the favoured
game during the Viking period (MacGregor, 1985, 135); the latter reached northern Europe in the later
tenth century and rose in popularity through the medieval period (Chapman 2005, 1). In the context of
chess the piece might represent a stylised knight.The top of the piece, with a conical tip above a faceted
section, resembles a helmeted head. As well as the more realistically carved chess pieces, most famously
represented by the thirteenth-century Lewis Chessmen from the Western Isles of Scotland (Caldwell et
al. 2009), there was a strong tradition of simple stylised pieces produced for the lower end of the market
(Chapman 2005). Though the horse is often an integral part of the knight (ibid., figs 2.7, 2.11), several
examples from Lewis show just an armed and helmeted figure (Caldwell et al. 2009, fig. 6).
93E195:228. Gaming piece (Ills 3.8, 3.9). Small,
upstanding artefact carved in one piece, with
cylindrical lower part and pointed upper part.Top
part takes form of a cone, tapering below to a
narrow, square-sectioned centre, formed by four
facets cut into sides. This sits on top of a wider
cylinder with a flat base, which has split down the
middle. Antler. H 30, Dia 12. Redeposited soil
(F634(5)) in lower monks’ garden.
Ill. 3.9—Antler
gaming piece
93E195:228
(Con Brogan,
DAHG).
Pin
A fish bone possibly utilised as a pin was
recovered from the lower monks’ garden
(93E195:231). The degree of polish on the shaft
suggests that it was in use for some time. Similarshaped objects have been identified as pin beaters
(small pointed tools used to manipulate threads
during weaving), but typically these are sturdier
and over twice as long (Hurley 1997a, 670–4). It
is possible that this example was used for very
delicate weaving work or may have been used to
secure a shroud or light clothing.
93E195:231. Pin (Ill. 3.8). Length of fish bone, little altered from its natural form.Tip polished from use
as a pin. Caudal vertebra from neural spine of large fish. L 43, W 4. Redeposited soil (F636(5)) in lower
monks’ garden.
356
THE FINDS
3.7.5 STONE
Monastic period (seventh–thirteenth centuries)
Stone dish/plate
The concavo-convex profile of this stone object suggests that it is part of a small plate or dish, and though
the shape is slightly irregular the diameter is estimated at just 135mm. It would thus not have made a
practical dish from which to eat. Dining vessels at this date would have been chiefly wooden bowls and
trenchers. The fine workmanship, polish and delicacy of the artefact, which is only 3mm thick (possibly
thinner in the centre), suggest a more refined use.
Its function is unclear and further research will be needed in order to find comparanda for it. It has
been suggested that it could have been used to grind pigment for ink-making, although no comparanda
have been discovered so far. The stone bowls from Nendrum are far cruder and thicker (Lawlor 1925,
136–7, pl. X) and would have been used for heavy grinding. So are the illustrated objects from excavations
on the Inishkeas in the 1950s, where dye/pigment-making seems to have been carried out (Henry 1952,
172–3).
A second suggestion that it could have been used as a paten seems unlikely. There was a progressive
move towards the use of only precious materials in a Communion context throughout the early medieval
period. Pope Zephyrinus, who died c. AD 217, ordered that wine should be consecrated not, as heretofore,
in a wooden vessel but in a glass vessel (Cronin 1963, 238). The use of wood and horn was banned by
the Council of Cealinth, and Pope Leo IV banned the use of wood, lead or glass (Smith and Cheetham
1827–1908, vol. 1, 337). Patens of the period tended to be larger, as they normally used a Communion
cake/loaf rather than the wafers which are made expressly for Communion today, and while stone was
used it tended to be of an exotic nature such as onyx or semi-precious stones (ibid., vol. 2 1570–3).
Nevertheless, stone paten fragments have been found in Merovingian contexts in Bordeaux. They are
not of precious stone, but do carry religious symbols (M.F. Ryan, pers. comm.).
While there was a conservatism about Irish monasticism, which is shown by St Gall refusing to
abandon the use of bronze chalices in favour of silver, his excuse being that St Columbanus used bronze
on the grounds that our Saviour was affixed to the Cross with brazen nails (ibid., vol. 2, 339), the general
principle of using precious materials seems to have applied (Bourke 1994, 174–5).
93E195:252. Dish/plate fragment (Ills 3.10, 3.11). Part of thin, concavo-convex disc with largely flat base,
curving up at sides. Gently bulging rounded rim that forms a flat border before curving downwards to
a gently dished interior. Surfaces smoothed to a fine polish. Fine, light purple siltstone, Skellig. Dia 135,
T 3–8. Lower monks’ garden, lower collapse, F638(1&3).
Crosses (Ill. 3.12)
Approximately eighteen stone crosses and many fragments of worked stone that may have belonged to
crosses were recovered from the excavations.
The stones have been classified according to terminology outlined by Fisher (2001, 11–18). Two of
the crosses were incised on slabs of stone of similar shape and size, with sides tapering towards a domed
top, which were probably grave-markers. Both are linear crosses, though 93E195:291 is made with
considerably more care and attention than 93E195:289. This type of cross is common in the west of
Ireland, and also in the Isle of Man, Wales and western Scotland. Such simple crosses are particularly
associated with monastic sites and burial grounds and are generally dated to between the seventh and
ninth centuries, although they have been carved at many different periods (Fisher 2001, 12, 28–31).
Similar simple crosses have been found at Inishmurray (O’Sullivan and Ó Carragáin 2008, fig. 24,
357
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 3.10—Stone
artefacts.
358
THE FINDS
Ill. 3.11—Stone dish? fragment 93E195:252
(Con Brogan, DAHG).
Ill. 3.12—Selection of stone crosses and
cross fragments (Con Brogan, DAHG).
93E195:307
93E195:308
E338:27
93E195:291
93E195:282
001.50A), High Island (Marshall and Rourke 2000, 156) and Church Island (O’Kelly 1959, fig. 8:1–2)
and are still standing in the Skellig graveyard.
Other crosses are typically very simple, falling under the heading of cruciform stones. These are all
simple cross shapes cut out of the stone, with typically straight, square-cut arms and no adornment (Fisher
2001, 17, 56–7). This type of cross has been found at Scottish sites such as Iona but is generally limited
to the more ascetic sites of the west coast of Ireland (ibid., 16, 56).They have been found at Illaunloughan
(Marshall and Walsh 2005, 9), Church Island (O’Kelly 1959, 103–4, fig. 10J) and High Island (Marshall
and Rourke 2000, 157, fig. 118–19). Other sites may have had simple crosses made of wood and it may
359
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
have been merely the lack of wood at these island sites that led to the tradition in stone. Of all these
island sites, the avoidance of decoratively carved stone seems particularly severe at Skellig, in this
assemblage in particular but also in other stones found and still standing on the island (Horn et al. 1990).
The larger examples of these cruciform stones may have been monumental standing crosses, while
smaller examples would have served as grave-markers and processional crosses. The very smallest would
have been hand-held versions and may also have been interred within graves to accompany the dead
(Thomas 1971, 112–23; Marshall and Rourke 2000, 139).
The crosses were recovered from different excavated areas—the lower monks’ garden, the leacht area
to the east of St Michael’s Church, the small oratory terrace, and from beneath the threshold of the
structure at the base of the east steps. On the South Peak they were recovered from the oratory terrace
and the various terraces and traverses.They no doubt all had individual life stories. Some may have broken
during manufacture; cross-slab 93E195:291, for example, appears to be a grave-marker but shows no
trace of weathering consistent with being erected above ground. Others may have broken during use or
have been disturbed from earlier burials.
93E195:291. Linear incised cross (Ill. 3.12). Linear incised cross with barred terminals, pecked into surface
on front face; wide arms, curving in section.Thick stone slab, surface levelled on front, left roughly hewn
on back. Both edges straight, tapering slightly towards top, which may have been rounded, though one
corner broken. Broken at lower edge. No trace of weathering on surfaces. Purple, flaggy fine sandstone,
possibly imported. L 119, W 106, T 24. Lower monks’ garden, unstratified.
93E195:289. Linear incised cross. Simple, linear, equal-armed cross, roughly scratched into surface with
deep incised lines.Thin slate with tapering sides towards rounded top. Broken at lower edge. Fine purple
slate, Skellig. L 161, W 124, T 11. Leacht area, F577, possibly contemporary with Burials 1 and 2.
E338:27. Cruciform stone. Fragment of a stone cross. Circular hollow at the crossing of arms, forming one
large expanded arm and one short arm, probably the shaft and side respectively. Rough, uneven surfaces.
Rhyolite, Skellig. L 289, W 146, T 45. Small oratory terrace, Cutting 2, F219, nineteenth/twentiethcentury deposit overlying paving on terrace.
E338:41. Cruciform stone (Ill. 3.12). Head of a stone cross with broken shaft. Short arms, one rounded,
the other squared, and a wide, slightly curving top. Pale grey, fine micaceous sandstone, Skellig. L 313,W
280, T 25. Small oratory terrace, in bottom of clay/rubble layer F219.
93E195:308. Cruciform stone (Ill. 3.12). Fragment of stone cross. Simple squared arms, one wider than
the other, probably the top and a side arm. Surfaces very rough, stained brown around break. Rhyolite,
Skellig. L 238, W 230, T 45. Under threshold of structure (F694) at base of east steps.
93E195:287. Cruciform stone. Fragment of stone cross. Simple squared arms, two opposing arms
represented and stump of a third. Roughly worked. Hackley fine sandstone, Skellig. L 201, W 96, T 39.
Lower monks’ garden, unstratified.
93E195:307. Cruciform stone (Ill. 3.12). Fragment of stone cross. Simple squared arms.Two perpendicular
arms of similar size. Interbedded silt/fine sandstone, Skellig. L 195, W 184, T 29. Lower monks garden,
F631, in campion.
93E195:285. Cruciform stone. Fragment of stone cross. Possibly part of two perpendicular arms. Grey
360
THE FINDS
sandstone, Skellig. L 193, W 108, T 28. Lower monks’ garden, unstratified.
93E195:305B. Cruciform stone. Fragment of stone cross. Simple squared arms. Two perpendicular arms,
one a little wider and longer than the other. Fine grey siltstone, Skellig. L 123, W 117. Leacht area, in
campion layer (F561).
93E195:288. Cruciform stone. Fragment of stone cross. Rounded expanding arms. Remains of two
perpendicular arms. Flaggy fine sandstone, reddish hue, possible import. L 102,W 83,T 17. Lower monks’
garden, unstratified.
93E195:282. Cruciform stone (Ill. 3.12). Fragment of stone cross. Simple squared arms. Part of two
perpendicular arms. Siltstone, Skellig. L 90,W 90,T 8. Redeposited soil (635(7)) in lower monks’ garden.
Slates
A large number of slate fragments were recovered.These are small fragments that are either featureless or
ambiguous in terms of function. Several have notches in one or more edges which could be the remains
of broken suspension holes. There are no fragments that can be positively identified as roof slates. The
most likely example is E338:36 from the small oratory terrace, which is ovoid in shape with a notch at
one end (Ill. 3.10). The notch is more than likely the remains of a broken perforation that served as a
suspension hole. As argued above (Section 2.2.1 above), however, its recovery from the fill of the drain
underlying the north wall of the small oratory, a context that pre-dates St Michael’s Church, would seem
to preclude its use as a roofing slate and its precise function remains obscure.
E338:36. Ovoid slate with perforation (Ill. 3.10). Slate, roughly oval in shape with one straight end and the
other broken, with the remains of a peg hole or perforation. Purple slate, Skellig. L 1780, W 890 (max.),
T 7, Dia of perforation 10. Small oratory terrace, fill of drain F224, coeval with construction of oratory.
Incised slates
There are two slate fragments on which patterns have been incised. These appear to be fragments of
waste slate. The first (93E195:306K), with its wide cross-hatched pattern, appears to be a practice piece.
It is reminiscent of several finds from the Scottish island monastery site of Inchmarnock in the Firth of
Clyde. Slate was readily available there and it seems to have been widely used for everything from gaming
boards to writing practice, children’s drawings and practice pieces (Lowe 2008, 114–75).
The second slate (93E195:305D) is harder to decipher. It may be a casual doodle or part of a larger
pattern.The linear design is reminiscent of a sketch-map and the lines may represent some of the island’s
many paths and steps, though it does not contain enough detail to match it to any specific location. It is
probable that both pieces are of medieval date.
93E195:306K. Incised slate (Ill. 3.10). Small slate fragment, lightly incised with a wide, cross-hatched band.
Grey slate, Skellig. L 68, W 48. Leacht area, F577, possibly contemporary with Burials 1 and 2.
93E195:305D. Incised slate (Ill. 3.10). Small slate fragment; lightly incised linear band with two
perpendicular bands, area of cross-hatching and square at end. Grey slate, Skellig. L 107,W 52,T 4. Leacht
area, in campion, F561.
Lignite
Two lignite artefacts were found in the large oratory.The ring is very thick but with an internal diameter
361
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 3.13—Stone
artefacts.
of just 26mm. It is too large for a finger-ring but too small for a bracelet, even for a small child. When
rings of this size are found, use as hair ornaments or amulets is suggested, but there is no direct evidence
for how they were utilised (Johns 1996, 123; Hencken 1950, 150). The polished inner curved surface
suggests wear through use, perhaps by the drawing through of fabric. This could suggest use as a tie for
the mouth of a cloth or leather bag (J. Sheehan, pers. comm.) or for a heavy woollen garment.The early
medieval ivory rings found in Britain (which average 15cm or more in diameter) are described as ‘bag
rings’ and it is suggested that they were attached to the mouth of a leather or cloth pouch (MacGregor
1985, 110–12).The small size of the Skellig lignite ring, however, would preclude its use in such a manner.
The bead is also unusual. An assemblage of stone disc beads was found in a late tenth- to mideleventh-century context in York (Mainman and Rogers 2000, 2598), though these were typically much
smaller, the largest being 18mm. Beads could be strung together to make necklaces or rosaries, but the
size and shape of this bead suggest that this is unlikely. They could also have been used singly to adorn
small objects (ibid.).
Given their context, it is possible that rather than being personal adornments these were somehow
used to decorate the oratory or objects within it. No other lignite artefacts or fragments were found
elsewhere on the island.
E338:24. Ring (Ills 3.10, 3.14). Half of ring, deep D-shaped section but damage to flat surface masks
original profile.Well polished on curved side, particularly on interior, implying wear through use. Lignite,
imported. External Dia 59, internal Dia 26, T 13. Large oratory, south-east quadrant, F302, disturbed
floor deposit, just south of F305.
362
THE FINDS
Ill. 3.14—Lignite ring, E338:24
and lignite bead, E338:28 (Con
Brogan, DAHG).
E338:28. Bead (Ills 3.10, 3.14). Disc bead. Slightly irregular round shape, off-centre perforation cut from
one side. Flat polished sides. Edge smoothed and polished to ridge. Lignite, imported. Dia 20,T 3.1. Large
oratory, south-east quadrant, F306, undisturbed floor deposit (probably early medieval), close to east wall.
Querns
The fragments of two rotary flat querns (93E195:290; 93E195:312) are both of fine conglomerate from
Valentia Island.Traces of wear indicate that both have been used.This type of quern was used principally
for grinding grain for flour, but also malt for brewing, mustard and other spices, and raw materials for
pottery and metal production. Irish querns of the early medieval period are typically flat, particularly in
the latter part of the period (Watts 2006). During the later medieval period, towards the end of the
monastery occupation, the use of querns was widely prohibited in favour of the manorial mIll. It seems
likely, however, that the Skellig monks would have continued to mill their own flour, at least from the
cereals they were able to grow on the island.
93E195:290. Quern (Ill. 3.15). Rotary flat disc quern, upper stone fragment. Slightly domed top and flat
underside, thickening towards middle. Slight traces of wear on top, particularly towards thicker centre,
suggesting a change of use. Two concentric bands, c. 15mm wide, pecked into surface, one 50mm from
edge forming a raised border, the other towards centre, 155mm from outer edge. Flat underside worn
smooth. No remains of central hole or handle hole. Fine green conglomerate, Reenadrolaun, Valentia
Island. Dia 565, T 35–40. Lower monks’ garden, surface find.
93E195:312. Quern (Ill. 3.15). Rotary flat disc quern, half upper stone. Domed top and concave
underside, thickening slightly towards middle. Tapering handle hole, widened at top from use, at which
363
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 3.15—
Quernstones
(Con Brogan,
DAHG).
93E195:312
93E195:312
point quern is broken. Decorated with radial line between central hole and handle hole, flanked by two
further radial lines on either side. Pecked recessed border around central hole. Central hole of hourglass
section. Some traces of wear around middle of underside. Fine red conglomerate, Reenadrolaun,Valentia
Island. Dia 475, central hole Dia 87, handle hole Dia 28–42, T 55–73. Incorporated in paving outside
Cell B; removed for safekeeping.
Spindle-whorls
Neither of the two spindle-whorls recovered is from a secure context but it seems likely that both are of
medieval date. Whorls were used to weight the end of a hand-spindle to preserve its momentum when
in use. Both are simply made and, at 46g and possibly as much as 180g, very heavy. Whorls over 40g are
exceptional and may have been used for plying two or more yarns together (Oakley and Hall 1979).The
assemblage of whorls from medieval Waterford includes similar examples but all fall within the range 8–
34g (McCutcheon 1997, 404–7). A group of whorls from London, however, included a lead example of
78g and a giant stone whorl of 150+g (Egan 1998, 255–61).The monks were presumably spinning coarse
fibres for their own clothing.
93E195:263. Spindle-whorl (Ill. 3.13). Complete discoid whorl. Simply made with flat faces and rectangular
364
THE FINDS
section. Hourglass-shaped perforation, slightly off-centre, with some traces of wear. Light purple fine
sandstone, Skellig. Dia 52, hole Dia 8, T 8–11, weight 46g. Found in spoil heap by lighthouse toilet;
originally from lower monks’ garden.
93E195:264. Spindle-whorl (Ill. 3.13). Segment from large stone disc, representing about a third of the
whole. Flat faces, rectangular section. Central hole with rounded section, wider at one end. Light purple
fine sandstone, Skellig. Dia 83, hole Dia 12, T 10, weight 61g, estimated complete weight 180g. East
entrance, nineteenth-century pit (F503).
Cobble tool (rubber/hammer stone)
The cobble tool (93E195:335) could have served a number of functions, e.g. smoothing, polishing,
grinding, hammering etc.
93E195:335. Rubber/hammer stone. Large cobble tool, broken along one side. Well worn to smooth
concave surface on one side.Traces of smoothing wear also on one end of opposing side. Pitting on both
ends consistent with hammering. Small area of pitting also on main smoothed concave face. Multi-use
cobble tool, probably for grinding and hammering. Light purple fine sandstone, probably sourced from
mainland beach. L 163, W 78, T 55. Lower monks’ garden, nineteenth-century redeposited soil
(F632(1A)).
Whetstones
A small number of stone fragments, including beach pebbles, show signs of smoothed surfaces possibly
resulting from use as whetstones. Two examples (E338:20; 93E195:250) were more specifically used for
sharpening metal blades and tools. The complete example (E338:20) is well shaped and is very flat with
unusual bowed edges. It is of unusual shape compared to other medieval whetstones (cf. O’Connor 1991;
Marshall and Walsh 2005, 200; Fanning 1981, 131), which might suggest a more recent date. It is simply
made, however, and there is no particular reason why it should not be medieval. It was found in rubble on
the small oratory terrace ledge and may belong to the time of construction of the terrace retaining wall. It
does not appear to have been particularly well used and was probably lost before the end of its natural life.
E338:20. Whetstone (Ill. 3.13). Flat stone shaped into rough rectangle; long sides convex. Traces of wear
on both long sides, particularly on one, which appears to have been worn flatter than the opposing side.
Also slight traces of wear on both flat faces. Fits well into hand, particularly for edge use. Deep purple
fine-grained micaceous sandstone, probably from mainland. L 102, W 64, T 18. The small oratory ledge,
Trial-pit 1, bottom of rubble F104, probably monastic.
93E195:250. Whetstone (Ill. 3.13). Fragment of whetstone, representing part of one face and edge. Split
along middle, and broken at both ends. Face worn smooth from use. Edge also a little worn, with two
separate areas of dishing. Fine micaceous sandstone, Skellig. L 70+, W 19+, T 8+. Redeposited soil
(F635(5)) in lower monks’ garden.
Chipped stone (flint)
Nine small pieces of flint were recovered from the large oratory, the small oratory terrace and ledge, and
south entrance 2, inner enclosure. They include part of a nodule with cortex still adhering, tiny waste
flakes and pebbles. Two retouched pieces (E338:10; E338:25) could have been used as scrapers but are
not sufficiently diagnostic to be assigned to any particular period.The flint, which ranges from dark grey
to honey-coloured, had to have been brought to the island, and was probably sourced from mainland
365
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
beaches. Small assemblages of flint and chert have been recorded at several early medieval ecclesiastical
sites, including Illaunloughan (Marshall and Walsh 2005, 210), Reask (Fanning 1981, 136) and Inishmurray
(O’Sullivan and Ó Carragáin 2008, 310).
E338:25. Worked flint. Flake of dark grey flint retouched along two sides to create a rounded ‘point’.
Could have functioned as a scraper. L 31, W 30. Large oratory, disturbed floor deposit F302.
E338:10. Worked flint. Small trapezoidal flake of dark grey flint with cortex still adhering to one side.
Retouch along two sides, could have been used as a scraper. L 13, W 18 (max.). South entrance 2 (inner
enclosure), nineteenth-century dump deposit F12.
3.7.6 GLASS (not illustrated)
Lighthouse period and later (1820–present)
Bottle glass
Fifty sherds of bottle glass were found during the excavations, representing a minimum of eight bottles.
The majority were of green wine bottles, represented by 45 sherds from a minimum of six bottles.These
were all of cylindrical form, mostly handmade, though at least one bottle was mould-blown. Moulded
bottles replaced handmade forms during the 1820s (Dumbrell 1983, 115) and thus this decade is the
most likely date for the deposition of this assemblage. This ties in exactly with the period of occupation
by the lighthouse-builders.
The largest concentration of sherds came from the area just to the east of St Michael’s Church. The
25 sherds represent at least three bottles, two of which (93E195:153; 93E195:176) appear to have been
quite old when deposited.They are base sherds of a form that is unlikely to have been made after c. 1800.
They are, however, associated with a neck from a bottle that dates from after c. 1820 (93E195:152). The
area is relatively sheltered and hidden from the rest of the settlement and thus might have been a good
spot for illicit drinking. Alternatively, it may simply have been a convenient area for the dumping of
rubbish.
Most of the rest of the glass was found in the lower monks’ garden, particularly around Cell G.There
were fewer sherds (only eleven) but again representing a minimum of three bottles.The wider scattering
suggests redeposition and rubbish disposal rather than in situ drinking.
One sherd (93E195:338) has had letters scratched into its surface using a sharp tool such as a nail or
the tip of a knife. The exterior of the sherd has a clear if slightly wobbly ‘A’, which overlies an upsidedown and fainter possible ‘T’.The interior has a faint and elongated ‘V’, possibly a first attempt at an ‘A’.
The scratches on the interior indicate that the marks were made on the sherd rather than on the original
bottle.They may represent practice at letters using materials to hand, or the casual scratching of a person’s
initials.
The remains of a clear polygonal bottle were found in the upper levels within Cell G. This appears
to be of twentieth-century date and was probably left by a recent visitor. Another clear, colourless bottle
was found in the campion at the east end of the lower monks’ garden. This is of unusual form and may
date from the lighthouse-building period or from more recent activity. Some sherds recovered from the
archaeologists’ hut following the fire were unidentifiable.
93E195:338. Incised wine bottle sherd. Green wine bottle sherd with scratched initials on surface. On
exterior, a clear if slightly wobbly ‘A’, with a fainter and smaller, possibly earlier ‘T’, upside down in
relation to A. On interior is a very elongated and wobbly ‘V’, possibly a first attempt at an ‘A’. Lines
366
THE FINDS
scratched with a sharp point, possibly a knife tip. L 49,W 39. Lower monks’ garden, campion layer F631(3)
Vessel glass
Two sherds from the same glass vessel were found in the campion layer in the lower monks’ garden,
F631(1) (93E195:343; 93E195:344). It is a modern glass tumbler and was probably left by recent visitors
picnicking in this spot.
Bead
Plain, round glass beads are difficult to date.Turquoise beads would have been in use for the entire period
of the island’s occupation, but the even roundness, symmetry and straight-bored perforation of this bead
suggest that it is more likely to be a modern machine-made bead rather than one of medieval date. It
was found in the campion layer of the lower monks’ garden (F631(3)), associated with several sherds of
nineteenth-century pottery and bottle glass.
93E195:341. Bead. Translucent turquoise bead with faint pale encircling streaks. Rounded profile. Dia
11, W 8, perforation 2.5. Lower monks’ garden, campion layer F631(3).
3.7.7 LEATHER (not illustrated)
Mairéad Dunlevy and Julie Franklin
Lighthouse period and later (1820–present)
Shoes
Pieces of two shoes were found. The first (E338:29) was found in the modern revetment wall at south
entrance 2 (inner enclosure) and is represented by part of the insole and upper.The heavy leather stitching
in the welting, as well as the substantial tacks, suggests that it was locally made. The piercing from the
stitching around the back, because of its slight irregularity, suggests the same, but it is pierced in a manner
that emulates machine sewing. The heavy stitching across the waist of the shoe is surprising, particularly
as it does not seem to suggest the joining of two pieces of leather. It appears to be a man’s shoe, sewn by
a local shoemaker, and because of the machine-sewing influence could be late nineteenth- or early
twentieth-century.
The second (93E195:245), from a modern context in the lower monks’ garden, comprises two
fragments, possibly from the upper of a shoe. Both fragments have stitch holes along the edges and one
piece has three short slashes that have fully penetrated the leather.
Glove fragment?
A single leather (pigskin) artefact (93E195:238) was recovered from the fill of the cistern. It is an oval
piece folded along its length (L 145), with stitch holes around the edge. Cistern fill (F615), mixed deposit,
mostly modern.
3.7.8 MISCELLANEOUS (not illustrated)
Cordage
Two lengths of strapping knotted together (E338:37) were found on the small oratory ledge in a deposit
of material dumped over the terrace wall. It was a plain woven braid of coarse fibre. One part was knotted
367
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
into a loop, of approximate wrist size (circumference c. 220mm), and its presence on the ledge suggests
some kind of basic safety equipment. It may date from the lighthouse-building period or later.
Plastic
There was a surface find of a clear plastic inset fitting, inscribed with the message ‘Remember the Mission’
around a heart topped with a cross. It had presumably been dropped by a recent pilgrim.
368
4. THE RADIOCARBON DATES
4.1 INTRODUCTION
One of the primary research questions governing all aspects of work on Skellig Michael was the
establishment of a chronological framework for the development of the monastic settlement.The general
lack of stratified deposits, owing to periodic collapses of drystone structures (together with their associated
deposits), the redeposition of soils to backfill and cover areas of collapse and the disturbance caused by
burrowing birds and rabbits, has meant that a greater reliance has had to be placed on radiocarbon dating
in an effort to establish chronological sequences. This too was not without its problems, however, as
samples suitable for such dating, particularly from meaningful contexts, were in short supply. Only one
sample from the entire South Peak excavations, for example, was considered to be potential dating
material. On analysis, however, this black soil from underneath the leacht on the oratory terrace did not
contain any organic material and therefore could not be dated. Every sample from the monastery
excavations that was deemed suitable was submitted for radiocarbon dating. Each of the human burials
has been dated, in addition to any animal bone and charcoal from a secure context. A sample of limpet
shells used to block one of the enclosure entrances was also dated, taking into account the marine reservoir
effect. The charcoal selected was from short-lived species, although two samples of yew (Taxus) wood, a
slow-growing softwood, have been included.The latter were identified as coming from young branches.
All samples were processed at the 14CHRONO Centre, Queen’s University, Belfast. (See tables
overleaf.)
369
Table 5—Radiocarbon dates from the lower monks’ garden. All dates were calibrated using radiocarbon calibration program CALIB REV6.0.0 (Stuiver and Reimer 1993).
Material Location Radiocarbon Cal. age range, Cal. age range, Associated
age BP
68.3%
95.4%
artefacts
confidence
confidence
UBA16034
Sheep
vertebra
LMG
F633
962±24
AD 1024–1149 AD 1020–1154
UBA15963
Wood,
Taxus
LMG
F634
1207± 21
AD 777–866
AD 726–887
UBA15964
Wood,
Taxus
LMG
F634
1182 ± 27
AD 782–886
AD 773–946
UBA16035
Sheep
LMG
mandible F634
1237 ± 21
AD 694–852
AD 689–870
UBA15965
Charcoal, LMG
Corylus F635
915 ± 19
AD 1046–1159 AD 1037–1167
UBA16031
Sheep
vertebra
1189 ± 29
AD 781–882
UBA15966
Charcoal, LMG
Corylus F636
964 ± 19
AD 1024–1147 AD 1021–1153
LMG
F642
AD 723–943
Combined
date range
Ham Green B, 11th–mid-13th
late 12th to mid- century
13th century
Context
F633: Uppermost layer of redeposited
medieval soil in LMG. This context was
from above the walls in the LMG.
8th–9th century F634: Layer of redeposited medieval soil
underlying F633. This layer has no pottery
associated with it but it has 14C dates
ranging from the 7th to the 10th century
(see UBA-15964 and UBA-16035 below).
8th–10th century F634: Layer of redeposited medieval soil
underlying F633. See UBA-15963 above.
7th–9th century F634: Layer of redeposited medieval soil
underlying F633. See UBA-15963 above.
Local medieval
pot,
13th/14th
century;
Orléans-type
ware, late
12th/13th
century
11th–14th
century
F635: Layer of redeposited medieval soil
underlying F634. Here the dates for the
pottery closely match the date for the soil,
suggesting a terminus post quem of the
13th/14th century for the deposit.
8th–10th century F642: Layer of redeposited medieval soil,
transition between F635 and F636, in
LMG. The date from this layer is earlier
than the layer above, but also earlier than
the layer (F636) below.
11th–12th
F636: Layer of redeposited medieval soil
century
underlying F642/F635.
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
370
Sample
Table 6—Radiocarbon dates from the leacht area, the east entrance and the small oratory terrace. All dates were calibrated using radiocarbon calibration program CALIB REV6.0.0 (Stuiver and
Reimer 1993). The date of the limpet shells (UBA-16033) was calibrated with reference to the marine reservoir correction database (Reimer et al. 2009; Ascough et al. 2009). LMG = lower
monks’ garden; SOT = small oratory terrace.
Sample Material Location
Radiocarbon Cal. age range,
Cal. age range,
Date range
Age BP
68.3% confidence 95.4% confidence
Context
Late 10th–
12th century
8th–mid-10th
century
Late 8th–mid10th century
Burial aligned N–S against east wall of St Michael’s
Church. Male, 50–60. No visible grave-cut.
Layer south and east of leacht, overlay F583, earlier than
Burials 1 and 2.
Layer to south and east of leacht, underlay F581. This was
the basal deposit above secondary steps of south entrance.
AD 779–970
Late 8th–10th
century
AD 1035–1151
AD 1030–1155
11th–mid12th century
UBA- Sheep
E. entrance, 1232 ± 22
16032 vertebra F508
AD 713–859
AD 691–876
Late 7th–9th
century
UBA- Limpet
16037 shells
E. entrance, 1515 ± 24
F508
AD 702–800
AD 675–871
Late 7th–9th
century
South-west of leacht. Layer of charcoal-rich clay abutting a
stone setting. Contained a fragment of a human skull and
a concentration of water-rolled pebbles.
Upper layer in drain beneath leacht. This may be disturbed
as storm petrels had got into the drain. This is most likely,
as the other layers above have produced completely
consistent dating to the 8th–10th century.
Material from elsewhere in the monastery used to backfill
the entrance. Obviously mixed and does not date the
backfilling.
Lens of limpet shells in backfill F508. Obviously mixed
and does not date the backfilling.
UBA- Cattle
16033 tibia
E. entrance, 1167 ± 21
F509
AD 782–932
AD 778–948
Late 8th–mid10th century
AD 685–778
AD 672–869
Late 7th–9th
century
Skeleton 1 955 ± 44
AD 1024–1153
AD 995–1180
Skeleton 2 826 ± 54
AD 1169–1261
AD 1045–1280
UB4712
UBA15968
UBA15969
Human Skeleton 3 968 ± 44
bone
Charcoal, Leacht area 1181 ± 25
Betula
F581
Charcoal, Leacht area 1170 ± 20
Corylus F583
AD 1020–1152
AD 990–1165
AD 782–887
AD 775–941
AD 782–893
AD 778–942
UBA- Charcoal, Leacht area 1156 ± 27
15970 Betula
F588
AD 783–949
UBA- Charcoal, Leacht area 941 ± 19
15967 Corylus F590
UBA- Cattle
SOT
16036 horn core F229
1255 ± 35
Layer of backfill in entrance, directly beneath F508. Date
later than the layer above, indicating presence of material
redeposited from elsewhere in the monastery.
F229 is primary construction layer of terrace. Horn core
may, however, be residual. Date could relate to
construction of terrace or be a terminus post quem.
RADIOCARBON DATES
Burial aligned E–W south of leacht and east of St
Michael’s Church. Juvenile, 9–11. Slight grave-cut.
Burial aligned E–W south of leacht, east of St Michael’s
Church. Probable male, 25–30. No visible grave-cut.
Human
bone
Human
bone
THE
371
Late 10th–
12th century
Mid-11th–
13th century
UB4710
UB4711
5. THE HUMAN REMAINS
Linda G. Lynch
5.1 INTRODUCTION
The human skeletal remains recovered from Skellig Michael comprise both articulated burials and
disarticulated bones.This study examines all of the human skeletal remains.The project had a number of
aims:
• to provide a detailed catalogue of all of the human skeletal remains recovered from the island to
date;
• to examine the condition of the bones and therefore the level of preservation of human remains on
the island;
• to establish a minimum number of individuals (MNI) represented by those remains;
• to provide a detailed osteoarchaeological assessment of the bones.
5.2 THE REMAINS
The skeletal remains recovered on Skellig Michael consist of three articulated skeletons and a quantity
of disarticulated remains.The bones were retrieved from four discrete areas within the monastery (Table
7). Most of the bones came from the area to the east of St Michael’s Church. Disarticulated remains were
also recovered from the monks’ graveyard, from the small oratory terrace and from the lower monks’
garden. Plans and in situ photographs were made available to the writer for the compilation of this report.
Most of the bones were processed by the writer, following the established standards (see Buckley et al.
1999). The exceptions were some of the bones recovered when the writer was not on site. When
preservation allowed, these had been cleaned following the recognised standards.
Table 7—Distribution of human skeletal remains by area.
Area
Articulated skeletal remains
Disarticulated skeletal remains
East of St Michael’s Church
Monks’ graveyard
Small oratory terrace
Lower monks’ garden
SK1, SK2, SK3
ID nos 1–155
ID nos 173–178
ID nos 156–172
ID no. 179
In general, the bones recovered from the area to the east of St Michael’s Church and to the south of
the large oratory (all the in situ burials and most of the disarticulated bone) were well preserved, with
some fragmentation. None of the burials was complete, particularly Skeletons 1 and 2, and all had some
level of taphonomic disturbance. Skeletons 1 and 2 had been significantly disturbed by the digging of a
pit in modern times (E. Bourke, pers. comm.). In addition, many bones had collapsed into a puffin burrow,
which ran underneath the burials. These remains were excavated by E. Bourke in the seasons between
1997 and 2000. The remains of a leacht, which appeared to have been rebuilt in the nineteenth century
(E. Bourke, pers. comm.), were also present in the immediate vicinity.The writer was employed on Skellig
Michael in the 2000 season to supervise the recording and lifting of human skeletal remains from the
area to the east of St Michael’s Church. During this latter season, a small quantity of disarticulated human
remains were recovered, as well as Skeleton 3. The relatively good preservation of most of the skeletal
373
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
remains recovered from this area ensured that they could be washed with soft brushes and allowed to
dry naturally. More delicate fragments were allowed to dry naturally and were then dry-brushed with
soft brushes.
A small quantity of disarticulated human bone (ID 156–172) was excavated by Ann Lynch in 1986
from underneath a small slab on the small oratory terrace. These bones and teeth were in a very poor
state of preservation, with extensive erosion of the bone and severe fragmentation of the dental remains,
including the enamel, the hardest substance in the human body. In addition, the bone was flaking, as if it
had been exposed to the elements, although some of this may have occurred as a result of post-excavation
deterioration. The bones were received by the writer in an already-dry state, and there were no residual
earth fragments adhering to the remains. Further dry-brushing was deemed unnecessary and, in any case,
the bones were too delicate.
A very small quantity of disarticulated bone (ID 173–178) was retrieved from the monks’ graveyard
by E. Bourke in the summer of 2001 during conservation works. Although the bone initially appeared
to be in a good state of preservation, on further analysis significant post-mortem damage was apparent.
Portions of the internal structures of the bones had completely eroded away, leaving only a very thin,
paper-like, outer layer of bone. There were also breaks evident on the bones, particularly on the femur
(ID 174). The distal posterior metaphysis of the latter had breaks that probably occurred when the bone
was still relatively fresh, but not at the time of death. This suggests that the burial was truncated some
time relatively soon after burial, when the bones were still quite organically fresh.
Finally, a single bone was recovered from the lower monks’ garden by E. Bourke (ID 179). This was
the lower portion of a femur (thigh bone) and was relatively well preserved.The bone was disarticulated.
Indeed, no evidence of in situ burials has been identified in the lower monks’ garden.
5.3 METHODOLOGY
The age, sex and stature of an individual may be estimated through the examination of a variety of
skeletal elements.The age determination of adult individuals is based on the rates of degeneration of the
skeleton.These include morphological changes in the pubic symphysis, using the Suchey–Brooks method,
and changes in the auricular surface of the ilium (Lovejoy et al. 1985). In addition, the rates of dental
attrition (Brothwell 1981, 71–2) may be utilised (although only in skeletal remains that pre-date the
post-medieval period). Archaeological adult skeletons cannot be aged very accurately and are usually
assigned into broad age categories.These are ‘young adult’ (17–24 years), ‘middle adult’ (25–44 years) and
‘old adult’ (45+ years). The age of an infant or juvenile may be determined with more accuracy, using
the known rates of growth and development of various parts of the skeleton. Primarily, the methods
used are the calcification and eruption of the teeth (Moorrees et al. 1963a; 1963b; Ubelaker 1989, 64).
In addition, the rates of fusion of the epiphyseal plates (Schwartz 1995) and the diaphyseal lengths of the
long bones (Ubelaker 1989, 70–1; Scheuer et al. 1980) may be examined. The teeth are generally
considered to be more reliable indicators of age at death in juveniles as, unlike other skeletal indicators,
they are not so susceptible to such factors as poor diet and disease processes.
The sex of adults is determined by assessing morphological differences in the skeleton between
females and males, particularly using the pelvis and skull. In general, females tend to be slender and small,
with marked particular traits in the pelvis for the birthing process. Males tend to be larger and more
robust (Ubelaker 1989, 52–4; Buikstra and Ubelaker 1994, 16–20).The sex of infants and juveniles cannot,
as yet, be accurately determined using macroscopic methods.
The stature of the adult individuals is estimated from complete long bones (Trotter and Glesier
1952). Priority is given to the use of the lower limb bones, as these are considered more reliable in the
374
THE HUMAN REMAINS
calculation of stature, but upper limb bones are used where necessary. The actual bones used in the
calculation of the living statures of the individuals in this population are noted in the catalogue (Section
5.6). The final calculations are rounded off to one decimal place. The standard deviation figures are
presented to two decimal places, as these are the figures provided by Trotter and Glesier (1952).
The permanent teeth are recorded using the standard chart below:
18 17
48 47
right
16
46
15
45
14
44
13
43
12
42
11
41
21
31
22
32
23
33
24
34
25
35
26
36
27
37
28
38
left
The upper row represents the maxilla and the lower row represents the mandible. These are further
subdivided into left and right quadrants. Each permanent tooth (1–8) is prefixed by the number of the
quadrant to which it belongs (1–4). A different chart is generally used for deciduous teeth, but no in situ
deciduous teeth were recovered from Skellig Michael. A single deciduous tooth was recovered in the
disarticulated sample (ID 78).
The following symbols are used to record the teeth:
P = tooth present
E = tooth erupting
U = tooth unerupted
CA = tooth congenitally absent
12 = socket absent
B = tooth broken post-mortem
PM = tooth lost post-mortem
AM = tooth lost ante-mortem
R = root only
All incidences of calculus, carious lesions, hypoplastic defects and abscesses are recorded.
In addition, a limited number of pathological conditions are present and these are recorded below
(see Section 5.4.4).
A small quantity of disarticulated human skeletal material was recovered from the site. A catalogue
of all the examined disarticulated human remains is provided in Appendix III. The ‘code’ of the actual
portion of bone surviving is after Chamberlain and Witkin 2000. In the following report, Skeleton 1
will be referred to as SK1, Skeleton 2 as SK2 and Skeleton 3 as SK3. A term such as ‘ID 1’ will refer to
a bone in the catalogue of disarticulated bone.
5.4 ANALYSIS
5.4.1 DEMOGRAPHY
Minimum number of individuals
The human remains recovered from Skellig Michael represent three articulated individuals and a quantity
of disarticulated remains. The disarticulated remains were analysed in detail, with emphasis placed on
identifying specific skeletal elements in order to determine the minimum number of individuals (MNI)
represented in the sample.The calculation of this figure is based on the fact that each individual skeleton
has only one of each element—for example, each individual will only have one right proximal humeral
joint—and therefore a minimum number of individuals may be determined by counting the numbers
of each specific skeletal element in an assemblage.The results from the examination of the disarticulated
remains were compared with visual drawings of the three articulated burials from the writer’s osteological
375
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
recording sheets. In addition, all securely identifiable dental remains (that is, for example, ‘an upper right
first permanent molar’ as opposed to an ‘upper first/second? permanent molar’) from both articulated
and disarticulated remains were counted. The above information was further analysed on the basis of
differing contexts/areas, and on the age at death and the sex profiles.
The results of the above analysis indicate a minimum number of nine individuals in the total
assemblage: five adults and four juveniles (Table 8).
Table 8—Minimum number of individuals.
Area
Articulated burials
Disarticulated remains
East of St Michael’s Church
2 adults (SK2, SK3);
1 juvenile (SK1)
1 adult; 2 juveniles
Monks’ graveyard
Small oratory terrace
Lower monks’ garden
1 adult
1 juvenile
1 adult
Two of the adults were articulated burials (SK2 and SK3, both east of St Michael’s Church), while
at least one other adult was represented in the disarticulated remains, again recovered from the area to
the east of St Michael’s Church.This was evidenced by the remains of an additional proximal radius head
(ID 129) and an additional distal right humerus (ID 126, 127) (that is, ‘additional’ to the bones already
recovered in the articulated SK2 and SK3). It was also noted that fragments of five right temporals (from
the cranium) were recovered to the east of St Michael’s Church.While this may indicate an MNI of five
adults from this area (including articulated burials and disarticulated material), in reality it is likely that
at least some of the temporal bones were from the juvenile individuals (see below). It can often be difficult
to separate older juvenile temporal bones from adult remains. Therefore the MNI for the adults in the
area to the east of St Michael’s Church remains at two adult burials, with the remains of another adult
in the disarticulated material. The remains of the fourth adult were identified in the disarticulated adult
remains (ID 173–178) in the area of the monks’ graveyard. This fourth adult was confirmed by the
presence of four distal right humeri in the articulated and disarticulated assemblage. Finally, a fifth adult
was identified by a single disarticulated bone (ID 179) from the lower monks’ garden.
There were also the remains of at least four juveniles, all of approximately the same age (9–12 years).
There was one articulated juvenile burial (SK1), while the remains of two other juveniles were recovered
from the disarticulated bones, all in the area to the east of St Michael’s Church.The bones clearly indicated
the presence of a second juvenile in this area; for example, additional to the bones of SK1 there was a
right scapula (ID 133), a proximal right ulna (ID 131, 132), and a right femur (ID 110, 111, 112). Analysis
of the teeth, however, indicated the remains of a third juvenile in this area: there were three lower left
first and second permanent molars from juveniles of the same age.The remains of a fourth juvenile were
recovered in a quantity of disarticulated remains (ID 156–172) retrieved from the small oratory terrace.
Sex
As mentioned earlier, there are as yet no reliable macroscopic methods for determining the sex of juvenile
individuals. The sex of both of the articulated adult individuals was assessed. SK2 was determined to be
probably male, although the bones were noticeably quite slender.The skeleton of the second adult (SK3)
was more clearly a male.
The disarticulated remains revealed the presence of at least three other adults. With regard to the
remains of the disarticulated adult to the east of St Michael’s Church, the overall size of an additional
third right distal humerus (ID 126, 127; see above) suggested that this was a male individual. In addition,
376
THE HUMAN REMAINS
a disarticulated and unsided proximal femoral head was recovered (ID 151). SK3 had both of its proximal
femoral heads, while SK2 only had the right proximal femoral head.Technically the disarticulated femoral
head could have belonged to the latter skeleton, but its diameter (51.2mm) contrasted sharply with that
of the femoral head of SK2 (42.8mm) and it was deemed unlikely to belong to SK2. Therefore the
diameter of the disarticulated femoral head again indicated that the disarticulated adult to the east of St
Michael’s Church was a male individual (after Bass 1995).
The remains of the fourth adult from the monks’ graveyard and the fifth adult from the lower monks’
garden were too incomplete to allow determination of sex.
Therefore three of the five adults recovered from Skellig Michael were male individuals, while it
was not possible to determine the sex of the other two adults.
Age
It was possible to determine the age of both of the articulated adult individuals. SK2 (male) was aged
between 25 and 35 years, while SK3 (male) was aged between 50 and 60 years at the time of death. The
dental remains of the third disarticulated adult from the area to the east of St Michael’s Church (ID 12,
13, 17, 18, additional to the articulated dentitions) indicated the presence of a young adult aged 17–25
years at the time of death. It was not possible to determine an accurate age for the remains of the fourth
adult from the monks’ graveyard or the fifth adult from the lower monks’ garden.
The juvenile burial (SK1) was aged 9–11 years at the time of death. As mentioned earlier, the remains
of at least three other juveniles were recovered from the disarticulated remains. The two additional
juveniles recovered from the area to the east of St Michael’s Church were recognised through the dental
remains, which therefore also allowed the ages at death to be determined. One was aged approximately
10–11 years (ID 4, 5), while the other was aged 10.5–11.5 years (ID 7, 8). In addition, the overall size of
the bones and some of the dental remains recovered from the small oratory terrace (ID 156–172) indicate
that the fourth juvenile was again the same age as those already described. Therefore all four juveniles
represented in the assemblage were aged between approximately nine and twelve years at the time of
death.
5.4.2
STATURE
It was possible to determine the living statures of both of the in situ adult individuals.The stature of SK2
(middle adult male) was estimated as approximately 169.4cm, while the stature of SK3 (old adult male)
was calculated as 171.7cm, giving an average stature of 170.6cm. Table 9 provides a comparison of this
average stature with other contemporary male statures.
The statures of the Skellig males were comparable with some of their contemporaries in Cork and
Waterford, but were considerable smaller in stature than others in Cork and Limerick. They were taller
than the noticeably small medieval male adults buried in St Mary’s Cathedral in Tuam.
Table 9—Comparison of average male stature from Skellig Michael with contemporary populations.
Medieval site
Average male stature (cm)
Skellig Michael, Co. Kerry
St Mary’s Cathedral, Limerick (Power 1994)
Cove Street, Cork (Power 1994)
St Mary’s of the Isle, Cork (Power 1994)
St Peter’s Church, Waterford (Power 1994)
St Mary’s Cathedral, Tuam, Galway (Lynch 2005)
170.6
180.0
175.7
170.0
170.2
166.1
377
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
5.4.3
DENTAL ANALYSIS
The dental remains recovered from Skellig Michael represent both articulated and disarticulated remains.
‘Articulated teeth’ refers to the dental remains from the three articulated burials, while ‘disarticulated
teeth’ refers to teeth retrieved with other randomly recovered human remains. A total of 82 teeth were
recovered—46 articulated and 36 disarticulated—and all but one of these were permanent. One upper
left second deciduous molar (ID 78) was found associated with SK1 and is likely to have been part of
the dentition of this individual. The remains of at least eleven disarticulated, probably permanent, teeth
were recovered from the small oratory terrace (ID 160, 162–172). These were excluded from the dental
analysis below, however, because they had suffered extensive post-mortem damage, with only fragments
of enamel preserved in most cases. Any evidence of dental diseases has therefore been lost and the
inclusion of these teeth would distort the analysis. Only 44 of the articulated teeth were counted in the
analysis, as two of the teeth of SK1 were unerupted and the enamel was only partially formed.The typical
dental diseases could not have begun to form and the enamel was also too incomplete to observe dental
hypoplastic defects. Only 24 disarticulated teeth were included in the dental analysis, as one tooth was a
deciduous tooth and eleven were unidentified (see above). Dental diseases such as calculus, caries and
dental enamel hypoplastic defects were observed on a number of teeth (summarised in Table 10), and
these are examined in more detail below. It is noted that the numbers of teeth recovered and available
for analysis overall are very low and are not statistically viable. In addition, the poor dental health of the
old adult male (SK3; see Ill. 5.1) will have introduced bias to the percentages given below (Table 10).
Table 10—Prevalence of observed dental diseases in all observable human teeth.
Articulated teeth
Disarticulated teeth
All teeth
Calculus
Caries
Hypoplastic defects
68.2% (30/44)
50% (12/24)
61.8% (42/68)
0% (0/44)
4.2% (1/24)
1.5% (1/68)
22.7% (10/44)
8.3% (2/24)
17.6% (12/68)
Ill. 5.1—Mandible, anterior and
superior view, SK3 (male, old adult).
Ante-mortem loss of left second
molar, moderate to severe calculus
and heavy attrition rates.
378
THE HUMAN REMAINS
Calculus
Dental plaque can become mineralised as calculus, a hard, somewhat gritty substance that adheres to the
enamel and/or root of a tooth. It may be partially removed in the living individual through even the
most basic dental hygiene, using, for example, a small brush or stick.The deposits may also be inadvertently
removed through the consumption of grittier foods. Calculus deposits in a population may suggest both
poor oral hygiene and the consumption of a soft and sucrose-based diet (Roberts and Manchester 1995,
55). The deposits are very common in archaeological populations.
Calculus was present on 61.8% of all teeth recovered from Skellig Michael—68.2% articulated, 50%
disarticulated. In general the deposits varied from slight to moderate in severity, although severe deposits
were present in the dental remains of SK3 (male old adult) and on one disarticulated tooth (ID 98).
Dental caries
Carious lesions, or erosive lesions in the enamel of the teeth, usually occur as a result of a high
carbohydrate intake. Bacteria contained in plaque can metabolise certain carbohydrates into an acidic
waste that can dissolve the enamel of teeth (Mays 1998, 148). The pulp cavity can subsequently become
infected and this may lead to the formation of a dental abscess (Roberts and Manchester 1995, 50). Sugar
is highly cariogenic.
Only one tooth recovered from Skellig Michael bore evidence of caries. This was a single
disarticulated tooth (ID 99) that had a large carious lesion on the mesial aspect of the tooth at the
cemento-enamel junction. This represents 4.2% of disarticulated teeth or 1.5% of all teeth. In her study
of calculus in medieval Munster, Power (1994, 101) recorded a caries prevalence rate of 4.2% of all teeth.
These low rates typically reflect the unrefined medieval diet.
As mentioned above, carious lesions may lead to the formation of dental abscesses. No dental abscesses
were present in this population. This may be due to a combination of factors, such as the small size of
the sample, incomplete preservation, the age profile and the medieval date of the skeletons. Dental
abscesses may also occur as a result of severe attrition or trauma.
Dental enamel hypoplastic defects
These defects appear as a depressed line or series of lines or pits in the enamel of the tooth. They occur
as a result of a disturbance to the growth of the organic matrix, which is later mineralised to form enamel
and thus the defect is preserved in the enamel (Mays 1998, 156; Hillson 1986). The defects can occur as
a result of a number of diseases and/or nutritional deficiencies, including diarrhoea, parasitic infestations
of the gut, scurvy, rickets, allergic reactions, vitamin deficiencies and general malnutrition (Mays 1998,
158). As the calcification of teeth is confined to the years of childhood, the defects are a general reflection
of stresses suffered by an individual
in childhood.
Hypoplastic
defects
were
observed on a number of teeth,
affecting 22.7% of the articulated
teeth and 8.3% of the disarticulated
teeth (or 17.6% of the total). All of
the teeth affected by hypoplastic
defects in the articulated remains
came from a single individual (SK1,
9–11 years; see Ill. 5.2), indicating
that this juvenile had already
endured a series of physiological
379
Ill. 5.2—Right
mandibular
canine of SK1
(9–11 years)
with hypoplastic
defect (L.
Lynch).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
stresses to the body in the years preceding her/his death. It is unknown what these stresses were, but the
defects suggest that this individual may have been in poor health for much of her/his short life and the
apparent general state of ill health may have contributed to an early death. The locations of the lesions
on the teeth indicated that the physiological stresses occurred between 3–4.5 years and 5.5–6.5 years.
Interestingly, this individual also had indications of iron deficiency (see below), which may have been
linked with the hypoplastic defects.The dental enamel hypoplastic defects observed in the disarticulated
remains were on two teeth (ID 21, 23), both from adult dentitions.
5.4.4 PATHOLOGICAL
CONDITIONS
Observed pathological conditions were classified and examined using the following headings: metabolic
disorders, degenerative joint disease, trauma and non-specific infections.
Metabolic disorders
Certain porotic lesions of the eye orbits (cribra orbitalia) and skull vault (porotic hyperostosis) are readily
identifiable as specific pathological lesions owing to the process of their formation. They are indicative
of a metabolic disorder relating to iron deficiency.This condition occurs when, as a result of a deficiency
of iron, the body’s marrow increases its output of iron (Mays 1998, 142). The middle layer of the bone
expands and there is a corresponding thinning of the outer surface of the bone. This can result in the
diagnostic appearance of small holes or foramina on the outer surface of the bone. Although it is
frequently assumed that these lesions are indicative of iron-deficiency anaemia, studies have indicated
Ill. 5.3—Cribra
orbitalia in right
orbit of SK1 (9–11
years) (L. Lynch).
that when a body is under stress from an invading organism (such as a parasitic infestation of the gut) the
system increases its output of iron in order to counteract the stress. Thus this pathological process may
actually be a sign of a healthy defence system (Stuart-Macadam 1991, 105; Roberts and Manchester
1995, 166–7).
Evidence of this condition was observed on the skeletal remains of one individual recovered from
Skellig Michael. Traces of both porotic hyperostosis and cribra orbitalia (Ill. 5.3) were present on the
380
THE HUMAN REMAINS
remains of SK1 (9–11 years). The lesions were healing at the time of death. It is probable that the child
had suffered from some physiological insult/s that affected the metabolic system sometime prior to
her/his death. As noted above, the dentition also indicated that this child had suffered a series of stresses
in early childhood, all of which may have contributed to a shortened lifespan.
There was no evidence of metabolic disorders in the two articulated adult burials (SK2, SK3), nor
in the disarticulated remains.
Degenerative joint disease
Degenerative joint disease (DJD) is one of the most commonly observed pathological processes on the
human skeleton. The onset of the disease is typically age-related, as it appears to occur primarily as a
result of repeated ‘wear and tear’ on the joints, with resultant degeneration of the articular cartilage
(Ortner and Putschar 1981, 419–20). The disease can be accelerated by occupational activities and can
be brought on by trauma. The evidence of joint degeneration is manifested in the form of porosity or
pitting of the joint surface and/or additional bone growths known as osteophytes. In more advanced
cases, eburnation or polishing of the bone can occur as the bones of the joint rub against each other.
Perhaps not surprisingly, no joint disease was present on the remains of the juvenile individual (SK1).
Mild degeneration in the form of osteophytic growths and porosity was present on a limited number of
some of the vertebral or spinal joints of SK2 (male middle adult). In addition, a single Schmorl’s node
was observed on a lumbar vertebra of this individual. These depressions in the disc of the vertebra can
occur when the nucleus pulpous—the pulpous gelatinous core of the intervertebral disc—expands or
bursts into the adjacent vertebral body as a result of pressure (Mann and Murphy 1990, 52; Ortner and
Putschar 1981, 323). The pressure may be caused by a fall or by carrying heavy objects, and the lesions
are believed to occur during childhood when the bone is still relatively soft. More severe evidence of
degeneration was observed on the skeletal remains of the older male adult, SK3. Mild degeneration was
present on the right elbow, the shoulders and the knees, while mild to moderate degeneration was present
in the spinal column in the form of porosity, osteophytes and Schmorl’s nodes (Ill. 5.4). In addition, there
was compression of the body of the first lumbar vertebra (Ill. 5.5).This level of joint disease is not unusual,
given the estimated age at death of this individual.
Serious joint disease was observed in the
lower spinal column of SK3 (male old adult).
There was complete fusion of the fourth and
fifth lumbar vertebrae and of the first sacral
vertebra at both the bodies and the
apophyseal joints (Ills 5.5, 5.6). As discussed
above (SK3), the lesions were suggestive of a
number of disease processes, but it was not
possible to determine a more conclusive
diagnosis. The cause of the fusion remains
speculative.
Further evidence of DJD was observed
in a small number of fragments recovered
from the disarticulated remains. Schmorl’s
nodes and osteophytes were present on a
single thoracic/lumbar body fragment (ID
59). Slight degeneration was also observed in
a small number of lumbar vertebral fragments
associated with SK1 (ID 147, 148).
381
Ill. 5.4—T12,
superior view,
anterior to top,
SK3 (male, old
adult).
Schmorl’s node
with slight
osteophytic
growth at edge
of vertebral
body (L. Lynch).
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
[Left] Ill. 5.5—
T12 to S2, right
lateral view,
superior to top,
SK3 (male, old
adult). Note
compression
of body of L1
and total fusion
of L4, L5 and
S1 (L. Lynch).
[Right] Ill. 5.6—
L4 to S2,
anterior view,
superior to top,
SK3 (male, old
adult). Fusion of
fourth and
fifth lumbar
vertebrae with
first sacral
vertebrae.
Possible cloacae
arrowed (L.
Lynch).
Trauma
The diaphysis of the right fifth proximal foot phalanx of SK3
(male old adult) is abnormally shaped (Ill. 5.7). This is
indicative of a fracture, which had healed a considerable time
before death.The overall shape of the healed fracture suggests
that it occurred as a result of a twisting stress rather than, for
example, through compression or through tension stress
(after Ortner and Putschar 1981, 56). The joint of the
phalanx appears to have altered to
accommodate the different stress that
would have been put on the joint
through incomplete healing of the
fracture.
Ill. 5.7—Fifth right proximal foot
phalanx, dorsal view, distal to top, SK3
(male, old adult). Healed twisted
fracture with alteration of distal joint
(L. Lynch)
382
THE HUMAN REMAINS
Non-specific infections
Infection is an inflammatory process and is classically observed on the dry skeleton in the form of
osteomyelitis and/or periostitis. Osteomyelitis involves infection of the actual bone marrow and can be
quite distinct. Periostitis involves inflammation of the periosteum, the fibrous covering on the surface of
the bone. When the periosteum becomes inflamed, it rises owing to the accumulation of pus. A new
layer of grey striated or fibre bone (periostitis) may form underneath.With time, the lesion may heal and
become almost completely remodelled into lamellar bone.
Evidence of infectious lesions was observed in one disarticulated fragment. Healed periostitis was
present on a disarticulated juvenile right tibia (ID 156), of an individual aged 9–12 years at the time of
death.
As noted earlier, serious pathological lesions were observed in the lower spine of SK3 (old adult
male). The lower lumbar vertebrae had completely fused together and the fifth lumbar had also fused to
the first sacral vertebra. In addition to the joint disease noted above, there were lesions on the sacrum of
this individual that were suggestive of an infectious process. There is evidence of two possible sinuses in
the anterior of the lumbar bodies (Ill. 5.6). These are similar to lesions known as cloacae associated with
an infectious disease called osteomyelitis. As mentioned above, osteomyelitis occurs when the marrow of
a bone becomes infected, which can happen through a variety of incidents, either direct or indirect.
Cloacae or sinuses form, allowing the infectious pus and dead fragments of bone from the interior of the
bone to be extruded.These eventually work their way up to the surface of the skin (Ortner and Putschar
1981; Roberts and Manchester 1995). The holes observed in the vertebrae of this individual bear a
resemblance to the cloacae associated with the infectious process of osteomyelitis. The lumbar vertebrae
do not show any other indication of infection, however, and so the oval lesions cannot definitively be
attributed to osteomyelitis. In addition, the actual preservation of the spinal elements was quite poor,
which significantly hindered the assessment.
5.5
SUMMARY
The human skeletal remains excavated from Skellig Michael represent three articulated burials. The
presence of six other individuals was established through analysis of the disarticulated human skeletal
remains.
Two of the articulated burials were male adults. One was aged 25–35 years, while the other was
aged 50–60 years at the time of death. Analysis also indicates that a third adult (from east of St Michael’s
Church) was also a male, and was possibly aged 17–25 years at the time of death. The age and sex of the
fourth and fifth adults (from the monks’ graveyard and the lower monks’ garden) could not be determined.
The third articulated burial was that of a juvenile, aged approximately 9–11 years at the time of
death.Two other juveniles were also represented in the disarticulated dental remains from the area to the
east of St Michael’s Church. The age at death of both of these juveniles was estimated as being 10–12
years. The remains of a fourth juvenile of approximately the same age were recovered from the small
oratory terrace.
The stature of two of the adult burials could be determined and averaged 170.6cm, which is
comparable with contemporary males.
Analysis of the dental remains indicated the presence of calculus in particular. The rate of carious
lesions was very low, while no dental abscesses were observed. Dental enamel hypoplastic defects, while
observed in a limited number of disarticulated teeth, were primarily observed in the remains of the
articulated child burial. The latter also bore indications of other physiological stresses.
A limited number of pathological conditions were observed on the skeletal remains of this
383
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
population. Evidence of iron deficiency is present on the skeletal remains of the articulated child burial.
The lesions were healing at the time of death.
Evidence of degenerative joint disease was observed on the remains of both of the articulated adult
burials. The lesions are much more extensive on the remains of the older adult. On the younger adult
the lesions are confined to the axial skeleton, while in the older adult the appendicular skeleton is also
affected. In addition, ankylosis (fusion) of the lower spine is present in the older adult, although the
causative factors are unknown. Further traces of joint degeneration were observed in a limited number
of disarticulated spinal fragments.
One incident of possible trauma was observed in the form of a healed fracture of one of the bones
of the right fifth toe of the old adult male burial.
Evidence of healed infectious lesions was observed on the disarticulated tibia of a child. The lower
spine of the older articulated adult also had traces of possible osteomyelitis, although this cannot be
confirmed owing to post-mortem damage.
5.6 CATALOGUE OF ARTICULATED SKELETONS
The numbers of the skeletons below were assigned during the current analysis. When the articulated
burials were initially discovered they were given alternative numbers/identifications. SK1 below was
originally labelled ‘skeleton 2’ (as in ID 109–155) or ‘Meara’. SK2 was originally labelled ‘CJ’. SK3 was
uncovered while the author was on site and was not assigned any other label.
SK1
Age: Juvenile, 9–11 years (dentition and epiphyseal fusion).
Sex: Not determined.
Stature: Not determined.
Skeletal preservation: Poor. Incomplete and fragmented.
Skeletal position: Supine.
Skeletal attitude: Extended.
Orientation: South-west/north-east, head to south-west.
Associated skeleton/s: Directly underlay SK2.
Associated finds: None.
Bones present
Cranium and mandible. Scapulae, clavicles, humeri (minus the distal ends), ulnae, radii (distal half of left
radius absent). Manubrium, cervical and upper thoracic vertebrae. Proximal end of the right femur, distal
half of the left femur, proximal half of the left tibia.
Dental inventory
PM P
P
18 17 16
48 47 46
P
P
384
P
15
45
PM
PM
14
44
PM
E
13
43
E
P
12
42
P
P
11
41
P
21
31
P
22
32
P
E
23
33
P
P
24
34
P
P
25
35
U
P
26
75 36
PM P
P
27
37
P
U
28
38
U
THE HUMAN REMAINS
Dental pathology
Calculus—9/22 erupted/erupting teeth (16, 15, 26, 46, 43, 42, 32, 33, 36), slight to moderate in severity.
Hypoplastic defects—13, 12, 22, 23, 47, 43, 33 (at least two distinct periods of stress, 3.5–4.5 years
and 5.5–6.5 years).
Other—the lower left second premolar is unerupted. This may be due to the late retention of the
lower left second deciduous molar (lost post-mortem), the sockets of which are still visible in the
mandibular bone.
Skeletal pathology
Metabolic—bilateral healing cribra orbitalia and porotic hyperostosis of parietals and occipital.
Comments
The epiphyses suggest an age at death of approximately 12–14 years. This individual was radiocarbondated to AD 995–1174.
SK2
Age: Young middle adult, 25–35 years.
Sex: Male.
Stature: 169.4 ± 4.32cm (right radius).
Skeletal preservation: Good. Partially incomplete, bones are well preserved.
Skeletal position: Supine.
Skeletal attitude: Extended.
Orientation: South-west/north-east, head to south-west.
Associated skeleton/s: SK2 directly overlay SK1.
Associated finds: None.
Bones present
Fragments of the mandible. Right scapula, right clavicle, distal right humerus, right ulna (minus the distal
epiphysis), right radius, fragment of distal epiphysis of left humerus, proximal end of diaphysis of left
radius. Incomplete left and right carpals, left metacarpals, incomplete hand phalanges.Vertebrae include
C2, C6, C7, T1, T7–12, L1–L4, fragments of S1 and S2, right ilia, ischia and pubis, left pubis. Proximal
half of right femur, shaft of left femur.
Dental inventory
18 17
48 47
PM P
16
46
P
15
45
P
14
44
P
13
43
P
12
42
P
11
41
P
21
31
22
32
P
23
33
P
24
34
P
25
35
P
26
36
P
27
37
P
28
38
P
Dental pathology
Calculus—14/14, slight to moderate.
Other—more pronounced attrition on teeth on right side of mandible than on those on the left,
probably owing to individual chewing traits.
385
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Skeletal pathology
Joint disease—mild DJD of spine, mild to moderate porosity on the bodies of C6, C7 and T10, and mild
to moderate osteophytic growths on the bodies of L3 and L4. Single Schmorl’s node on the superior body
of L3.
Comments
This individual was radiocarbon-dated to AD 1040–1280.
SK3
Age: Old adult, 50–60 years (pelvis).
Sex: Male (pelvis).
Stature: 171.7 ± 2.99cm (right femur and tibia).
Skeletal preservation: Poor. Incomplete and fragmented.The burial comprised articulated lower body parts
and disarticulated upper body parts.The bones are believed to be those of a single individual on the basis
of the colour and size of the bones and on the non-duplication of skeletal elements. The upper body
(from approximately T12 upwards) had been disturbed and redeposited, apparently owing to the building
of a wall in later times. The skull—but not the mandible—appears to have been removed at this point
and may have been redeposited elsewhere. The rest of the disturbed bone was redeposited in a pile over
the original location of the abdomen and lower chest.The lower right ribs, the lower vertebrae, the pelvis
and the legs were all recovered in an articulated position.
Skeletal position: Supine.
Skeletal attitude: Extended. It is suggested that the arms and hands may originally have been crossed over
the abdomen or chest as opposed to being extended down either side of the body. None of the hand
bones were recovered in articulation, despite the fact that the lower back and pelvis were undisturbed by
the truncation, and this would suggest that they were originally over the chest area. The femora were at
an angle of approximately 45º to the pelvis so that the tibiae and feet lay at a higher level than the rest
of the body. This indicates that the body was not buried in a coffin. In addition, the legs and feet were
noticeably splayed, which indicates that if the body was buried in a shroud the wrapping was quite loose.
The right leg of the skeleton was recovered lying directly against the plinth of St Michael’s Church.
Orientation: South-east/north-west, head to south-east. Directly aligned with eastern gable of St Michael’s
Church.
Associated skeleton/s: None.
Associated finds: None.
Bones present
The bones present are the mandible, the scapulae, the clavicles and fragments of the sternum. The upper
limbs are represented by the humeri, radii and ulnae, and the hands are almost complete. The spine is
complete from C1 to S2, with the remaining sacrum being incomplete.The pelvis is complete.The lower
limbs are represented by the femora, tibiae and fibulae, and the feet are complete except for some of the
phalanges.
Dental inventory
P
18 17 16
48 47 46
CA PM P
386
15 14 13
45 44 43
PM PM P
12 11 21 22
42 41 31 32
PM PM PM P
23
33
P
24 25
34 35
PM P
26
36
P
27 28
37 38
AM CA
THE HUMAN REMAINS
Dental pathology
Calculus—7/7, slight to severe.
Skeletal pathology
Joint disease—mild DJD in right elbow (proximal ulna), left and right shoulders (glenoid cavities and
lateral end of right clavicle), and left and right knees (distal femora and proximal tibiae). Mild to moderate
DJD on most of the vertebrae. Moderate DJD on the articular dens of C1 and C2. Almost all of the
vertebrae from C5 to T9 are affected by DJD to some degree. Primarily the apophyseal joints are affected,
and the joint disease varies between mild and moderate in severity. Osteophytic growths present on the
bodies of T5 to T8, T10 to T12, and L3 and L4. There is also noticeable anterior compression of the first
lumbar vertebra (Ill. 5.5). Schmorl’s nodes are present on the bodies of T11 and T12 (Ill. 5.4).
Trauma—shaft of the fifth proximal right foot phalanx is abnormally twisted (Ill. 5.7), indicative of
a well-healed twisted fracture. The distal articular facet of this phalanx has been altered with slight
flattening. The articular facet is aligned more towards the dorsal surface. This suggests that the fracturing
of the phalanx and subsequent incorrect healing of the alignment led to a different mechanical stress
being put on the joint; the joint appears to have altered slightly to accommodate this.
Other—a more serious pathological condition was observed in the vertebrae of the lower spine.The
fourth and fifth lumbar vertebrae and the first sacral vertebra are fused at the bodies and the apophyseal
joints (Ills 5.5, 5.6).The inferior body and apophyseal joints of L4 are fused to the opposing joints of L5,
and the inferior body and apophyseal joints of L5 are fused to the opposing joints of S1. It is clear from
the remains that neither the centrum nor the arch of the third lumbar was involved in the process of
fusion.There is some post-mortem damage to the anterior of the bodies, but there is certainly no excessive
osteophytic growth associated with the fused bodies and the fusion is smooth and vertical. The joints—
both in the centra and the apophyseal joints—appear to be fused throughout the surface of the joint
area. In addition, despite the post-mortem damage, there is noticeable thinning of the cortical bone of
the anterior bones, with some exposure of the trabecular bone. There appear to be at least two small,
smooth openings, one in the anterior of the fourth lumbar and another on the left lateral side of the
anterior body of the fifth lumbar (near the inferior edge). These were particularly noticeable during
excavation. Close-up photographs were taken during excavation but these were unavailable at the time
of writing. Both holes have smooth oval openings with average dimensions of c. 7.7mm superior–inferior
and c. 3.1mm medial–lateral.
There are a number of conditions that can cause fusion or ankylosis of elements of the vertebrae. A
brief synopsis of some of those conditions is given below. It is noted that the pathological lesions observed
in the lower spine of this individual are unlikely to have occurred as a result of the following diseases.
Ankylosing spondylitis is a progressive inflammatory disease of unknown aetiology. It results in fusion
of both the bodies and the apophyseal joints of the spine. It typically progresses upwards through the
spine from the lumbar vertebrae, and is more frequent in males than in females. In addition, the fusion
of the bodies tends to result in a ‘squaring off ’ of the body—that is, there is no excessive osteophytic
growth. All of the above is similar to the findings from SK3. Ankylosing spondylitis also affects the sacroiliac joint, however (Ortner and Putschar 1981, 411; Roberts and Manchester 1995, 119; Rogers and
Waldron 2000, 175–6; Schwartz 1995, 249). In fact, the involvement of the sacro-iliac joint is considered
to be ‘the hallmark of ankylosing spondylitis’ (Roberts and Manchester 1995, 119, after Resnick and
Niwayama 1988, 1112), and is never absent if the lumbar vertebrae are involved (Ortner and Putschar
1981, 411). No pathological lesion was observed in the sacro-iliac joint of this individual. Thus, despite
initial appearances, the fusion observed in the spine of this individual is unlikely to have occurred as a
result of ankylosing spondylitis.
Diffuse idiopathic skeletal hyperostosis, or DISH, can also classically involve ankylosis of the spine
387
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
and typically affects the older male. The condition is believed to be associated with obesity and diabetes.
The vertebral bodies, in particular, can become fused by the growth of osteophytes, which take on the
appearance of ‘flowing wax’ (Mays 1998, 127; Roberts and Manchester 1995, 120; Rogers 2000, 170–1;
Schwartz 1995, 243). In DISH, however, the apophyseal joints are typically unaffected; it is usually the
thoracic vertebrae that are involved, and the fusion tends to occur on the right side of the bodies. In
addition, other parts of the skeletal structure would also be affected by enthesopathies or small bony
growths (Rogers 2000, 170–1; Schwartz 1995, 243). The lack of the ‘candle-wax’ osteophytic growths
on the fused vertebral bodies of SK3, the involvement of the apophyseal joints and the lack of
enthesopathies on the remainder of the skeleton militate against a diagnosis of DISH.
Calcium pyrophosphate deposition disease, or CPDD, which results from both mechanical problems
and inflammation, can also cause ankylosis of the vertebrae. This condition usually affects individuals over
the age of 30. Again, however, it does not involve ankylosis of the apophyseal joints (Schwartz 1995, 243).
Spondylosis deformans or spinal osteophytosis can also result in fusion in the spine and is relatively
commonly observed in older individuals (Ortner and Putschar 1981, 421; Rogers and Waldron 1995).
The fusion is caused by the outgrowth of osteophytes along the edges of the vertebral bodies. The
osteophytes, however, tend to form at almost right angles to the body itself, as ‘shelf-like bony protrusions’
(Ortner and Putschar 1981, 421), and, indeed, can resemble DISH. In addition, there is no involvement
of the apophyseal joints (Schwartz 1995, 239).
Other diseases that can lead to fusion of the spine include tuberculosis and osteoarthritis, but the
lesions observed in the lower spine of this individual bear no resemblance to skeletal manifestations of
those diseases. Therefore it is not possible to suggest what disease process may have caused the fusion of
these vertebral bodies.
As mentioned above, despite the post-mortem damage, there are at least two holes (ante-mortem)
in the anterior of the fourth and fifth lumbar vertebrae.These are similar to lesions known as cloacae (or
sinuses) associated with an infectious disease called osteomyelitis. Osteomyelitis occurs when the marrow
of a bone becomes infected, which can happen through a variety of means. The process involves both
bone destruction and bone formation. As the ordinary bone is destroyed by the infection, new layers of
bone are also built up, which, particularly in long bones, can lead to a very distinct thickening of the
bone. In addition, cloacae or sinuses form, allowing the infectious pus and dead fragments of bone from
the interior of the bone to be extruded; these eventually work their way up to the surface of the skin
(Ortner and Putschar 1981; Roberts and Manchester 1995).The holes described in the vertebrae of this
individual bear a resemblance to the cloacae associated with the infectious process of osteomyelitis. The
lumbar vertebrae do not show any other indication of infection, however, and so the oval lesions cannot
accurately be attributed to osteomyelitis and may in fact be a natural occurrence.
Comments
The bones of this adult are very noticeably robust, even for a male individual, with strong muscle markings
where the various muscles attach to the bones. Most of the bones from the upper body were disarticulated
(see below). This individual was radiocarbon-dated to AD 989–1167.
5.7 OSTEOLOGICAL MEASUREMENTS
5.7.1 ADULT
SKELETAL MEASUREMENTS
A limited number of osteological measurements could be taken on the adult human bones retrieved
from Skellig Michael.The measurements from the articulated burials are presented below and are divided
388
THE HUMAN REMAINS
Table 11—Adult cranial measurements.
Adult cranial measurements
Maximum length
Maximum breadth
Basion-bregma height
Porion-bregma height
Basion-porion height
Auricular height
Minimum frontal breadth
Total facial height
Upper facial height
Facial width
Nasal height
Nasal breadth
Orbital height
Orbital breadth
Maxillo-alveolar height
Maxillo-alveolar breath
Palatal length
Palatal breadth
Bicondylar breadth
Bigonial breadth
Height of ascending ramus
Min. breadth of ascending ramus
Symphysis height
Mandibular length
SK2
-
SK3
62.7mm
34.6mm
9.59mm
Table 12—Adult post-cranial measurements.
Adult post-cranial
measurements
GlL
GlB
ClL1
HuL1
HHD
HuBI.EPI.
HuART.
RaL1
UlL1
FeL1
FeHD1
FeD1
FeD2
FeE1
TiL1
TiD1
TiD2
TiE1
FiL1
SK2 (mm)
Left
-
Right
61.15
44.2
23.9
42.8
-
SK3 (mm)
Left
41.4
47.5
61.8
44.5
25.8
47.3
32.2
30
87.9
76.9
-
Right
43.1
33.2
25.4
456
47.9
31.8
27.8
37.8
37
27.3
81.4
-
389
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
into cranial and non-cranial measurements (after Bass 1995). Only a very limited selection of
measurements could be taken on the cranial remains—only on a mandible—owing to the incomplete
nature of the remains.The post-cranial measurements are equally scant. Any measurements that could be
taken on disarticulated adult bones are provided in the database (Appendix III).
5.7.2 JUVENILE
SKELETAL MEASUREMENTS
The remains of the articulated child burial (SK1) are incomplete, with significant fragmentation. The
cranial remains were too fragmentary to allow any measurements. It was not possible to measure the
length of any of the long bones. It was possible to measure the mid-shafts of a limited number of bones
and these are presented below. Any measurements that could be taken on disarticulated subadult remains
are presented in the relevant database fields in Appendix III.
Table 13—Juvenile skeletal measurements.
Bone
Mid-shaft measurement (mm)
Clavicle
Humerus
Radius
Ulna
Left
16.8
12.2
12.3
390
Right
10.2
18.2
12.2
12.8
6. PALAEOENVIRONMENTAL ANALYSES
6.1 INTRODUCTION
Throughout all phases of excavation, sampling for bioarchaeological analysis was undertaken from all
meaningful contexts. In some instances (e.g. wood) samples were hand-picked but, given the lack of
facilities for processing on site, in general bulk soil samples were taken; these were submitted in two
batches (in 2001 and 2008) to Margaret Gowen and Co. Ltd for preliminary analysis. This identified
those samples that had potential for further analysis for plant macrofossils, wood, charcoal and insects
(Johnston 2001; Allen et al. 2008). Following consultation with the archaeological directors, those samples
deemed worthy of full analysis were proceeded with in 2009 by independent environmental
archaeologists Ryan Allen, Lorna O’Donnell and Eileen Reilly.
This report details the results of plant macrofossil, wood/charcoal and insect analysis and assesses the
results in the context of contemporary findings from other monastic sites, where available. Integration of
the results, where appropriate, is also attempted, to give an overview of the environment and economy
of the settlement at Skellig Michael during the early medieval period. Further biogeographical
implications of the insect assemblages, in particular, are also discussed.
6.2 PLANT MACROFOSSILS
Ryan Allen
6.2.1 INTRODUCTION
Plant remains were sorted and analysed from eight samples from Skellig Michael.These were from three
different areas: the lower monks’ garden, the cistern and the leacht area. While few plant remains were
recovered, they did give a picture of the general plant economy present on site.
6.2.2 METHODOLOGY
Processing
The samples were processed using either bucket flotation or wet sieving, depending on the state of
preservation of the deposit. Bucket flotation, used for carbonised remains, is carried out by placing a fivelitre subsample into a container and then adding water. After gentle agitation any carbonised remains
float to the top. Excess water is poured through a 250µm mesh, and the resulting remains (called the
‘flot’) are dried and bagged for analysis. Anything left in the bucket (called the ‘retent’) is washed through
a 2mm mesh, and is then dried and bagged as well. Wet sieving is used for waterlogged samples and
involves washing a five-litre subsample through a stack of nesting sieves, in this instance with meshes of
1mm and 250µm.
Identification
Both the waterlogged and carbonised samples were sorted using a low-powered binocular microscope (7x–
45x magnification). Any remains found were identified using seed atlases (Berggren 1981; Cappers et. al.
2006) and a modern reference collection when possible. In the case of fragmentary grains or seeds, only
391
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
the embryo was counted in the final results.This is because the embryo is the most diagnostic part of these
remains; it also eliminates the problem of possibly counting multiple pieces of the same seed/grain. Identified
plant remains were stored in labelled specimen tubes. To facilitate reading, the plants’ common names are
used (with the scientific name in parenthesis after the first instance) in this report.
6.2.3 RESULTS
Seven samples from Skellig Michael contained plant remains: ID2, ID7, ID12, ID13, ID14, S77 and S88.
These were all from the lower monks’ garden, except ID2, which was from the leacht area east of St
Michael’s Church, and ID14 from Cistern 3. Sample S77 was a single barley grain attached to mortar
found in the lower monks’ garden. Full results can be found in Table 14.
Table 14—Plant remains recorded from Skellig Michael.
Sample ID
77
Fraction
14
14
1
0.25 1
13
13
12
7
7
2
2
88
88
0.25 1
12
0.25
1
0.25
1
0.25
1
0.25
(mm)
mL
600 100
150
100
300
150
300
100
100
100
50
50
mL picked
600 100
150
100
300
150
300
100
100
100
50
50
% picked
100 100
100
100
100
100
100
100
100
100
100
100
Silene dioica
1
cf. Alnus glutinosa
1
Chenopodium
13
1
2
1
2
1
1
album
Hordeum vulgare
1
Avena sp.
7
Poaceae
1
1
1
Leacht area, east of St Michael’s Church
Three seeds/grains were recovered and identified from sample ID2 (F594), which came from a drain fill
pre-dating the construction of the leacht and therefore is of secure early medieval date. These were from
barley (Hordeum vulgare), goosefoot (Chenopodium album) and an indeterminate grass seed (Poaceae).
Cistern 3
Two seeds were identified from the single cistern sample (ID14), one of campion (Silene dioica) and one
of alder (cf. Alnus glutinosa). This is a mixed, disturbed context with modern material.
Lower monks’ garden
Six samples from the lower monk’s garden yielded plant remains. These contained a range of taxa,
including barley, oats (Avena sp.), goosefoot and indeterminate grass. The oats could not be identified as
the domesticated type owing to a lack of diagnostic plant parts, but it is likely that they were purposely
grown for consumption. Likewise, the type of barley (naked/hulled, two-row/six-row) could not be
identified owing to its poor state of preservation.
(Following completion of the above report, a single charred seed from the lower monks’ garden
(labelled 93E0195:229) was presented for analysis to Meriel McClatchie and was identified as a grain
fragment of Avena sp. (oat).)
392
PALAEOENVIRONMENTAL ANALYSES
6.2.4 DISCUSSION
While the plant remains from Skellig Michael were sparse, they still shed some light on the agricultural
economy. The only ‘environmental’ plant remains present were those of alder and campion, which are
likely to be of modern origin given their recovery from the disturbed/mixed fill (F615) of the cistern
(E. Bourke, pers. comm.).
Both barley and oats were found in layers from the lower monks’ garden, although the low numbers
make it impossible to determine with any statistical relevance which cereal was utilised in greater
abundance. Five barley grains were identified in four samples, while seven oat grains were identified in
one. Both were important economic plants during the early medieval period for both human and animal
consumption. Unlike wheat, both barley and oats are tolerant of cooler and wetter climates (Feehan
1997). This is especially true of oats, which were a readily available and cheap crop used for making flat
oatcakes, porridges and gruels (Sexton 1998), as well as being useful as animal fodder (Monk et al. 1998).
Barley was also used for making porridges and gruels, along with malt for beer production, although the
latter doesn’t appear to have been the case on this site as there was no indication of germination on any
of the grains found.
The presence of oat and barley, along with the lack of wheat or rye, is in line with other findings
from early medieval sites in Ireland (Monk 1991; Monk et al. 1998; Fredengren et al. 2004). More
importantly, the same held true for both High Island, Co. Galway, and Illaunloughan, Co. Kerry, both
contemporary island sites (McClatchie 2008; Murray et al. 2004).
Goosefoot, found in six of the eight samples analysed, is a common invasive species generally
associated with plant cultivation. Its relative abundance in the Skellig Michael samples is suggestive of
cultivation taking place on the island, as opposed to the oats and barley being imported as grain. There
were likely other ‘weed species’ present, but their small seed size and thin coats make them liable to early
destruction in the carbonisation process as well as through taphonomic processes (Boardman and Jones
1990). This may also explain the lack of plant species that could be used as general environmental
indicators.
6.2.5 SUMMARY
While very few plant remains were recovered from Skellig Michael, they did suggest a plant economy
based on barley and oats. These findings fit in with the agricultural models found on monastic sites
throughout Ireland during the early medieval period, including those of High Island, Co. Galway, and
Illaunloughan, Co. Kerry. Goosefoot, present in six of the eight samples analysed, is a possible indicator
that the crops were grown on site.
6.3 CHARCOAL AND WOOD
Lorna O’Donnell
6.3.1 INTRODUCTION
Charcoal and wood remains were identified from two main areas at Skellig Michael, the lower monks’
garden and the leacht area. The aim of the charcoal and wood analysis was to assess the use of fuel on the
island and to examine possible habitats from which the wood may have been derived. It also served to
identify suitable material for radiocarbon dating.
393
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
6.3.2 METHODOLOGY
Processing
The samples examined from Skellig Michael were a mixture of bulk soil and hand-picked material. The
method of processing the bulk soil samples depended upon the nature and degree of preservation of the
deposit.
Waterlogged soil samples were processed using wet sieving. This involves washing a two-litre
subsample through a stack of meshed sieves, measuring 2mm, 1mm and 250µm. The remains (called
‘retent’) in each mesh are then stored separately in airtight, waterlogged conditions. Carbonised samples
were processed using a method called bucket flotation. A five-litre subsample is put into a bucket, water
is added, and the sample is agitated, allowing any carbonised remains to float to the top of the water.This
is then carefully poured over a 250µm sieve.The remains (called the ‘flot’) are dried and bagged. Anything
left in the bucket (called ‘residue’) is washed over a 2mm mesh, dried and bagged.
Wood and charcoal identification
Each wood and charcoal piece was identified by a first selection under a stereomicroscope, with low
magnification, varying from 30x to 75x. The transverse, tangential longitudinal and radial longitudinal
surfaces of each wood and charcoal piece were revealed. In the case of the wood fragments, this was
achieved by taking a section from each plane of wood (see Hather 2000) using a razor-blade and
mounting the section (typically only a few millimetres across and 50–75µm thick) onto a glass slide.
Analysis of thin sections of wood took place under a transmitted light microscope with magnifications
of 40x to 400x. Each charcoal piece was broken to reveal the necessary sections for identification. A
microscope with incident-light optics (40x–400x) was used to identify the charcoal. The identifications
were carried out using manuals (Hather 2000; Schweingruber 1978;Wheeler et al. 1989) and a modernday reference collection. It was aimed to identify 100 fragments of charcoal per sample, following Keepax
1988. If 100 identifiable fragments were not present in a sample, as many as possible were identified.
Fragments were both weighed and counted.
Details of charcoal recording
The general age group of each taxon per sample was recorded, and the growth rates were classified as
slow, medium, fast or mixed. Ring curvature of the pieces was also noted—for example, weakly curved
annual rings suggest the use of trunks or larger branches, while strongly curved annual rings indicate the
burning of smaller branches or trees (Ill. 6.1). Tyloses in vessels in species such as oak can denote the
presence of heartwood. These occur when adjacent parenchyma cells penetrate the vessel walls (via the
pitting), effectively blocking the vessels (Gale 2003). Insect infestation is usually recognised as round
holes.Their presence normally suggests the use of decayed degraded wood, which may have been gathered
from woodland floors or may have been stockpiled.
Ill. 6.1—Ring
curvature.
Weakly curved
rings indicate
the use of
trunks or large
branches
(Marguerite
and Hunot
2007, fig. 3).
394
PALAEOENVIRONMENTAL ANALYSES
6.3.3 RESULTS
Twenty-six charcoal and eleven wood samples were examined from Skellig Michael.The majority of the
samples were hand-picked.Thirteen wood taxa were identified (Table 18; Ill. 6.2), including alder (Alnus),
birch (Betula), hazel (Corylus avellana), ash (Fraxinus excelsior), holly (Ilex aquifolium), wild/bird cherry
(Prunus avium/padus), spruce/larch (Picea abies/Larix), Prunus sp., oak (Quercus), purging buckthorn
(Rhamnus catharticus), willow (Salix), yew (Taxus baccata) and elm (Ulmus). The results are dominated by
oak. Four pieces of uncarbonised yew wood were also identified from the site. Oak is present in nineteen
of the samples, hazel in twelve and ash in eleven. Willow was identified in five samples, alder and birch
in four.Yew is present in two samples, while buckthorn, holly, spruce, cherry, Prunus sp. and elm were all
only identified in one sample.
Ill. 6.2—Total
charcoal
identifications
from Skellig
Michael.
Ring curvature and growth details
General growth rates, the type of curvature of the annual rings, size, the range of annual rings, presence
of heartwood and any insect holes were noted on all the charcoal pieces (per taxon, per sample). This
can determine whether the material represents wood from mature woodlands, what sort of ground
conditions the parent trees were growing in and what condition the wood was in prior to burning (e.g.
decayed, wet, dry). The most frequent trees identified will be discussed in detail.
The alder fragments range in size from 5mm to 10mm, with two to nine growth rings remaining.
The growth rate is medium. Strongly curved annual rings and roundwoods present indicate the burning
of small branches, while weakly curved annual rings from F507 suggest the burning of larger pieces of
wood. Insect holes were noted in material from Sample 125 (F572), indicating that decayed branches
were burnt.The hazel pieces range in size from 4mm to 10mm, and overall growth is medium; all of the
pieces appear to be derived from small branches or rods, given the strongly curved nature of their annual
rings. The ring counts range from two to ten and no insect holes were noted.
In contrast to the alder and hazel fragments, growth within the ash fragments is quite mixed, ranging
from slow, through medium to fast. Ring curvature is predominantly strongly curved, indicating branches,
although it is weakly curved from Sample 16 (F581), Sample 17 (F583) and Sample 84 (F634(4)),
indicating larger branches or trunkwood.Tyloses were recorded in Sample 84, suggesting that heartwood
was burnt. The sizes range from 4mm to 20mm, while the ring counts range from two to 35.
The oak fragments range in size from 4mm to 12mm, while the ring counts range from two to
fifteen. In comparison to the ash, the oak growth is mixed, with evidence for slow, medium and fast
growth.The growth rings are predominantly weakly curved, with tyloses (indicating heartwood) in eight
of the samples (81, 131, 16, 17, 18, 83 and from F505). The presence of some strongly curved annual
395
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
rings suggests that, along with larger branches and trunkwood, small branches must also have been burnt.
Growth of the yew and spruce fragments would have been slow, as would be expected from
softwoods. The pieces range in size from 5mm to 10mm, and have between five and nine annual rings
remaining. The strongly curved nature of their annual rings indicates that they are all derived from
branches. The uncarbonised yew roundwoods are tightly grouped in size, ranging in width from 3mm
to 5mm. Their annual rings are strongly curved, indicating that they are derived from young branches;
between three and four annual rings remain.
6.3.4 CONTEXTUAL
RESULTS
Charcoal and wood samples were mainly derived from contexts relating to the lower monks’ garden and
the leacht area (see Table 18).
Lower monks’ garden
Charcoal and wood was examined from F634(4) (Samples 66–76, 81 and 84), F635(4) (Sample 83),
F636(4) (Sample 85), F634(5) (Samples 15–16 and 54), F635(5) (Samples 18–20) and F636(5) (Sample
17) of the lower monk’s garden. These layers were deposited/redeposited between the twelfth and
fourteenth centuries but contain material ranging in date from the seventh to the fourteenth century
(see section on radiocarbon dating above).Yew wood and oak, ash, hazel and birch charcoal were identified
from Cutting 4. Oak, ash, yew, hazel, alder, holly and willow charcoal was identified from Cutting 5. Oak
was the main wood identified.
Leacht area
Charcoal was identified from different contexts relating to activity around the leacht. Alder and oak
charcoal was identified from F572 (Sample 125), which dates from the post-medieval period. Oak only
was noted in a sample from an ash layer within the south-east corner of the leacht, east of St Michael’s
Church (Sample 131).
Hazel, willow and yew were identified from F579 (Sample 14), located against the base of the leacht,
although it had suffered some petrel/puffin disturbance. Hazel charcoal was noted from F580 (Sample
15), against the base of the leacht. Birch, ash and oak were noted in F581 (Sample 16), located underneath
F579. Birch, hazel, ash, spruce, oak and willow were identified from F583 (Sample 17), located to the
south-east of the leacht, overlying one of the steps relating to the secondary use of the east entrance. From
F588 and F589, located south-west of the leacht, alder, birch, oak and purging buckthorn were identified
(Samples 18 and 19). South of the leacht, oak charcoal only was identified from F590 (Sample 20). All of
the samples retrieved from the leacht area (just described), with the exception of Sample 14, are of secure
early medieval date. The charcoal from contexts associated with the leacht is dominated by oak, in
comparison to those from the lower monks’ garden.
It was not possible to identify a wood piece from F509 (Sample 41), which came from an early
medieval backfill layer in the east entrance. Neither was it possible to identify the charcoal from F685
(Sample 141), which came from a hearth in the structure at the base of the east steps.
East and south entrances
Three samples were examined from early medieval deposits relating to both the east entrance and south
entrance 1, inner enclosure. Hazel, ash and oak were identified from the three contexts F505, F507 and
F545. Willow was identified from the first two contexts, while Prunus sp. and elm were noted within
F505 and alder and cherry within F507.
396
PALAEOENVIRONMENTAL ANALYSES
6.3.5 DISCUSSION
One of the main principles of charcoal analysis is the assumption that people will gather fuel as close to
a site as possible and that the charcoal can reflect the local woody flora of the area. Abandoned structural
timbers or wood brought to the site for use in construction or other activities are also often reused as
firewood.
One of the fundamental questions to consider when interpreting the Skellig Michael charcoal is
where the wood came from. Was it gathered and burned on the island, or does the charcoal represent
wood brought in from different locations and transported to Skellig Michael as fuel or as structural wood
that was later burned for fuel?
Pollen was examined from F318, an old ground surface, probably contemporary with the
construction of the large oratory in the monastery (see pollen analysis below). Oak, hazel and birch
pollen was identified, although in such small quantities that it is interpreted as representing long-distance
transport from the mainland. The main pollen type representing wood is Sorbus (rowan/mountain ash).
The pollen data also indicate a completely open landscape. Skellig Michael is a craggy rock in the Atlantic,
exposed to extreme weather conditions and with a patchy, thin soil cover, and is therefore unlikely to
have supported any significant tree cover.The steep, rocky shoreline of the island also makes it improbable
that driftwood would have been gathered there for burning.
All of the trees identified within the wood and charcoal are native Irish species, with the exception
of spruce/larch.The results indicate that fuel was being gathered from different environments, including
tall canopy woodlands (oak, elm and ash), possible scrub (hazel, Prunus sp., cherry) and wetland (alder
and willow).
The wood from Skellig Michael, therefore, was most likely gathered from different types of
woodlands on the mainland. The oak present could be native pedunculate (Quercus robur), which prefers
wetter, heavier clays than sessile oak (Quercus petraea) (Beckett 1979). Ash is common on shallow, often
skeletal soils over limestone and chalk, though the moisture levels must not reach extremes of dryness in
summer. It is also common beside streams on soils of only moderate base status. Ash appears to be able
to obtain its nutrient requirements when moving eutrophic water is available (Cousens 1974). Elm trees
will grow well on rich, alluvial soils and prefer riverine habitats (Gale and Cutler 2000, 264).
Hazel is a very tolerant tree. It can grow in wet or dry conditions but not in waterlogged ones (Orme
and Coles 1985). It was once very common in Ireland; indeed, McCracken (1971) notes that it was once
widespread to an extent that is hard to imagine today. It can grow as a tree (frequently associated with
oak as an understorey) or can form hazel scrub. Other possible scrub trees identified within the charcoal
include Prunus sp.
Wild cherry (Prunus avium) needs light to grow, on or near woodland margins and on light, welldrained soils (Orme and Coles 1985, 11). Bird cherry (Prunus padus) occurs particularly in marginal
forests and is generally solitary (Stuijts 2005, 142). Blackthorn (Prunus spinosa) grows in woodland where
the canopy has been opened, on woodland margins, in scrub and along streams, where it may be found
with alder. It does not survive under heavy shade (Orme and Coles 1985, 11).
Alder is a wetland tree and can often be seen growing alongside rivers, lakes, on marshes or in fens.
It can form alder carr when its roots are in water. It is able to survive on these wet sites (which generally
lack the nitrates needed for growth) as its roots have nodules that contain nitrogen-fixing bacteria, which
extract nitrogen from the air (Lipscombe and Stokes 2008). Ireland’s native tree is the black or grey alder
(Alnus glutinosa).Willow will also grow in wet areas.The main Irish native willows are grey willow (Salix
cinerea), goat willow (Salix caprea) and eared willow (Salix aurita). Birches such as silver birch (Betula
pendula) and downy birch (Betula pubescens) are often found on the margins of bogs, lakes and rivers.They
are very good at colonising disturbed ground (Hickie 2002).
397
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Yew is an evergreen conifer that grows up to 20m. Older trees often divide into several distinct
trunks. It prefers well-drained and sheltered sites (Hickie 2002). Norway spruce likes moist soils and can
grow in cold, wet and shallow soils (Lipscombe and Stokes 2008). Spruce is not a native Irish tree.
McCracken (1971) writes that by the eighteenth century it was being imported into Ireland. The
spruce/larch identified from Skellig Michael may indicate post-medieval contamination of the sample.
Two shade-tolerant trees within the assemblage are buckthorn and holly. Purging buckthorn is a
shrub or small tree growing to 6–8m tall, with grey-brown bark and spiny branches. Holly is a hardy
tree and can be found on higher, exposed ground or growing underneath taller forest trees, forming
understorey (Hickie 2002, 59).
It is possible that other monastic settlements, such as the one tentatively identified on Bray Head,
Valentia Island, were supplying smaller monasteries such as Skellig Michael and Illaunloughan with fuel.
Hazel, alder, oak, yew and blackthorn were identified from Illaunloughan (O’Carroll 2005). This is very
similar to the results from Skellig Michael.
Skellig Michael was inhabited from approximately the seventh to the thirteenth century AD (see
interim conclusions below). The value of woodlands was recognised in the eighth-century Bretha
Comaithchesa or ‘laws of the neighbourhood’. Trees were classified into four groups, depending on the
economic value of the wood. This value could be calculated by its use as a structural timber, such as ash
or oak, or because of its dietary contributions, such as hazelnuts or apples.The classifications were ‘nobles
of the wood’, ‘commoners of the wood’, ‘lower divisions of the wood’ and ‘bushes of the wood’. Trees
like oak, holly and hazel were considered ‘nobles of the wood’, while the lowest classification (bushes of
the wood) included bracken, bog myrtle and gorse or furze. Fines were imposed on people caught cutting
branches or cutting at the base of the trees. These varied according to the classification of the tree, and
included confiscation of heifers and sheep (Kelly 1976).
A considerable amount of charcoal was identified from High Island, an early medieval monastic
settlement in County Galway, approximately 3km off the north-west coast of Connemara (Gale,
forthcoming). The main activity on the site is from the eighth to the thirteenth century AD. Similar to
Skellig Michael, oak and hazel were important fuels on the site. Other trees that were being burned on
the two islands include alder, birch, ash, holly, willow and yew. In contrast to Skellig Michael, juniper
charcoal was identified from High Island (Gale, forthcoming).
Excavations at the early medieval site of Kilgobbin, Co. Dublin, provided evidence for cropprocessing and metalworking. Fourteen trees were identified. Oak, hazel and alder dominated the charcoal
results. Most of the trees identified from Skellig Michael are present, with the exception of spruce and
purging buckthorn (O’Donnell 2005). Scrub trees such as hazel, elder and cherry dominated the charcoal
remains recovered from the early medieval cemetery (Site M) at Knowth, Co. Meath (Johnston and
O’Donnell 2008).
Charcoal results from Skellig Michael indicate that most of the fuel was being gathered from oak
and hazel woodlands, presumably located on the mainland. This has to be approached with caution,
however, as the charcoal from the site represents mainly hand-picked fragments and the overall counts
are relatively low.
The presence of some wood-borer species in the beetle remains from the lower monks’ garden could
indicate the stockpiling of wood or the burning of degraded wood (see insect remains, below). Insect
holes were, however, recorded infrequently in the charcoal.
6.3.6 RADIOCARBON
DATING
It is best to select short-lived wood species with strongly curved annual rings, such as hazel or willow,
398
PALAEOENVIRONMENTAL ANALYSES
for radiocarbon dating. Although yew can be a long-lived species, the strongly curved nature of the annual
rings of the yew wood identified indicated that they were derived from small branches and would be
suitable for radiocarbon dating. Table 18 includes a column with recommended material for dating.
6.3.7 SUMMARY
Charcoal and wood were analysed from 37 mainly hand-picked samples from the excavations on Skellig
Michael. Thirteen trees were identified, dominated by oak and hazel. It is likely that the wood was
gathered from different types of woodlands on the mainland, including oak, scrub and wet or carr
woodlands. Charcoal identifications are similar to those from the early medieval monastic island
settlements at Illaunloughan, Co. Kerry, and High Island, Co. Galway.
6.4 THE INSECT REMAINS
Eileen Reilly
6.4.1 METHODOLOGY
Table 15 lists the samples that were analysed for insect remains from among the samples scanned in 2001
and 2008.
Table 15—Samples analysed for insect remains.
Sample/ID no. Context
Description
S88
F634(4), lower monks’ garden
Blackish organic silty soil with frequent wood
inclusions
S87
F634(4), lower monks’ garden
ID 2
ID 3/4
F594, leacht area
F593, from beneath leacht
ID 5
F505, east entrance
Blackish organic silty soil with frequent charred
and uncharred wood
Grey-black silty soil with frequent charcoal
Silty deposit with high pebble content and
abundant feathers
Grey-black silty soil with high inorganic fraction
and charcoal
ID 8
F596, leacht area
Silty deposit with high pebble content and
abundant feathers
ID 11
Lower monks’ garden, Cutting
5, east end
Inorganic silty layer with charcoal and burnt
shell
ID 12
F638, lower monks’ garden
Blackish organic silty soil
ID 13
F638, lower monks’ garden
Blackish organic silty soil
ID 14
F615, Cistern 3
Moderately organic silty soil with frequent
undiagnostic plant remains
399
Table 16—Basic habitat statistics: MNI counts and percentage presence of each habitat group (% of total MNI excluding ‘varied’).
S88 (LMG)
F634(4)
S87 (LMG)
F634(4)
12 (LMG)
F638
13 (LMG)
F638
68
64
31
23
2
8
0
1
19
15
9
1
3
6
4
14
F615,
cistern
20
18
9
7
4
0
0
0
5
3
0
1
0
5
2
3/4
F593,
leacht area
181
166
21
7
2
12
1
1
7
3
2
2
4
132
15
8
F596,
leacht area
155
145
16
5
2
5
0
4
1
4
0
0
1
128
10
Total no. of individuals
Total no. of individuals exc. varied
Total no. of species
Index of diversity
Ground vegetation
Damp ground/mosses
Coastal/bare ground
Disturbed/cultivated ground
Generalist decomposers
Foul/decaying vegetation
Synanthropic (‘house’) fauna
Wood litter/woodland associates
Wood-borers
Carrion/birds’ nests
Varied biotopes
Percentage presence of habitats
Ground vegetation
Damp ground/mosses
Coastal/bare ground
Disturbed/cultivated ground
Generalist decomposers
Foul/decaying vegetation
Synanthropic (‘house’) fauna
Wood litter/woodland associates
Wood-borers
Carrion/birds’ nests
71
61
29
19
0
8
0
0
16
10
15
5
3
5
10
219
188
41
15
1
23
0
4
44
22
29
8
12
45
31
105
88
35
19
1
7
0
0
32
10
12
8
8
10
17
0.0
13.1
0.0
0.0
26.2
16.4
24.6
8.2
4.9
8.2
0.5
12.2
0.0
2.1
23.4
11.7
15.4
4.3
6.4
23.9
1.1
8.0
0.0
0.0
36.4
11.4
13.6
9.1
9.1
11.4
3.1
12.5
0.0
1.6
29.7
23.4
14.1
1.6
4.7
9.4
22.2
0.0
0.0
0.0
27.8
16.7
0.0
5.6
0.0
27.8
1.2
7.2
0.6
0.6
4.2
1.8
1.2
1.2
2.4
79.5
1.4
3.4
0.0
2.8
0.7
2.8
0.0
0.0
0.7
88.3
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
400
Sample/ID no.
Context
PALAEOENVIRONMENTAL ANALYSES
As the samples were processed in 2001 and 2008 using the wash-over method, the two or three
resulting flots per sample were recombined and subjected to paraffin flotation, the standard method used
by archaeo/palaeoentomologists for the extraction of insect remains from soil samples (Coope and
Osborne 1967; Kenward 1980; Kenward et al. 1986). The samples were disaggregated using water and
washed over a 300µm sieve. The retent in the sieve was then treated with paraffin and cold water added.
The paraffin concentrates the insects by adhering to the waxy cuticle of the insect exoskeleton. The flot
was then poured through a 300µm sieve, washed with detergent in hot water to remove the excess paraffin
and stored in 70% ethanol. All residues were kept. Processing took place at the laboratory of Dr Bettina
Stefanini. Insects were sorted under low-powered binocular magnification and all sclerites were extracted
onto wet filter paper.
Of the ten samples listed in Table 15, three had less than twenty insect fragments after full sorting
(ID 2, 5 and 11). These are excluded from any further analysis.
Identification of the insect remains was carried out using published keys, the writer’s own collection
of comparative material and the comparative collections of British Coleoptera housed in the Oxford
University Museum of Natural History (with thanks to Darren Mann). Habitat data were gleaned from
various published sources and the BUGS database, the coleopteran ecology package designed and updated
by Dr Phillip Buckland and Dr Paul Buckland (Buckland 2007; Buckland and Buckland 2006). A full
species list is presented in Table 19, with taxonomy following Böhme 2005.
Analysis of samples involved summing the individual insect taxa into ecologically related habitat
groups (Table 16; Ill. 6.3). The purpose of analysing the data in this way is to generate a general picture
of change in the relative proportion of each habitat group within each deposit, group of deposits and
site as a whole, in order to elucidate the character of the local site environment. Fisher’s alpha (a measure
of species diversity) was also calculated for each sample where N > 20 (N = number of individuals) (see
Table 16) (Fisher et al. 1943).
Ordination of samples from the lower monks’ garden was also carried out to assess the source of the
insects in those assemblages (see below). Ordination is the arrangement of items along a scale (axis) or
100
90
% presence of each habitat group
80
70
Ground vegetation
60
Damp ground/mosses
Coastal/bare ground
50
Disturbed/cultivated
ground
Generalist decomposers
Foul/decaying veg.
40
Synanthropic ('house')
fauna
Wood litter/woodland
associates
30
20
Wood borers
Carrion/bird's nests
10
0
LMG F634(4)
LMG F634(4)
LMG F638
LMG F638
Cistern 3 F615
Leacht F593
Leacht F596
S88
S87
12
13
14
3/4
8
sample/ID no and context
401
Ill. 6.3—Insect
habitat data for
main deposits
from Skellig
Michael.
LMG=lower
monks’
garden.
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
multiple axes. It is used to summarise complex relationships, extracting one or a number of dominant
patterns from an infinite number of possible patterns (McClune and Grace 2002). It is used here to
examine underlying variance/similarities in the lower monks’ garden assemblages and assemblages from
a selection of other sites in Ireland and one from Russia (Table 17; Ill. 6.4). Non-metric multidimensional
scaling (NMS) was used for this purpose, as it is well suited to data that are non-normal or on arbitrary,
discontinuous or otherwise questionable scales, particularly data sets that have upwards of 70% zeroes.
Ordinations were calculated using the PC-ORD programme (McClune and Mefford 1999).
6.4.2 ANALYSIS
Lower monks’ garden deposits
Four samples came from the lower monks’ garden, from two layers dated to the twelfth–fourteenth
centuries (F638 and F634(4)) which also contained material dated to the early medieval period
(seventh–tenth centuries) that was clearly redeposited (see Sections 2.27 and 4).
Sample ID 13, F638
The assemblage from this sample was moderately numerically rich, with 68 individuals representing 31
taxa. This produced a species diversity index of 23, which is quite high, suggesting mixed origins for the
insects in the assemblage (Tables 15 and 19). The habitat groups represented in the assemblage are
illustrated in Ill. 6.3. One notable aspect of the assemblage is the relatively low number of what might
be called ‘native’ island fauna with any degree of certainty. This is in comparison to the relatively high
number of species more than likely brought to the island by its human inhabitants.
‘Generalist decomposer’ and ‘foul/decaying vegetation’ species dominate the assemblage.These groups
represent species found in vegetation in various stages of decay, fungoid wood and dung.The commonest
species in these categories from this deposit are Xantholinus spp, Pteridium spp, Cercyon atricapillus and
Megasterum obscurum. These species are all found regularly on human habitation sites, particularly in the
medieval period. They favour the accumulated rubbish and dung in both occupation layers and pit fills
(Kenward and Hall 1995). Nevertheless, all are found in nature too, particularly in accumulating leaf
litter, forest floors, birds’ nests, rotting seaweed and animal dung (Hansen 1987; Koch 1989).
The proportion of synanthropic fauna (i.e. species that favour human occupation sites) is surprisingly
high for an island site like Skellig Michael, and is a common feature of all the lower monks’ garden
assemblages. Many such species are now extremely rare in nature, such was their dependence on the artificial
niches created by human occupation sites, particularly during the medieval period (Kenward 1997). Species
such as Aglenus brunneus and Mycetaea subterranaea, in particular, are strongly associated with human habitation
and are highly unlikely to have been part of the native island fauna (Kenward and Hall 1995).
Two other notable habitat groups represented in this assemblage are ‘carrion/birds’ nests’ and ‘woodborer’ species. The former may represent some native elements of the fauna, as the island is famous for
its bird colonies. It is likely that a range of carrion/detritus-feeding beetles became established on the
island over time, given the plentiful availability of suitable habitats. While this is certainly true of
Sciodrepoides watsoni (see Samples 3/4 and 8 below for further discussion), however, it is possible that
Omosita colon and Hister impressus are incorporated into the assemblage though the presence of food waste
in the soil matrix (Halstead 1963; Hinton 1945). It is likely, given the synanthropic element of the
assemblage, that general household waste from the cells—bedding material, stored foodstuffs, discarded
bones, possibly even human faeces—were used to manure and build up the garden soil. O. colon and H.
impressus may have been imported intermittently with carcasses and animal products and this may be the
reason for their presence in the deposit. Equally, they may have established temporary populations feeding
402
PALAEOENVIRONMENTAL ANALYSES
on butchered bird bones, with O. colon possibly also occurring on decaying seaweed being used as a
fertiliser (Backlund 1945). It is notable that only one true dung-beetle (Aphodius sp.) was found in any
of the deposits from Skellig Michael, suggesting that live animals were not kept on the island and that
dung was not used as a fertiliser. Given the constant supply of birds as a meat source, it was probably
deemed an unnecessary drain on food resources to husband live animals on the island.
The wood-boring insects are another surprising element in the fauna. As it is very unlikely that trees
of any significance existed on the island at the time of the monastery, all the wood-associated species
must have been imported to the island in wood or wood products. The two most significant species are
Rhopalomesites tardyi and Pseudophloeophagus aeneopiceus. Both are found in dead wood of a variety of tree
species, generally in woodlands, old parkland trees and occasionally from driftwood (Morris 2002). P.
aeneopiceus has been found very occasionally in structural wood in southern England but no such
association is known from Ireland, where it is generally very rare (ibid.). Both have a markedly southerly
and westerly coastal distribution in Britain, though R. tardyi is more widespread in Ireland and is known
from inland woodland situations (Alexander 2002; Reilly 2008). Their presence on Skellig in the lower
monks’ garden soils is intriguing and may be due to the stockpiling of collected driftwood or forestfloor wood from the mainland, brought out to the island for various uses.These species can emerge from
drying driftwood many months after they enter it (Darren Mann, pers. comm.).
Of the possible native island elements in the assemblage, a small number of damp ground, disturbed
ground and plant-feeding species are present, such as Agriotes sp. (general foliage, grassland), Lathrobium
spp, Stenus spp (both generally found in damp situations) and Trechus quadristriatus (a subterranean groundbeetle, found in disturbed ground, under rocks and stones and fairly ubiquitous in Ireland).
In summary, the insect assemblage appears to indicate two main sources for the species recorded: a
likely imported element of decomposer species (including synanthropes), wood-borers and possible
carrion species from bedding and foodstuffs used on the site and subsequently used to build up the
garden soil, and a small native element of decomposer, vegetation and wet-ground species. The former
element, however, forms the bulk of the assemblage.
Sample ID 12, F638
This sample was numerically richer than ID 13 (which was derived from the same context), with 105
individuals representing 31 taxa. The index of diversity was lower, however, at 19, suggesting a slightly
more restricted range of source habitats (Tables 15 and 19). While this sample came from the same layer
as ID 13, it displays some independent trends within the insect assemblage, suggesting that the garden
soil is not a homogeneous deposit.
Ill. 6.3 indicates that ‘generalist decomposers’ are again proportionally the largest habitat group
represented. Once again, species common in fungoid wood and decaying vegetation, as well as decaying
seaweed, are more frequently represented in this group. Some of these species may well be native to the
island but it is likely that a proportion have been imported with bedding, fertilising material and
foodstuffs.The synanthropic fauna is again well represented, with eight examples of Mycetaea subterranaea,
amongst others, present. Decaying animal matter was also incorporated into the soil here, as Omosita
colon, Sciodrepoides watsoni and Hister impressus are once again recorded.
Compared to sample ID 13, this assemblage has proportionally higher wood litter associates (Ill. 6.3)
and numerically more wood-boring beetles present, although the latter habitat group is mostly
represented by the same species as sample ID 13. It suggests that part of the matrix of the soil in this
location is derived from wood debris. This may have been wood used for other purposes on site and
then discarded, or wood litter gathered on the mainland and used directly to build up the soil layer.
Anobium punctatum, the well-known wood-borer generally referred to as the ‘woodworm’ beetle, is present
here and is generally a pest of structural wood (Alexander 2002). A. punctatum is still found in natural
403
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
woodland situations, however, especially in the Killarney woods (Reilly 2008). Its presence, therefore,
does not necessarily confirm that the wood debris in Layer G had a structural source.
Background or ‘natural’ elements in the fauna from this sample are quite low, with a small number
of damp ground/moss-feeding species occurring but a very low representation of plant-feeding species
(Table 15).
In summary, even excluding the generalist decomposer element of this assemblage as possibly native
to the island (and not all can be excluded on this basis), almost half of the assemblage is associated with
human habitation or woodland environments and originated in imported material.
Sample 87, F634(4)
This sample is from a layer which contained material dated from the seventh to the ninth century but
which was redeposited some time around the fourteenth century. It is numerically the richest assemblage
of all the samples, with 219 individuals representing 41 species. The index of diversity is quite low at 15,
suggesting that a small number of species are dominant.
In many respects, the assemblage from Sample 87 mirrors those from F638 described above.
Generalist decomposers, synanthropic species and ‘carrion/bird’s nest’ species are the three dominant
habitat groups, with particular taxa being quite numerous. Omosita colon, represented by 44 individuals,
suggests the presence of dead birds, other carcasses, bones or decaying seaweed in the soil matrix. The
small rove-beetle Metopsia clypeata is very numerous and is generally indicative of decaying vegetation
and leaf litter (Koch 1989). The synanthrope Mycetaea subterranea, represented by seventeen individuals,
suggests the presence of straw, hay or other bedding material (ibid.). At the fouler end of the decomposing
spectrum, Megasternum obscurum and Cercyon littoralis occur, indicating the possible presence of dung and
decaying seaweed respectively (Hansen 1987).The synanthropic group is somewhat more species-diverse
than the earlier layer, which may be indicative of developing in situ breeding populations. This is an
observed phenomenon on long-lived human habitation sites elsewhere and occurs as long as suitable
habitats remain available (Kenward 1997; Kenward and Allison 1994a).
Wood litter and wood-boring insects again are well represented, suggesting a continued presence of
gathered or stockpiled wood and perhaps wood litter on site. Interestingly, there are also proportionally
higher numbers of moss-dwelling/damp ground insects, which may come from the surrounding natural
landscape. Some, however, may be casualties brought on site in wet, fungoid or moss-covered wood or
with moss used for bedding and toilet purposes.
Two ground-dwelling beetles are also represented. Trechus quadristriatus, mentioned above, occurs in
disturbed ground and under stones, while Calathus fuscipes is common in cultivated, arable and garden
soils (Luff 2007). Both were probably living within the soil matrix itself and may be native to the island.
The assemblage clearly indicates that F634, as well as F638, was a deliberately built-up soil layer
largely comprised of gathered/imported material.
Sample 88, F634(4)
This assemblage is very similar to the foregoing but is somewhat numerically poor, 71 individuals
representing 29 species. The index of diversity is slightly higher than Sample 87 at 19.
The assemblage overall has a very similar profile to Sample 87, with generalist decomposers and
synanthropes both strongly represented but a slightly lower representation of carrion beetles. Woodboring and wood litter species are also proportionally well represented and there is negligible ground
vegetation, bare/coastal or disturbed ground species presence.The wet ground/moss species may, however,
have been derived locally (Table 15; Ill. 6.3).This indicates that there is very little incorporation of natural
elements of the local environment in the soil matrix.
404
PALAEOENVIRONMENTAL ANALYSES
Leacht area, east of St Michael’s Church
Two samples from material excavated beneath the leacht, east of St Michael’s Church, were examined for
insect remains—ID 3/4 (combined flots of samples ID 3 and 4) from F593 and ID 8 from F596. These
deposits were remarkable as they were largely made up of decaying bird’s feathers. It appears that storm
petrels and puffins built nests under the leacht at a later date, which contaminated these otherwise early
medieval deposits (Ed Bourke, pers. comm.). For this reason little of archaeological value can be derived
from the assemblages. Nevertheless, a short summary below outlines the main findings and some parallels
that can be drawn between these deposits and those from the lower monks’ garden.
F593, F596; Samples ID 3/4 and 8
Both assemblages were numerically rich, with 188 and 155 individuals respectively. Both samples were
species-poor, however, with 21 and 16 species each, giving diversity indices of 7 and 5 respectively. This
was mainly due to the total dominance of both samples by one species, Sciodrepoides watsoni (132 and
128 individuals each).
This small catopid beetle is almost exclusively tied to birds’ nests, animal burrows and the carcasses
of animals/birds in general (Harde 1984; Johnson 1965). It is found in association with bird colonies
along the western and northern coasts of Scotland and as far north as the Faroe Islands (Bengston 1981;
Berry 1985; Moore 1997;Waterston 1981). Its presence on Skelligs is therefore unsurprising. Nevertheless,
although included on the current Irish list of Coleoptera, there are no recent records of this species for
Ireland (Anderson et al. 1997). The insect sclerites were recently disarticulated and included many
immature species, clearly indicating an in situ breeding population. It is likely that they are wholly
representative of the later nest-building of the petrels/puffins at this location.
Of the remaining insects of both samples, some overlap with the lower monks’ garden deposits and
may be part of the original deposit’s insect death assemblage. In particular, both samples have a small
number of Rhopalomesites tardyi and Pseudophloeophagus aeneopiceus, which are likely to be contemporary
with the lower monks’ garden soils.
A very small number of beetles are unique to these samples and are probably indicative of local island
habitats. Trechus fulvus is known from coastal cliffs, bare, rocky terrain and coastlines (Luff 2007).
Otiorhynchus arcticus is generally found on low vegetation in uplands, while Rhinoncus ?castor is found on
weeds in fallow fields and disturbed ground (Bullock 1993). The latter may have been living on the
disturbed ground generated by construction of the monastery or in the garden area after it was
abandoned.
Unfortunately, owing to the mixed nature of the assemblage and the contamination by modern
insects, little more can be derived archaeologically from these assemblages. They do, however, give some
interesting insights into aspects of the biogeography of the island and these will be revisited later (Section
6.4.3).
Cistern 3
A small assemblage of beetles was recovered from the fill (F615) of the cistern—twenty individuals
representing nine species. Unfortunately, it is unlikely that this material is of archaeological significance
as the cistern contained material dumped in the nineteenth and twentieth centuries. Indeed, the
assemblage is dominated by fly puparia, in particular the species Coproica ?vagans, which breeds primarily
in animal dung and human excrement (Smith 1989).
The beetle assemblage is quite mixed and has a small number of ground vegetation-dwelling species,
likely to be locally derived accidental inclusions, and some generalist/foul decomposers. Interestingly,
there are no synanthropic species represented in the assemblage, suggesting that the cistern was probably
not used for dumping organic waste or that there was little organic waste (other than human excreta)
405
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
present on the island in the nineteenth century. It also suggests that any synanthropic fauna present on
the site during the monks’ occupation probably died out once suitable habitats were no longer maintained
(for further discussion see Section 6.4.3).
A note about fly and ant species from the Skellig samples
All samples had a moderate representation of ants and fly puparia. They have not been included in any
of the statistics or calculations as fly puparia, in particular, are not systematically extracted from samples
in the same way as beetle sclerites.
They are, however, a useful complementary or additional indication of the range of habitats
represented in the beetle assemblages.The four lower monks’ garden samples contain similar numbers of
the ant species Tetramorium caespitum and Stenamma westwoodi.The former is likely to be a native inhabitant
of the Skelligs, as its natural habitat is heath, upland coastal cliffs and sand (Bolton and Collingwood
1975). The origin of the latter is less certain. S. westwoodi is more often found in woodland, hedgerows
and the humus layer of soils (ibid.). It was recovered in forest hollows in Killarney and has, along with T.
caespitum, a generally southerly and westerly distribution in Ireland (Reilly 2008). It is possible that S.
westwoodi was brought to the Skelligs with dead wood or other wood debris and established a breeding
population in the garden soil of the lower monks’ garden during the period of monastic occupation. It
is notable that, apart from one example in one of the leacht samples (ID 8), it does not occur in the other
largely ‘modern’ assemblages. T. caespitum is represented in every deposit, however.
The range of fly puparia recorded is interesting but perhaps a little less informative. As flies are highly
mobile, they can establish temporary populations for short periods when favourable habitats are present,
and re-establish those populations again at later times. This is perhaps best illustrated by Coproica ?vagans,
which was recorded in the lower monks’ garden soils but also in the cistern and leacht deposits. This fly
is fairly ubiquitous, given its habitat preferences (cow dung and human excreta, in particular), so its
presence on Skellig is not surprising (Smith 1989). It does, however, tentatively suggest that human
excreta was being used to fertilise the monks’ garden, as there are no other true dung-beetles present.
Dung-beetles are also highly mobile, though heavier than flies, but could possibly have made it to the
Skelligs from the mainland, especially if dung was being stockpiled or used in large amounts on the island.
Notwithstanding that, if large quantities of cow dung were being transported out to the island to fertilise
the gardens, dung-beetles should have been incorporated in the material prior to transportation and
subsequently deposited in the monastic garden soils.
All other fly species are represented by only one or two puparia each. Calliphora vicina/vomitoria (the
‘bluebottle’) is a blowfly that occurs on recently dead animals and humans and could have been present
on dead birds (Smith 1989). Fannia sp. and Musca sp. (neither identified to species) both occur in the
cistern deposit. Members of the genus Fannia occur in rotting vegetables and fruit but, most notably, also
in latrines, being attracted to the smell of human urine (ibid.).The Musca genus contains the well-known
‘house fly’ and are general scavengers and decomposers, being especially found in dung and other foul
habitats. One example of Lonchaea sp. occurs in one of the lower monks’ garden deposits. Members of
this genus are generally found under bark and in dead wood of various tree species (ibid.). Again, it is
highly unlikely that this was a member of the native island fauna and it was probably brought to the site
in dead or structural wood.
Ordination of the lower monks’ garden assemblages
In order to tease out fully the nature of the lower monks’ garden insect assemblages and their possible
origins, an ordination of these four assemblages along with assemblages from a number of sites around
Ireland and further afield was carried out. The aim of ordination is to examine any underlying
variance/similarities in the data and to identify any possible patterns.
406
PALAEOENVIRONMENTAL ANALYSES
Two aspects of the lower monks’ garden assemblages were particularly surprising and interesting,
given Skellig Michael’s size, isolation and early medieval date—the relatively high proportion of
synanthropic and wood-dependent beetles. A continuously occupied, mainland, medieval-dated site
surrounded by farmland, woodlands and many potential source habitats might be expected to produce
such assemblages. Therefore a number of assemblages from sites of early to later medieval date with
moderately rich synanthropic and/or wood-dependent elements were selected for comparative purposes
(Table 17). To explore the woodland element of the assemblage further, two samples from a woodland
hollow in Derrycunihy Wood, Killarney, were selected. These samples also represent the geographically
closest ‘natural’ insect assemblages that could be sourced.
Site and Feature
Skellig Michael - Low er Monks Garden soils
Roestow n - Ditch fill, 8-10th century AD
Essex St - House floor layer, 10th century AD
Essex St - Pit fills, 13th century AD
Back Lane - Backfill house layer, 13/14th century AD
Back Lane - House floor layers, early 13th century AD
Novgorod - House floor layers, early 13th century AD
Novgorod - Outdoor 'midden' layer, 13/14th century D
Derrycunihy Wood - Forest hollow peat, 4th C BC/16th C AD
Ill. 6.4—Ordination of insect assemblages from lower monks’ garden and a range of other
assemblages from Ireland and Russia.
407
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Table 17—List of sites used in ordination, showing total MNI, percentage synanthropes and percentage wood-associated
beetles.
Site/sample no.
MNI
72
219
105
68
168
Synanthropic
fauna1
21.1
13.2
11.4
13.2
20.1
Wood-associated
fauna1, 2
11.3
9.1
15.2
5.9
3.0
SK88
SK87
SK12
SK13
RS194
TB613
68
17.6
1.5
TB479
87
34.5
14.9
TB480
84
40.5
13.1
BL8
77
41.6
7.8
BL56
72
20.8
5.6
BL58
103
52.4
21.4
ND49
130
20.8
5.4
ND48
100
19.0
10.0
DC10
53
0.0
39.6
DC54
63
0.0
17.5
1 As percentage of overall MNI, included ‘varied’ or ‘various biotopes’ categories.
2 May include species counted in the ‘synanthropic’ group, e.g. Anobium punctatum.
3 Reilly 2006.
4 Reilly 2003.
5 Reilly, forthcoming.
6 Reilly 2008.
408
Feature type
Skellig Michael: LMG
Skellig Michael: LMG
Skellig Michael: LMG
Skellig Michael: LMG
Roestown, Co. Meath: ditch fill,
rural medieval settlement, 8th–
10th century AD3
Essex St., Dublin: house floor
layer, urban settlement, 10th
century AD4
Essex St., Dublin: pit fill, urban
settlement, 13th century AD4
Essex St., Dublin: pit fill, urban
settlement, 13th century AD4
Back Lane, Dublin: backfill
‘house’ floor, urban settlement,
13th/14th century AD4
Back Lane, Dublin: house floor
layer, urban settlement, early
13th century AD4
Back Lane, Dublin: house floor
layer, urban settlement, early
13th century AD4
Novgorod, Russia: house floor
layer, urban settlement, early
13th century AD5
Novgorod, Russia: outdoor
deposit beside property
boundary, urban settlement, late
13th/early 14th century AD5
Derrycunihy Wood, Co. Kerry:
forest hollow peat deposit, 16th
century AD6
Derrycunihy Wood, Co. Kerry:
forest hollow peat deposit,
approx. 4th century BC6
PALAEOENVIRONMENTAL ANALYSES
This left a distance matrix of fifteen samples and 217 species. Four successive runs of NMS were
performed, producing similar and stable two-dimensional solutions with consistent stress levels of 9.04
(less than 10 is considered a ‘good’ solution). A Sorenson (Bray-Curtis) city block distance measure was
used for all calculations, since Euclidian distance measures are deemed unsuitable for data sets that contain
many empty cells (McClune and Grace 2002).
In ordination space the most similar samples are grouped closest together and the axes describe the
gradients of highest variance. Ill. 6.4 shows the ordination of all the samples listed in Table 17. A number
of obvious patterns emerge. The two Derrycunihy samples group away from all the others, suggesting
that there is little underlying similarity between these assemblages and all the others. The lower monks’
garden samples also form a discrete group, which is unsurprising given the essential similarity between
these four assemblages.
Axis 1 appears to describe the greatest variance within the assemblages as a whole, as it is along this
axis that samples are most spread out. Interestingly, some affinity exists between the Derrcunihy samples,
particularly DC54, and the lower monks’ garden samples. This may have to do with woodland species
similarity. On Axis 2, the majority of samples are more or less grouped together, with only the
Derrycunihy samples forming a discrete group. Roestown displays some independence, which may be
related to the higher ‘outdoor’ and ‘aquatic’ element of this assemblage compared to all but the
Derrycunihy samples. This sample came from the ditch fill of a rural enclosed settlement site, but the
ditch was clearly a repository for ‘household’ waste (Reilly 2006). Therefore the graph would appear to
confirm that the Skellig samples have more in common with all the assemblages from continuously
occupied, mainland medieval sites, particularly the ‘urban’ settlement locations of Back Lane and Essex
Street West, Dublin. As the graph shows, geography does not necessarily play a role in assemblage
differentiation, as the Novgorod samples also show affinity to the Back Lane/Essex Street samples. The
‘independent’ trend in the two Russian assemblages is undoubtedly due to the unique elements of the
fauna, primarily the wood-dependent beetles, which do not occur in any of the Irish assemblages.
Synanthropic fauna, particularly during the medieval period, was common to all settlement sites, regardless
of isolation, but clearly depended on particular human behaviours, which will be discussed further in
Section 6.4.3.
6.4.3 DISCUSSION
Comparisons with other contemporary monastic sites
The analysis of insect remains from Skellig Michael is one of the unique elements of the post-excavation
analysis. As far as the author is aware, there are no published examples of insect analysis carried out on
any other early monastic site in Ireland or Britain. Such work may exist in unpublished ‘grey’ literature
or otherwise inaccessible archives. Nevertheless, evidence from other biological analyses may give insights
into the origins of the Skellig Michael insect assemblages.
A small number of monastic sites around Ireland have had animal/fish/bird bone, wood/charcoal
and plant macrofossil analyses completed. Close by, and particularly relevant to Skellig Michael, is the
island monastic site of Illaunloughan (Marshall and Walsh 2005). The monastery lies on a tiny island in
the Portmagee Channel, between Portmagee and Valentia Island, and was occupied between the seventh
and ninth centuries, contemporary with the early phase of occupation at Skellig Michael. Analysis of the
bioarchaeological remains showed a mixture of locally sourced foodstuffs, mainly in the form of birds,
fish and molluscs, and foodstuffs clearly sourced from further afield (Murray and McCormick 2005).The
island was too small to sustain arable farming, so the arable food crops identified, including oats and
barley, must have come from the mainland or possibly from Valentia Island (ibid.; Murray et al. 2004).
409
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
There is a suggestion that the extensive medieval settlement on Bray Head, Valentia Island, acted as a
‘mother-house’ to Illaunloughan, and perhaps also to Skellig Michael (Hayden 1998). Excavations and
field survey have identified multiple house sites, two extant grain kilns and an extensive early medieval
field system. Plant remains from Skellig Michael were sparse and poorly preserved, with barley being the
main cultivated crop, similar to Illaunloughan (see plant remains, above). No chaff was recovered but
goosefoot, a common arable weed plant, was found in F638, one of the early medieval garden soils.
Overland, in her analysis of pollen from the possible original ground surface on Skellig at the time of
construction of the large oratory, tentatively suggests arable cultivation on the island (see pollen analysis,
below), but there is nothing in the insect evidence to corroborate this.There are very few phytophagous
beetles and none specifically indicative of arable cultivation. No grain pests were recovered either, but
this is less surprising as it is unlikely that these non-native beetles arrived in Ireland before the late Viking
or early Anglo-Norman period and may not have been present in the local area prior to the late thirteenth
century (Reilly 2003). It is possible that goosefoot seeds, and indeed pollen of arable plants, were caught
up in hay and other bedding material brought out to the island and subsequently used for fertilising the
garden soil.
Most of the wood taxa identified in the charcoal remains at Illaunloughan were also identified on
Skellig Michael (O’Carroll 2005; O’Donnell, this report). The overlap in terms of species is notable,
especially the dominance of oak and hazel but also the presence of yew. Extensive yew- and oakdominated woodlands exist at Derrycunihy and Reenadinna near Killarney, and would have been even
more extensive in early medieval times (Mitchell 1988). Indeed, given the underlying geology of the
immediate area, acidic Old Red Sandstone, it is unlikely that the yew wood was sourced from nearby
and may very well have come from Reenadinna. It is likely, however, that most of the other wood gathered
for use as fuel was sourced much closer to home, perhaps in sheltered copses or small woodlands on the
Iveragh peninsula. The wood-boring insects would suggest a ‘natural’ rather than a structural origin for
much of the fuel wood, which generally agrees with the range of wood taxa found at Skellig and
Illaunloughan.
Animal bone from Illaunloughan consisted mainly of cattle and sheep/goat, with a high incidence
of very young animals.This could indicate culling strategies owing to grazing pressure or, possibly, support
from the local lay population in the form of dues or tithes of the first-born calf, lamb etc. (McCormick
1998; Murray and McCormick 2005). Insects from the lower monks’ garden layers in particular suggest
that animal butchery waste—possibly discarded bone or skins—were incorporated into the soil matrix.
This is not surprising, since the mammal remains recovered indicate that the meat of goat, sheep, pig and
cattle was consumed by the monks, in addition to birds (predominantly Manx shearwater and puffin)
and locally caught fish such as sea bream.
The ‘midden’ layer on Illaunloughan, from which much of the bioarchaeological material was
recovered, showed a deliberate attempt to build up a garden soil, possibly for growing vegetables and
medicinal plants, but was much less extensive than the monks’ gardens on Skellig (Murray et al. 2004). It
is unfortunate that insects were not analysed from this deposit, but in other respects the midden mirrors
the ‘made-up’ nature of the monks’ garden soil layers. Indeed, excess waste material from this settlement
may have been brought out to Skellig for use in the garden there and may be one source of the
synanthropic fauna.The settlement at Bray Head,Valentia, may have been another source, given the more
extensive, and possibly intensive, nature of the settlement there.
At High Island, Co. Galway, an early medieval monastic site occupied around the same time as Skellig
Michael, charcoal, animal bone and plant remains were analysed (Scally, forthcoming). As O’Donnell
notes in this report, oak and hazel again appear to be the main fuel woods (Gale, forthcoming). Large
quantities of grain were also recovered, along with somewhat more concrete evidence for crop-processing
and perhaps arable cultivation on site (McClatchie 2008). Animal bone included cattle, sheep and pig,
410
PALAEOENVIRONMENTAL ANALYSES
while fish and bird bone were also present (Scally 1999). While High Island is a larger, flatter island than
Skellig and could support some animal husbandry and crop production, it is likely that certain food
produce, bedding and other organic material was physically brought to the island from supporting
monastic or lay communities on the mainland. In the absence of insect remains, however, it is difficult to
get a picture of the nature of the occupation layers and the signatures that such activities would have left
behind.
A thick, humus-rich ‘midden’ beside the monastic settlement on Iona, off the western Scottish coast,
was extensively sampled for animal, fish and bird bone (Murray et al. 2004). Unfortunately, no
archaeobotanical analysis was carried out but a pollen core was taken from the vallum ditch (Bohncke
1981). The combined evidence suggested a mixed food base of animal products and cereal cultivation,
with evidence for the growing of crops on the island (Murray et al. 2004). Interestingly, despite the
somewhat larger size of this island and the possibility of supporting animal husbandry, the animal bone
evidence suggested importation of much of the meat consumed on the island. Again, as with Skellig
Michael and Illaunloughan, it indicates the degree to which these settlements were supported by ‘outside’
communities and gives some insights into the sources of the imported elements of the insect fauna on
Skellig Michael.
Comparisons with other contemporary ‘isolated’ medieval sites
While no comparative insect studies of monastic sites exist for Ireland and Britain, there is a small corpus
of work from early medieval ‘isolated’ settlement sites. The term ‘isolated’ is taken here to mean a smallscale settlement delimited by ditch or water and/or separated by some distance from neighbouring
settlements. This distinction is important, as it helps to illustrate the remarkable nature of the insect
assemblages from Skellig Michael, in particular the synanthropic and wood-associated elements.
Theoretically, any human habitation site that is occupied continuously over many decades or
centuries will accrete a rich insect fauna. In general, the synanthropic species are associated with decaying
plant, wood and animal matter. The gathering of plants for food, floor coverings, bedding and animal
fodder, the processing of animal products, the cleaning of skins and woodworking all have signature insect
associations, which, over time, become closely associated with human habitation sites. The mechanisms
by which species end up on habitation sites are manifold but undoubtedly human transport has a large
role to play (Kenward and Allison 1994b). Direct invasion from surrounding habitats to ‘artificial’ ones
that mimic their natural requirements is also important but obviously requires direct access.
‘Mainland’ sites, such as Deer Park Farms, Co. Antrim, or Roestown, Co. Meath, probably attained
their synanthropic fauna via both pathways (Kenward and Allison 1994a; Reilly 2006). At Deer Park
Farms, the remarkably rich synanthropic insect fauna led Kenward to speculate that wholesale importation
of some elements of the fauna from other existing centres of occupation must have occurred at different
times, and/or that the site was continuously occupied for many centuries (Kenward 1997). The richness
of the fauna was really only comparable, despite obvious limitations, to long-lived Anglo-Scandinavian
and medieval deposits at Coppergate,York (ibid.). While the sampling contexts at Skellig Michael differ
from Deer Park Farms (garden soils vs house floor layers) and the richness of the faunas are not directly
comparable, there are clear parallels. Wholesale elements of the synanthropic fauna were probably
imported onto the Deer Park Farm site in its earliest phase, as it is perhaps too ‘complete’ or rich to have
accreted naturally. The nature of the lower monks’ garden soils would also suggest initial importation of
the synanthropic element with a gradually developing in situ breeding population over time, probably
from the addition of bedding and other material from the cells on the island (see Section 6.4.2).
Roestown had a much more modest synanthropic fauna, as the insects were recovered in a secondary
deposit, e.g. a ditch fill. Admixture of different habitat elements, most notably naturally occurring species
living within the wet ditch and on plants growing there, were dominant (Reilly 2006). Nevertheless, it
411
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
suggests that living conditions within the habitation zone of the site were similar to what one might
expect to see in Hiberno-Norse or medieval Dublin (Reilly 2003). The key difference between the
Roestown and Skellig Michael assemblages is the rich dung and ‘stable manure’ fauna evident in the
former (Reilly 2006). This illustrates the more extreme isolation of the Skellig Michael community
compared to Roestown, which was clearly surrounded by pastureland and may even have had animals
present within the enclosure.
Crannogs might be considered a better analogue for Skellig Michael, although these island settlements
were much more accessible from the mainland and potential habitat sources for the insects recorded in
excavated deposits. Unfortunately few insect studies from crannogs exist, despite the obvious potential
they present. A midden feature excavated on Coolure Demesne, Co. Westmeath, was sampled for insects
and plant macrofossils (Johnston and Reilly 2007). The plant remains were well preserved and plentiful,
indicating a background wetland element and possible anthropogenic activity.The presence of carbonised
cereal grains and fruit stones suggested the dumping of organic domestic waste. The insect remains,
however, were poorly preserved, fragmentary and mainly indicative of foul conditions, possibly human
faeces and animal dung. Despite its location within the occupation zone of the crannog, no true
synanthropic fauna were present. This suggests that the midden was more akin to a cesspit deposit than
a dump of household waste. Buiston crannog, in Ayrshire, Scotland, produced a similar picture, with a
minor synanthropic element within the occupation zone (Kenward et al. 2000). It was suggested there
that the poor nature of this aspect of the fauna was due to a number of factors, including intermittent or
discontinuous occupation, the use of ‘fresh’ wood and the waste disposal habits of the occupants, which
meant that household waste was not dumped within the occupation zone (ibid.). The first of these is
considered the most important, however, and the key difference perhaps between crannog sites like
Buiston or Coolure and continuously occupied sites like Deer Park Farms, Roestown or even Skellig
Michael.
Biogeographical implications of the insect assemblages
The insect assemblages from Skellig Michael raise some interesting biogeographical issues in relation to
small islands and their native flora and fauna. A large element of the recorded assemblages from all contexts
on Skellig Michael is probably imported. This is surmised on the basis that the habitats required to
maintain many of the recorded species, most notably the wood-associated, synanthropic and some
generalist decomposer species, do not exist in the present day on the island. The latter two groups, in
particular, require the continuous presence of large amounts of decaying plant matter, e.g. leaf litter,
mouldy hay etc., and other decomposing material.
Sadler (1999) notes that islands are ‘bounded ecosystems’, with their flora and fauna largely
representing a microcosm of the ‘mainland’ and mostly derived through chance dispersal. Species generally
arrive over a long period of time via wind, precipitation, ice/soil/wood-rafting and carriage by birds. In
general, research shows that contact with humans often has a detrimental effect on the local biodiversity
of islands, causing extinctions of plant and animal species, most notably birds (Vitousek 1988).
Palaeoecological studies, however, particularly in the north Atlantic region, have also shown little or no
impact in terms of extinctions but a dramatic increase in new arrivals of plants, animals and insects into
otherwise relatively impoverished ecosystems (Buckland et al. 1991; McGovern et al. 1983; Sadler 1990).
Changes in land management practices since the first Norse settlements on these islands have resulted
in local extirpation of some species (Sadler 1999). Over half of the present fauna of Iceland and the Faroe
Islands, however, is thought to be a direct result of human importation.These species are most commonly
associated with farm site habitats—stored hay, stored foodstuffs, waste disposal of faeces, dung and carrion.
Continued human occupation of islands like Iceland and Faroe has maintained these populations, while
the abandonment of European settlements on Greenland, for example, led to the extinction of elements
412
PALAEOENVIRONMENTAL ANALYSES
of this imported synanthropic fauna before recolonisation of the island in the mid-eighteenth century
(ibid.).
The pre-Landnam and post-Landnam insect fauna is well understood on these islands because of the
extensive palaeoecological work that has been carried out there. The situation on Skellig Michael is of
course not so clearly understood. We have no baseline understanding of the present island insect fauna;
indeed, no systematic baseline survey of the beetle fauna of any of our coastal islands exists. No longterm palaeoecological study of the insect fauna from this or any other island exists either. Both lines of
enquiry would be highly beneficial to the future management of the island and would also place in
clearer context the remarkable nature of the settlement-phase insect fauna. It would more clearly
disentangle the ‘native’ from the ‘imported’ elements and would establish whether any of the imported
elements have subsequently found a foothold on the island, in the absence of direct human occupation.
6.4.4 SUMMARY
Seven samples from the monastic settlement at Skellig Michael produced large insect assemblages.Three
were largely modern—two from under the leacht, contaminated by petrel/puffin nests, and one from the
cistern. The four samples from the lower monks’ garden soil layers, dated to the early medieval period
but in part redeposited in the twelfth–fourteenth centuries, produced remarkable assemblages, however,
giving insights into the local site environment and economy. They suggest that much of the deposited
material was made up of discarded ‘household’ waste, probably from the cells on the island but
supplemented or originally supplied from imported organic material from the mainland.The assemblages
also produced two wood-boring beetles, one of which, Pseudophloeophagus aeneopiceus, is very rare in
Ireland today. Neither species could be native to the island and together they suggest the gathering of
either mainland-sourced driftwood, kindling from local forest floors or possibly even secondary-use wood
from mainland occupation sites.
The building up and maintaining of the garden area was clearly a very important element of monastic
life and has clear parallels on other contemporary early medieval sites, such as Illaunloughan, Co. Kerry,
High Island, Co. Galway, and Iona in Scotland. The insects do not indicate that animal dung was used as
a fertiliser but seaweed may have been a contributor, as well as human excrement and animal/bird bones.
Unfortunately, no clear evidence from the insect assemblages emerges about what might have been grown
in the garden soils.
Acknowledgements
Ryan Allen, Lorna O’Donnell and Eileen Reilly wish to acknowledge the help and assistance of the
following: Margaret Gowen and Co. Ltd, Dr Darren Mann, Oxford University Museum of Natural
History, Dr Bettina Stefanini, Dr Meriel McClatchie, Ellen O’Carroll, Georgina Scally, Alan Hayden and
Dr Ingelise Stuijts.
413
Table 18—Charcoal and wood identification details from Skellig Michael.
Sample
Sample
description
Context
no.
Context description
Cutting
Level
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1995
1996
1996
1996
1996
1996
1996
1996
1996
1996
1996
1996
1996
1996
1996
1996
S54
S54
S66
S67
S68
S69
S70
S71
S72
S73
S74
S75
S76
S81
S83
S83
S83
S84
S85
S85
S15
S15
S15
S16
S16
S16
S17
S17
S18
S18
S18
S18
S18
S19
S20
Charcoal
Charcoal
Wood sample, tiny #1
Wood sample, tiny #2
Wood sample, tiny #3
Wood sample, tiny #4
Wood sample, tiny #5
Wood sample, tiny #6
Wood sample, tiny #7
Wood sample, tiny #8
Wood sample, tiny #9
Wood sample, tiny #10
Wood sample, tiny #11
Charcoal #4
Charcoal #2
Charcoal #2
Charcoal #2
Charcoal #3
Charcoal #1
Charcoal #1
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
F634(5)
F634(5)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F634(4)
F635(4)
F635(4)
F635(4)
F634(4)
F636(4)
F635(4)
F634(5)
F634(5)
F634(5)
F634(5)
F634(5)
F634(5)
F636(5)
F636(5)
F635(5)
F635(5)
F635(5)
F635(5)
F635(5)
F635(5)
F635(5)
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
5
5
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
6
6
3
3
3
1
1
1
3
3
5
5
5
5
5
3
1
Layer
Identification
Fragment
count
Weight
(g)
15
1
2.45
1.68
D
D
D
D
D
D
D
D
D
D
D
D
E
E
E
D
F
F
D
D
D
D
D
D
F
F
E
E
E
E
E
E
E
Quercus sp. (oak)
Salix sp. (willow)
Unidentified
Unidentified
Unidentified
Unidentified
Unidentified
Unidentified
Unidentified
Taxus baccata (yew)
Taxus baccata (yew)
Taxus baccata (yew)
Taxus baccata (yew)
Corylus avellana (hazel)
Betula sp. (birch)
Corylus avellana (hazel)
Quercus sp. (oak)
Fraxinus sp. (ash)
Corylus avellana (hazel)
Quercus sp. (oak)
Corylus avellana (hazel)
Fraxinus sp. (ash)
Quercus sp. (oak)
Fraxinus sp. (ash)
Quercus sp. (oak)
Taxus baccata (yew)
Fraxinus sp. (ash)
Quercus sp. (oak)
Alnus sp. (alder)
Corylus avellana (hazel)
Fraxinus sp. (ash)
Ilex aquifolium (holly)
Quercus sp. (oak)
Quercus sp. (oak)
Corylus avellana (hazel)
2
3
1
9
10
4
3
1
1
6
3
13
1
1
9
4
1
6
4
11
3
2
0.71
0.57
0.53
1.46
3.27
0.5
0.99
0.03
0.01
0.1
0.68
0.7
0.11
0.01
0.31
0.3
0.02
0.73
0.25
0.97
1.65
0.52
Radio- Growth
carbon
dating
Medium
Yes
Medium
Ring
curvature
Dimensions Ring
(mm)
count
Strongly curved
Strongly curved
8–10
10
4–5
20
Yes
Yes
Yes
Yes
Yes
Strongly curved
Strongly curved
Strongly curved
Strongly curved
Strongly curved
42 x 15 x 3
11 x 9
102 x 10 x 7
70 x 15 x 5
8
7–8
5
10
18–20
7–8
5–9
4–5
4–5
4–5
8
10
8
4–5
7–8
10
10
10
10
10
10
7
3–4
3
3–4
3–4
3
2–3
8
5–7
35
10
6
2
2–3
2–3
7–8
7–8
9
3
8
4
4–5
4–5
4–5
5–6
5
4
Slow
Slow
Slow
Slow
Medium
Medium
Yes
Yes
Slow
Slow
Medium
Medium
Yes
Yes
Yes
Medium
Medium
Slow
Medium
Slow
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Roundwood
Mixture
Weakly curved
Strongly curved
Weakly curved
Strongly curved
Strongly curved
Strongly curved
Strongly curved
Strongly curved
Strongly curved
Weakly curved
Strongly curved
Strongly curved
Strongly curved
Strongly curved
Weakly curved
Strongly curved
Strongly curved
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
414
Year
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
F635(5)
F505
F505
F505
F505
F505
F505
F507
F507
F507
F507
Lower monks’ garden 5
Behind east entrance
Behind east entrance
Behind east entrance
Behind east entrance
Behind east entrance
Behind east entrance
Low level behind east entrance
Low level behind east entrance
Low level behind east entrance
Low level behind east entrance
1
E
1999
1999
1999
Charcoal
Charcoal
Charcoal
F507
F507
F545
1999
Charcoal
F545
1999
Charcoal
F545
2000 S125
Charcoal
F572
2000 S125
Charcoal
F572
2000 S131
Charcoal
F561
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
S14
S14
S14
S15
S16
S16
S16
S17
S17
S17
S17
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
F579
F579
F579
F580
F581
F581
F581
F583
F583
F583
F583
Low level behind east entrance
Low level behind east entrance
Sample from south entrance, lower blocking.
In ash and brown organic material.
Sample from south entrance, lower blocking.
In ash and brown organic material.
Sample from south entrance, lower blocking.
In ash and brown organic material.
From leacht in west end of lower monks’ garden
from loose mid-brown sandy silt
From leacht in west end of lower monks’ garden
from loose mid-brown sandy silt
Area to east of St Michael’s Church from ash layer
in SE corner of leacht
Base of leacht, disturbed by puffins
Base of leacht, disturbed by puffins
Base of leacht, disturbed by puffins
Base of leacht
Base of leacht, under C00:09
Base of leacht, under C00:09
Base of leacht, under C00:09
SE of leacht, overlying large step
SE of leacht, overlying large step
SE of leacht, overlying large step
SE of leacht, overlying large step
2000 S17
Charcoal
F583
SE of leacht, overlying large step
Quercus sp. (oak)
Corylus avellana (hazel)
Fraxinus sp. (ash)
Prunus sp.
Quercus sp. (oak)
Salix sp. (willow)
Ulmus sp. (elm)
Alnus sp. (alder)
Corylus avellana (hazel)
Fraxinus sp. (ash)
Prunus avium/padus
(wild/bird cherry)
Quercus sp. (oak)
Salix sp. (willow)
Corylus avellana (hazel)
5
13
3
3
2
7
1
2
16
44
2
0.53
0.6
0.4
0.29
0.14
0.44
0.02
0.48
1.05
2.29
0.24
20
16
49
0.91
1.38
5.35
Fraxinus sp. (ash)
6
Quercus sp. (oak)
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Strongly curved
Strongly curved
Strongly curved
Strongly curved
Weakly curved
Strongly curved
Strongly curved
Weakly curved
Strongly curved
Strongly curved
Strongly curved
8
5–10
5–10
5–10
5–10
5–10
5–10
5–10
5–10
5–10
5–10
2–3
2–8
2–10
5
5
1–5
5
5
2–10
5–16
2–8
Medium
Medium
Medium
Weakly curved
Strongly curved
Strongly curved
5–10
5–10
5–10
2–15
2–8
2–9
1.27
Medium
Strongly curved
5–10
5–10
45
6.66
Medium
Weakly curved
5–10
5–15
Alnus sp. (alder)
2
0.68
Medium
Strongly curved
9
4–9
Quercus sp. (oak)
8
0.75
Medium
Strongly curved
6–8
3
Quercus sp. (oak)
1
3.33
Corylus avellana (hazel)
Salix sp. (willow)
Taxus baccata (yew)
Corylus avellana (hazel)
Betula sp. (birch)
Fraxinus sp. (ash)
Quercus sp. (oak)
Betula sp. (birch)
Corylus avellana (hazel)
Fraxinus sp. (ash)
Picea/Larix (spruce/
larch)
Quercus sp. (oak)
4
5
2
3
6
6
6
1
3
4
3
0.39
0.21
0.69
0.1
0.48
1.27
0.42
0.2
0.17
0.2
0.23
3–4
3–5
5
3
3
16
4
10
1.48
Yes
Yes
Yes
Yes
Weakly curved
Yes
Yes
Yes
Yes
Medium
Medium
Slow
Medium
Slow
Fast
Slow
Strongly curved
Weakly curved
Weakly curved
9
5
15
7
9
15
6
Medium
Slow
Slow
Strongly curved
Weakly curved
Strongly curved
5–8
10
5–10
5–6
10
8
7–12
2–7
Strongly curved
Strongly curved
415
Mixed
Weakly curved
(slow and fast)
PALAEOENVIRONMENTAL ANALYSES
1996 S20
1999
1999
1999
1999
1999
1999
1999
1999
1999
1999
Year
Sample
Sample
description
Context
no.
Context description
Cutting
2000
2000
2000
2000
2000
S17
S18
S18
S18
S19
Charcoal
Charcoal
Charcoal
Charcoal
Charcoal
F583
F588
F588
F588
F589
SE of leacht, overlying large step
SW of leacht
SW of leacht
SW of leacht
SW of leacht
2000 S20
2000 S41
Charcoal
Piece of timber
F590
F509
2002 S141
S81
S81
Charcoal
Charcoal #4
Charcoal #4
F684
F634(4)
F634(4)
S of leacht
East entrance above stones
collapsed onto steps
Bulk sample from 3b hearth
Lower monks’ garden 4
Lower monks’ garden 4
Level
Layer
D
D
Identification
Fragment
count
Weight
(g)
Salix sp. (willow)
Alnus sp. (alder)
Betula sp. (birch)
Quercus sp. (oak)
Rhamnus catharticus
(purging buckthorn)
Quercus sp. (oak)
Unidentified
2
4
1
2
1
0.36
0.13
0.05
0.07
0.9
1
0.19
3
3
0.28
0.31
Unidentified
Fraxinus sp. (ash)
Quercus sp. (oak)
Radio- Growth
carbon
dating
Fast
Medium
Yes
Medium
Medium
Yes
Ring
curvature
Dimensions Ring
(mm)
count
Strongly curved
6
5
5
5
2
2–3
2–3
3–4
8
10
4
3
Weakly curved
Slow
Medium
Strongly curved
Weakly curved
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
416
Table 18 (cont.)—Charcoal and wood identification details from Skellig Michael.
Table 19—Species list (nomenclature after Böhme 2005). LMG = lower monks’ garden.
Sample/ID no.
Context
Genus/species
Carabidae
Nebria brevicollis (F.)
Notiophilus biguttatus (F.)
Trechus fulvus Dej.
T. quadristriatus (Schrank)
Trechus sp.
?Bembidion (Ocys) harpaloides Serv.
Pterostichus strenuus (Panz.)
P. nigrita (Payk.)
Pterostichus spp
Calathus fuscipes (Goeze)
417
Scydmaenidae
Neuraphes sp.
Stenichus sp.
Ptilidae
Ptenidium spp
Acrotrichus spp
8
Habitat
F596,
leacht
area
Distribution
1
1
-
1
2
1
Woods, woodland edges, gardens
Woods, gardens, in decaying vegetation
Coastal sites, rocky cliffs, shorelines
Disturbed ground, litter, under stones
Varied biotopes
In leaf litter, under bark, in damp wood
In damp woodland, marshes
Grassland, marshes, wet woodland
Varied biotopes
Disturbed ground, arable and cultivated
soils
In bogs, marshes, wet woodland
Widespread
Widespread
Widespread but mainly coastal around Ireland
Widespread
Varied distribution
Widespread
Widespread
Widespread
Varied distribution
Widespread
Aquatic habitats
Varied distribution
In tidal refuse and seaweed
Widespread, around the coasts
In damp compost, litter, dung
Dung, decaying vegetation
Varied biotopes but mainly foul
Dung, decaying vegetation
Compost and dung
Widspread but localised
Widespread
Varied distribution
Widespread
Widespread
Carrion and dung, mainly
Varied distribution
In dead birds, animals, birds’ nests
Westerly distribution in Ireland and UK (esp. Scottish islands)
Decaying wood, vegetation on mould/
fungi
Varied distribuion
Wood litter mainly
Under bark, in moss, in woodlands
Varied distribution
Varied distribution
2
2
2
1
1
-
1
1
-
-
1
1
1
1
1
1
1
1
3
1
1
5
1
2
2
2
6
10
3
6
5
2
1
1
1
2
2
3
1
2
5
1
5
4
1
7
3
5
4
3
2
1
3
5
4
1
Widespread
132
2
1
128
In rotting vegetation, seaweed and wood Varied distribution
In dung and decaying vegetation mainly Varied distribution
PALAEOENVIRONMENTAL ANALYSES
Agonum fuliginosum (Panz.)
Hydraenidae
Ochthebius sp.
Hydrophilidae
Cercyon littoralis (Gyll.)
Cercyon lateralis (Marsh.)
Cercyon atricapillus (Marsh.)
Cercyon analis (Payk.)
Cercyon sp.
Megasternum obscurum (Marsh.)
Cryptopleurum minutum
Histeridae
Hister impressus F.
Catopidae
Sciodrepoides watsoni (Spence)
Clambidae
Clambus sp.
S88
S87
12
13
14
3/4
F634(4), F634(4), F638, F638, F615, F593,
LMG
LMG
LMG LMG cistern leacht
area
Sample/ID no.
Context
S88
S87
12
13
14
3/4
F634(4), F634(4), F638, F638, F615, F593,
LMG
LMG
LMG LMG cistern leacht
area
Genus/species
Staphylinidae
Micropeplus staphylinoides (Marsham) 2
Metopsia clypeata (Mull.)
Phyllodrepa sp.
1
Omalium sp.
Lesteva spp
Syntomium aeneum (Mull.)
Carpelimus spp
Anotylus tetracarinatus Block
Stenus spp
Euaesthetus sp.
Lathrobium spp
Leptacinus sp.
Xantholinus spp
Othius spp
2
2
1
2
1
4
12
3
4
2
2
2
4
2
1
3
1
2
1
3
3
1
3
2
2
8
2
4
2
1
9
7
1
2
7
4
2
Gabrius spp
4
4
Ocypus olens Mull.
1
Tasgius ater Grav.
1
1
Quedius spp
Philonthus/Quedius spp
Sepedophilus sp.
Crataraea suturalis (Mann.)
3
Aleocharinae sp. indet.
Pselaphidae
Euplectus sp.
5
Bryaxis bulbifer (Reich.)
Pselaphus heisei (Hbst.)
8
Habitat
F596,
leacht
area
1
3
6
6
2
5
6
20
7
2
2
3
2
2
10
1
2
1
1
1
1
4
8
4
5
1
Fungoid decaying vegetation
In moss, leaves, decaying vegetation
Varied biotopes
Varied biotopes but often in decaying
seaweed, fungi, carrion
Varied biotopes but generally wet
In wood debris, decaying vegetation
Varied biotopes but generally in
decaying vegetation
In rotting vegetation and carrion
Varied biotopes but generally wet
In damp places, wet moss, litter
Varied biotopes but generally wet
In compost, dung, nests of ants
Varied biotopes
Varied biotopes but generally in
decaying vegetation
Varied biotopes but often in decaying
seaweed, fungi, carrion
Decaying vegetation, generally foul
habitats
Damp litter, near the sea
(can be synanthropic)
Varied biotopes
Varied biotopes
Decaying wood/vegetation, mould, fungi
In decaying vegetation
(often synanthropic)
Varied biotopes
In moss, decaying vegetation
and under bark
In damp moss, in fens, alder carr
In damp moss, meadows, marshes
Distribution
Widespread but local
Widespread but local
Varied distribution
Varied distribution
Varied distribution
Widespread but not generally abundant
Varied distribution
Widespread
Varied distribution
Widespread but local
Varied distribution
Varied distribution
Varied distribution
Varied distribution
Varied distribution
Widespread
Mainly coastal distribution
Varied distribution
Varied distribution
Varied distribution
Varied distribution
Varied distribution
Widespread but local
Widespread
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
418
Table 19 (cont.)—Species list (nomenclature after Böhme 2005). LMG = lower monks’ garden.
Sample/ID no.
Context
Scarabaeidae
Aphodius sp.
Curculionidae
Otiorhynchus arcticus (O. Fab.)
Apion sp.
Pseudophloephagus
aeneopiceus (Bohe.)
Rhopalomesites tardyi (Curtis)
1
1
1
1
3
2
8
Habitat
F596,
leacht
area
2
1
1
44
7
1
2
3
1
2
1
3
Varied distribution
Widespread
2
In varied biotopes but generally mould-feeders (often synanthropic)
In varied biotopes but generally mould-feeders (often synanthropic)
In varied biotopes but generally mould-feeders (often synanthropic)
Varied distribution
Varied distribution
Varied distribution
In varied biotopes but generally mould-feeders (often synanthropic)
In varied biotopes but generally mould-feeders (often synanthropic)
In varied biotopes but generally mould-feeders (often synanthropic)
Varied distribution
Varied distribution
Varied distribution
1
In mould on decaying plant and animal matter (often synanthropic)
In dead wood, under bark
Very local
Varied distribution
6
In plant and animal debris, rotting hay, foodstuffs (often synanthropic)
Widespread
Wood-borer in structural and natural wood (often synanthropic)
Widespread,
especially indoors
Generally in animal dung
Varied distribution
Ground-living, low vegetation, upland
Widespread but more
generally coastal than
montane in Ireland
Varied distribution
Rare in Ireland
1
2
17
8
2
2
Varied distribution
On various flowering herbs and shrubs
Carrion, other animal remains, rotting
1
11
Widespread but local
Varied distribution
Widespread
3
1
2
Dead wood, humus layer in woodlands
On bushes, foliage, grassy places
On shrubs, trees, herbs and flowers
Varied biotopes
Distribution
1
1
1
1
1
2
3
1
2
2
2
2
1
7
4
1
2
On various herbs and ground vegetation
Damp rotting timber of deciduous tree species in coastal woodlands,
driftwood, rare in buildings
In dead wood of wide range of deciduous trees, not assoc. with buildings
419
Marked western/coastal
distrib. in Britain, slightly
more widespread in Ireland
PALAEOENVIRONMENTAL ANALYSES
Genus/species
Elateridae
Dalopius marginatus (L.)
Agriotes sp.
Athous ?haemorrhoidalis (F.)
Athous sp.
Nitidulidae
Meligethes sp.
Omosita colon (L.)
Cryptophagidae
Cryptophagus dentatus grp.
Cryptophagus sp.
Atomaria spp
Lathridiidae
Lathridius minutus grp.
Dienerella sp.
Corticaria sp.
Colydidae
Aglenus brunneus (Gyll.)
Cerylon sp.
Endomychidae
Mycetaea subterranea (Marsh.)
Anobiidae
Anobium punctatum (Deg.)
S88
S87
12
13
14
3/4
F634(4), F634(4), F638, F638, F615, F593,
LMG
LMG
LMG LMG cistern leacht
area
Sample/ID no.
Context
S88
S87
12
13
14
3/4
F634(4), F634(4), F638, F638, F615, F593,
LMG
LMG
LMG LMG cistern leacht
area
Genus/species
Rhinoncus ?castor (F.)
Rhynchaenus sp.
?Rhamphus sp.
Non-Coleoptera
Hymenoptera (ants)
Tetramorium caespitum (L.)
6
17
5
4
Stenamma westwoodi Westwood
3
16
3
3
6
1
6
Diptera (true flies)
Coproica ?vagans (Haliday)
Lonchaea sp.
Calliphora ?vomitoria (L.)
?Fannia sp. (larvae)
?Musca sp.
1
8
Habitat
F596,
leacht
area
Distribution
1
On Rumex, fallow fields, banks etc.
Leaf-miner on various deciduous tree species
Widespread
Varied distribution
3
Coastal sand or cliffs, heaths
1
Under large stones in woodland or hedgerows
Mainly southerly distribution
in Ireland
Mainly southerly distribution
in Ireland
39
Mainly in cow dung but also human excrement
Under bark in dead wood
Freshly dead animals inc. humans
Wood debris, rotting vegetables/fruit, latrines (urine)
Dung of animals, humans, other foul conditions
1
1
13
8
2
103
1
1
1
2
-
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
420
Table 19 (cont.)—Species list (nomenclature after Böhme 2005). LMG = lower monks’ garden.
PALAEOENVIRONMENTAL ANALYSES
6.5 PAST ENVIRONMENT AND LAND USE ON SKELLIG MICHAEL IN THE MONASTIC
PERIOD—THE POLLEN ANALYTICAL EVIDENCE
Anette Overland and Michael O’Connell
6.5.1 INTRODUCTION AND
METHODS
A minerogenic soil (F318, coarse and fine samples combined) from the large oratory was subsampled
and prepared for pollen analysis. The context from which it derived was an old ground surface that was
probably contemporary with the construction of the large oratory in the monastery (A. Lynch, pers.
comm.).The sample (2cm3) was prepared for pollen analysis using standard procedures. Lycopodium spores,
in tablet form, were added to facilitate estimation of pollen concentration. After KOH treatment, sieving,
HF treatment and acetolysis, a fine sieve (mesh size 5µm), together with ultrasonic treatment, was used
to remove fine debris. The sample was mounted in glycerol and counted under phase contrast using a
Leica DM LB2 microscope fitted with a Planapo 63/1.4 objective and oculars that gave an overall
magnification of x788. Pollen and spore identification mainly followed Fægri and Iversen 1989. Other
authorities consulted include Moore et al. 1991 and Beug 2004. Cereal-type pollen was distinguished
following the criteria in Beug 2004. It should be noted, however, that grains of grass pollen less than
40µm in maximum length have been placed in Poaceae, irrespective of annulus characteristics.
Microscopic charcoal (>30µm) was also counted. In calculating percentage pollen values a total terrestrial
pollen sum (TTP) was used. Crumpled, corroded and unknown pollen grains (categorised as
Unidentified) and micro-charcoal were calculated relative to the TTP + Unidentified and micro-charcoal
counts respectively.
The material retained in the 100µm-mesh sieve during pollen preparation, i.e. after KOH treatment,
was checked for macrofossils (none noted) and macro-charcoal (a few small fragments were recorded).
6.5.2 RESULTS AND
INTERPRETATION
Pollen was exceptionally abundant (4.4 x 106 grains per cm3) and well preserved, though somewhat
crumpled, which made certain identification difficult in some instances. The number of such grains,
however, is not so high as to compromise the data set (see Unidentified in Table 20; this category consists
mainly of crumbled grains). Crumpling also made it difficult to determine the size of Poaceae pollen,
which is important for distinguishing the pollen of cereals from that of non-cultivated grasses. Since only
Poaceae pollen grains that were at least 40µm long and with substantial annulus and pore were classified
as cereal-type, cereal-type pollen has probably been underestimated.
The results are presented graphically in Ill. 6.5. Arboreal pollen (AP) accounted for only 0.6% and
consisted of Quercus, Corylus and Betula. Given the extremely low representation, it is assumed that these
represent long-distance transport from the mainland, where oak, hazel and birch were presumably
common at this time. Interestingly, pine, which produces pollen that is well dispersed, was not recorded.
This indicates that the pollen is no older than c. 1000 BC (after this date Pinus pollen values are invariably
low in south-west Ireland) nor younger than c. AD 1750, after which date Pinus pollen again appears in
Irish pollen records as a result of the widespread planting of Scots pine (cf. Overland and O’Connell
2008).
The main woody taxon is Sorbus. This attains 1%, which is substantial given that Sorbus is
underrepresented in pollen records. On ecological grounds, it is assumed that this is the pollen of S.
aucuparia (rowan/mountain ash) rather than the rarer, broad-leaved whitebeam (Sorbus) species (ecological
421
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 6.5—Piechart showing
percentage
pollen
representation,
Skellig Michael.
preferences, etc., are based largely on personal observations; see also Webb et al. 1996; Uí Chonchubhair
1998). The relatively strong representation suggests local presence within, or at least near, the monastic
settlement.
The pollen data, in general, suggest a completely open landscape with grassy swards that supported
a wide variety of herbaceous species. Poaceae (grasses) and Rumex (dock) constitute 74% of the pollen.
Such high values are only possible if there is local presence even, as in this instance, where grasses and
dock are involved, which are high pollen-producers with good dispersal properties. Rather surprisingly,
Plantago has only modest representation (Ill. 6.5). Apart from a single Plantago grain that could not be
identified with confidence to species level, this pollen taxon consists exclusively of P. lanceolata (ribwort
plantain). The main forb is Jasione (sheep’s-bit scabious; the pollen values are exceptionally high), which
was probably favoured by the marine-influenced conditions. Chickweed (Cerastium-type), dandelion and
hawkweed (Liguliflorae) were also common.
While the taxa referred to above are mainly indicative of grassland, including possibly pasture, there
was probably also arable activity. The uncertainties associated with interpreting cereal-type pollen are
compounded here by the difficulties in measuring the pollen (see above). As already indicated, the values
for cereal-type probably underestimate cereal-type representation. In view of this and also the
underrepresentation of most cereals (apart from Secale, which was not recorded) in pollen records, the
overall evidence points to cereal-growing at the monastic settlement. This is not unexpected, given the
increasing importance of cereal cultivation at about this time in medieval Ireland (Overland and
O’Connell 2011). Hornungia-type (this includes many small weedy species of the Brassicaceae (cabbage)
family), Chenopodiaceae (goosefoot family) and Spergula (corn spurrey) represent ruderal and arableweed species. Chenopodiaceae include also species of salt-influenced habitats and so, in this situation,
the pollen probably does not derive exclusively from plants growing in arable/ruderal contexts. Overall,
422
PALAEOENVIRONMENTAL ANALYSES
however, these pollen taxa support the idea of arable farming during the monastic period.
The records for Sedum are of interest in that this pollen is seldom recorded in pollen diagrams. S.
anglicum (English stonecrop) and/or S. acre (yellow stonecrop) are probably represented. Both species
occur on cliffs and rock outcrops, especially near the sea; local presence at/near the monastic site, as well
as elsewhere on the island, is assumed. Armeria maritima (sea pink) is also associated with salt-influenced
habitats and was probably common but apparently not abundant in the immediate vicinity of the
monastery (a single pollen grain only was recorded).
The high representation of micro-charcoal (also macro-charcoal) suggests fires. Fires were presumably
required for cooking and heating, though it is difficult to envisage the source of fuel for such fires, given
the paucity of woody species. Firing of herbaceous vegetation, with a view to burning off old growth
and stimulating new growth in the spring, may also have been a source. Given the biologically
indestructible properties of charcoal, it may derive from a time when the island was not as devoid of
woody vegetation as suggested by the pollen preserved in this sample; i.e. the charcoal fragments may be
older than the pollen.We do not know whether the island was ever wooded. Small islands with little soil
cover such as Inis Oírr (Inisheer) had substantial woodland cover for most of the Holocene (Molloy and
O’Connell 2007); on the other hand, the island of St Kilda carried little tall woody vegetation (Walker
1984). Skellig Michael is much more remote and the local north Atlantic environment much harsher.
Table 20—Pollen analytical data: counts and percentage values.
Taxa
Betula
Quercus
Corylus
Sorbus
Counts
1
3
2
10
% values
0.1
0.3
0.2
1.0
Poaceae
Plantago lanceolata
Plantago p.p.
Rumex acetosa-type
Liguliflorae
Cerastium-type
417
18
1
342
14
25
40.6
1.8
0.1
33.3
1.4
2.4
Sedum
Jasione
Armeria (B-line)
5
141
1
0.5
13.7
0.1
3
11
30
2
0.3
1.1
2.9
0.2
Cereal-type (40–44µm)
Hornungia-type
Chenopodiaceae
Spergula arvensis
Pollen sum
Unidentified
Charcoal (>30µm)
Volume (cm3):
Lycopodium added:
Lycopodium counted:
1026
33
521
3.1
33.7
2
74000
8
423
7. THE FAUNAL REMAINS
7.1 THE MAMMAL BONES
Emily Murray
7.1.1 PROVENANCE
OF THE FAUNAL MATERIAL
A small assemblage of animal bones, comprising mammal, bird and fish bones (the latter two are discussed
separately by Sheila Hamilton-Dyer) and marine molluscs (see Murray, below), was recovered from
excavations at the monastery on Skellig Michael, Co. Kerry, between 1986 and 2000.The phasing of the
majority of the contexts from which the bones derive is poor, as many of the layers excavated were found
to be redeposited. Additional disturbance was also caused by the burrowing activities of puffins and storm
petrels, and to a lesser extent by rabbits (Bourke 2005, 134–5).
The largest assemblage was recovered from the lower monks’ garden (layers F633, F634, F635, F636
and F638), with faunal material also found in and around the leacht, the eastern and southern entrances
and Cell G. Animal bones from the lower monks’ garden returned radiocarbon dates ranging from the
seventh to the twelfth century AD but were found with twelfth- to fourteenth-century pottery. This
indicates that the excavated faunal material had been redeposited from older garden or midden contexts
elsewhere on the island. A similar problem was also encountered at the east entrance, where material
returned radiocarbon dates ranging from the seventh to the tenth century but was found with later
medieval pottery. In the absence of any nineteenth-century material in either of these assemblages,
material from the garden layers and east entrance have been broadly phased as ‘early medieval’ (seventh–
fourteenth centuries). Faunal material was also found associated with medieval burials (tenth–thirteenth
centuries) located in and around the leacht. Much of the material recovered from the leacht area and in
the vicinity of St Michael’s Church was, however, badly disturbed by burrowing. Material from these
deposits, along with bone recovered from the upper levels of the lower monks’ garden and material from
Cell G, have all been designated as ‘post-medieval’ in date. The assemblage can therefore essentially be
divided into ‘medieval’ and ‘post-medieval’ material and it is discussed in these terms (Table 27).
7.1.2 METHODOLOGY
The mammal bones were quantified by the number of ‘countable’ identifiable specimens (NISP) present
(see McCormick and Murray 2007, 9–11).The differentiation of bones of sheep and goat was attempted
on distal metapodials, distal tibiae, distal humeri, astragali, calcanei and deciduous fourth premolars (dP4),
using the criteria of Boessneck (1969), Kratochvil (1969) and Payne (1969; 1985). Sheep and goat horn
cores were also differentiated to species, and sex, where possible. Evidence for age at death was determined
by the epiphyseal fusion data and stage of dentition. Tooth eruption and tooth wear stages for cattle and
pigs were recorded after Grant (1982) and Payne (1973; 1987) for sheep and goat. Measurements were
made on fused (adult) bones following, for the most part, the criteria of von den Driesch (1976; see also
McCormick and Murray 2007, table A1:4.1).
425
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
7.1.3 CONDITION
OF THE BONES
The assemblage was generally well preserved though the majority of bones were in a fragmentary state,
with just six ‘countable’ specimens being complete (all small lower limb bones of sheep/goat and pig).
No evidence for gnawing or chewing by rodents and/or carnivores was noted. There were just ten
‘countable’ burnt bones; all were heavily calcined, of sheep and goat and derived from the same context
in the lower monks’ garden (S050 from F635(6)). These calcined bones were represented by a range of
cranial and post-cranial elements and it seems probable that the material may originally have derived
from a hearth or cooking area. One other burnt (‘blackened’) bone, a piece of cattle scapula (‘noncountable’; F579), was also recovered from an area of burning east of the leacht.
7.1.4 RANGE
OF SPECIES
The range of mammal species recorded comprises cattle, sheep, goat, pig, horse, seal and whale, while
two neonate bones recovered from the leacht are most probably of rabbit. The identified specimens are
represented by a total of 160 ‘countable’ elements and an MNI of 26, with the majority (71% NISP)
deriving from early medieval deposits (Tables 21 and 22).
The rabbit is a recent introduction to Skellig Michael (OPW 2008, 15); grey seal, rabbit and house
mouse comprise the current list of mammals recorded for the island (ibid., 110). Rabbit burrows were
noted during the course of the excavation and conservation works at the monastery (Bourke 2005, 125,
130). The relative absence of rabbit bones (NISP 2) in the excavated faunal assemblage would suggest,
however, that rabbit burrowing has caused limited disturbance of the excavated layers, as many more
rabbit bones would otherwise be expected.
No bones of any terrestrial carnivores were found (e.g. cat, dog or fox) and their absence is also
supported by negative indirect evidence, as there were no signs of any gnawed or chewed bones. The
avifaunal evidence (see Hamilton-Dyer, below) indicates that sea birds were exploited in medieval times,
suggesting that birds must also have burrowed and nested on the island in earlier centuries. The
introduction of cats and/or dogs to the island could only have been as pets, as no herding or rodentcatching was required, and would potentially have put this valuable resource under threat. The absence
of these animals is therefore not surprising.
Horse was represented by a single bone fragment—the distal end of an adult metapodial from the
lower monks’ garden (S031 from F633(5E)). Given the steep slopes of Skellig Michael, it is not likely
that horses ever lived on the island. As at the nearby medieval island settlements on Illaunloughan and
Church Island, where horse bones were also found (Murray and McCormick 2005, 73), the obvious
explanation is that the bones made their way to the island in joints of meat. Despite the prohibition of
the eating of horsemeat by the early medieval church (Kelly 1997, 353), it was clearly something that
was nevertheless undertaken on occasion.
7.1.5 STOCK ANIMALS: CATTLE, SHEEP, GOAT AND
PIG
Stock or farm animals were represented by bones and teeth of cattle, sheep, goat and pig, and the available
metrical and ageing data for the assemblage were limited (Tables 25 and 26). Sheep and goat dominated,
accounting for 71% of the total site NISP and half of the MNIs (Tables 21 and 22). A relatively small
proportion of the caprine assemblage (25% medieval; 10% post-medieval) could be identified to species
and of these the majority were of goat (Table 23).
426
THE FAUNAL REMAINS
Goats would have been an important source of fresh produce (milk, and meat on occasion) and the
most practical animals to keep on the vertiginous island. In the early twentieth century it was noted that
the lighthouse-keepers kept goats on the island and that they were ‘generally located in some almost
inaccessible position at milking time. With bated breath I have watched the men retrieve these perverse
animals . . . there is no place in the scheme of nature for a giddy goat . . .’ (Thomas Mason, quoted in
Lavelle 1993, 61–2).
Both male and female goat horn cores were represented in the early medieval assemblage, with MNIs
of two and three respectively, along with one of each in the post-medieval assemblage. The presence of
pregnant ewes/nanny goats in both phases is also indicated by the occurrence of a minimum of two
neo- or perinatal caprines (NISP 7; 1-Ext. and S032 from F584). The tooth eruption and wear data for
caprines (Table 26) suggest that animals were killed in two age brackets: in their second year (probably
for meat) and then more mature or older animals.
Three of the goat horn cores, two male and one female, had multiple chop-marks at their bases, at
the core/skull juncture, clearly indicating that they had been deliberately chopped from the skull. They
were probably removed to retrieve the keratinous horn that sheaths the core. Two cervical vertebrae (an
atlas and axis) from medieval deposits in the lower monks’ garden (S032 from F633(5E)) had mediolateral chop-marks on their ventral side, which would have been caused by the decapitation of the carcass.
One metatarsal (S123 from F572) had been split longitudinally, presumably to access the marrow, and
another had cut-marks across the distal condoyles (1-Ext.), probably caused by the skinning of the animal’s
foot.The butchery evidence therefore indicates that the animals were fully exploited for their horn, skin,
meat, marrow and probably offal. The few estimated withers heights, all from the early medieval phase
(Table 25), fall within the range recorded for the early medieval period (McCormick and Murray 2007,
185).
Pig was represented by a small assemblage, mainly skull/mandibular and lower limb bones, along
with the distal half of a humerus (S049 from F634(6)) that had been split longitudinally. The low
frequency of pig at Atlantic coastal sites has been noted elsewhere (Murray and McCormick 2005, 68–
71), which it has been suggested is due to the lack of suitable fodder and grounds for grazing. One
proximal phalange (1-Ext.) was relatively large (GL 38.7mm; Bp 20.2mm) and may have derived from a
wild animal. Ageing data were limited to one mandible (S089 from F635(5)) with its third molar in
eruption, indicating an age of death at around 17–21 months (after Higham 1967, 105).
Cattle bones were also uncommon, with a total NISP of just sixteen. The low occurrence of these
animals is not unexpected, given the steep landscape of the Skelligs, though that did not necessarily deter
all inhabitants from attempting to keep cattle on the island; Lavelle (1993, 61) recounts that an assistant
lighthouse-keeper on Skellig Michael, Michael Wishart, died in October 1821 when he fell over the cliff
while cutting grass for his cow! Loose teeth, horn cores and upper and lower limb bones were all
represented in the excavated assemblage, suggesting that complete cattle carcasses were butchered on the
island, but whether or not they were also kept there is impossible to say. No ageing data were available
and only a couple of bones were suitable to measure (Table 25).
7.1.6 WHALE
Marine mammals were represented by bones of seals and fragments of the bone of a large cetacean. The
latter was recovered near Cell G, within wall collapse dated to the nineteenth century (Ill. 7.1). It has
been identified (J. Herman, pers. comm.) as part of the skull of either a sperm whale, the largest of the
odontoceti or toothed whales, or of a baleen whale (e.g. humpback, minke, right, fin and blue whales).
The largest fragment is a flattish piece of bone with a slight curve; it is smooth on the exterior with a
427
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
shallow ridge on the opposing side, and measures approximately 0.18m by 0.12m by 0.01m thick.
The two lighthouses on Skellig Michael were built in the 1820s and, according to Lavelle (1993,
62), the Argand oil-lamps used ‘sperm oil’, which was later replaced, in the 1840s, by rape-seed oil. The
term ‘sperm oil’ used by Lavelle may specifically refer to sperm whale oil or be a shorthand term for
‘whale oil’ (also commonly known as ‘train oil’). Whale oil could be extracted from the liver or blubber
of whales (Fairley 1981, 100), while ‘spermaceti’ is oil obtained exclusively from the head of sperm whales
(it was originally thought that the spermaceti was the seminal fluid of these whales, thus the origin of
their name). Whale oil was primarily used as a fuel for lamps, in making candles and as a lubricant in
engineering (Sharkey 1985, 34), and Customs Documents from 1697 to 1829, examined by Fairley (1981,
100), indicated that Ireland both exported and imported whale oil. It could be speculated that the whale
skull fragment recovered from nineteenth-century deposits on Skellig Michael is of a sperm whale from
which spermaceti was extracted and potentially used on the island, though it seems unlikely. The Blakes
of County Galway (Whelan 1995, 91) refer to the stranding of a sperm whale at Roundstone, and they
wrote that, as the whale had teeth as opposed to whalebone (i.e. baleen), the only thing of use was the
oil, although ‘The value of this, which was £1200, would have been much greater, but for the ignorance
of the people who suffered a considerable quantity to escape before they were aware of its value’. Other
accounts by Thompson (1856, 40–60) of strandings, both natural and deliberate, refer primarily to the
exploitation of the animals for their blubber in order to extract oil.
It is also possible that the bone is residual and derives from earlier medieval layers. Professional
whaling enterprises only emerged in the early twentieth century; before that whales were generally
exploited on an infrequent, ad hoc basis.Whale bones and whale teeth have been found on a number of
early medieval sites, including the nearby island monastery of Illaunloughan off Valentia Island
(McCormick and Murray 2007, 78).
7.1.7 SEALS
Seal was represented by eighteen ‘countable’ specimens (NISP), representing a minimum of four immature
animals and one adult. The element distribution (Table 24) indicates that parts of the head, torso and
limbs are represented.
The majority of the seal bones from Skellig Michael were of immature specimens and several
displayed evidence for butchery.This included two mandibles, one mature (S121 from F634(5)) and one
immature (S115 from F635(5)), both of which had very fine knife cuts on the lingual (internal) side of
the mandible ramus. The latter had also been chopped across the hinge and on the buccal (external,
‘cheek’) side of the ramus (Ill. 7.2). An ulna (S129 from F576) also displayed evidence of butchery: it had
been chopped, medio-laterally proximal to the olecranon process (Ill. 7.3).This probably occurred during
an attempt to remove the forelimb from the carcass. An immature pelvis also displayed cut-marks across
the ischium (S032 from F633(5E)).
This butchery evidence provides confirmation of the deliberate exploitation of these animals. The
mature specimens, specifically an ulna, humerus and mandible, were all of grey seal (Halichoerus grypus)
and, although it was not possible to identify the immature specimens to species with certainty, it seems
probable that these too are of grey seal. Grey seals are present in large numbers along the exposed western
coast of Ireland, especially uninhabited rocky islands, sand bars, caves and beaches (see Lockley 1966,
142). Grey seals haul out on rocky ledges around Skellig Michael, in particular in the summer months,
on Washerwoman Rock and Little Skellig, which has some low-lying flat ledges typical of grey seal
summer habitat (O. Ó Cadhla, pers. comm.).The recorded numbers of grey seals for the Skelligs are not
significant, however (OPW 2008, 15; Cronin et al. 2004, 38), with a count of 50 in 1989 being the
428
THE FAUNAL REMAINS
greatest number noted by Lavelle (1993, 87).
The frequency of immature specimens in the excavated assemblage would suggest that a rookery
was targeted at pupping time. A national breeding population assessment in 2005 did not find evidence
of breeding or pupping on the Skelligs (Ó Cadhla et al. 2008), but one of the most important grey seal
colonies in Ireland is c. 30km north, on and around the Great Blasket Island (ibid., 35), where the peak
in pupping occurs from late September to early October (ibid., 29, and tables IV and V). It is possible,
therefore, that the Blaskets, including the islands of Beginish and Inishvickillane, were targeted on occasion
in the autumn by the inhabitants of the Skelligs.
Grey seals were historically hunted, and accounts of seal-hunting are recorded by natives of the
Blasket Islands—Thomas O’Crohan and Maurice O’Sullivan—at the turn of the century. To hunt for
seals, according to O’Crohan (1978, 97), you needed calm weather and a good spring tide. The hunters
on the Blasket Islands would head to the sea caves and bring with them ‘Ropes to drag them [the seals]
out of the cave when they should be dead, and a big stout club with a thick end to it—we should want
that right enough to lay them low’ (ibid.). On reaching the cave by boat, the swimmer would go in
‘carrying the end of the rope in his mouth, the slaughtering stick under his oxter, a candle and matches
in his cap and the cap on his head’.The seals would then be dragged out of the caves with the ropes and
thrown into the boats in between swells (ibid., 98–9). O’Sullivan (1983, 233) also mentions the use of a
stick to hunt a young seal in a cove on the Blaskets, while Smith (1756, 85) gave the following account
of the methods used to capture seals in County Cork in the eighteenth century:
‘They [seals] are sometimes taken in the caverns among the rocks and particularly the young seals.
This is done in the moon-light nights; but the old ones fight and bite most furiously in defence of
their young; and it has been affirmed that they never let go their hold until they hear whatever they
fasten on crunch between their teeth; and for this reason seal catchers have bags, with quilted charcoal
in them fixed on their arms by way of defence.’
The majority of the seal bones from Skellig Michael were recovered from redeposited horizons in
the lower monks’ garden, suggesting that they all derive from the early medieval monastic occupation of
the island. Little is known about pre-1960s grey seal habitat and distribution (O. Ó Cadhla, pers. comm.),
but seal bones have been recorded at other contemporary early medieval monastic sites, including
Illaunloughan and Church Island, Co. Kerry, and Inishkea North, Co. Mayo (McCormick and Murray
2007, 78). Contemporary medieval references to the hunting of seal are given in Adamnán’s Life of
Columba and the Life of St Brigid, and it has been suggested that seal meat may have been a food permitted
for consumption by monks during periods of fasting on the grounds that seals were considered more
fish than mammal (Murray and McCormick 2005, 78).
7.1.8 SUMMARY AND
CONCLUSION
Excavations at Skellig Michael have recovered a small but informative assemblage of mammal bones.
Unfortunately, the medieval terracing and gardening, together with periodic collapses and later
interference by birds and other building work, has caused a mixing of deposits that has compromised the
stratigraphic integrity of the faunal material. Nevertheless, the range of species present gives some insight
into the diet, farming and hunting practices of the early medieval monks.
The range of species identified comprises cattle, sheep, goat, pig, horse, rabbit, seal and whale, with
the greater proportion (73% NISP) dating from early medieval horizons and the remainder assigned to
post-medieval/nineteenth-century activity. Sheep and goat were the principal animals exploited (71%
429
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
NISP) and were probably kept on the island, while it is suggested that horse, pig and cattle bones were
most likely imported as joints of meat or with complete carcasses. The seal bone assemblage included
both immature animals (MNI 4) and mature animals and several butchered elements.This presents strong
evidence for the deliberate hunting of these sea mammals, possibly at pupping time in autumn in and
around the Blasket Islands, where there is one of the largest grey seal colonies in Ireland. Whale was
represented by a single skull fragment of a large whale, possibly sperm whale or a large baleen whale,
while intrusive rabbit bones were limited to a small number of immature specimens recovered from the
leacht. There is no evidence for any pets, rodents or other commensal animals on the island.
Acknowledgements
I would like to thank Jerry Herman, Senior Curator of the Mammals Department of Natural Sciences
in the National Museums Scotland, Edinburgh, who kindly identified the cetacean bones, and Oliver Ó
Cadhla of the National Parks and Wildlife Service (NPWS) of Ireland for information on seals in the
Skelligs area.
Table 21—Distribution of species (NISP) by phase. The sheep/goat total includes those identified as sheep and goat. * =
represented by a ‘non-countable’ specimen.
Species/phase
Early medieval
NISP
%
Post-medieval
NISP
%
Total %
NISP NISP
Cattle
Sheep/goat
Sheep
Goat
Pig
Horse
Seal
Whale
Rabbit
Total NISP
% NISP
11
84
4
17
9
1
12
–
–
117
73
5
29
–
3
1
–
6
*
2
43
27
16
113
4
20
10
1
18
0
2
160
10
74
–
–
8
1
11
–
–
11
62
–
–
2
–
13
–
4
10
71
–
–
6
1
11
–
1
Table 22—Distribution of species (MNI) by phase.
Species/phase
Early medieval
MNI
%
Post-medieval
MNI
%
Total %
MNI MNI
Cattle
Sheep/goat
Pig
Horse
Seal
Whale
Rabbit
Total MNI
% MNI
1
11
1
1
3
–
–
17
65
2
3
1
–
2
1
1
10
38
3
14
2
1
5
1
1
27
430
7
73
7
7
20
–
–
18
27
9
–
18
9
9
12
54
8
4
19
4
4
THE FAUNAL REMAINS
Table 23—Distribution of sheep/goat and goat elements by phase. MNI values for each element (taking sides into
consideration) are given in parentheses.
Element
Horn core
Skull
LXT
LMT
Mandible
VT1
VT2
Scapula
Humerus
Radius
Ulna
Metacarpal
Pelvis
Femur
Tibia
Patella
Astragalus
Calcaneum
Metatarsal
Metapodial
Tarsal/carpal
Phalanges
Early medieval
Sheep & Goat
sheep/goat
–
8 (4)
7 (4)
–
12
–
2
–
12 (7)
2 (2)
2
–
2
–
3 (2)
–
4 (3)
3 (3)
2 (2)
1
1
–
1
1
2 (2)
–
7 (3)
–
4 (2)
–
1
–
–
–
–
–
4 (2)
2 (1)
–
–
–
–
1
–
NISP: 84, MNI: 11
(incl. 4 goat)
Post-medieval
Sheep &
Goat
sheep/goat
1
2 (1)
1
–
1
–
2
–
3 (2)
–
–
–
–
–
1
–
2 (2)
–
4 (2)
–
1
–
–
–
–
–
1
–
2 (1)
–
–
–
1
–
–
–
1
1
1
–
2
–
2
–
NISP: 29, MNI: 3
(incl. 1 goat)
431
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Table 24—Element distribution for seal bones from the Skellig Michael excavations. * = represented by a ‘non-countable’
specimen.
Seal
Lower monks’ garden
Early medieval
Post-medieval
Leacht area
Post-medieval
Skull
Loose tooth
Mandible
Ribs
Vertebrae
Scapula
Humerus
Radius
Ulna
Pelvis
Femur
Tibia
Patella
Astragalus
Calcaneum
Metapodial
Tarsal/carpal
Phalanges
NISP
MNI
*
1
6
*
–
–
1
–
–
2
–
1
–
–
–
1
–
–
12
3
–
–
–
*
–
–
1
–
1
2
–
–
–
–
–
–
–
–
4
1
432
–
2
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
2
1
THE FAUNAL REMAINS
Table 25—Metrical data from Skellig Michael for cattle (B), goat (CAH), sheep/goat (O) and sheep (OVA). Measurements are in
millimetres, with the exception of EWH (estimated withers height), and follow von den Driesch 1976. RA = radius; SC = scapula; HC =
horn core; HU = humerus; MC1 = metacarpal; MT1 = metatarsal; TI = tibia; R = right; L = left; U = side unknown. The EWHs were
calculated using the values of Teichert, quoted in von den Driesch and Boessneck 1974, 339.
Phase
S No. Sp.
Element Side
GL
Bp SD
Bd BT
HTC GLP
W W L
SLC min. max.
–
–
–
–
–
–
–
–
112.1
110.5
–
139.6
–
–
130.7
–
–
72.2 –
– –
– –
– –
– –
– –
– –
– –
– 16.6
20.5 12.7
19.1 14.2
19.1 14.2
18.7 –
29.0 –
– 17.5
27.4 –
39.2 –
–
–
–
–
–
–
–
–
27.2
23.5
24.9
23
–
–
–
–
–
–
–
–
–
12.3
13
14.6
12.1
–
–
–
–
–
–
–
–
–
–
36.7
27.3
54.3
–
–
–
–
–
–
–
–
–
–
–
–
–
EWH
(47)
(cm)
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
Post-med.
E. med.
E. med.
E. med.
E. med.
Post-med.
E. med.
E. med.
Post-med.
E. med.
S050
S049
S089
S045
S050
S050
S034
S026
1 EXT
S046
1 EXT
1 EXT
S026
S050
S082
S129
S050
B
B
CAH
CAH
CAH
CAH
CAH
O
CAH
OVA
CAH
OVA
CAH
CAH
OVA
O
O
RA
SC
HC
HC –
HU
HU
HU
HU
MC1
MC1
MT1
MT1
MT1
RA
RA
RA
TI
R
R
–
–
L
L
L
R
L
R
L
L
L
L
L
L
L
–
–
–
–
28.6
28.6
31.6
28.3
–
–
–
–
–
–
–
–
–
–
54.6
–
34.1
–
–
–
–
–
–
–
–
–
–
–
–
–
(46)
–
–
45.6
210
–
–
–
–
–
–
–
–
–
–
–
–
–
(45)
–
–
155
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
54.8
54.0
–
63.4
–
–
52.5
–
–
433
Phase
S No. Sp.
Elem. SIDE dP2
dP3
dP4
P2
P3
P4
M1
M2
M1/2 M3
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
E. med.
Post-med.
Post-med.
Post-med.
Post-med.
Post-med.
S090 CAH
S050 CAH
S050 O
S050 O
S082 O
S043 O
S043 O
S043 O
S043 O
S040 O
1 EXT O
S082 O
S026 O
S026 O
S004 O
S044 O
S026 O
MN
MN
MN
MN
MN
MN
MN
MN
MN
MN
LMT
LMT
MN
MN
MN
MN
LMT
p
p
p
p
–
–
–
–
–
–
–
–
–
–
–
–
–
16L
16L
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
A
–
A
A
A
A
A
–
–
X
X
–
a
–
–
–
x
P
x
p
p
p
p
p
–
–
p
p
–
p
–
–
–
7A
4A
5A
12S
12S
14A
14A
15A
–
–
12S
12S
X
15A
–
5A
8A
9A
X
9A
12A
9A
12A
14A
15A
–
–
9A
9A
12A
14A
–
–
2A
8A
X
9A
9A
9A
9A
9A
12A
P
–
X
9A
9A
9A
–
–
–
–
–
–
–
–
–
–
–
10A
8A
–
–
–
–
–
L
L
L
L
L
L
L
R
R
R
U
U
L
R
R
L
R
p
p
p
p
–
–
–
–
–
–
–
–
–
–
–
–
–
C
C
H
X
2A
X
X
11A
12G
X
–
–
X
X
11G
11G
U
Higham
MWS
13
13
13
–
14
14
14
17
17–18
17–18+
–
–
14
14
17
17
–
Est. age in months
21–24
21–24
21–24
25–26
25–26
25–26
Adult
Adult/old
Old
–
–
25–26
25–26
Adult
Adult
–
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
434
Table 26—Tooth eruption and tooth wear data for sheep/goat (O) and goat (CAH). Sp. = element; MN = mandible; LMT = loose mandibular molar; dP = deciduous premolar;
P = premolar; M = molar; MWS = mandible wear stage.
THE FAUNAL REMAINS
Ill. 7.1—Part of the skull of a
sperm whale or baleen whale
(e.g. humpback, minke, fin and
blue whales) recovered from
wall collapse dated to the
nineteenth century, near Cell G
(Con Brogan, DAHG).
Ill. 7.2—Immature seal mandible with multiple cutmarks across the hinge (lower monks’
garden, F635(5)) (Con Brogan, DAHG).
Ill. 7.3—Seal ulna with chop-mark located above (to the
proximal end of) the olecranon process
(F576 south-east of leacht) (Con Brogan DAHG).
435
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Table 27—Sample numbers, location and phasing for features/contexts that produced animal bone and/or shells.
S No.
Location
Date
1 EXT
S004
S006
S008
S024
S025
S026
S031
S032
S033
S034
S037
S038
S039
S040
S043
S044
S045
S046
S048
S049
S050
S051
S055
S082
S089
S090
S091
S113
S114
S115
S116
S117
S118
S119
S120
S121
S122
S123
S128
S129
S130
S131
S132
S400
S401
Small oratory terrace (F229)
South entrance 2, inner enclosure (F12)
Leacht area
Leacht area
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
East entrance
East entrance
East entrance
East entrance
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Leacht area
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Leacht area
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Lower monks’ garden
Leacht area
Leacht area
Leacht area
Leacht area
Cell G
Leacht area
South entrance 1
Leacht area
Early medieval
Early medieval
Post-medieval
Early medieval
Post-medieval
Post-medieval
Post-medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Post-medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Early medieval
Post-medieval
Early medieval
Early medieval
Post-medieval
Post-medieval
Post-medieval
Post-medieval
Post-medieval
Post-medieval
Post-medieval
Post-medieval
436
THE FAUNAL REMAINS
7.2 BIRD AND FISH BONES
Sheila Hamilton-Dyer
7.2.1 INTRODUCTION AND
METHODOLOGY
Bird and fish bones were hand-collected during excavation along with the mammal bones. Some dry
sieving was employed but wet sieving over fine mesh was not possible owing to the lack of a fresh water
supply.The bones were separated out during the mammal bone recording and made available for analysis.
Taxonomic identifications were made using the author’s modern comparative collections. All
fragments were identified to taxon and element where reasonably possible. Measurements mainly follow
von den Driesch 1976 for birds and Morales and Rosenlund 1979 for fish and are in millimetres unless
otherwise stated. The archive includes metrical, condition and other details of individual specimens not
presented in the text.
The remains have been analysed in three phase groups: Phase 1, early medieval, mostly from levels
in the lower monks’ garden; Phase 2, post-medieval, including a collapse deposit from that area; Phase 3,
nineteenth-century and later, including material from revetment cuts and a recent puffin burrow.
Although this division is rather broad, it enables analysis of the largest groups possible and avoids some
of the bias inherent in small samples and possible contamination. All of the bone was recorded by sample
number or context group and remains separate in the archive. Comparison of the bone condition (Table
28) shows that the two later phases are less well preserved and have no burnt bones.The last group (Phase
3) is likely to be biased by small sample size.
7.2.2 BIRDS
Over a thousand bird bones were available for analysis, the majority from early medieval levels in the
lower monks’ garden area. There are eight species present in the assemblage, dominated by the remains
of Manx shearwaters, Puffinus puffinus. Auks are common, with puffin, Fraticula arctica, the most frequent.
Some of the auk bones could not be distinguished between guillemot, Uria aalge, and razorbill, Alca torda,
and have been recorded as indeterminate auk. The majority of the phalanges, vertebrae, ribs and other
indeterminate fragments are probably from these four species. There are a few bones of gannet, Sula
bassana, shag, Phalacrocorax aristotelis, and gull (probably kittiwake, Rissa tridactyla) and a single small
passerine bone. It is not possible to determine the species on this one bone but the size, in combination
with the locality, is suggestive of rock pipit. The distribution of the remains is summarised in Table 29.
The anatomical distribution of the most frequent species includes all body areas. Some of the most
fragile elements and the smallest ones are underrepresented as expected, although the number of the
small elements that are present is a testament to the careful collection of the material (Tables 30 and 31).
There is further bias in the Manx shearwater bones, with lower leg bones the most frequent elements.
The remains of wild sea birds at coastal sites, especially where they breed, could be attributable to nonanthropogenic causes. Analysis of the anatomical distribution might indicate the source of the bones.
Taphonomic analysis of gull-predated Manx shearwaters shows distinctive damage and anatomical pattern
(Serjeantson et al. 1993). The clearest evidence is beak damage on the sternum but this was not seen on
the few that were identified in this assemblage. In gull predation the wing bones and pectoral girdle
remain together longer than other elements as the carcass decomposes and becomes dispersed. In the
archaeological remains, however, it is the tarsometatarsus and tibiotarsus of the leg that are most frequent.
A study by Ericson (1987) showed that wing bones were more frequent in natural accumulations of bird
437
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
bones, while leg bones were more frequent in food remains. This can be modified by the biological
anatomy of the species; larger or sturdier bones would tend to be preferentially preserved and collected
in excavation. In the case of the Manx shearwater, however, the humerus and ulna of the wing are much
larger than the leg bones. The implication is that the bird bones from Skellig Michael are mainly food
remains, a conclusion supported by processing evidence on several bones.The humeri are often snapped
in half, and in some cases a blade mark can be observed at the break point (Ills 7.4, 7.5). Other bones
have very clear cut-marks (Ills 7.6, 7.7); those on tarsometatarsi (Ills 7.8, 7.9) indicate the removal of the
feet as waste (Table 32). The puffin and razorbill bones have similar breaks and the puffin bones are also
biased in favour of the lower leg. Some of the bones are charred or heat-affected. Two razorbill humeri,
which are chopped/broken in half, are partially charred and discoloured, showing where the lower part
of the bone was exposed but the rest was protected by the flesh (Ill. 7.10).This, together with the number
of similarly broken/chopped humeri of several species, probably indicates that the birds were roasted
over a fire after the end of the wing was broken off. The deposits appear to contain both trimmings and
the final food waste, the discrepancy between leg and wing bones perhaps indicating that some of the
food waste was discarded away from the midden deposits. In the smaller assemblages from the postmedieval and later layers there are also several broken and/or chopped bones but the anatomical
distribution is less biased. Most of the bones from the post-medieval group are from a collapse layer in
the lower monks’ garden; this context group has many similarities with the earlier material and should
perhaps be considered with caution as the bones are probably redeposited. Bones from the
puffin/shearwater burrow may include the partial remains of a natural mortality but there are bones
present of three different species, including the non-burrowing guillemot, and there are also two
butchered bones. It is probable that these burrowing birds caused some stratigraphic disturbance in the
same manner as rabbits (also present today).
Almost all of the bird remains represent a seasonal resource exploited during the summer breeding
season. The Manx shearwater in particular is a migratory bird only seen in the north Atlantic region
during summer. Puffins are resident but during winter remain at sea, as do the other sea birds identified.
Puffin and Manx shearwater both breed in burrows that they excavate themselves or in rabbit burrows.
These small, fat birds have long been used for food, and puffins are still eaten in Iceland and the Faroes.
The young and sitting adults can be dug out of the burrows; adults can also be caught in long-handled
nets, as can razorbills (see Serjeantson 2009). Almost a quarter of the Manx shearwater bones are from
immature birds (see Table 30); these are not young chicks but half-grown and near full-sized birds. The
young birds would be at their maximum fatness just before fledging, and could be taken while still
flightless around the end of August. Immature puffin bones are also present but less frequent than for the
Manx shearwater; all are from the medieval levels rather than the post-medieval and modern ones. The
other, less frequently found, bird species tend to nest on cliff ledges. Perhaps eggs were also collected in
addition to the birds themselves, though there is no direct evidence for this in the finds.
The negligible amount of gannet is of interest, considering that Little Skellig is today home to the
world’s second-largest breeding colony (Wanless et al. 2005). Although there is, as yet, no archaeological
date for the colony, it was certainly present by 1700 (Gurney 1913), but it has suffered from severe
fluctuation (Fisher and Vevers 1943; 1944) and the birds were heavily harvested in the nineteenth century
(Lavelle 1993). Also known as ‘Solan Goose’, the unfledged birds are still exploited in Ness, Isle of Lewis,
from the colony on Sula Sgeir. Gannets do not breed on Skellig Michael despite protection of all birds
there today and do not seem to have done so in the past. Of the few gannet bones, half (four) are
immature but full-size; none of these are from medieval levels. Little Skellig is a difficult place to land a
boat; if there were gannets breeding at that time, perhaps the monks were not equipped to visit the island,
or perhaps they were deliberately restricted by choice or instruction to remain on Skellig Michael.
The bird bone assemblage from Illaunloughan is similar in species and distribution but without
438
THE FAUNAL REMAINS
razorbill, only a few guillemot and with some additional species such as cormorant, waders and single
bones of species such as pigeon, goose and crow (O’Sullivan 2005). Unlike Skellig, there are bones of
domestic fowl, albeit just two. The balance of the species is slightly different; Manx shearwater is again
dominant, but the secondary species are more evenly spread between puffin, cormorant, shag and
kittiwake, perhaps reflecting the difference in environment. Gannet is again present but rare. The lack of
domestic poultry is not surprising, given the extreme location and the availability of wild birds. Domestic
fowl bones are found at non-coastal sites such as Lough Gur and Lagore but do not dominate as they do
in urban and castle assemblages; although mentioned in law-tracts (Kelly 1998), the widespread use of
domestic fowl appears to be an Anglo-Norman interest (Hamilton-Dyer 2007). At the much larger
establishment on Iona, domestic poultry are more frequent in the medieval deposits; only one bone was
identified in the Early Christian levels (McCormick 1993; Coy and Hamilton-Dyer 1993).
7.2.3 FISH
Fish remains were not recovered in such numbers as those of birds; a total of 273 bones have been
recorded.The majority of the remains again come from early medieval levels in the lower monks’ garden.
At least seven species are present, with bones of sea breams the most frequent of the identified bones
at 50 specimens (Table 33). Some of the sea bream bones can be identified as the red sea bream, Pagellus
bogaraveo, and most bones represent fish of around 40cm in length. In the medieval deposits pollack,
Pollachius pollachius, is the next most frequent species at twelve bones, followed by whiting, Merlangus
merlangus, and wrasse at seven bones each. The pollack are mainly from fish of about 50cm, with one
considerably larger. The wrasse, probably all Ballan, Labrus bergylta, are from fish of about 30–45cm in
length. There are two bones of scad, Trachurus trachurus, and one each of grey gurnard, Eutrigla gurnardus,
and cod, Gadus morhua. This last is from the small oratory terrace and is a large precaudal vertebra. This
is of interest as the only other cod bones are a group of 26 vertebrae from a post-medieval deposit in the
leacht area. The sample size is too small for detailed analysis of anatomical distribution but the sea bream,
pollack and wrasse bones do contain head elements and vertebrae suggesting the use of whole fish (Table
34). Apart from the cod, all of these species are fish typical of the area that could have been caught directly
from the island on lines, or from local boats. Cod over 1m are less common catches from the immediate
area; the post-medieval vertebrae may represent the remains of a piece of store cod.
At nearby Illaunloughan the fish assemblage was of considerable size (over 4,000 bones) and much
larger than the bird assemblage from that site (Hamilton-Dyer 2005). Part of the reason must be the
routine sieving at the site; while most bird bones are large enough for hand-recovery, only the largest
fish bones are usually visible on excavation. At Illaunloughan sea breams were also the dominant fish,
with wrasse second and hake and pollack the next most frequent species. As at Skellig, cod bones were
comparatively rare. Although the lack of sieving at Skellig will inevitably bias the assemblage in favour
of the larger bones and species, the remains do seem broadly similar.This pattern of inshore, rocky-coast
fishing is also seen at Doonloughan (Hamilton-Dyer 2000). Other west coast sites appear to be similar
but were not sieved and have too few fish bones to be certain. This exploitation pattern contrasts with
the evidence from later and urban sites, which are dominated by large Gadidae (Hamilton-Dyer 2007;
in prep.). On Iona hand-collected fish bones are mainly of large cod and ling, while whiting, gurnard
and flatfish dominate the sieved material (Coy and Hamilton-Dyer 1993). The smaller fish are inshore
species, but the cod and ling would probably have been caught offshore on a long line.The combination
of different location and status may explain the large gadids at Iona.
439
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
7.2.4 COMMENT
In considering the diet at the Atlantic island monasteries it is suggested that Illaunloughan may have
been an outpost dependent on a larger house (Murray et al. 2004). If so, the nearby Skellig Michael would
have been an even more isolated community. Most of the meat for the monks would have been supplied
from the mainland, as shown by the mammal bones; the birds and fish would, however, have been a useful
complementary resource. Being of the sea, the birds may well have been classed as fish for days when
meat was not permitted; indeed, this still appears to have been the case for inhabitants of Kerry in 1765
(Smith 1765). Wildfowling would have taken place mainly during the summer, although it is possible
that some of the birds might have been salted and pickled or smoked for later use. Fishing is likely to
have been restricted to periods of good weather and might therefore also have been a mainly summer
pursuit.
Bone condition counts
butchered
250
ivoried
200
calcined
150
charred
100
eroded
gnawed
50
fragmented
0
Table 28—Bird
and fish bone
condition:
counts and
percentages.
1
2
3
Bone condition percentages
100%
butchered
ivoried
80%
calcined
60%
charred
40%
eroded
20%
gnawed
fragmented
0%
1
440
2
3
Table 29.—Bird species: context totals.
Phase Site sub
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
L and St M
LMG
LMG
Burrow
Cell G
L and St M
LMG
LMG
25
10
6
7
8
9
12
13
14
16
17
19
1
Gannet
Shag
Gulls RazorbillGuillemot
2
8
1
3
46
45
1
Puffin
Indet. Small passerine Indet. bird
auks
1
1
1
1
4
3
3
25
82
2
1
2
1
2
2
3
9
1
5
6
20
2
1
73
5
1
231
41
2
1
107
82
1
1
1
2
1
2
1
1
1
1
Total
Percentage
% excl. unid.
377
44.9
70.9
2
0.2
0.4
24
2
4
7
27
3
Total
Percentage
% excl. unid.
34
17.3
20.7
3
1.5
1.8
20
5
23
21
22
2
2
2
11
3
Total
Percentage
% excl. unid.
17
22.1
23.0
3
3.9
4.1
2
0.2
0.4
0
0
0
28
3.3
5.3
9
1.1
1.7
103
12.3
19.4
10
1.2
1.9
2
1
27
2
80
14
2
1.0
1.2
28
14.2
17.1
2
1.0
1.2
81
41.1
49.4
14
7.1
8.5
1
1
13
3
9
3
22
11
14.3
14.9
4
5.2
5.4
35
45.5
47.3
1
0.1
1
0
0
0
0
0
839
2
31
10
1
186
33
197
3
20
2
3
1
51
3
77
1
0
0
0
4
5.2
5.4
0
0
3
19
13
5
81
137
1
438
130
1
5
3
3
307
1
0
0
0
Total
THE FAUNAL REMAINS
441
3
3
3
3
3
East ent.
L and St M
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
SOT
Sample Manx shearwater
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Table 30—Manx shearwater and auk: element counts.
Manx
Juv. Manx
Razorbill
Guillemot
442
Anatomy
Skull
Premaxilla
Mandible
Humerus
Radius
Ulna
Scapula
Coracoid
Furcula
Pelvis
Femur
Tibia/tibiotarsus
Carpometacarpus
Tarsometatarsus
Major wing phalanx
Synsacrum
Sternebra/sternum
Phase 1
6
16
17
22
17
20
12
14
19
5
11
34
15
54
7
7
8
Phase 2
Phase 3
7
1
4
3
1
1
4
3
4
4
6
2
2
1
Anatomy
Humerus
Radius
Ulna
Coracoid
Tibia/tibiotarsus
Carpometacarpus
Tarsometatarsus
Major wing phalanx
Phase 1
19
8
15
1
5
7
35
3
Phase 2
5
Phase 3
Anatomy
Skull
Quadrate
Premaxilla
Humerus
Radius
Ulna
Coracoid
Furcula
Femur
Tibia/tibiotarsus
Tibia/tibiotarsus
Carpometacarpus
Tarsometatarsus
Synsacrum
Sternebra/sternum
Phase 1
Phase 2
8
1
5
2
3
3
1
3
8
1
1
1
1
1
Anatomy
Skull
Mandible
Humerus
Furcula
Femur
Tibia/tibiotarsus
Sternebra/sternum
1
1
1
Phase 3
1
4
2
2
8
2
1
1
1
1
2
Phase 1
1
Phase 2
1
2
Phase 3
1
1
2
3
1
1
1
1
1
1
1
THE FAUNAL REMAINS
Table 30 (cont.)—Manx shearwater and auk: element counts.
Puffin
Anatomy
Premaxilla
Mandible
Humerus
Humerus
Radius
Radius
Ulna
Scapula
Coracoid
Furcula
Pelvis
Femur
Femur
Tibia/tibiotarsus
Tibia/tibiotarsus
Carpometacarpus
Carpometacarpus
Tarsometatarsus
Major wing phalanx
Synsacrum
Sternebra/sternum
Indet. Auk Anatomy
Skull
Mandible
Humerus
Ulna
Scapula
Coracoid
Furcula
Femur
Synsacrum
Phase 1
1
8
9
2
1
1
3
7
5
6
3
9
1
14
2
Phase 2
Phase 3
2
9
7
4
18
4
10
2
6
3
4
3
1
11
6
4
1
13
1
5
11
Phase 1
1
4
1
1
1
2
3
8
8
1
2
Phase 2
Phase 3
1
6
4
1
2
3
1
Ill. 7.4—Manx
shearwater
broken humeri,
proximal halves,
cranial view (S.
HamiltonDyer).
443
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Table 31—Other birds: element counts.
Gannet
Anatomy
Mandible
Humerus
Radius
Ulna
Femur
Tibia/tibiotarsus
Sternebra/sternum
Phase 1
Shag
Anatomy
Femur
Tibia/tibiotarsus
Phase 1
1
1
Phase 2
Phase 3
Gull
Anatomy
Radius
Ulna
Phase 1
Phase 2
1
1
Phase 3
Passerine
Anatomy
Ulna
Phase 1
1
Phase 2
Phase 3
Anatomy
Humerus
Ulna
Furcula
Furcula
Femur
Tibia/tibiotarsus
Tibia/tibiotarsus
Fibula
Carpal
Tarsometatarsus
Tarsometatarsus
Foot phalanx (bird claw)
Foot phalanx 1
Atlas
Axis
Precaudal vertebra 1
Cervical vertebra
Cervical vertebra
Vertebra not assigned
Rib
Limb shaft fragment
Limb shaft fragment
Phase 1
2
3
3
1
2
Phase 2
Phase 3
Indet.
444
Phase 2
1
Phase 3
1
Juv.
1
1
2
Juv.
1
1
Juv.
Juv.
Juv.
1
1
4
2
1
1
4
112
1
4
1
22
13
61
64
5
Juv.
Juv.
2
1
30
Juv.
2
Juv.
THE FAUNAL REMAINS
Table 32—Butchery: species anatomy.
Phase
1
1
1
1
1
1
1
1
Common name
Manx shearwater
Manx shearwater
Manx shearwater
Manx shearwater
Manx shearwater
Razorbill
Razorbill
Puffin
NISP
9
5
4
1
3
5
2
6
Anatomy
Humerus
Ulna
Coracoid
Femur
Tarsometatarsus
Humerus
Tibia/tibiotarsus
Humerus
2
2
2
2
2
2
2
2
Manx shearwater
Manx shearwater
Gannet
Gannet
Razorbill
Razorbill
Puffin
Auks
11
1
1
2
1
1
9
2
Humerus
Furcula
Mandible
Femur
Humerus
Ulna
Humerus
Humerus
3
3
3
3
Manx shearwater
Razorbill
Guillemot
Puffin
3
3
1
5
Humerus
Humerus
Humerus
Humerus
Ill. 7.5—Manx
shearwater
broken humeri,
proximal halves,
caudal view (S.
Hamilton-Dyer).
445
Table 33—Fish species: context totals.
Site Sub
East ent.
L and St M
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
SOT
Sample
25
10
6
8
9
12
13
14
16
18
19
1
Total
Percentage
% excl.
2
2
L and St M
LMG
24
3
Total
Percentage
% excl.
3
LMG
22
Cod
Pollack
3
Whiting
2
Gadid
3
Gurnard
1
9
1
5
Seabream
Wrasse
1
3
8
9
1
11
4
3
Scad
Indet.
15
22
10
26
2
26
20
2
1
8
2
2
23
4
2
0.9
2.5
148
64.9
228
11
41
1
11
26.2
42
1
1
0.4
1.3
12
5.3
15.0
26
26
61.9
83.9
0
0
0.0
7
3.1
8.8
4
1.8
5.0
1
1
1
2.4
3.2
1
2.4
3.2
1
0.4
1.3
0
0
0
46
20.2
57.5
7
3.1
8.8
1
1
1
1
2.4
3.2
2
4.8
6.5
3
Total
21
2
25
18
40
2
50
29
1
33
6
1
0
0
0
3
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
446
Phase
1
1
1
1
1
1
1
1
1
1
1
1
THE FAUNAL REMAINS
Table 34—Fish species: element counts.
Cod
Whiting
pollack
Gadid
Gurnard
Scad
Seabream
Wrasse
Indet.
Anatomy
Precaudal vertebra
Caudal vertebra
Anatomy
Vomer
Precaudal vertebra
Caudal vertebra
Anatomy
Post temporal
Articular
Hyomandibular
Vomer
Premaxilla
Mandible
Precaudal vertebra 1
Precaudal vertebra
Caudal vertebra
Anatomy
Articular
Precaudal vertebra
Caudal vertebra
Anatomy
Spine
Anatomy
Caudal vertebra
Anatomy
Skull
Skull fragment
Face area (fish)
Post temporal
Articular
Hyomandibular
Preoperculum
0perculum
Mandible
Scapula
Precaudal vertebra
caudal vertebra
Anatomy
Operculum
Inferior pharyngeal
Premaxilla
Face area (fish)
Caudal vertebra
Anatomy
Vertebra not assigned
Ray/spine/frag
Scale
Fragment
Phase 1
1
Phase 1
3
4
Phase 1
1
2
1
1
1
1
1
2
2
Phase 1
1
3
Phase 1
1
Phase 1
2
Phase 1
1
1
2
2
2
2
1
Phase 2
13
13
Phase 2
1
Phase 3
Phase 2
Phase 3
Phase 2
1
Phase 3
Phase 2
Phase 3
Phase 2
Phase 3
Phase 2
Phase 3
Phase 3
1
1
1
10
25
Phase 1
1
Phase 2
1
Phase 3
3
1
1
3
Phase 1
1
109
27
11
Phase 2
Phase 3
7
4
447
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Ill. 7.6—Manx
shearwater
coracoid
showing clear
cut mark near
humeral
articulation (S.
HamiltonDyer).
Ill. 7.7—Manx
shearwater
furcula showing
several cutmarks (S.
HamiltonDyer).
448
THE FAUNAL REMAINS
Ill. 7.8—Manx
shearwater
tarsometatarsus
with cut-mark
(S. HamiltonDyer).
Ill. 7.9—Manx
shearwater
tarsometatarsus
with proximal
cut through (S.
HamiltonDyer).
Ill. 7.10—
Razorbill
humeri
chopped/
broken in half
and partially
charred (S.
HamiltonDyer).
449
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
7.3 THE MARINE
Emily Murray
MOLLUSCS
A small number of marine molluscs were recovered during excavations at Skellig Michael and these
comprised limpet, oyster and scallop shells (Table 35). All of the shells were poorly preserved and
fragmentary.The scallops were quantified by the number of umbones or hinges present, and as they were
all of the left, flat valve the total of nine also represents the minimum number of individuals (i.e. no pairs
of left and right valves were represented). Limpets were quantified by the number of apices present.
The scallop shells were all of the great scallop (Pecten maximus) and all were found associated with
human skeletal remains. It seems probable that they were deliberately placed with the stones covering
these burials, especially as they comprise exclusively left flat valves and are apparently absent from other
midden and garden deposits on the island.The shells were all broken, with uneven, indented margins, so
it was not possible to determine whether any had been artificially perforated, as has been recorded on
medieval sites elsewhere, including at the nearby monastic island settlement on Illaunloughan, Co. Kerry
(Marshall and Walsh 2005, 89). Scallop shells have been widely used in ceremonial and symbolic imagery
(ibid., 89–96) and are famously associated with St James and the pilgrimage to Santiago de Compostela
in northern Spain, where the adoption of the scallop as a pilgrim badge appears to date from around the
twelfth century (ibid., 95). Giraldus Cambrensis states that by the late twelfth century the monks of Skellig
Michael had left the island and moved to Ballinskelligs (Bourke 2005, 125). The tradition of pilgrimage
to the island continued, however, in particular during the sixteenth and seventeenth centuries and through
to the present (ibid.). It is possible that the scallop shells were placed on the graves on Skellig Michael at
the time of burial, sometime in the tenth–twelfth centuries, or, if the graves were accessible and marked,
they may have been left by pilgrims visiting the island at some later date.
Oysters were represented by just one right, cupped valve and, like the scallop shells, it was badly
eroded and incomplete. Unlike the scallop shells, it is more probable that the oyster may have been
exploited for its meat, possibly by the lighthouse-keepers, given its discovery in contexts that had been
disturbed in the nineteenth century. Alternatively, given the occurrence of just one shell, its introduction
to the island may have been more esoteric—as a keepsake, as a mistaken scallop or through the activities
of birds.
The third and most common species represented (94% NISP) is the limpet (Patella sp.). Limpets are
ubiquitous on rocks all around the Irish coastline, and could easily have been harvested from the rocky
margins of the island at falling or low tide.The recovery of dumps of limpet shells, in particular from the
backfill of the east entrance, would suggest that they were eaten and processed in one spot, either as food
or as a source of bait for fishing lines.
450
THE FAUNAL REMAINS
Table 35—Range and number of marine molluscs from Skellig Michael by context and phase.
S no. Location
–
–
S001
S002
S003
S004
S005
S006
S033
S038
S039
S046
S046
S049
S051
S091
Leacht (93E0195:235)
Phase
Great scallop
Flat oyster
Limpets
(Pecten maximus) (Ostrea edulis) (Patella sp.)
–
1
–
Post-medieval
(19thC disturbance)
South entrance
–
–
1
1
From among and
1
below stones covering
1
skeletal material
4
1
Lower monks’ garden Early medieval
–
Backfill in east entrance Early medieval
–
(5th–6thC)
East entrance
Early medieval
–
Lower monks’ garden Early medieval
–
Lower monks’ garden Early medieval
–
Lower monks’ garden Early medieval
–
Lower monks’ garden Early medieval
–
Leacht area
Early medieval
–
Total (MNI)
9
–
–
–
–
–
–
–
–
–
5
–
–
–
–
–
–
1
74
–
–
–
–
–
–
1
56
1
1
1
3
2
144
451
8. INTERIM CONCLUSIONS
The following interim conclusions are presented as initial thoughts on the implications of the results of
archaeological excavations carried out to date. It is intended to obtain a number of additional radiocarbon
dates and, if deemed suitable, the human remains will also be subjected to further scientific analysis (e.g.
strontium isotope analysis), all of which will hopefully contribute further to the story of Skellig Michael.
The archaeological results will also need to be fully assessed in the context of the ongoing geological,
architectural and historical researches to provide an integrated account and better understanding of the
form and development of the monastic settlement and its place in a wider early medieval context.
8.1 THE FOUNDATION OF THE MONASTIC SETTLEMENT
The date of the establishment of the monastic settlement on Skellig Michael is still a matter of conjecture.
The earliest definite historical reference to an ecclesiastical foundation on the island is an annal entry
dated AD 824, which records a Norse raid on the site. The dearth of securely stratified deposits in the
areas excavated has also hindered the establishment of a chronology for the early phases of occupation
of the island. The radiocarbon dates obtained from excavated material indicate activity on the island no
earlier than the late seventh century. The material dated, charcoal and animal bone, was mostly from
redeposited contexts and so cannot be tied into specific building phases.
It should be remembered that in the monastery, in particular, the areas excavated were limited to the
upper levels of revetment walls and terrace deposits, with the lower (earlier) levels left largely undisturbed.
In the area of the east entrance (inner enclosure), for instance, up to 3m of deposits remain unexcavated
below the Phase 3 entrance. None of the beehive huts were excavated and, with the exception of a
quadrant of the large oratory, the ecclesiastical buildings were not investigated. A chronology of building
can be postulated based on stratigraphic relationships and masonry styles (see Section 8.2 below), but an
accurate dating of the foundation and subsequent development of the monastery will require further
investigation and dating programmes.
On present evidence, therefore, it can be stated that the monastic settlement on Skellig Michael was
established by the late seventh/early eighth century AD.There is no evidence, to date, to corroborate the
suggested association with St Fionán in the sixth century. Neither has any evidence of pre-monastic
occupation of the island come to light.
8.2 THE EVOLUTION OF THE MONASTERY
8.2.1 THE
CELLS
There are seven cells in the monastery, six in the inner enclosure and one in the lower monks’ garden.
The first six are described by O’Sullivan and Sheehan (1996, 284–5) and the sixth is described above in
the account of the excavation of the lower monks’ garden. The stratigraphic relationship between Cells
B, C and D has been established, with D being the earliest, B the second-earliest and C the latest. Cell
D is the only cell with a circular internal ground-plan, and the only cell with an orthostatic base to its
inner drystone wall. Cell C was built on the ruins of Cell D after its roof had collapsed. Cell E can be
shown to post-date Cell D, as the platform on which it is built post-dates the platform for Cell D. Cells
B and D are taller and narrower than the other cells, with well-finished interior walls and a square plan
453
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
internally, but neither have an offset base. It can be shown that Cell B pre-dates Cell C, but their
similarities in form and the fact that they are both built on the one platform indicate that they were
roughly contemporary. Cells A and E share certain constructional features: a square internal plan,
projecting external slabs and a base which is offset externally. Both are set on separate platforms, with
Cell A pre-dating Cell B and Cell E post-dating Cell D. Cell F, which is also on a separate platform,
stands on an offset base and has wall niches similar to those in Cell A, but it cannot be physically related
to the other cells, except on the basis of form.
Cell G, at the eastern end of the lower monks’ garden, is the simplest of the cells, with no internal
features and a simple circular plan. It is not tied to any of the other cells stratigraphically and was
reoccupied in the nineteenth century. No material for dating its construction was found, but its simpler
plan may place it earlier in the sequence than some of the cells in the inner enclosure.
The nearby island monastery of Illaunloughan has a hut (Hut D), with similarities to Cell G, which
has been dated to between AD 775 and 961. This hut also had projecting stones on the outside similar
to those from Cell A, but not the elaborate plan of that cell (Marshall and Walsh 2005, 37–42).
Based on stratigraphic relationships and limited comparanda, a suggested sequence of building of
cells on Skellig Michael is: G, D, A, F, B, C, E.
8.2.2 THE
ORATORIES
As the stratigraphic relationship between the large and small oratories is not known, it is difficult to
differentiate them by date. It seems most likely that the leacht and small terraces on the small oratory
terrace are roughly contemporary with the small oratory.
Rourke (in Marshall and Walsh 2005, 112–21) has postulated a typological series for the drystone
oratory, based on building-style evolution, which places the oratory on Illaunloughan in the eighth
century, the large oratory on Skellig Michael slightly later and the small oratory in the ninth century.
This scheme, which shows the oratories gradually adapting from structures with rounded corners and
roofs to structures entirely rectangular in shape, has been generally accepted (Ó Carragain 2010, 49–52).
The date (AD 672–869) from the make-up of the small oratory terrace does not contradict the scheme
as proposed. Unfortunately the excavations did not provide any dates relating to the construction of the
large oratory.
8.2.3 ST MICHAEL’S CHURCH
This was the last building constructed within the monastery. At some period, either between 950 and
1040, based on historical evidence (Horn et al. 1990, 10), or between 690 and 880, based on a 14C date
(Berger 1995, 168), a mortared church was constructed which blocked the decorated face of the south
cross and leacht and the west doorway of the large oratory from view. This church was subsequently
enlarged some time between 990 and 1280 (ibid., 169). On architectural grounds, this most likely occurred
in the twelfth century.
Part of the foundations of the early phase of the east wall was revealed during the excavation in the
leacht area. A burial (Skeleton 3), dated to between AD 990 and 1165, was interred against these
foundations, providing a terminus ante quem for the early phase of construction of St Michael’s Church.
A date of AD 779–970 obtained for the fill of the nearby cist, which was disturbed by the construction
of the church foundations, provides a terminus post quem for the construction.This evidence, taken together,
implies a late tenth-century date for the first phase of construction of St Michael’s Church.
454
INTERIM CONCLUSIONS
8.2.4 THE
CIST
Beneath the paving that underlay the leacht south of the large oratory was a structure (F599) composed
of upright stones. The western end had been disturbed, but it originally formed a long cist, divided in
two by uprights at its centre. This structure is remarkably similar to the two cists (F104 and F24) from
nearby Illaunloughan. A radiocarbon date of between AD 779 and 970 was obtained for the fill of this
feature. Similar cists in Illaunloughan were lined with quartz. This did not occur here, but seven waterrolled pebbles were found in its fill (F588). Similar features have also recently been excavated on Church
Island (Valentia), where cists and a charnel-pit have been interpreted as part of the ‘translation’ of human
remains into shrines (A. Hayden, pers. comm.).
At a later stage a large cross (see below) was inserted into this cist, the area was paved over and a
leacht was built up against the east face of the cross.
8.2.5 THE
BURIAL PLATFORMS
While it is not possible to date all these structures, it is possible to indicate the order in which they were
constructed.The earliest of the burial platforms is the one to the north of and partly underlying the large
oratory.The north wall of the large oratory is built on top of this burial platform, which supports a large
cross. Excavation of the nearby monks’ graveyard has shown that it started its life as two similar
platforms—a larger southern platform with a later, smaller platform added at its northern end. The
excavation did not remove any of the interior of the original platforms and therefore no dating evidence
can be offered for this structure.
8.2.6 THE
LEACHTA
The leacht to the south of the large oratory is built on paving stones that form the surface of the lower
plinth beneath the oratory. This lower plinth does not respect the orientation of the south wall of the
oratory and may in fact be a plinth for an earlier structure. There was no datable material in the leacht,
but two dates were derived from material later than the leacht—from layer F581 above the paving on
which the leacht was built (AD 775–941) and from layer F583 (AD 778–942). As the date from the cist,
which pre-dates the leacht, and the two contexts that post-date it are so similar, a late eighth- to tenthcentury date seems likely for the structure itself.
The leacht on the small oratory terrace appears to be contemporary with the oratory, as both lie
directly on the paved surface. There is one date from the construction layer of this terrace of AD 672–
869, which would provide a terminus post quem for this feature between the late seventh and ninth
centuries.
The construction of leachta would appear to fit most comfortably into a late ninth- to eleventhcentury context (O’Sullivan and Ó Carragáin 2008, 320–1) and, if correct, this would place the two
leachta on Skellig Michael early in the series.
8.2.7 THE
LARGE CROSSES
It has recently been noted by G.D. Rourke (pers. comm.) that the stone forming the west end of the
leacht to the south of the large oratory is in fact the base of a second cross, almost identical in shape and
455
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
decoration to the cross set in the burial platform to the north of the oratory.
The leacht and the burial platform, both supporting monumental crosses and flanking the large
oratory, would have made an impressive east end to a public open area in front of the large oratory.
8.2.8 THE
BURIALS
Three burials that post-dated the leacht were revealed in the area south of the large oratory. These were
interred in the late tenth to late twelfth century; as their dates overlap and their relative stratigraphical
position is known, it is hoped to conduct Bayesian analysis to give tighter dates for each burial.
8.2.9 THE
PAVEMENTS
The pavement in the small oratory terrace appears to be original and forms part of the original layout
of the terrace, with both the leacht and the oratory built directly on it. The pavement between the large
oratory and the platforms supporting the cells appears to be largely original, but the pavement
surrounding the monks’ graveyard is nineteenth-century.
8.2.10 THE
CISTERNS
There are two cisterns in the main area of the monastery, one in the base of the platform south of Cell
E and one in the base of the platform south of Cell B. A third cistern was excavated c. 18m west of Cell
A. All three appear to follow a similar pattern, utilising the geology of the island to collect rainwater (see
above, Section 1.2.2). A fourth structure outside the outer enclosure may also have been for water
collection, but was collecting water filtered through the lower monks’ garden. Collection of water was
of great importance for the monks, as there is no natural spring on the island. It is likely that cisterns for
the collection and storage of water would have been essential from the earliest days of the monastic
settlement.
8.2.11 THE
ENCLOSING WALLS AND ENTRANCES
Inner enclosure
The earliest phase of the inner enclosure walls is in the south-east corner.
South wall
An earlier entrance was uncovered in the southern wall in 1999 (Illus. 2.36) that pre-dates the large
oratory by a long period. This can be stated with confidence, as the plinth for this oratory lies on up to
1.5m of earth built up inside the entrance (Illus. 2.40). This wall is unusual, as it is constructed with an
inner face of stones running parallel to the line of the wall and an outer face comprising large angular
stones with their long axis at right angles to the line of the wall.
West of this segment of the south wall, the base of another early wall (Phase 2) can be seen below
St Michael’s Church. This wall is comprised of large, roughly square blocks. It collapsed at some time in
the early medieval period, taking the west end of the Phase 1 wall with it, and was replaced by Phase 3,
which survived until it collapsed and was repaired in the 1890s by the OPW. Phase 3 helped to support
456
INTERIM CONCLUSIONS
St Michael’s Church, and its collapse undermined the church, causing the collapse of its south-west
corner (Westropp 1897, 307, 311). It was the collapse of the Phase 2 wall that was later dealt with by
redepositing earth in the middle part of the enclosure to make the lower monks’ garden in the fourteenth
century. The Phase 4 wall is the nineteenth-century repair, and there is an area that was repaired in the
1970s (Phase 5).
East wall
Here the base of the wall (Phase 1) appears, owing to its depth beneath the present interior ground
surface, to be of the same date or earlier than Phase 1 of the south wall. As the south wall is built on a
step in the bedrock, however, there is no bonding between the south and east walls at this point. The
southern portion of the Phase 1 wall partially collapsed, between Cell F and the south-east corner, in
the early medieval period, and the wall was rebuilt, stepping back slightly to prevent further collapse
(Phase 2). This was not successful, as a further collapse took place, necessitating the same remedial work
(Phase 3).This repair, which included the construction of an entrance at the south-east corner, was more
successful and the entrance still stands, although blocked following a collapse within the enclosure and
the raising of the level to build the large oratory. Once again it can be shown that this pre-dated the
construction of the oratory by a long time, as the plinth of the oratory is c. 3m higher than the base of
the entrance. Three radiocarbon dates were obtained for the backfilling of this entrance. F508, a backfill
layer, produced dates of AD 691–876 and AD 675–871. A date of AD 778–948 was obtained from the
layer below (F509), which is interpreted as the collapse of a structure into the entrance. These dates
indicate collapse and later backfilling in the eighth/ninth centuries. The entrance was again repaired in
the nineteenth century, when another wall was built over the lintels (Phase 4).
North of the Phase 1 masonry, the small oratory terrace is enclosed by a wall constructed in an
entirely different style, with both the inner and outer faces formed by stones running parallel to the line
of the wall and the top of the wall being considerably narrower.This wall appears later on stylistic grounds
but the exact relationship between the two has not been established. It is argued above that the small
oratory terrace was built as one operation and survived relatively intact. The late seventh- to ninthcentury date would probably put this later than the first three phases of the south wall.
Outer enclosure walls
Excavation of part of the upper monks’ garden revealed that there was an original outer wall, built on
natural boulders, underlying the lighthouse period rebuilding. None of the original surface survived
below nineteenth-century and modern soils, and no date could be assigned to the monastic wall.
The monastic wall and wall-walk in the lower monks’ garden survived at the western end of the
garden, but had been removed from the east end of the garden when a major collapse took place in the
nineteenth century. The entrance at the western end of the garden, which had been blocked up in the
nineteenth century, was reopened in 1988 and conserved. This wall appears to be a single phase of
construction, although there are some alterations visible in the wall-walk on the inside of the wall. None
of the stratigraphy contemporary with the wall survives.
In the absence of direct stratigraphic relationships between the inner and outer enclosing walls and
datable contexts contemporary with their construction, it is not possible to define a sequence of enclosure
or to date the original construction of either wall.
8.2.12 PRIVATE AND
PUBLIC SPACES
If one reconstructs the leacht supporting a large cross south of the large oratory, combined with the burial
457
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
platform and cross north of it, the result is a very narrow access to the area east of the oratory. It is
tempting to see this as a deliberate strategy. Before St Michael’s Church was built, there would have been
a large open space west of the large oratory, which could have been for pilgrims. The area east of the
oratory may have been restricted territory. The layout of the small oratory terrace is even more private
and may have been regarded as an even more sanctified space.This idea of sanctus, sanctior and sanctissimus
has been used to describe the outer, middle and inner enclosures in early medieval ecclesiastical sites
such as Armagh. On Skelligs it is tempting to see this represented as a linear phenomenon.
8.2.13 THE
LOWER MONKS’ GARDEN
At the east end of this terrace there are two phases of early medieval occupation, both of which are above
the level of the wall-walk of the outer enclosing wall. The second phase of early medieval occupation
consists of a paved area, a wall and Cell G. This cell was still standing in the nineteenth century, when it
was occupied by the builders of the lighthouse.
Further west, the inner enclosure wall collapsed at some time in the fourteenth century and filled
that area of terrace with rubble and very large building stones. It must have proved too difficult for the
monks to completely clear this collapse out of the outer enclosure, and so they redeposited midden
material from elsewhere on the site and seem to have used the middle section as a garden. This midden
material produced radiocarbon dates ranging between the seventh and twelfth centuries, while the finds
included medieval pottery dating from the twelfth to fourteenth centuries. This area remained a garden,
while the area to the east remained occupied by buildings and paving. This begs the question as to
whether there had been buildings in this part of the ‘garden’ prior to the collapse and redeposition. The
redeposited material sloped downwards to both the south and east, and a small pathway led through it
towards the paved area at the east end of the garden. The only other features were walls that appear to
have been built in order to terrace part of the garden, probably for cultivation. There were no structures
relating to the lighthouse occupation at this end of the garden.
8.2.14 THE
STRUCTURE AT THE END OF THE EAST STEPS
This enigmatic structure is at the current base of the east steps. It has an orthostatic base to its inner wall
and is built into the east steps. Material was sampled but did not provide enough charcoal for a
radiocarbon date. This may have been a storehouse for material brought ashore at the east landing and
appears to be monastic in date.
8.2.15 A
•
•
•
•
•
•
•
•
458
HISTORY OF STRUCTURAL WEAKNESSES/COLLAPSE WITHIN THE MONASTERY
Pre-tenth century: Phase 1 east wall (inner enclosure)
Pre-tenth century: Phase 2 east wall (inner enclosure)
Pre-tenth century: blocking of entrance in Phase 3 east wall (inner enclosure)
C. 1300: collapse of Phase 2 south wall (inner enclosure)
Nineteenth century or earlier: collapse of south wall (outer enclosure), upper monks’ garden
C. 1820: movement in Phase 3 east wall (inner enclosure)
C. 1820: collapse of Phase 1 south wall (inner enclosure).
C. 1820: collapse of east end of outer enclosure, lower monks’ garden
INTERIM CONCLUSIONS
• C. 1870: collapse of south wall (inner enclosure) below St Michael’s Church
• C. 1978: small collapse west of south entrance 2 (inner enclosure)
8.2.16 DATING
The dates obtained from the redeposited layers in the lower monks’ garden show that, while the artefacts
mostly date from the high medieval period, the soil that was redeposited in the garden contained charcoal
dating from the seventh to twelfth centuries. This indicates that the soil was probably redeposited from
midden material sourced from elsewhere in the monastery. In the leacht area the three burials have
overlapping dates and have a direct stratigraphical relationship; the same is true of the dates from above
and below the leacht and from the east entrance. It is hoped to conduct Bayesian analysis in order to
tighten up these dates and give greater precision to our understanding of the site.
8.3 THE SOUTH PEAK
8.3.1 EROSION
While it is inevitable that the passage of time will have an effect on any built structures, especially those
erected in an environment as exposed and harsh as that of the South Peak, the recent works dramatically
demonstrated that the surviving structures were generally in a worse state of preservation than had been
anticipated by the ongoing monitoring.This was a result of so much being hidden beneath plant growth,
and the earlier difficulty of access without ropes and scaffolding. The works clearly showed that much
has been lost, that the surviving remains were often highly vulnerable and that erosion and loss were
continuing. The south and east sides of the oratory terrace, the shrine terrace, the outer terrace and the
south end of the ‘garden’ terrace were actively eroding and deteriorating, and demonstrably had lost
material since the 1980s/90s. Even in some areas where structures initially appeared to be well preserved,
e.g. the upper traverse and the west end of the ‘garden’ terrace, they proved to be in a much poorer state
of preservation when they were examined in detail, which was often only possible after a certain amount
of excavation. The lower part of the route up the Peak was also very vulnerable, as there only the very
last remnants of previously unknown and formerly substantial structures survived.
The works revealed, not unexpectedly, that a host of different factors were having a detrimental
effect on the surviving remains and, if not dealt with, would result in even further loss. Some of the
erosional factors, such as weathering and falling rocks, are unstoppable and will always continue; others,
however, could be slowed or halted. There was clear evidence in places that of course these effects also
occurred during the time when the South Peak was used by the monks of the monastery.
The very basic structure of the Peak itself contributes to the erosion of anything built upon it. The
relatively soft nature of the sandstone of which it is composed, allied to its fractured, folded and faulted
state, renders it vulnerable to erosion. The natural fracturing of the rock of the Peak along fault lines, for
example, resulted in the breaking away of a substantial area of rock at the base of the gully and a section
of the cliff at the junction of the short broad and ragged ledges. This erosion, indeed, could have been
the cause of the abandonment of the gully as a route to the top of the Peak, and in the other example it
severed access to the Peak and led the lighthouse-builders to blast away a section of the cliff to restore it.
On a smaller scale, the erosion of the exposed ends of the fracture planes in the bedrock has led to the
break-up or disappearance of several rock-cut steps, particularly those on the western-facing slopes of
the Peak.
Of course, without such natural erosion the ledges, which the routes up the Peak follow and on
459
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
which the various terraces and traverses were constructed, would not have formed out of the natural
bedding planes.The erosion of gullies that had formed along fault lines also provided natural access routes,
while the presence of fracture lines provided naturally broken-off rock and allowed the relatively easy
quarrying of rock for use as building material. Indeed, it is clear that the underlying geology had a most
fundamental influence on what was erected on the Peak, where it could be erected and how well it
survived. The monks clearly had a very intimate familiarity with and knowledge of their natural
environment, the uses to which they could put it, how they could alter and shape it, but also the limits
of its use.
Naturally falling rocks constitute an ongoing and unstoppable erosive force. Their dramatic effects
can be seen in particular all along the lower part of the route and along the southern traverses of the
north-west passage.There are several large stones in the gully at the north-west end of the broad terrace,
for example, which will fall at some time in the relatively near future.
Erosion caused by rainwater gathering and flowing down the Peak was also well attested.The features
in the gully were all heavily eroded by water and the material washing down in it. The bridging stone
beneath the oratory was partly rotted away by the rainwater that gathered and flowed down the gully
dividing the oratory and shrine terraces. This erosion ultimately led to the collapse of the oratory, but
also to the loss of the wall of the shrine terrace and the undermining of the shrine itself. There was a
substantial debris fan on the first section of the north-west passage at the base of a natural gathering
point for water. The monks themselves built a wall there to try to hold back this material. Narrow ‘drip
gullies’ also formed at the base of cliffs as a result of water trickling down the cliff faces, for example on
the upper traverse and the broad ledge. In the case of the latter, the gathering water flowed down and
out across the ledge, eroding and removing much material and depositing it over the cliff. The monks
themselves utilised the water-gathering effects of the high cliffs to collect water to feed cisterns on the
oratory terrace.
Rainwater also had a detrimental effect on virtually all the built structures on the Peak, as it washed
out the finer particles in the quarry waste used to infill and level the terraces. This caused many of the
terrace walls to move inwards and some eventually to fail and collapse.The loss of the walls at the southern
end of the ‘garden’ terrace and the northern end of the outer terrace was largely due to water draining
down to the lower ends of these terraces and undermining and destabilising their lower walls. Once the
lower ends of these terraces collapsed, the terrace infill began to erode ever more quickly and would
eventually have led to the collapse of most of these terraces. As the outer walls moved inwards, the stones
in the base of the outer face gradually moved to an ever-steeper angle until they slid out and brought
the whole wall down. This process was dramatically illustrated by the wall at the west side of the outer
terrace, which would have fallen relatively soon if not dealt with. The monks tried to halt the inward
collapse of the wall of the ‘garden’ terrace by the construction of an inner reinforcing wall. A similar
reaction to such a situation is evidenced in the monastery, and also along the inner side of the enclosing
wall on nearby Church Island in Valentia Harbour. Even walls built close to rock faces were vulnerable
to erosion and undermining by water running down the rock faces behind them. The high terrace on
the fourth section of the north-west passage, the wall of the outer passage and the west wall of the outer
terrace were all good examples of this effect.
The exposure of the walls to sun for long hours during the summer months, particularly on the
southern side of the Peak, may also have had a weakening effect on the stones, by causing them to expand
and contract and eventually to break up. The fractured nature of the stone revealed in the walls of the
‘garden’ and outer terraces may have been partly due to this effect. In addition, the opening of joints in
the walls and the fracturing of the edges of the individual stones could be caused by a combination of
sun and rainfall. The very finely jointed nature of a small section of the outer wall of the south end of
the ‘garden’ terrace that was protected by soil contrasted dramatically to the open-jointed nature of the
460
INTERIM CONCLUSIONS
more exposed walling revealed elsewhere on the Peak.
Even wind had a demonstrably damaging effect on the structures on the South Peak. As well as
driving water into the walls and blowing away the finer particles from the terrace infill, strong winds
have actually blown stones off the tops of walls. This was clearly demonstrated at the south end of the
upper traverse, where east to south-easterly winds funnelled by the cliffs at the top of the south side of
the Peak blew away the conserved inner face of the top of the wall. Wind damage from a westerly wind
was also documented on the lower and upper platforms and the small enclosure above the Needle’s Eye.
Winds blowing up the cliffs create unusual vortices on the terraces, resulting in the removal of fine
material (and small stones) from the inner side of the terrace and its deposition against the inner face of
the outer wall, where it forms a low mound. This also noticeably affects the lighthouse roadway.
Plant growth, especially sea campion and to a lesser extent sea pinks, also served to open joints in
stonework and actually lifted and separated stones in structures, thereby weakening and destabilising
them.
Burrowing birds and animals also contributed to undermining walls and weakening structures, and
were responsible for the destruction of the interior of the leacht on the oratory terrace, for example.
Indeed, some birds were even more destructive. A herring gull was witnessed throwing stones with its
beak off the top of the wall of the small enclosure in an attempt to get at petrels nesting in the walls.
Rabbits are to be found all over the South Peak: an active burrow was revealed halfway up the nearvertical lower part of the gully, for example.
Even humans cannot escape blame for erosion, whether deliberate or accidental, although, thankfully,
deliberate acts of vandalism have not occurred on the Peak. Climbers jumping down onto masonry
structures, where the erosion of previous steps forced such actions, also damaged structures such as the
west end of the lower traverse. Indeed, just by their presence humans have eroded some fragile rock-cut
steps and have had a particularly detrimental effect on the looser soils, cutting deep scars into the scree
slope, for example.
Some of the very construction methods employed by the monks contributed to the vulnerability
and destruction of the structures they erected. The fact that the terraces were infilled with fine, loose
material and were not constructed from solid stonework weakened them considerably. The tendency to
build the inner face of the outer terrace walls with very small stones—a testimony to the difficulty of
winning sufficient stone for construction—also weakened the structures. The erection of the inner face
of the wall of the ‘garden’ terrace on loose infill and the building of some terrace walls on sloping but
smooth bedding plane ledges on the north and south sides of the Peak also led to instability. Only on the
oratory terrace, where the slopes of the bedding plane were at its steepest, did the monks cut notches
into the rock to provide a better footing for the walling.
8.3.2 CONSTRUCTION
The recent works on the South Peak have clearly evidenced the great degree to which the monks altered
the South Peak. There is in fact hardly any place where the climber of either of the access routes stands
on a natural surface. Many areas where the cliffs were substantially cut back to facilitate or improve access
were recorded. The monks also quarried or removed stone from many areas, and sometimes moved it
considerable distances for construction work. In some areas direct evidence was uncovered of how the
stone was raised to build structures: for example, the hauling platform beneath the oratory terrace, the
tripod holes on the ledge with fallen stones, and rope-worn notches that survived on the outer terrace.
The carefully worked ledges that created a secure footing for the builders outside the outer passage
and oratory terrace demonstrated how the monks utilised natural features to help their building works.
461
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
The monks also displayed several construction techniques by which they sought to use natural
features or overcome obstacles. Walls were sometimes keyed into fissures in the rock (as at the east end
of the ‘garden’ terrace), or tucked behind high ribs of bedrock (as at the north-east and south-east ends
of the outer terrace) to give them greater stability. Gullies were bridged by placing large stones across
them (as evidenced beneath the oratory on the oratory terrace, and the north and south platforms below
the ‘garden’ and oratory terraces). Narrow passages were widened by building walls outside them on
lower ledges (as on the narrow ledges on either side of the blind corner, and where the route crosses the
gullies on the southern traverses of the north-west passage). Stones were also sometimes jammed vertically
down into fissures or gullies in the bedrock or vertically down and transversely across them to provide a
secure footing for walling (as in the gully leading up to the oratory terrace, and at the junction between
the northern stairway and the northern traverse on the north-west passage). Some of these techniques
can be paralleled elsewhere on the island and further afield. The reinforcing wall on the ‘garden’ terrace
has a clear parallel with that used to counteract the inward movement of the cashel wall on Church
Island, near Valentia, Co. Kerry.Walls built on lower ledges to widen access can also be seen on the recently
discovered early route to the monastery.
The monks created rock-cut steps where construction was impossible or where altering a bedrock
ledge could most easily create steps. Climbing the steps makes it readily apparent that they were very
carefully laid out. There are clear right foot/left foot sequences, and the steps are always exactly where
they are needed. On some of the more awkward climbs and passages, the climber naturally reaches out
to hold onto adjacent rock and almost invariably finds carefully placed rock-cut handholds there.
The excavation and conservation works revealed that there would have been very major practical
issues with storage space when building the various terraces. The rock lying on ledges would have had
to be broken up, removed and stored before being used for construction, as would the quarrying debris
used to fill the interior of the terrace.
We very often do not know how much rock was quarried away or removed to allow the construction
of the larger terraces and traverses on the Peak, as the bedrock beneath these structures was generally not
revealed in total. For example, only some of the rock lying on the second section of the southern traverse
of the north-west passage was removed; that lying beneath the intended finished level of the terracing
was left in situ. In this case, however, leaving some of the fallen rock in place may indicate that there was
a more than adequate supply of building stone in this area, which was very prone to falling stone. In
other areas the use of very small stones in the inner faces of walls shows that good building stone was in
short supply.
8.3.3 THE TWO ACCESS
ROUTES
The most important finding of the recent works has undoubtedly been the discovery of the new and
earlier route to the top of the South Peak and the terraces and structures associated with it.
The archaeological evidence could not clearly demonstrate how this route evolved into the
previously recorded final route. Because the two routes intersect twice and there is virtually no dating
evidence for most of the structures, several different scenarios are possible. All hinge on the relationship
in time between the opening of the Needle’s Eye and the building of the oratory and ‘garden’ terraces
and the upper traverse.
It is possible that the upper terraces may all have been built before the Needle’s Eye was opened.
The route then could have run across the north-west passage, up the lower half of the gully and then on
up to the ‘garden’ terrace. While this could explain the elaborate terracing of the north-west passage (its
presence when the route simply extended up the full length of the gully seems almost superfluous when
462
INTERIM CONCLUSIONS
the difficult climb up the full height of the gully had to be faced), it does seem like an unlikely conclusion.
The second suggested scenario could have seen the Needle’s Eye being opened to bypass only the lower
half of the gully. This was the most difficult part of the climb. Because of the presence of intersecting
faults, the lower part of the gully was very prone to large-scale erosion and so could have been rendered
very difficult to climb at an early stage. Above the Needle’s Eye the route could have run back into the
upper half of the gully and straight on up to the top of the Peak. As we do not know whether the area
where the upper traverse was constructed would have been passable prior to its construction, it is possible
that the upper part of the route in this second scenario could have extended up what later became the
upper part of the final route. The ‘garden’ and oratory terraces need not necessarily have been built, as
the upper part of this route can involve not actually setting foot on either of them. The most likely
scenario suggested, however, would appear to involve the opening of the Needle’s Eye and the building
of the ‘garden’ and oratory terraces and the upper traverse as part of one single development.The opening
of the Needle’s Eye was a considerable and difficult undertaking. It required quarrying and the cutting
of many rock-cut steps just to get to its base, and then, of course, there was the substantial amount of
quarrying involved in opening the Needle’s Eye itself. The upper part of the cleft leading up to the
‘garden’ terrace also had to be substantially quarried back to allow access to this terrace.
The interrelationship of the ‘garden’ and oratory terraces and the upper traverse is somewhat
complicated by the fact that there are two worked routes linking the three structures. While the route
via the outer passage from the ‘garden’ terrace to the oratory terrace is little used today, this is largely
because of the collapse of the upper part of the outer wall at the west end of the oratory terrace. Although
more exposed and vertiginous, this route is actually an easier climb than that up from the north-west
end of the ‘garden’ terrace. The route to the upper traverse up the cliff at the north side of the oratory
terrace is also the lesser-used today of the two routes connecting these structures. It is a very easy climb,
at least upwards. The placement of the large stone across the drip gully on the upper traverse, which
probably facilitated access to this route, suggests that it was not just a route used during construction and
then abandoned. While we will probably never know for certain what these different routes signify and
how they were used, it is possible that the routes represent something of a ‘round’, allowing for a different
route to be followed on the descent from the top of the Peak than was taken on the ascent. One of these
routes would have involved visiting the oratory terrace, while the other could have bypassed it.
8.3.4 THE
OUTER TERRACE
The presence of the outer terrace complicates matters even more, as it is accessible from both the primary
and final routes and we have no dating or stratigraphic evidence to show when it was built. It is also an
enigmatic structure but its interpretation could be vital to understanding the history of the development
of the site.
We do not know, for example, whether the outer terrace was even completed, such was the degree of
erosion to its lower and eastern sides. If completed it would have been a substantial structure; its lower
(north) wall probably would have been the highest built on the Peak.The alignment of the terrace is similar
to that of the oratory terrace and afforded a fine view eastwards of the monastery and Little Skellig beyond.
That this view was significant is clear from the orientation of the oratory on the oratory terrace and St
Michael’s Church in the monastery. It is therefore not impossible that a small oratory could have stood, or
there could have been the intention to erect one, on the outer terrace, although no evidence remains of
such a structure.The presence of the probable leacht on the outer terrace could also suggest that this terrace
was, or was intended to be, a version of the oratory terrace, possibly earlier than it but never finished. The
leacht, of course, could be later and was perhaps intended to mark an earlier sacred spot.
463
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
8.3.5 NEW
STRUCTURES ON THE LOWER PART OF THE ACCESS ROUTE
The partial and poorly preserved remains of previously unrecorded and substantial structures were
uncovered or detected all along the lower part of the route. They were generally in very poor condition
and represented only the very last remnants. They do, however, show that there was hardly anywhere on
the climb that was not improved by the monks. Unfortunately it is not possible to determine when these
structures were built relative to the usage of two different routes higher up.
8.3.6 DATING
It has generally and perhaps correctly been suggested that the structures on the South Peak would most
likely have been built only after the monks established a permanent presence in the monastery on the
island. This does not, however, preclude the early use of the South Peak. John Windele’s description of
the South Peak refers to a ‘stone of Don’ on the South Peak (Horn et al. 1990, 16). The Leabhar Gabhála
Érenn records that Don, killed during the battles between the Tuatha Dé Danann and the Milesians for
the control of Ireland, requested burial on an island, supposed to have been Skellig, and that his brother
Amairgen declared that the high rock, Tech Donn, should be visited by people (quoted in Horn et al.
1990, 16). This myth raises the possibility of a pre-Christian presence on Skellig. It seems unlikely, given
the conspicuousness and brooding presence of the Skelligs, that the islands would not have had a human
presence in early times, even if only a temporary or seasonal one. No evidence for such early occupation
has been recovered as yet. Neither is it clear whether the primary route, which gives access to the South
Peak, would have been negotiable before the monks cut back the rock faces and constructed the various
terraces, raising the issue of whether the South Peak could have been accessed in pre-monastic times.
The date of commencement of use of the South Peak by the monks themselves is also not
determinable. Unfortunately the samples of burnt material from the oratory and outer terraces only
contained oxidised clay and were not suitable for radiocarbon dating.The only dating evidence therefore
relies on a comparison of the structures on the South Peak with those found elsewhere.
The cross-slab found on the oratory terrace in the 1980s is the only reasonably closely datable object
that was found on the South Peak. Marshall has argued that it belonged within a ninth-century tradition
(Horn et al. 1990, 80–3). The oratory there also fits comfortably into a similar time period. As noted
above, however, the oratory was built after the rock-cut water basins were constructed and so it is possible
that this area could have been in use for some time before the small church was built.The shrine associated
with the oratory terrace is unlikely to be earlier than the eighth or ninth century, as recent excavations
of other shrines in Iveragh have shown.The relative dating of the construction of the shrine and oratory
terraces is also unknown.
It is not possible to determine on present evidence how much earlier the primary route may have
been than the final route.
Some of the structures on the South Peak have general parallels with parts of the very earliest set of
steps to the monastery (which were only discovered in 2010), and also with the monastery itself.
It should be remembered that the individual structures uncovered on the South Peak represent their
final form. Owing to, for example, natural erosion and damage by falling stones many of these structures
must have been repaired or rebuilt on a number of occasions. Unfortunately, owing to the nature of these
structures, with the exception of the reinforcing wall on the ‘garden’ terrace and the raising of the interior
of the upper traverse we cannot distinguish with any certainty between what might be original and what
might be a repair or replacement. It is also impossible to determine in what order the structures on each
route may have been built. One would expect that construction of the various terraces and climbs would
464
INTERIM CONCLUSIONS
have been undertaken over a protracted period, and that there would also have been a natural urge to
enhance and improve the access over a considerable time. There are places where access originally may
have been possible but where at a later time additional structures may have been built to improve it. A
good example of this is the additional walling on lower ledges below the narrow ledges on either side of
the blind corner on the lower part of the route. It is not clear whether this masonry was primary or
added to widen the route at a later time. It is also unclear whether the rock-cut steps there were intended
originally as steps for feet or were primarily intended or secondarily used to secure masonry. This
constructional trick, however, has a parallel on the earliest steps built to the monastery. The masonry
stairway on the broad ledge is another example. It was a large construction project and could conceivably
have been undertaken at a later time to improve access.There are also several structures on the Peak that
were not necessary to the climb—the platforms and the small enclosure above the Needle’s Eye and the
‘garden’ terrace, for example. These could have been later additions. The crosses and leachta that dot the
routes up the Peak could also have been added at any time. We have to remain aware that pilgrimage to
the South Peak continued until the seventeenth/eighteenth century and that some structures could have
been repaired, rebuilt or even added at a relatively late date. For example, the kneeling stone before the
altar in the ruined oratory on the oratory terrace appears to have been moved, relaid and had crosses
scratched on it after the collapse of the oratory at an unknown date.
The hole drilled for explosives at the junction of the short broad and ragged ledges also clearly shows
that the lighthouse-builders were active on the South Peak. They may have used the lower part of the
route combined with the southern traverses of the north-west passage (the masonry of the fourth section
of which they may have repaired) as a short cut between the monastery (where they lived) and the upper
lighthouse. We also know that later in the nineteenth century the lighthouse-keepers were also familiar
with the South Peak, as they first brought the structures there to antiquarian attention.
8.3.7 FUNCTION
The final access route, with its oratory, shrine, leachta and crosses, was obviously used for pilgrimage. We
know little about early pilgrimage to Skellig Michael, however; one of the first references to it is in 1543
(Harbison 1991, 87), while the first historical mention of pilgrimage to the South Peak itself only
occurred several centuries later.
The monks went to great lengths to allow the circumambulation of the oratory, reminiscent of the
layout of the small oratory terrace within the monastery itself. Such circumnavigation of structures is
well attested in medieval pilgrimages, such as that to Lough Derg (ibid., 62), but is also evidenced in the
construction of the seventh/eighth-century shrines on nearby Illaunloughan and Church Island (near
Valentia). The presence of the shrine on the oratory terrace is also a sure indication of pilgrimage. The
small leachta and crosses at several points along the routes up the Peak would also appear to indicate
pilgrimage and were probably used as ‘stations’ where prayers were said. There are also a number of
places—e.g. section 1 of the southern traverses of the north-west passage, the upper and lower platforms
above the Needle’s Eye and the ‘garden’ terrace—where the terracing appeared excessively wide, and
these locations could have been used as gathering points or stations. The small enclosure above the
Needle’s Eye and the extension at the north-west end of the ‘garden’ terrace could also have functioned
as small prayer stations.
The primary access route appears to have run directly to the top of the Peak but may also have
been a pilgrimage route.The presence of the Outer Terrace and the fact that we cannot relate its building
to the use of either of the two routes complicate matters somewhat.
Horn et al. (1990) suggested that the South Peak was also a hermitage. They postulated that either
465
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
the ‘garden’ terrace or the outer terrace might have been the site of a hermit’s residence. The presence
of the reinforcing wall on the ‘garden’ terrace precludes the presence of a cell there, but the possible
shelter at the end of the outer terrace could have been the putative hermit’s residence. This, however, is
the most exposed and coldest terrace on the Peak and would have been a very challenging place to stay
for extended periods. If the outer terrace was built at an early date, it could alternatively have been the
focus for early pilgrimage, as it would have been the only terrace on the upper part of the Peak. This
shelter therefore could have been intended for pilgrims. If the outer terrace was built later when the
final access route was in use, it could have been a possible residence for a hermit, as it lay isolated from
the main route.The presence of a leacht there means, however, that the terrace could also have functioned
as part of the pilgrimage. Unfortunately, we do not know when this leacht was built. Of course, any
putative hermit associated with the final access route might have lived in the oratory on the oratory
terrace.
As with much else on the South Peak, the surviving archaeological evidence is more than ambiguous,
and interpretation is not helped by the lack of relative and absolute dating.
8.4 THE MONASTIC COMMUNITY
The excavations have thrown some light on the lifestyle of the monks who toiled, prayed and died on
Skellig Michael.
8.4.1 THE
LOCAL ENVIRONMENT
The analysis of pollen from the old ground surface at the time of construction of the large oratory has
indicated the possibility that pockets of Sorbus (Sorbus aucuparia?—rowan or mountain ash) were growing
on the island. Even though there is no record of trees growing on Skellig Michael within living memory,
there is no reason why Sorbus could not have grown there in the past, as it is a tree which is found at the
highest level on our mountains and can grow from crevices in rocks. Otherwise, as expected, the pollen
analysis indicated that the local environment was one of open grassy swards, with species such as Sedum
that occur on sea cliffs and rock outcrops.The burning off of herbaceous vegetation before construction
of the large oratory is also suggested by the presence of charcoal in the old ground surface.
8.4.2 CROP
CULTIVATION
The evidence from both pollen and plant macrofossil remains indicates that barley and oats were cultivated
by the monks on Skellig Michael. The presence of fat hen (Chenopodium album), an invasive species
generally associated with plant cultivation, supports the theory of on-site cultivation, although it could
also be argued that the Chenopodium was gathered together with the sheaves of barley and oats on the
mainland and that its seeds (and pollen) were released during processing on the island. The fragments of
quernstones found during excavation confirm that crop-processing took place within the monastery.
Neither cereal species could be identified to type, nor does the very small sample size allow any
conclusions to be drawn about the relative importance or spatial distribution of the cereals.
The presence of barley and oats is not surprising, given that these cereals were important plants in
the early medieval period and both are tolerant of cooler and wetter climates. This is especially true of
oats, which can also withstand high coastal winds and which was a cheap and readily available crop used
466
INTERIM CONCLUSIONS
for making flat oatcakes, porridges and gruels (Monk 2011, 37). Its high nutritional value has led to its
being referred to as the ‘superfood’ of the early medieval period (ibid., 36). Barley was also used for
making bread, porridge and gruels, along with malt for beer production, but there was no evidence for
the latter at Skellig Michael. The cultivation of barley and oats on Skellig Michael fits well with the
evidence from contemporary eccelesiastical sites, such as nearby Illaunloughan (where the cereals were
sourced from the mainland) (Marshall and Walsh 2005, 130), Caherlehillan, near Cahirciveen (J. Sheehan,
pers. comm.), and High Island, Co. Galway (G. Scally, pers. comm.). Research by Monk (1991) has also
shown that these two cereals are the dominant species on secular sites of the early medieval period.
Perhaps what is most surprising is that cereal cultivation was attempted at all on the craggy rock that
is Skellig Michael. The monastery and South Peak structures were constructed on terraces supported by
substantial revetment walls. Several of these terraces do not appear to have supported structures of any
kind and consequently are referred to as ‘garden terraces’. The largest of these are the upper and lower
monks’ gardens within the outer enclosure of the monastery, although excavation revealed that at some
stage in its use the lower monks’ garden supported a cell (Cell G) and associated paving. Built on a southfacing slope and sheltered by the retaining wall, these areas are well suited to gardening and cultivation,
as demonstrated in the 1980s, when a range of vegetables were grown as an experiment in the upper
monks’ garden (G.D. Rourke, pers. comm.). The insect remains analysed from the lower monks’ garden
indicate that seaweed may have been used as a fertiliser, in addition to human excrement and animal/bird
bones. The possible use of seaweed is interesting, given that it is used to mitigate the lack of nutrient
nitrogen that can affect the production of oats (Marshall and Walsh 2005, 132).
8.4.3 EXPLOITATION
OF MAMMALS, BIRDS AND FISH
Bearing in mind the limitations imposed by sample size and the broad date range (from the late seventh
century to the fourteenth century), the excavated evidence allows us to hypothesise that the diet of the
Skellig Michael monks was relatively broad-based. In addition to using cereal-based foods and presumably
vegetables, which do not normally leave a trace in the archaeological record, the monks also exploited a
diverse range of terrestrial and marine resources.
Goat and sheep dominated the small assemblage of mammal bones and would have been important
sources of milk and meat. Both could have been kept on the island but no evidence of the presence of
animal dung could be inferred from the insect remains. Joints of pig and cattle meat must have been
imported from the mainland, and there is evidence that complete cattle carcasses were butchered on the
island. Seal meat was also consumed and may have been a food permitted during periods of fasting. The
single horse bone found must have made its way to the island in a joint of meat.
Birds, predominantly Manx shearwater and the auk family, were exploited mainly during the summer
breeding season, and it is possible that some of them may have been salted, pickled or smoked for later
use. Both trimmings and final food waste were retrieved during excavation. Chop-marks indicate that
the feet were removed as waste and there is evidence that the birds were roasted over a fire after the ends
of the wings were broken off.
At least seven species of fish were recorded in the bone assemblage, with sea bream (probably red sea
bream) the most frequent. All except cod could have been caught locally, either directly from the island
on lines or from local boats. The cod, which favours deep, cooler waters, must have been caught well
offshore. Marine molluscs are represented by a cache of limpet shells and a small number of scallop shells
found with the human remains.
The dietary evidence, restricted as it is, reflects much richer fare than the penitential diets described
in early documentary sources and certainly does not suggest an ascetic regime commensurate with the
467
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
extreme monasticism indicated by the choice of location. Perhaps pragmatism ruled the day, and monks
who must have had to maintain their health and physical fitness to construct and maintain their monastery
were allowed a balanced and varied diet. The nature of that diet reflects the monastery’s remote coastal
location and the need to exploit local resources as a means towards establishing an element of selfsufficiency.
8.4.4 WOOD AND
FIRE
Charcoal and wood analysis has thrown some light on the wide range of wood brought onto the island
and used for fuel and possibly for structural purposes, e.g. the construction of temporary shelters,
scaffolding etc. Nine different trees were identified as having been utilised during the early medieval
period, dominated by oak and also including ash, hazel, birch, alder, willow, holly, yew and purging
buckthorn. All are native species and must have been collected from a range of different environments
on the mainland, indicating strong external contacts.
8.4.5 MATERIAL
CULTURE
The limited number of artefacts recovered allow us further glimpses of the monastic way of life and
contacts with the outside world. Of the nineteen sherds of medieval pottery found, thirteen are imports—
from near Bristol, England, and the Saintonge and Orléans areas of France. Dating predominantly from
the late twelfth and early thirteenth centuries, these could perhaps be attributed to offerings by pilgrims,
if not from abroad then possibly from the nearby urban centres of Limerick and Cork, where such pottery
would have been in plentiful supply.
A single writing lead, for use on parchment or vellum, hints at scholastic activities.The two fragments
of incised slates, whilst not having clearly discernible patterns or subject-matter, could also perhaps relate
to instruction or training, as evidenced at the contemporary island monastic site of Inchmarnock, Scotland
(Lowe 2008). The bodies of juveniles aged 9–12 years at death on Skellig Michael also suggests the
presence of novices.The practice of fostering out young children, between the ages of seven and seventeen
for boys and fourteen for girls, was widespread in Ireland in the early medieval period, both in the secular
world and the religious world, where training began in boyhood (ibid., 262). Fosterage is also suggested
as an explanation for the presence of three juveniles buried in the monastic settlement of Illaunloughan
(Marshall and Walsh 2005, 84–5).This theme of fosterage will be further researched in relation to Skellig
Michael and it is hoped that additional analytical research will be carried out on the juvenile burials,
which may shed light on their genetic make-up and cultural backgrounds.
The antler gaming piece, lignite ring and disc bead and bronze pins all provide insights into the
more personal side of the monastic community.
8.4.6 CONCLUSION
In spite of the limitations imposed by the scale of the excavations, the nature of the deposits and the
small sample sizes, important information has been gleaned about the lifestyle of the Skellig Michael
monks. Perhaps the most significant picture that is emerging is that in many respects the Skellig Michael
monastery is little different in terms of diet and economy from other contemporary island and mainland
ecclesiastical sites. Given the precipitous nature of the island, it would be unrealistic to expect the
468
INTERIM CONCLUSIONS
monastery to have been self-sufficient, but the extent of external provisioning suggested by the findings
is somewhat surprising and hints at a degree of external contact not to be expected for an eremitical
community.The monastery would of course have developed and changed over the centuries, but towards
the end of the early medieval period perhaps we should think of Skellig Michael as offering both a
communal experience for those in the monastery and the opportunity for a more eremitical way of life
on the South Peak.
8.5 ABANDONMENT OF THE ISLAND MONASTERY
The departure of the monks from Skellig Michael is generally dated to the thirteenth century, with
deteriorating weather conditions and the influence of the twelfth-century church reform movement
cited as contributing factors. A tradition of pilgrimage to the island appears to have been well established
by the thirteenth century, a factor which may also have contributed to the monks’ decision to relocate
to Ballinskelligs Abbey on the mainland. It is likely that the abandonment of the island monastery took
place gradually over a period of time. Once the Augustinian priory was established in Ballinskelligs, the
monks may no longer have occupied Skellig Michael on a year-round basis but instead used it as a place
of ‘retreat’ during the summer months.
The radiocarbon dates obtained from habitation material (animal bone, shells and charcoal) indicate
activity on the island up to the early twelfth century. Burials 1 and 3 have been dated to the period
between the early eleventh and early twelfth centuries, while burial 2 produced the wide date range of
early eleventh–late thirteenth century. The latter is the only radiocarbon date that extends (at 2 sigma)
into the thirteenth century.
The medieval pottery tells a slightly different story and, if the single sherd of Saintonge sgraffito
(which dates from the late thirteenth/early fourteenth century) is excluded, the assemblage could belong
to a 50-year period in the later twelfth/early thirteenth century. Those sherds which came from secure
contexts indicate that substantial works, including the rebuilding of walls and building up of terrace
deposits (particularly in the lower monks’ garden), continued into the late twelfth/early thirteenth century.
The Saintonge sgraffito sherd does not come from a secure context.
In conclusion, then, the results of the excavations indicate that substantial works were being carried
out on the structural remains on Skellig Michael into the early thirteenth century. We do not know
whether the monks were in full-time occupation of the monastery at this time, but on present evidence
it appears that the final abandonment of the island probably took place in the mid–late thirteenth century.
469
BIBLIOGRAPHY
Abbreviations
JCHAS = Journal of the Cork Historical and Archaeological Society
JIA = Journal of Irish Archaeology
JRSAI = Journal of the Royal Society of Antiquaries of Ireland
PRIA = Proceedings of the Royal Irish Academy
PSAS = Proceedings of the Society of Antiquaries of Scotland
UJA = Ulster Journal of Archaeology
References
Alexander, K.N.A. 2002 The invertebrates of living and decaying timber in Britain and Ireland: a provisional
annotated checklist. English Nature Research Reports No. 467. Peterborough.
Allan, J.P. 1984 Medieval and post-medieval finds from Exeter 1971–1980. Exeter.
Allen, R., O’Donnell, L., Reilly, E. and Scully, S. 2008 Skellig Michael: assessment of environmental
and artefactual remains. Unpublished technical report for Margaret Gowen and Co. Ltd on behalf
of the Department of the Environment, Heritage and Local Government.
Anderson, R., Nash, R. and O’Connor, J.P. 1997 Irish Coleoptera: a revised and annotated list. Belfast.
Archdall, M. 1786 Monasticon Hibernicum (3 vols; reprinted 1876). Dublin.
Ascough, P.L., Cook, G.T. and Dugmore, A.J. 2009 North Atlantic marine C-14 reservoir effects:
implications for late-Holocene chronological studies. Quaternary Geochronology 4, 171–80.
Backlund, H.O. 1945 Wrack fauna of Sweden and Finland. Opuscula Entomologica Supplement 5.
Barton, K.J. 1963 The medieval pottery kiln at Ham Green, Bristol. Transactions of the Bristol and
Gloucestershire Archaeological Society 82, 95–126.
Bass,W.M. 1995 Human osteology. A laboratory and field manual (4th edn). Missouri Archaeological Society.
Beckett, J.K. 1979 Planting native trees and shrubs. Norwich.
Bengtson, S.-A. 1981 Terrestrial invertebrates of the Faroe Islands: III. Beetles (Coleoptera): checklist,
distribution, and habitats. Fauna Norvegica B28, 52–82.
Berger, R. 1995 Radiocarbon dating of early medieval Irish monuments. PRIA 95C, 159–74.
Berggren, G. 1981 Atlas of seeds and small fruits of northwest-European plant species (Part 3: Salicaceae–
Cruciferae). Stockholm.
Berry, R.J. 1985 The natural history of Orkney. London.
Best, R.I. and Lawlor, H.J. 1931 The Martyrology of Tallaght. London.
Best, R.I., Bergin, O. and O’Brien, M.A. (eds) 1954 The Book of Leinster: formerly Lebar na Núachongbála,
Vol. 1. Dublin.
Beug, H.-J. 2004 Leitfaden der Pollenbestimmung für Mitteleuropa und angrenzende Gebiete. Munich.
Bhreathnach, E. 1995 Tara, a select bibliography. Dublin.
Biddle, M. and Brown, D. 1990 Writing equipment. In M. Biddle (ed.), Object and economy in medieval
Winchester, 729–53. Winchester Studies 7(ii). Oxford.
Blake, H. and Davey, P. (eds) 1983 Guidelines for the processing and publication of medieval pottery from
excavations. London.
Bliss, W.H. and Twenlow, J.A. 1902 Calendar of entries in the papal registers relating to Britain and Ireland:
papal letters,Vol. 4. London.
Boardman, S. and Jones, G. 1990 Experiments on the effects of charring on cereal plant components.
Journal of Archaeological Science 17, 1–11.
Boessneck, J.A. 1969 Osteological differences between sheep (Ovis aries L.) and goat (Capra hircus L.).
471
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
In D. Brothwell and E. Higgs (eds), Science and archaeology, 331–58. London.
Boggard, A., Heaton, T.H.E., Poulton, O. and Merbach, I. 2007 The impact of manuring on nitrogen
isotope ratios in cereals: archaeological implications for reconstruction of diet and crop management
practices. Journal of Archaeological Sciences 34, 335–43.
Böhme, J. 2005 Die Käfer Mitteleuropas. K. Katalog (Faunistiche Übersicht) (2nd edn). Munich.
Bohncke, S. 1981 The pollen diagram from Ditch 1. In J.W. Barber, ‘Excavations on Iona, 1979’. PSAS
111, 282–380, at 346–8.
Bolton, B. and Collingwood, C.A. 1975 Hymenoptera: Formicidae. Handbooks for the Identification of
British Insects 4, Part 3(c). London.
Bourke, E. 1994 Glass vessels of the first nine centuries AD in Ireland. JRSAI 124, 163-209.
Bourke, E. 2005 A preliminary analysis of the inner enclosure of Skellig Michael, Co. Kerry. In T. Condit
and C. Corlett (eds), Above and beyond. Essays in memory of Leo Swan, 121–38. Bray.
Breatnach, P. 1977–8 Medieval traditions from West Munster. Studia Hibernica 17/18, 58–70.
Brooks, S.T. and Suchey, J.M. 1990 Skeletal age determination based on the os pubis: a comparison of
the Acsádi–Nemeskéri and Suchey–Brooks methods. Human Evolution 5, 227–38.
Brothwell, D.R. 1981 Digging up bones (3rd edn). Oxford.
Brown, D. 2002 Pottery in medieval Southampton, c. 1066–1510. CBA Research Report 133.York.
Brugge, A. 1905 Caithreim Cellachain Caisil. Cristiania.
Buckland, P.C., Dugmore, A.J. and Sadler, J.P. 1991 Faunal change or taphonomic problem? A comparison
of modern and fossil insect faunas from southeast Iceland. In J. Maizels and C. Caseldine (eds),
Environmental change in Iceland, 127–46. Dordtrecht.
Buckland, P.I. 2007 The development and implementation of software for palaeoenvironmental and
palaeoclimatological research: the Bugs Coleopteran Ecology Package (BugsCEP). Umea.
Buckland, P.I. and Buckland, P.C. 2006 Bugs Coleopteran Ecology Package. Downloaded/CDROM June
2007 (www.bugscep.com).
Buckley, L., Murphy, E. and Ó Donnabháin, B. 1999 The treatment of human remains: technical paper for
archaeologists. Dublin.
Buikstra, J.E. and Ubelaker, D.H. 1994 Standards for data collection from human skeletal remains. Arkansas
Archaeological Survey Research Series No. 44. Arkansas.
Bullock, J.A. 1993 Host plants of British beetles: a list of recorded associations. Amateur Entomologist
11a, 1–24.
Butler, R., Green, C. and Payne, N. 2009 Cast copper-alloy cooking vessels. Finds Research Group 700–
1700 Datasheet 41.York.
Brown, D. 2002 Pottery in medieval Southampton, c. 1066–1510. CBA Research Report 133.York.
Byrne, F.J. 2001 Irish kings and high kings (2nd edn). Dublin.
Caldwell, D.H., Hall, M.A. and Wilkinson, C.M. 2009 The Lewis hoard of gaming pieces: a reexamination of their context, meanings, discovery and manufacture. Medieval Archaeology 53, 155–203.
Cappers, R.T.J., Bekker, R.M. and Jans, J.E.A. 2006 Digital seed atlas of the Netherlands. Groningen.
Carey, J. 1993 A new introduction to Lebor Gabála Érenn: the Book of the Taking of Ireland. Dublin.
Carter, A. 1974 The bayonet: a history of knife and sword bayonets 1850–1970 based on the ‘Guns Review’
articles. London.
Chamberlain, A.T. and Witkin, A. 2000 Human skeletal remains from Cloghermore Cave, County Kerry,
Ireland. Unpublished report on human remains from first season’s excavation at Cloghermore Cave..
Chapman, A. 2005 Medieval stylised chess pieces. Finds Research Group 700–1700 Datasheet 35. York.
Charles-Edwards, T. 2000 Early Christian Ireland. Cambridge.
Coope, G.R. and Osborne, P.J. 1967 Report on the coleopterous fauna of the Roman Well at Barnsley
Park, Gloucestershire. Philosophical Transactions of the Royal Society of London B280, 234–41.
472
BIBLIOGRAPHY
Cousens, J. 1974 An introduction to woodland ecology. Edinburgh.
Cowgill, J., de Neergaard, M. and Griffiths, N. 1987 Medieval finds from excavations in London, 1: knives
and scabbards. London.
Coy, J.P. and Hamilton-Dyer, S. 1993 The bird and fish bone. In F. McCormick, ‘Excavations at Iona,
1988’. UJA 56, 100–1.
Cronin, M., Duck, C., Ó Cadhla, O., Nairn, R., Strong, D. and O’Keeffe, C. 2004 Harbour seal population
assessment in the Republic of Ireland, August 2003 (Irish Wildlife Manuals No. 11). NPWS
(http://www.npws.ie/en/media/NPWS/Publications/IrishWildlifeManuals/Media,4584,en.pdf,
accessed 8 February 2011).
Cronin,V. 1963 A calendar of saints. London.
Dennehy, E.A. and Lynch, L.G. 2001 Unearthed secrets: a clandestine burial ground. Archaeology Ireland
15 (4), 20–3.
De Paor, L. 1955 A survey of Sceilg Mhichíl. JRSAI 85, 174–87.
Dumbrell, R. 1983 Understanding antique wine bottles. Woodbridge.
Dunlevy, M. 1988 A classification of early Irish combs. PRIA 88C, 341–422.
Dunning, G.C. 1964 Pottery. In H.E. O’Neil, ‘Excavation of a Celtic hermitage on St Helens, Isles of
Scilly, 1956–58’. Archaeological Journal 121, 55–66.
Dunning, G.C. 1968 The trade in medieval pottery around the North Sea. In J.G. Renaud (ed.),
Rotterdam Papers I, 35–58. Rotterdam.
Dunning, P.J. 1945 The Arroasian Order in medieval Ireland. Irish Historical Studies 4 (16), 297–315.
Dunraven, E.R. 1875–7 Notes on Irish architecture (ed. M. Stokes) (2 vols). London.
Dwyer, P. 1981 Célí Dé. Spiritual reform in Ireland 750–900. Dublin.
Edwards, N. 1990 The archaeology of early medieval Ireland. London.
Egan, G. 1998 Medieval finds from excavations in London, 6: the medieval household, daily living c.1150–
c.1450. London.
Emmerson, R. 1991 Church plate. London.
Ericson, P.G.P. 1987 Interpretations of archaeological bird remains: a taphonomic approach. Journal of
Archaeological Science 14, 66–75.
Etchingham, C. 1999 Church organisation in Ireland AD 650–1000. Maynooth.
Fægri, K. and Iversen, J. 1989 Textbook of pollen analysis (4th edn, by K. Fægri, P.E. Kaland and K.
Krzywinski). New York.
Fairley, J. 1981 Irish whales and whaling. Belfast.
Fanning, T. 1981 Excavation of an Early Christian cemetery and settlement at Reask, County Kerry.
PRIA 81C, 67–172.
Fanning, T. 1994 Viking Age ringed pins from Dublin. Medieval Dublin Excavations 1962–81, Ser. B,Vol.
4. Dublin.
Feehan, J. 1997 The heritage of the rocks. In J.Wilson Foster (ed.), Nature in Ireland: a scientific and cultural
study, 3–23. Dublin.
Fellows-Jenson, G. 1992 Scandinavian placenames of the Irish Sea Province. In J. Graham-Campbell
(ed.), Viking treasure from the north west.The Cuerdale hoard in its context, 31–42. Liverpool.
Fisher, I. 2001 Early medieval sculpture in the West Highlands and Islands. Royal Commission on the Ancient
and Historical Monuments of Scotland and the Society of Antiquaries of Scotland Monograph Series
1. Edinburgh.
Fisher, J. and Vevers, H.G. 1943 The breeding distribution, history and population of the North Atlantic
gannet (Sula bassana). Journal of Animal Ecology 12, 173–213.
Fisher, J. and Vevers, H.G. 1944 The breeding distribution, history and population of the North Atlantic
gannet (Sula bassana). Journal of Animal Ecology 13, 49–62.
473
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Fisher, R.A., Corbett, A.S. and Williams, C.B. 1943 The relation between number of species and the
number of individuals in a random sample of an animal population. Journal of Animal Ecology 12, 42–
58.
Follett, W. 2006 Céli Dé in Ireland: monastic writing and identity in the early Middle Ages. Woodbridge.
Fredengren, C., McClatchie, M. and Stuijts, I. 2004 Reconsidering crannogs in early medieval Ireland:
alternative approaches in the investigation of social and agricultural systems. Environmental Archaeology
9 (2), 161–6.
Gahan, A. and McCutcheon, C. 1997 Medieval pottery. In M.F. Hurley and O.M.B. Scully, Late Viking
Age and medieval Waterford: excavations 1986–1992, 285–336. Waterford.
Gahan, A., McCutcheon, C. and Twohig, D.C. 1997 Medieval pottery. In R.M. Cleary, M.F. Hurley and
E. Shee-Twohig (eds), Skiddy’s Castle and Christ Church, Cork: excavations 1974–77 by D.C. Twohig,
108–29. Cork.
Gale, R. 2003 Wood-based industrial fuels and their environmental impact in lowland Britain. In P.
Murphy and P.E.J. Wiltshire (eds), The environmental archaeology of industry, 30–47. Oxford.
Gale, R. (forthcoming) Charcoal analysis from High Island. In G. Scally, High Island—Ard Oiléan,
Connemara, Co. Galway: excavation of an early medieval monastery on an Atlantic Irish island.
Gale, R. and Cutler, D. 2000 Plants in archaeology. Identification of vegetative plant materials used in Europe
and the southern Mediterranean to c. 1500. West Yorkshire.
Glenn,V. 2009 The later medieval ecclesiastical metalwork and related items. In C. Lowe, ‘Clothing for
the soul divine’: burials at the tomb of St Ninian. Excavations at Whithorn Priory, 1957–67, 82–97. Historic
Scotland Archaeology Report No. 3. Edinburgh.
Graham-Campbell, J. (ed.) 1992 Viking treasure from the north west. The Cuerdale hoard in its context.
Liverpool.
Grant, A. 1982 The use of tooth wear as a guide to the age of domestic ungulates. In B. Wilson, C.
Grigson and S. Payne (eds), Ageing and sexing animal bones from archaeological sites, 91–108. British
Archaeological Reports, British Series 109. Oxford.
Gurney, J.H. 1913 The gannet: a bird with a history. London.
Gwynn, A. 1992 The Irish church in the eleventh and twelfth centuries. Dublin.
Gwynn, A. and Hadcock, R.N. 1970 Medieval religious houses: Ireland; with an appendix to early sites.
London.
Halpin, A. and Buckley, L. 1995 Archaeological excavations at the Dominican Priory, Drogheda, Co.
Louth. PRIA 95C, 175–253.
Halstead, D.G.H. 1963 Coleoptera: Histeridae. Handbooks for the Identification of British Insects 9, Part
10. London.
Hamilton-Dyer, S. 2000 Fish bones. In E.V. Murray and F. McCormick, ‘Excavations of two Early
Christian sandhill sites at Doonloughan, Slyne Head, Co. Galway’. Unpublished report for Queen’s
University, Belfast.
Hamilton-Dyer, S. 2005 Fish bone. In J. Marshall and C. Walsh, Illaunloughaun Island: an early medieval
monastery in County Kerry. Bray. (Supporting archive only, http://wordwellbooks.com
/ftp/Wordwell/Illaunloughan/.)
Hamilton-Dyer, S. 2007 Exploitation of birds and fish in historic Ireland: a brief review of the evidence.
In E.M. Murphy and N.J. Whitehouse, Environmental archaeology in Ireland, 102–18. Oxford.
Hamilton-Dyer, S. (in prep.) Exploring the contrasts: fishbone assemblages in historic Ireland. In J.
Barrett and C. Johnstone, Cod and herring: the chronology, causes and consequences of medieval sea fishing.
Hamlin, A. and Foley, C. 1983 A women’s graveyard at Carrickmore, County Tyrone, and the separate
burial of women. UJA 46, 41–6.
Harbison, P. 1991 Pilgrimage in Ireland.The monuments and people. London.
474
BIBLIOGRAPHY
Harde, K.W. 1984 A field guide in colour to beetles. London.
Hansen, M. 1987 The Hydrophiloidea (Coleoptera) of Fennoscandia and Denmark. Fauna Entomologica
Scandinavica 18. Leiden.
Hather, J.G. 2000 The identification of the northern European woods: a guide for archaeologists and conservators.
London.
Hayden, A. 1998 Bray Head, Valentia Island. In I. Bennett (ed.), Excavations 1997: summary accounts of
archaeological excavations in Ireland, 81–2. Bray.
Hayden, A.R. 2011 A reassessment of Church Island, near Valentia, in the light of recent excavation.
Report submitted to DAHG and accepted for publication by the Royal Irish Academy.
Hencken, H.O’N. 1950 Lagore Crannog: an Irish royal residence of the 7th to 10th centuries AD. PRIA
53C, 1–247.
Henry, F. 1952 A wooden hut on Inishkea North, Co. Mayo. JRSAI 82, 163–78.
Henry, F. 1957 Early monasteries, beehive huts, and drystone houses in the neighbourhood of
Caherciveen and Waterville, Co. Kerry. PRIA 58C, 45–166.
Hickie, D. 2002 Native trees and forests of Ireland. Dublin.
Higgs, K. 2009 The geology of the Iveragh Peninsula. In J. Crowley and J. Sheehan, The Iveragh Peninsula.
A cultural atlas of the Ring of Kerry, 16–20. Cork.
Higham, C.F.W. 1967 Stock-rearing as a cultural factor in prehistoric Europe. Proceedings of the Prehistoric
Society 33, 84–106.
Hillson, S. 1986 Teeth. Cambridge.
Hinton, H.E. 1945 A monograph of the beetles associated with stored products, I. London.
Hogan, E. 1910 Onomasticon goedelicum, locorum et tribuum hiberniae et scotiae. Dublin.
Horn, W., Marshall, J.W. and Rourke, G.D. 1990 The forgotten hermitage of Skellig Michael. Los Angeles.
Horn, W., Marshall, J.W. and Rourke, G.D. 2002 The forgotten hermitage of Skellig Michael (2nd edn). Los
Angeles.
Hull,V. 1947 Conall Corc and the Corcu Loigde. Publications of the Modern Language Association of America
62 (4), 887–909.
Hunter, F. 2008 Jet and related materials in Viking Scotland. Medieval Archaeology 52, 103–18.
Hurley, M.F. 1997a Artefacts of skeletal material. In M.F. Hurley, O.M.B. Scully and S.W.J. McCutcheon,
Late Viking Age and medieval Waterford: excavations 1986–1992, 650–99. Waterford.
Hurley, M.F. 1997b Artefacts of skeletal material. In R.M. Cleary, M.F. Hurley and E. Shee Twohig (eds),
Skiddy’s Castle and Christ Church, Cork: excavations 1974–77 by D.C.Twohig, 239–73. Cork.
Husi, P. (ed.) 2003 La céramique médiévale et moderne du Centre-Ouest de la France (11e–17e siècle). Tours.
Irish Placenames Commission
2010
Placenames database of Ireland [online]. Available at
http://www.logainm.ie/ (accessed 27 July 2010).
Johns, C. 1996 The jewellery of Roman Britain: Celtic and Classical traditions. London.
Johnson, C. 1965 An introduction to the British Catopinae. Handbook of the North Western Naturalists’
Union, 15–22.
Johnston, P. 2001 Assessment of environmental remains, Skellig Michael, Co. Kerry. Unpublished
technical report for Margaret Gowen and Co. Ltd on behalf of the National Monuments Service.
Johnston, P. and O’Donnell, L. 2008 Analysis of the charred plant remains and charcoal. In G. Stout and
M. Stout (eds), Excavation of an early medieval secular cemetery at Knowth Site M, County Meath, 138–48.
Bray.
Johnston, P. and Reilly, E. 2007 Plant and insect remains. In A. O’Sullivan, R. Sands and E.P. Kelly,
Coolure Demesne crannog: an introduction to its archaeology and landscapes, 55–62 (with Appendix, 90–2).
Bray.
Joyce, P.W. 1975 The origin and history of Irish names of places (4th edn). Dublin.
475
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Keepax, C.A. 1988 Charcoal analysis with particular reference to archaeological sites in Britain.
Unpublished Ph.D thesis, University of London.
Kelly, F. 1976 The old Irish tree-list. Celtica 11, 107–24.
Kelly, F. 1997 Early Irish farming. Dublin.
Kelly, F. 1998 Early Irish farming. Early Irish Law Series Vol. IV. Dublin.
Kenny, J.F. 1929 The sources for the early history of Ireland:Volume 1, ecclesiastical (1993 reprint). Dublin.
Kenward, H.K. 1980 A tested set of techniques for the extraction of plant and animal macrofossils from
waterlogged archaeological deposits. Science and Archaeology 22, 3–15.
Kenward, H.K. 1997 Synanthropic decomposer insects and the size, remoteness and longevity of
archaeological occupation sites: applying concepts from biogeography to past ‘islands’ of human
occupation. In P.C. Ashworth, P.C. Buckland and J.P. Sadler (eds), Studies in Quaternary entomology: an
inordinate fondness for insects, 135–52. Quaternary Proceedings 5. Chichester.
Kenward, H.K. and Allison, E.P. 1994a Rural origins of the urban insect fauna. In A.R. Hall and H.K.
Kenward (eds), Urban–rural connexions: perspectives from environmental archaeology, 55–79. Symposia of
the Association of Environmental Archaeology No. 12. Oxford.
Kenward, H.K. and Allison, E.P. 1994b A preliminary view of the insect assemblages from the Early
Christian rath site at Deer Park Farms, Northern Ireland. In D.J. Rackham (ed.), Environment and
economy in Anglo-Saxon England, 87–107. CBA Research Report 89.York.
Kenward, H.K. and Hall, A.R. 1995 Biological evidence from Anglo-Scandinavian deposits at 16–22 Coppergate.
The Archaeology of York 14/7. London.
Kenward, H.K., Engleman, C., Robertson, A. and Large, F. 1986 Rapid scanning of urban archaeological
deposits for insect remains. Circaea 3, 163–72.
Kenward, H., Hill, M., Jaques, D., Kroupa, A. and Large, F. 2000 Evidence from beetles and other insects.
In A. Crone (ed.), The history of a Scottish lowland crannog: excavations at Buiston, Ayrshire, 1989–90, 76–
8 and 99–101. Edinburgh.
Koch, K. 1989 Die Käfer Mitteleuropas. Ökologie,Vols 1 and 2. Krefeld.
Kratochvil, Z. 1969 Das Fehlen des 2. Prämolaren beim europäischen Reh (Capreolus capreolus L.) aus
der jungeren Steinzei. Zeitschrift für Jagdwissenschaft 32, 248–51.
Kuijt, I., Lash, R., Gibbons, M., Higgins, J., Goodale, N. and O’Neill, J. 2010 Reconsidering early
medieval seascapes: new insights from Inis Airc, Co. Galway. JIA 19, 51–71.
Lavelle, D. 1977 Skellig Island. Outpost of Europe. Dublin.
Lavelle, D. 1993 The Skellig story. Dublin.
Lawlor, H.C. 1925 The monastery of St Mochaoi of Nendrum. Belfast.
Lionard, P. and Henry, F. 1961 Early Irish grave slabs. PRIA 61C, 95–169.
Lipscombe, M. and Stokes, J. 2008 Trees and how to grow them. London.
Lockley, R.M. 1966 The distribution of grey and common seals on the coasts of Ireland. The Irish
Naturalists’ Journal 15, 136–43.
Lovejoy, C.O., Meindl, R.S., Pryzbeck, T.R. and Mensforth, R.P. 1985 Chronological metamorphosis
of the auricular surface of the ilium: a new method for the determination of adult skeletal age at
death. American Journal of Physical Anthropology 68, 15–28.
Lowe, C. 2008 Inchmarnock: an early historic island monastery and its archaeological landscape. Edinburgh.
Luff, M.L. 2007 The Carabidae (ground beetles) of Britain and Ireland (2nd edn). Handbooks for the
Identification of British Insects 4, Part 2. London.
Lynch, A. 2010 Tintern Abbey, Co. Wexford; Cistercians and Colcloughs. Excavations 1982–2007.
Archaeological Monograph Series: 5. Dublin.
Lynch, L.G. 2001 Osteo-archaeological report on human skeletal remains excavated at Sceilig Mhichíl,
County Kerry. Unpublished report for Aegis Archaeology Ltd.
476
BIBLIOGRAPHY
Lynch, L.G. 2005 Osteoarchaeological report on human skeletal remains excavated at St Mary’s
Cathedral, Tuam, Co. Galway. Unpublished report for Aegis Archaeology Ltd.
Mac Airt, S. 1951 The annals of Inisfallen. Dublin.
Mac Airt, S. and Mac Niocaill, G. 1983 The annals of Ulster (to AD 1131). Dublin.
Macalister, R.A.S. 1956 Lebor Gabála Érenn.The Book of the Taking of Ireland, Part 5. Dublin.
McClatchie, M. 2008 High Island, Co. Galway: analysis of non-wood plant macro-remains. Unpublished
report for the Department of the Enviroment, Heritage and Local Government.
McClune, B. and Grace, J.B. 2002 Analysis of ecological communities. MjM Software Design, Gleneden
Beach, Oregon.
McClune, B. and Mefford, M.J. 1999 PC-ORD: Multivariate analysis of ecological data. Version 4.0. MjM
Software Design, Gleneden Beach, Oregon.
McCormick, F. 1993 Excavations at Iona, 1988. UJA 56, 78–108.
McCormick, F. 1998 Calf slaughter as a response to marginality. In C.M. Mills and G. Coles (eds), Life
on the edge: human settlement and marginality, 49–53. Oxford.
McCormick, F. and Murray, E.V. 2007 Knowth and the zooarchaeology of Early Christian Ireland. Dublin.
MacCotter, P. 2008 The history of Corcu Duibne. In T. Ó Carragáin and J. Sheehan, Making Christian
landscapes: settlement, society and regionality in early medieval Ireland—Phase 1. Report by Dept of
Archaeology, UCC, for the Heritage Council—Irish National Strategic Archaeological Research
(INSTAR) Programme.
McCracken, E. 1971 The Irish woods since Tudor times: distribution and exploitation. Devon.
McCutcheon, C. 2003 Pottery. In R.M. Cleary and M.F. Hurley (eds), Excavations in Cork City 1984–
2000, 197–235. Cork.
McCutcheon, C. 2006 Medieval pottery from Wood Quay, Dublin: the 1974–6 waterfront excavations. Dublin.
McCutcheon, S.W.J. 1997 The stone artefacts. In M. Hurley, O. Scully and S. McCutcheon, Late Viking
Age and medieval Waterford: excavations 1986–1992, 404–32. Waterford.
McGovern,T.H., Buckland, P.C., Savory, D., Sveinbjarnardottir, G., Andreasen, C. and Skidmore, P. 1983
A study of the faunal and floral remains from two Norse farms in the Western Settlement, Greenland.
Arctic Anthropology 20 (2), 93–120.
MacGregor, A. 1985 Bone, antler, ivory and horn: the technology of skeletal materials since the Roman period.
London.
Mainman, A.J. and Rogers, N.S.H. 2000 Craft, industry and everyday life: finds from Anglo-Scandinavian
York. The Archaeology of York 17/14.York.
Mann, R.W. and Murphy, S.P. 1990 Regional atlas of bone disease. A guide to pathological and normal variation
in the human skeleton. Smithsonian Contributions to Anthropology No. 28. Washington.
Marguerie, D. and Hunot, J.Y. 2007 Charcoal analysis and dendrology: data from archaeological sites in
north-western France. Journal of Archaeological Science 34, 1417–33.
Marshall, J.W. and Rourke, G.D. 2000 High Island: an Irish monastery in the Atlantic. Dublin.
Marshall, J.W. and Walsh, C. 2005 Illaunloughan Island: an early medieval monastery in County Kerry. Bray.
Mays, S. 1998 The archaeology of human bones. London.
Meyer, K. 1885 Cath Finntrága. Medieval and Modern Series,Vol. 1, Part 4. Oxford.
Meyer, K. 1910 Conall Corc and the Corco Luigde. In O. Bergin, R.I. Best, K. Meyer and J.G. O’Keefe
(eds), Anecdota from Irish manuscripts,Vol. 3, 57–63. Dublin.
Meyer, K. 1912 The Laud genealogies and tribal histories. Zeitschift für Celtische Philologie 8, 291–338.
Mitchell, F.J.G. 1988 The vegetational history of the Killarney oakwoods, SW Ireland: evidence from
fine spatial resolution pollen analysis. Journal of Ecology 76, 415–36.
Molloy, K. and O’Connell, M. 2007 Fresh insights into long-term environmental change on the Aran
Islands based on palaeoecological investigations of lake sediments from Inis Oírr. Journal of the Galway
477
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Archaeological and Historical Society 59, 1–17.
Monk, M. 1991 The archaeobotanical evidence for field crop plants in early historic Ireland. In J.
Renfrew (ed.), New light on early farming: recent developments in palaeoethnobotany, 315–28. Edinburgh.
Monk, M. 2011 Oats: the superfood of early medieval Ireland. Archaeology Ireland 25 (1), 36–9.
Monk, M., Tierney, J. and Hannon, M. 1998 Archaeobotanical studies and early medieval Munster. In
M. Monk and J. Sheehan (eds), Early medieval Munster: archaeology, history and society, 65–75. Cork.
Moore, M. 1984 Irish cresset-stones. JRSAI 114, 98–116.
Moore, P.D., Webb, J.A. and Collinson, M.E. 1991 Pollen analysis. London.
Moore, R. 1997 More notes on the Coleoptera of the Island of Raasay, Scotland. Entomologist’s Monthly
Magazine 133, 179–80.
Moorrees, C.F.A., Fanning, E.A. and Hunt, E.E. Jr 1963a Age variation of formation stages for ten
permanent teeth. Journal of Dental Research 42 (6), 1490–502.
Moorrees, C.F.A., Fanning, E.A. and Hunt, E.E. Jr 1963b Formation and resorption of three deciduous
teeth in children. American Journal of Physical Anthropology 21, 205–13.
Morales, A. and Rosenlund, K. 1979 Fish bone measurements. Copenhagen.
Morris, M.G. 1997 Broad-nosed weevils. Coleoptera: Curculionidae (Entiminae). Handbook for the
Identification of British Insects 5, Part 17a. London.
Morris, M.G. 2002 True weevils, Part 1. Coleoptera: Curculionidae (subfamilies Raymondionyminae to
Smicronychinae). Handbooks for the Identification of British Insects 5, Part 17b. London.
Murray, E. and McCormick, F. 2005 Environmental analysis and the food supply. In J.W. Marshall and
C. Walsh (eds), Illaunloughan Island: an early medieval monastery in County Kerry, 67–81. Bray.
Murray, E., McCormick, F. and Plunkett, G. 2004 The food economies of Atlantic island monasteries:
the documentary and archaeo-environmental evidence. Environmental Archaeology 9 (2), 179–88.
Murray, H.J.R. 1951 A history of board-games other than chess. Oxford.
Nicol, D.M. 1963 Meteora: the rock monasteries of Thessaly. London.
Ní Mhaonaigh, M. 1996 ‘Cogad Gáedel Re Gallaib’ and the Annals: a comparison. Ériu 47, 101–26.
Oakley, G.E. and Hall, A.D. 1979 The spindle whorls. In J.H. Williams, St Peter’s St., Northampton:
excavations 1973–1976, 286–9. Northampton Development Corporation Archaeological Monograph
2. Northampton.
O’Brien, M.A. 1976 Corpus genealogiarum Hiberniae,Vol. 1. Dublin.
Ó Buachalla, L. 1951 Contributions towards the political history of Munster, 450–800 AD. JCHAS 56,
87–90.
Ó Buachalla, L. 1952 Contributions towards the political history of Munster, 450–800 AD. JCHAS 57,
67–86.
Ó Cadhla, O., Strong, D., O’Keeffe, C. et al. 2008 An assessment of the breeding population of grey seals in
the Republic of Ireland, 2005. NPWS (http://www.npws.ie/en/PublicationsLiterature/
IrishWildlifeManuals/2006-2008/, accessed 8 February 2011).
Ó Carragáin,T. 2008 Preliminary landscape analysis. In T. Ó Carragáin and J. Sheehan, Making Christian
landscapes: settlement, society and regionality in early medieval Ireland—Phase 1. Report by the Dept of
Archaeology, UCC, for the Heritage Council—Irish National Strategic Archaeological Research
(INSTAR) Programme.
Ó Carragáin, T. 2010 Churches in early medieval Ireland. New Haven and London.
Ó Carragáin, T. and Sheehan, J. 2008 Making Christian landscapes: settlement, society and regionality in early
medieval Ireland—Phase 1. Report by the Dept of Archaeology, UCC, for the Heritage Council—
Irish National Strategic Archaeological Research (INSTAR) Programme.
Ó Carragáin, T. and Sheehan, J. 2009 Making Christian landscapes: settlement, society and regionality in early
medieval Ireland—Phase 2. Report by the Dept of Archaeology, UCC, for the Heritage Council—
478
BIBLIOGRAPHY
Irish National Strategic Archaeological Research (INSTAR) Programme.
O’Carroll, E. 2005 Charcoal analysis. In J.W. Marshall and C. Walsh (eds), Illaunloughan Island: an early
medieval monastery in County Kerry, 29. Bray.
O’Connor, L. 1991 Irish Iron Age and Early Christian whetstones. JRSAI 121, 45–76.
Ó Corráin, D. 1999 Vikings IV: is sceillec Old Norse? Peritia 13, 310–11.
Ó Crohan, T. 1978 The Islandman (trans. R. Flower). Oxford.
Ó Cróinín, D. 2005 Ireland, 400–800. In D. Ó Cróinín (ed.), A new history of Ireland.Volume 1: prehistoric
and early historic Ireland, 182–234. Oxford.
Ó Danachair, C. 1960 The holy wells of Corkaguiney, Co. Kerry. JRSAI 90 (1), 67–78.
O’Donnell, L. 2005 Analysis of the charcoal from Kilgobbin (04E1373), Co. Dublin. Unpublished report
for Margaret Gowen and Co. Ltd.
O’Donnell, L. 2007 The wood and charcoal. In E. Grogan, L. O’Donnell and P. Johnston (eds), The
Bronze Age landscapes of the Pipeline to the West: an integrated archaeological and environmental assessment,
27–69. Bray.
O’Donovan, J. 1851 Annals of the Four Masters (7 vols). Dublin.
Oftedal, M. 1976 Scandinavian place-names in Ireland. In B. Almqvist and D. Greene (eds), Proceedings
of the Seventh Viking Congress, Dublin, 15–21 August 1973, 125–33. Dublin.
O’Keeffe, J.G. 1931 Dál Caladbruig. In J. Fraser, P. Grosjean and J.G. O’Keeffe (eds), Irish texts, 19–21.
London.
O’Kelly, M.J. 1959 Church Island near Valencia, Co. Kerry. PRIA 59C, 57–136.
Oman, C. 1957 English church plate, 597–1830. London.
Oman, C. 1990 Chalices and patens: background and typology. In M. Biddle (ed.), Object and economy
in medieval Winchester, 789–91. Winchester Studies 7(ii). Oxford.
O’Meara, J.J. 1982 Giraldus Cambrensis: the history and topography of Ireland (revised edn). Mountrath.
OPW 2008 Skellig Michael World Heritage Site Management Plan 2008–2018 (http://www.environ.ie/
en/Heritage/WorldHeritage/SkelligMichael/ManagementandManagementPlan).
O’Rahilly, C. 1962 Cath Finntrágha. Dublin.
O’Rahilly, C. 1998 A classification of bronze stick pins from the Dublin excavations 1962–72. In C.
Manning (ed.), Dublin and beyond the Pale: studies in honour of Patrick Healy, 23–33. Bray.
Ó Riain, P. 1985 Corpus genealogiarum sanctorum Hiberniae. Dublin.
Ó Riain, P. 1990 The Tallaght martyrologies redated. Cambridge Medieval Celtic Studies 20, 21–38.
Ó Riain, P. 2009 Fionán of Iveragh. In J. Crowley and J. Sheehan (eds), The Iveragh Peninsula. A cultural
atlas of the Ring of Kerry, 126–8. Cork.
Ó Riain-Raedel, D. 2009 Skellig Michael: the German connection. In J. Crowley and J. Sheehan (eds),
The Iveragh Peninsula. A cultural atlas of the Ring of Kerry, 136–7. Cork.
Orme, B.J. and Coles, J.M. 1985 Prehistoric woodworking from the Somerset levels 2: species selection
and prehistoric woodlands. Somerset Levels Papers 11, 7–24.
Orser, C. 2006 Unearthing hidden Ireland. Bray.
Ortner, D.J. and Putschar, W.G.J. 1981 Identification of pathological conditions in human skeletal remains.
Washington.
O’Sullivan, A. and Sheehan, J. 1996 The Iveragh peninsula. An archaeological survey of south Kerry. Cork.
O’Sullivan, J. and Ó Carragáin, T. 2008 Inishmurray: monks and pilgrims in an Atlantic landscape, Vol. 1.
Archaeological survey and excavations 1997–2000. Cork.
O’Sullivan, M. 1983 Twenty years a-growing (translated by M.L. Davies). Oxford.
O’Sullivan, T. 2005 Bird bones. In J.W. Marshall and C. Walsh, Illaunloughaun Island: an early medieval
monastery in County Kerry (ftp://clayton.hostireland.com/www_old/ftp/Wordwell/
Illaunloughan/Appendices/.
479
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Overland, A. and O’Connell, M. 2008 Fine-spatial palaeoecological investigations towards reconstructing
late Holocene environmental change, landscape evolution and farming activity in Barrees, Beara
Peninsula, southwestern Ireland. Journal of the North Atlantic 1, 37–73.
Overland, A. and O’Connell, M. 2011 New insights into late Holocene farming and woodland dynamics
in western Ireland with particular reference to the early medieval horizontal watermill at Kilbegly,
Co. Roscommon. Review of Palaeobotany and Palynology 163, 205–26.
Parker Pearson, N., Sharples, N. and Symonds, J. 2004 South Uist: archaeology and history of a Hebridean
island. Stroud.
Payne, S. 1969 The metrical distinction between sheep and goat metacarpals. In P.J. Ucko and G.W.
Dimbleby (eds), The domestication and exploitation of plants and animals, 295–303. London.
Payne, S. 1973 Kill-off patterns in sheep and goats: the mandibles from A van Kale. Anatolian Studies 23,
281–303.
Payne, S. 1985 Morphological distinctions between the mandibular cheek teeth of young sheep, Ovis,
and goats, Capra. Journal of Archaeological Science 12, 139–47.
Payne, S. 1987 Reference codes for wear states in the mandibular cheek teeth of sheep and goat. Journal
of Archaeological Science 14, 609–14.
Plummer, C. 1910 Vitae sanctorum Hiberniae (2 vols). Oxford.
Plummer, C. 1922 Brethada na náem nÉrenn: Lives of Irish saints (2 vols). Oxford.
Ponsford, M. 1991 Dendrochronological dates from Dundas Wharf, Bristol, and the dating of Ham
Green and other medieval pottery. In E. Lewis (ed.), Custom and ceramics, 81–103. Wickham.
Power, C. 1994 A demographic study of the human skeletal populations from historic Munster. UJA
57, 95–118.
Quin, E.G. (ed.) 1953 Contributions to a dictionary of the Irish language. Fascicle S. Dublin.
Reilly, E. 2003 The contribution of insect remains to an understanding of the environment of Vikingage and medieval Dublin. In S. Duffy (ed.), Medieval Dublin IV, 40–63. Dublin.
Reilly, E. 2006 Analysis of insect remains from Roestown 2, M3, Co. Meath. Unpublished technical
report for Archaeological Consultancy Services Ltd on behalf of Meath County Council.
Reilly, E. 2008 Understanding Late Holocene woodland dynamics in southwestern and western Ireland
through the analysis of sub-fossil insect remains. Unpublished Ph.D thesis, University of Dublin,
Trinity College.
Reilly, E. (forthcoming) Fair and foul: analysis of sub-fossil insect remains from Troitsky XI–XIII,
medieval Novgorod excavations 1996–2002. In M. Brisbane, N. Makarov and E. Nosov, The
archaeology of medieval Novgorod in its wider context: a study of centre/periphery relations. Oxford.
Reimer, P.J., Baillie, M., Bard, E. et al. 2009 IntCal09 and marine09 radiocarbon age calibration curves,
0–5000 cal. BP. Radiocarbon 51, 1111–50.
Resnick, D. and Niwayama, G. 1988 Diagnosis of bone and joint disorders (2nd edn). Philadelphia.
Roberts, C. and Manchester, K. 1995 The archaeology of disease (2nd edn). New York.
Roe, H. 1976 The cult of St Michael in Ireland. In C. Ó Danachair (ed.), Folk and farm, 251–64. Dublin.
Rogers, J. 2000 The palaeopathology of joint disease. In M. Cox and S. Mays (eds), Human osteology in
archaeology and forensic science, 163–82. London.
Rogers, J. and Waldron, T. 1995 A field guide to joint disease in archaeology. Chichester.
Ryan, M. 1983 The Derrynaflan Hoard: a preliminary account. Dublin.
Sadler, J.P. 1990 Beetles, boats and biogeography: insect invaders of the North Atlantic islands. Acta
Archaeologia 61, 199–212.
Sadler, J.P. 1999 Biodiversity on oceanic islands: a palaeoecological assessment. Journal of Biogeography 26
(1), 75–87.
Scally, G. 1999 The early monastery of High Island. Archaeology Ireland 13 (1), 24–8.
480
BIBLIOGRAPHY
Scally, G. (forthcoming) High Island—Ard Oiléan, Connemara, Co. Galway: excavation of an early medieval
monastery on an Atlantic Irish island.
Scheuer, J.L., Musgrave, J.H. and Evans, S.P. 1980 The estimation of late fetal and perinatal age from
limb bone length by linear and logarithmic regression. Annals of Human Biology 7 (3), 257–65.
Schwartz, J.H. 1995 Skeleton keys. An introduction to human skeletal morphology, development and analysis.
Oxford.
Schweingruber, F.H. 1978 Microscopic wood anatomy. Birmensdorf.
Scully, O.M.B. 2007 The non-ferrous artefacts. In M. Clyne, Kells Priory, Co. Kilkenny: archaeological
excavations by T. Fanning and M. Clyne, 383–405. Archaeological Monograph Series 3. Dublin.
Serjeantson, D. 2009 Birds. Cambridge Manuals in Archaeology. Cambridge.
Serjeantson, D., Irving, B. and Hamilton-Dyer, S. 1993 Bird bone taphonomy from the inside out: the
evidence of gull predation on the Manx shearwater Puffinus puffinus. In A. Morales,
‘Archaeornithology: Birds and the Archaeological Record. Proceedings of the First Meeting of the
ICAZ Bird Working Group, Madrid 1992’. Archaeofauna 2, 191–204.
Sexton, R. 1998 Porridges, gruels and breads: the cereal foodstuffs of early medieval Ireland. In M.
Monk and J. Sheehan (eds), Early medieval Munster: archaeology, history and society, 76–86. Cork.
Sharkey, O. 1985 Old days, old ways. Dublin.
Sheehan, J. 2009 A peacock’s tale: excavations at Caherlehillan, Ivereagh, Ireland. In N. Edwards (ed.),
The archaeology of the early medieval Celtic churches, 191–206. Society for Medieval Archaeology
Monograph 21. Leeds.
S.M. 1913 Skelligs. Kerry Archaeological Magazine 2 (11), 163–72.
Smith, C. 1756 The ancient and present state of the county of Kerry (reprinted 1979). Dublin.
Smith, K.G.V. 1989 An introduction to the immature stages of British flies: Diptera larvae, with notes on eggs,
puparia and pupae. Handbooks for the Identification of British Insects 10, Part 14. London.
Smith, W. and Cheetham, S. (eds) 1827–1908 A dictionary of Christian antiquities. Being a continuation of
the ‘Dictionary of the Bible’ (2 vols). London.
Stokes, W. 1993 The Annals of Tigernach (2 vols; reprinted from Revue Celtique 1895/6). Felinfach.
Stroud, G. and Kemp, R.L. 1993 Cemeteries of the Church and Priory of St Andrew, Fishergate.York.
Stuart-Macadam, P. 1991 Anaemia in Roman Britain: Poundbury Camp. In H. Bush and M. Zvelebil
(eds), Health in past societies. Biocultural interpretations of human skeletal remains in archaeological context,
101–13. British Archaeological Reports, British Series 567. Oxford.
Stuijts, I. 2005 Wood and charcoal identification. In M. Gowen, J. Ó Neill and M. Philips (eds), The
Lisheen Mine Archaeological Project 1996–8, 137–86. Dublin.
Stuiver, M. and Reimer, P.J. 1993 Extended C14 database and revised CALIB 3.0 C14 age calibration
programme. Radiocarbon 35, 215–30.
Sweetman, H.S. and Handcock, G.F. 1886 Calendar of documents relating to Ireland, preserved in Her Majesty’s
Public Record Office, London. Vol. 5: 1302–7. London.
Thomas, C. 1971 The Early Christian archaeology of north Britain. London.
Thompson,W. 1856 The natural history of Ireland. Vol. 4: Mammalia, reptiles, fishes and invertebrates. London.
Thomson, R. and Brown, D. 1991 On some earthenware curiosities from the Saintonge. In E. Lewis
(ed.), Custom and ceramics, 62–80. Wickham.
Todd, J.H. 1867 Cogadh Gaedhel re Gallaibh. The war of the Gaedhil with the Gaill. Rerum Britannicarum
Medii Aevi Scriptores No. 48. London.
Toner, G., Fomin, M.,Torma,T. and Bondarenko, G. 2007 Electronic dictionary of the Irish language (eDIL)
(http://www.dil.ie/, accessed 27 July 2010).
Trotter, M. and Glesier, G.L. 1952 Estimation of stature from the long bones of American whites and
negroes. American Journal of Physical Anthropology 10, 463–514.
481
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Turner, D.J. and Orton, C.R. 1979 199 Borough High Street, Southwark: excavations in 1962. Surrey
Archaeological Society Research 7, 1–25.
Ubelaker, D.H. 1989 Human skeletal remains. Excavation, analysis, interpretation (2nd edn). Washington.
Uí Chonchubhair, M. 1998 Flóra Chorca Dhuibhne. Aspects of the flora of Corca Dhuibhne (2nd edn). Baile
an Fheirtéaraigh.
Valente, M. 2008 The Vikings in Ireland. Settlement, trade and urbanisation. Dublin.
Vitousek, P.M. 1988 Diversity and biological invasions of oceanic islands. In E.O. Wilson (ed.),
Biodiversity, 181–9. Washington.
von den Driesch, A. 1976 A guide to the measurement of animal bones from archaeological sites. Peabody
Museum Bulletin 1. Cambridge, Mass.
von den Driesch, A. and Boessneck, J.A. 1974 Kritische Anmerkungen zur Widerristhoherberechnung
aus Langermassen vor- und fruhgeschichtlicher Tierknochen. Saugetierkundliche Mitteilungen 22, 325–
48.
Wakeman, W.F. 1885 Inis Muirdaich, now Inishmurray, and its antiquities. JRSAI (4th ser.) 7, 175–332.
Walker, M.J.C. 1984 A pollen diagram from St Kilda, Outer Hebrides, Scotland. New Phytologist 97,
99–113.
Wanless, S., Murray, S. and Harris, M.P. 2005 The status of northern gannet in Britain and Ireland
2003/4. British Birds 98, 280–94.
Waterston, A.R. 1981 Present knowledge of the non-marine invertebrate fauna of the Outer Hebrides.
Proceedings of the Royal Society of Edinburgh 798, 215–321.
Watts, S. 2006 Rotary querns c. 700–1700. Finds Research Group AD 700–1700 Datasheet 38.York.
Webb, D.A., Parnell, J. and Doogue, D. 1996 An Irish flora (7th edn). Dundalk.
Westropp, T.J. 1897 Cruise in connection with the Munster meeting, a descriptive sketch of places
visited. JRSAI 17, 265–358.
Wheeler, E.A., Bass, P. and Gasson, P.E. 1989 IAWA list of microscopic features for hardwood
identification. IAWA Bulletin 10 (3), 219–332.
Whelan, K. (ed.) 1995 Letters from the Irish highlands of Connemara: the Blake family of Renvyle House
(1823–1824). Clifden.
White, T.D. and Folkens, P.A. 1991 Human osteology. San Diego.
482
APPENDIX I—FINDS CATALOGUE
483
No.
Type
Description
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
1
2
3
4
5
6
Pottery
Pottery
Pottery
Clay pipe
Clay pipe
Pottery, medieval
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Pottery, medieval
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery, unglazed
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Area
Location
Context
Ham Green A/B transitional ware LMG
Ham Green A/B transitional ware LMG
Saintonge
LMG
LMG
LMG
Local medieval ware
LMG
Layer C, Cutting 5
Layer C, Cutting 5
Layer C/E, Cutting 5
Layer B, Cutting 7
Layer B, Cutting 6
Top of steps, Cutting 6
Local medieval ware
Transfer-printed
Transfer-printed
Transfer-printed
White earthenware
White earthenware
White earthenware
White earthenware
Earthenware (Staffordshire)
Banded ware
Porcelain?
Ham Green A/B transitional ware
White earthenware
Shell-edged (green)
Transfer-printed
White earthenware
Transfer-printed
White earthenware
Earthenware (Staffordshire)
Banded ware
Earthenware (Staffordshire)
Pearlware
White earthenware
Banded ware
Surface find
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Layer B, Cutting 6
Main fill
Layer A, Cuttings 6 & 7
Layer A, Cuttings 6 & 7
Layer A, Cuttings 6 & 7
Stones over skeletal material
Layer B, Cutting 5
Layer B, Cutting 5
Layer B, Cutting 5
Layer B, Cutting 5
Layer B, Cutting 5
Topsoil, Cutting 7
Topsoil, Cutting 7
Topsoil, Cutting 7
F633(5)
F633(5)
F650(5)
F632(7)
F632(6)
Top of steps,
Cutting 6
Surface find
F632(6)
F632(6)
F632(6)
F632(6)
F632(6)
F632(6)
F632(6)
F632(6)
F632(6)
F632(6)
F615
F631(6&7)
F631(6&7)
F631(6&7)
F562
F632(5)
F632(5)
F632(5)
F632(5)
F632(5)
F631(7)
F631(7)
F631(7)
East steps
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
Cistern 3
LMG
LMG
LMG
Leacht area
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
484
E No.
No.
Type
Description
Area
Location
Context
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
31
32
33
34
35
36
Pottery
Pottery, medieval
Pottery, medieval
Pottery, medieval
Pottery, medieval
Pottery
Black-glazed ware
Orléans-type ware
Ham Green A/B transitional ware
Local medieval ware
Local medieval ware
19th-century stoneware
LMG
LMG
LMG
LMG
LMG
Leacht area
F631(7)
F635(6)
F637
F635(6)
F1014
F561
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Clay pipe bowl
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
c. 1800–30
Shell-edged (blue)
White earthenware
White earthenware
Shell-edged (blue)
White earthenware
Banded ware
Banded ware
Unidentified
White earthenware
Banded ware
Glazed red earthenware
19th-century stoneware
White earthenware
Transfer-printed
White earthenware
White earthenware
White earthenware
19th-century stoneware
White earthenware
White earthenware
White earthenware
White earthenware
Banded ware
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
Leacht area
Leacht area
Leacht area
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
Topsoil, Cutting 7
4th level, Cutting 6
N/S wall, Cutting 5
Layer E, Cutting 6
Inside Cell G, upper level
South of large oratory, between
leacht and wall
Cell G, upper levels
Layer B, Cutting 7
Layer B, Cutting 7
Layer B, Cutting 7
Campion above and west of Cell G
Campion above and west of Cell G
Campion above and west of Cell G
Campion above and west of Cell G
Campion above and west of Cell G
Campion above and west of Cell G
Campion above and west of Cell G
Campion above and west of Cell G
Campion above and west of Cell G
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
East of Cell G
East of Cell G
East of Cell G
Just inside Cell G
Just inside Cell G
Just inside Cell G
Just inside Cell G
Just inside Cell G
F1014
F632(7)
F632(7)
F632(7)
F1012
F1012
F1012
F1012
F1012
F1012
F1012
F1012
F1012
F562
F562
F562
F1013
F1013
F1013
F1014
F1014
F1014
F1014
F1014
APPENDICES
485
E No.
No.
Type
Description
Area
Location
Context
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Clay pipe stem
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Clay pipe bowl
Pottery
Shell-edged (blue)
Shell-edged (blue)
Shell-edged (blue)
Shell-edged
Shell-edged
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
Just inside Cell G
Just inside Cell G
Just inside Cell G
E and N of Cell G
E and N of Cell G
E and N of Cell G
E and N of Cell G
E and N of Cell G
E and N of Cell G
E and N of Cell G
E and N of Cell G
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
East end in campion, upper levels
F1014
F1014
F1014
F1013
F1013
F1013
F1013
F1013
F1013
F1013
F1013
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
F1011
Pearlware
Pearlware
White earthenware
Transfer-printed
Transfer-printed
Transfer-printed
Transfer-printed
Transfer-printed
Transfer-printed
Transfer-printed
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
c. 1800–30
19th-century stoneware
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
486
E No.
No.
Type
Description
Area
Location
Context
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery, medieval
Pottery, medieval
Pottery, medieval
Mocha ware
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
Transfer-printed
Black-glazed ware
19th-century stoneware
White earthenware
Local medieval ware
Local medieval ware
Saintonge green-glazed
LMG
Cistern 3
Cistern 3
Cistern 3
Cistern 3
Cistern 3
Cistern 3
Cistern 3
Cistern 3
Cistern 3
Cistern 3
LMG
LMG
LMG
LMG
LMG
East end in campion, upper levels
Main fill
Main fill
Main fill
Main fill
Main fill
Main fill
Main fill
Main fill
Main fill
Main fill
In Cell G
Layer B, Cutting 7
Layer E, Cuttings 6 & 7
Layer E, Cuttings 6 & 7
Cleaning of east steps (F10?)
93E0195:
108
Pottery
North Devon gravel-free
LMG
Cleaning of east steps (F10?)
93E0195:
109
Pottery
North Devon gravel-free
LMG
Cleaning of east steps (F10?)
93E0195:
93E0195:
110
111
Pottery, medieval
Pottery
Ham Green B ware
White earthenware
LMG
Steps, cleaning
Spoil heap, soil derived from LMG
Unknown
93E0195:
112
Pottery
White earthenware
Steps, cleaning
Unknown
93E0195:
113
Pottery
White earthenware
Steps, cleaning
Unknown
93E0195:
114
Pottery
Shell-edged (blue)
Steps, cleaning
Unknown
93E0195:
115
Pottery
Transfer-printed
Steps, cleaning
Unknown
F1011
F615
F615
F615
F615
F615
F615
F615
F615
F615
F615
F1015
F632(7)
F635(6&7)
F635(6&7)
Not from
excavation
Not from
excavation
Not from
excavation
Spoil heap
Not from
excavation
Not from
excavation
Not from
excavation
Not from
excavation
Not from
excavation
APPENDICES
487
E No.
No.
Type
Description
Area
Location
Context
93E0195:
116
Pottery
Transfer-printed
Steps, cleaning
Unknown
93E0195:
117
Clay pipe stem
Steps, cleaning
Unknown
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
Pottery
Pottery
Pottery
Clay pipe stem
Clay pipe stem
Pottery
Clay pipe stem
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Not from
excavation
Not from
excavation
F504
Surface find
F1016
F1016
F1016
F1017
F1017
F1017
F1017
F1017
F1017
F1017
F1017
F1018
F1018
F1018
F1018
F1018
F1018
F1018
F1018
F1018
F1018
F598
F598
F598
Not from
excavation
White earthenware
Black-glazed ware
19th-century stoneware
Shell-edged (blue)
Banded ware
Hand-painted fine ware
19th-century stoneware
Glazed red earthenware
Glazed red earthenware
Glazed red earthenware
Transfer-printed
White earthenware
White earthenware
White earthenware
White earthenware
White earthenware
Transfer-printed
White earthenware
Shell-edged (green)
Shell-edged (green)
White earthenware
Unidentified
Transfer-printed
19th-century stoneware
East entrance
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
Leacht area
Leacht area
Leacht area
Wailing Woman
Found on surface
Just outside Cell G to west
Just outside Cell G to west
Just outside Cell G to west
Fill of Cell G
Fill of Cell G
Fill of Cell G
Fill of Cell G
Fill of Cell G
Fill of Cell G
Fill of Cell G
Fill of Cell G
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Cell G, dark layer above pavement
Puffin burrow
Puffin burrow
Puffin burrow
Stray find on surface
(not from excavation)
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
488
E No.
No.
Type
Description
Area
Location
Context
93E0195:
145
Pottery
19th-century stoneware
Wailing Woman
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
146
147
148
149
150
151
152
153
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass bottle
Glass bottle
Glass bottle
Glass bottle
Not from
excavation
F598
F1016
F1013
F1013
F1013
F1013
F562
F562
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
Glass bottle
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Leacht area
LMG
LMG
LMG
LMG
LMG
Neck sherd, post-1820
Leacht area
Sherds from bottle made pre-1800 Leacht area
but sherds worn
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leachtarea
LMG
Stray find on surface
(not from excavation)
Puffin burrow
Just outside Cell G to west
East of Cell G
East of Cell G
East of Cell G
East of Cell G
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
19th-century disturbance SW of leacht
In N/S wall, Cutting 5E
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F562
F637(5E)
APPENDICES
489
E No.
No.
Type
93E0195:
174
93E0195:
Description
Area
Location
Glass, modern
Leacht area
175
Glass, modern
Leacht area
93E0195:
176
Glass bottle
From among stones covering skeletal
F563
material (F2?)
From among stones covering skeletal
F563
material (F2?)
Outside east wall of St Michaels Church (F6?) F562
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Glass, modern
Pottery
Pottery
Pottery
Pottery
Pottery, medieval
Pottery, medieval
Pottery
Pottery, medieval
Pottery, medieval
Pottery
Sherds from bottle made pre-1800 Leacht area
but sherds worn
Leacht area
Leach tarea
Leacht area
LMG
LMG
LMG
LMG
LMG
LMG
East entrance
LMG
?LMG?
?LMG?
?LMG?
?LMG?
Shell-edged
?LMG?
Shell-edged
?LMG?
White earthenware
?LMG?
White earthenware
?LMG?
Saintonge green-glazed
?LMG?
Ham Green A ware
?LMG?
White earthenware
?LMG?
Local medieval ware
?LMG?
Orléans-type ware
?LMG?
White earthenware
?LMG?
Context
Outside east wall of St Michaels Church (F6?) F562
Outside east wall of St Michaels Church (F6?) F562
Outside east wall of St Michaels Church (F6?) F562
Top layer, Cuttings 6 & 7
F631(6&7)
Top layer, Cuttings 6 & 7
F631(6&7)
Inside Cell G, upper level
F1014
Inside Cell G, upper level
F1014
Inside Cell G, upper level
F1014
Inside Cell G, upper level
F1014
F504
East end in campion, upper levels
F631
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Fire-damaged (from Cell G?)
F1019
Fire-damaged (from Cell G?)
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
Very badly damaged in fire in site hut
F1019
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
490
E No.
E No.
No.
Type
Description
93E0195:
93E0195:
93E0195:
202
203
204
Pottery, medieval
Pottery
Pottery
Saintonge green-glazed
?LMG?
Unidentified
?LMG?
Glazed earthenware; may have
?LMG?
been reused as a gaming piece
Rounded spatulate head, decorated LMG
with crosses, c. 1130–1200
c. 1900–59
LMG
205
Copper-alloy
stick-pin
Brass and wood
crucifix
Writing lead?
93E0195:
206
93E0195:
207
93E0195:
208
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
209
210
211
212
213
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
214
215
216
217
218
219
Slag
Slag
Slag
Slag
Slag
Iron
93E0195:
93E0195:
220
221
Fused metal
Slag
Possibly a writing lead; possibly
13th-century
Bent, undiagnostic, found with
bottle glass
Uncertain function and date
Copper or
bronze wire
Iron rod
Metal fragments
Iron nail
Iron strip
Copper-alloy ringed pin Crutch-headed with moulded
cross decoration, c. 1000–70
Leacht area
LMG
LMG
Leacht area
Leacht area
Leacht area
South entrance
491
LMG
LMG
LMG
LMG
LMG
Thin-walled, possibly from a can; LMG
19th-century or later
LMG
With possible hammerscale
LMG
attached
Location
Context
Very badly damaged in fire in site hut
Very badly damaged in fire in site hut
Very badly damaged in fire in site hut
F1019
F1019
Cutting 3 in lower collapse c. 2m
from wall
Under campion roots, directly below
large oratory in LMG
From clearance to S of leacht and N
of wall
Layer C, Cutting 5E
Layer D, Cutting 5E
East of St Michaels Church
East of St Michaels Church
19th-century disturbance SW of leacht
Pre-excavation find, found by workmen
in the west jamb of
the south entrance
Layer E, Cutting 5E (00:25)
East end in campion, upper levels
Layer D, Cutting 5 (00:24)
Layer C/D, Cutting 5 (00:26)
Layer E, Cuttings 6 & 7 (00:39)
Layer D, Cutting 5 surface (00:44)
Layer E, Cutting 5 (00:27)
Layer E, Cutting 6 (00:28)
F1019
F636(3)
F631
F561
F633(5E)
F634(5E)
F579
F577
F562
Pre-excavation
find
F635(5E)
F1011
F634(5)
F639(5)
F635(6&7)
F634(5)
F635(5)
F635(6)
APPENDICES
93E0195:
Area
No.
Type
Description
Area
Location
Context
93E0195:
93E0195:
93E0195:
222
223
224
Iron
Iron fused onto stone
Metal fused onto stone
93E0195:
225
Metal fused onto stone
East step structure
Possible hinge strap
LMG
Thin-walled, possibly from a can; Leacht area
19th-century or later
Leacht area
From wall of structure
Layer E, Cutting 7 (00:41)
From outside St Michaels Church (F7?)
F693
F635(7)
F561
F594
Leacht area
LMG
LMG
LMG
LMG
LMG
LMG
LMG
Leacht area
East steps
E of St Michaels Church, in cavity
under leacht
Layer E, Cutting 5, overlying
new steps (00:29)
Puffin burrow
Layer D, Cutting 5E
Layer D, Cutting 5 (00:21)
Layer F, Cutting 5
Layer F, Cutting 5
Area E of Cell G
Area E of Cell G
Just inside Cell G upper levels
19th-century disturbance SW of leacht
From cleaning steps
93E0195:
226
Iron fused onto stone
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
227
228
229
230
231
232
233
234
235
236
Bone, animal/bird
Antler gaming piece
Charred seed
Fish bone/bead
Fish-bone pin
Oyster shell
Coal?
Coal?
Oyster shell
Wooden object
93E0195:
237
Bone fragment, triangular
East steps
From cleaning steps
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
238
239
240
241
242
243
244
245
Leather stitched object
Brick fragment
Brick
Mortar
Mortar
Mortar?
Mortar?
Leather object
Cistern 3
LMG
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
LMG
246
247
Stone (lighthouse?)
Stone, red
Main fill
Cell G, upper levels
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Organic layer below campion,
below Cell G
Puffin burrow
Puffin burrow
93E0195:
93E0195:
LMG
Date uncertain; EC or medieval
Date uncertain; EC or medieval
Possible shoe upper
Leacht area
Leacht area
F635(5)
F598
F634(5E)
F634(5)
F636(5)
F636(5)
F1013
F1013
F1014
F562
Not from
excavation
Not from
excavation
F615
F1014
F598
F598
F598
F598
F598
F1021
F598
F598
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
492
E No.
No.
Type
93E0195:
93E0195:
93E0195:
248
249
250
Stone, red
Water-rolled pebble
Whetstone fragment
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
251
252
253
254
255
256
257
258
259
260
261
262
263
Stone disc
Stone dish/plate
Water-rolled pebble
Water-rolled pebble
Flint nodule
Rock crystal
Quartz (milky)
Water-rolled pebble
Water-rolled pebble
Water-rolled pebble
Water-rolled pebble
Water-rolled pebble
Spindle-whorl
93E0195:
264
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
265
266
267
268
269
270
271
272
273
93E0195:
274
Net-weight/
spindle-whorl
Hone stone?
Water-rolled pebble
Water-rolled pebble
Water-rolled pebble
Water-rolled pebble
Water-rolled pebble
Water-rolled pebble 7
Water-rolled pebble 5
? Two incompatible
descriptions
Water-rolled pebble
Description
One face and edge; fine
micaceous sandstone, Skellig
Complete discoid; hourglass
perforation
Hourglass perforation, slightly
off-centre
Area
Location
Context
Leacht area
Leacht area
LMG
Puffin burrow
Puffin burrow
Layer E, Cutting 5
F598
F598
F635(5)
Leacht area
LMG
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Leacht area
Spoil by
lighthouse toilet
East entrance
Stony layer between leacht and S wall
Boulder layer, Cuttings 3 and 1
West of broken cross base
Burnt area SE of leacht
Within leacht
Burnt area SE of leacht
Dark stony layer S of leacht, N of S wall
Loose dark stony layer
Loose dark stony layer
Loose dark stony layer
Loose dark stony layer
Origin LMG
F577
F638(1&3)
F575
F576
F590
F576
F577
F579
F577
F577
F577
F577
Spoil heap
Uppermost cut
F503
Layer B, Cutting 7 (00:14)
Loose dark stony layer 00:10
Loose dark stony layer 00:08
Prayer station
Prayer station
Prayer station
E of St Michaels Church
Mortar and soil layer above Skeleton 3
Surface find E of St Michaels Church
around leacht
Surface find E of St Michaels Church
around leacht
F632(7)
F577
F577
LMG
Leacht area
Leacht area
South Peak
South Peak
South Peak
Leacht area
Leacht area
Leacht area
Leacht area
F588
F566
F561
F561
APPENDICES
493
E No.
No.
Type
93E0195:
275
93E0195:
Description
Area
Location
Context
Water-rolled pebble
Leacht area
F561
276
Water-rolled pebble
Leacht area
93E0195:
277
Water-rolled pebble
Leacht area
93E0195:
278
Water-rolled pebble
Leacht area
93E0195:
279
Water-rolled pebble
Leacht area
93E0195:
280
Water-rolled pebble
Leacht area
93E0195:
281
Water-rolled pebble
Leachtarea
93E0195:
282
Cross fragment
Surface find E of St Michaels Church
around leacht
Surface find E of St Michaels Church
around leacht
Surface find E of St Michaels Church
around leacht
Surface find E of St Michaels Church
around leacht
Surface find E of St Michaels Church
around leacht
Surface find E of St Michaels Church
around leacht
Surface find E of St Michaels Church
around leacht
Layer E, Cutting 7 (00:32)
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
283
284
285
286
287
288
289
Roof slate
Roof slate
Cross fragment
Cross fragment
Cross fragment
Cross fragment
Slate (cross-inscribed)
93E0195:
290
93E0195:
93E0195:
93E0195:
291
292
293
Small cross, simple squared arms;
Valentia slate
LMG
LMG
East entrance
Grey sandstone, Skellig
None
None
Simple squared arms
None
Rounded expanding arms; Skellig? None
Simple cross roughly scratched
Leacht area
into surface;Valentia?
Quernstone fragment
Two concentric bands pecked
LMG
into surface; fine green
conglomerate, Reenadrolaun,
Valentia
Cross-incised stone
Linear cross pecked into front face LMG
Stone object (sandstone)
LMG
Water-rolled pebble
LMG
Layer D, Cutting 5
F561
F561
F561
F561
F561
F561
F635(7)
From spoil heap
From spoil heap
From spoil heap
From spoil heap
Dark stony layer S of leacht, N of S wall
F634(5)
F504
Spoil heap
Spoil heap
Spoil heap
Spoil heap
F577
Surface find
Surface find
Found in stone stack
Layer B, Cutting 6 (S47)
Layer B, Cutting 6 (S47)
Stone stack
F632(6)
F632(6)
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
494
E No.
E No.
No.
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
294
295
296
297
298
299
300
301
302
303
304
305
305B
305D
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
Description
Roof slate
Stone, non-local
Stone, non-local
Stone, pinkish
Stone, granite
Stone and mortar
Stone and burnt mortar
Red brick
Yellow sandstone
Pottery, medieval
Glass
Bottle
Roof slate
Stone cross fragment
Simple squared arms
Slate
Incised linear band and
cross-hatching
306 Roof slates (32)
306K Incised slate
Lightly incised with crosshatched band
307 Stone cross fragment
Simple squared arms
308 Stone cross fragment
Rhyolite, Skellig
309 Animal tooth
310 Slag
311 Ceramic furnace lining Possible Early Christian smithing
312 Quernstone fragment
Domed top, concave underside;
tapering handle hole; pecked radial
lines at handle hole; fine red
conglomerate, Reenadrolaun,
Valentia
313 Pottery
White earthenware
314 Pottery
Banded ware (cats-eye)
315 Pottery
White earthenware
Area
Location
Context
Leacht area
East entrance
East entrance
LMG
LMG
LMG
LMG
East entrance
East entrance
East entrance
East entrance
Leacht area
Leacht area
Leacht area
Within leacht over stone 88 (S35)
F560
Taking down wall above lintel to entranceF500
Taking down wall above lintel to entranceF500
West end in campion
F631
West end in campion
F631
Layer D, Cutting 5 (S53)
F634(5)
Layer D, Cutting 5 (S52)
F634(5)
Taking down wall above lintel to entranceF500
Taking down wall above lintel to entranceF500
Taking down wall above lintel to entranceF500
Taking down wall above lintel to entranceF500
E of St Michaels Church (S36)
F561
E of St Michaels Church (S36)
F561
E of St Michaels Church (S36)
F561
Leacht area
Loose dark stony layer
East steps structure Beneath doorway of structure
F577
F694
LMG
East steps structure
Stray find
LMG
LMG
Inner enclosure
In campion
Beneath doorway of structure
From 1994
In campion above Cell G and to W
Outside Cell G to E and N
In pathway outside Cell B
F631
F694
Stray find
F1012
F1020
Lifted from
pavement
for safety
South entrance
LMG
LMG
High in stone blocking
Cutting 3 inside outermost wall
Cutting 3 inside outermost wall
F500
F631(3)
F631(3)
APPENDICES
495
93E0195:
93E0195:
93E0195:
Type
No.
Type
Description
Area
Location
Context
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Iron
White earthenware
Shell-edged (green)
Banded ware
Transfer-printed
White earthenware
Transfer-printed
Transfer-printed
White earthenware
White earthenware
White earthenware
White earthenware
Banded ware
White earthenware
Banded ware
Hand-painted fine ware
White earthenware
19th-century stoneware
Object
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
F631(3)
F631(3)
F632(1A)
F632(1A)
F632(1A)
F632(1A)
F632(1A)
F632(1A)
F632(1A)
F631(3)
F631(3)
F631(3)
F631(3)
F631(3)
F631(3)
F631(3)
F631(3)
F638(1E)
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
93E0195:
334
335
336
337
338
339
340
341
342
343
344
345
Slag
Rubber/hammer stone
Iron cauldron sherd
Glass
Glass bottle
Glass
Clay pipe bowl
Glass bead
Furnace bottom
Glass vessel
Glass vessel
Pottery, medieval
Cutting 3 inside outermost wall
Cutting 3 inside outermost wall
Cutting 1A in black organic and wall
Cutting 1A in black organic and wall
Cutting 1A in black organic and wall
Cutting 1A in black organic and wall
Cutting 1A in black organic and wall
Cutting 1A in black organic and wall
Cutting 1A in black organic and wall
Cutting 3
Cutting 3
Cutting 3
Cutting 3
Cutting 3
Cutting 3
Cutting 3
Cutting 3
Cutting 1 extension, lower collapse,
black organic
In collapse of Cell G
Cutting 1A in black organic and wall
Cutting 3
Cutting 3
Cutting 3
Cutting 3
Cutting 3
Cutting 3
In collapse of Cell G
Cutting 1
Cutting 1
Cutting 1 extension, lower collapse
Modified cobble
18th-century or later
Bottle
Letters scratched into surface
Bottle
c. 1800–30
Turquoise, rounded profile
Modern tumbler
Modern tumbler
Saintonge sgraffito
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
F1014
F632(1A)
F631(3)
F631(3)
F631(3)
F631(3)
F630
F631(3)
F1014
F631(1)
F631(1)
638(1E)
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
496
E No.
No.
Type
Description
Area
Location
Context
93E0195:
93E0195:
93E0195:
93E0195:
346
347
348
349
Pottery
Pottery
Bone
Composite
LMG
LMG
LMG
?
Cutting 1 extension, lower collapse F638(1E)
Cutting 1 extension lower collapse F638(1E)
F638(1E)
Unprovenanced
93E0195:
350
Iron
White earthenware
Hand-painted fine ware
Part of sideplate for handle?
Italian bayonet —ScabiolaBaionette, M70, 1870–87
Fragments of a possible object
In campion above lintel
Surface find
93E0195:
93E0195:
93E0195:
93E0195:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
E338:
351
352
353
354
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Pottery
Clay pipe
Flint
Clay pipe
Pottery
Pottery
Pottery
Flint
Bone comb
Flint
Flint
Iron
Whetstone
Creamware sherd
Creamware sherd
Creamware sherd
Creamware sherd
3 sherds shell-edged (blue), 1820–30
2 sherds shell-edged (green), c. 1860
2 rim sherds, white earthenware, plate
Rim sherd, white earthenware, bowl?
Base sherd, white earthenware, bowl?
Base sherd, white earthenware, plate?
Sherd of transfer-printed ware
Sherd hand-painted fineware
Stem, early 19th-century
Flake with secondary retouch
Stem, early 19th-century
Body sherd, transfer-printed ware
Body sherd, unidentified
Base sherd, white earthenware
Small waste flake
Fragment of decorated sideplate
Small nodule
Nodule
Possible part of strap hinge
Subrectangular, wear on all sides
S entrance 1,
inner enclosure
LMG
LMG
LMG
LMG
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
S entrance 2, inner enclosure
SOT ledge
SOT ledge
SOT ledge
SOT ledge
St Michaels Church
SOT ledge
SOT ledge
SOT ledge
SOT ledge
In collapse of Cell G
In collapse of Cell G
In collapse of Cell G
In collapse of Cell G
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 1
Cutting 2
Cutting 2
Cutting 1
Cutting 2
Surface find
Cutting 1
Cutting 2
Cutting 1
Trial-pit 1
F1014
F1014
F1014
F1014
F4
F4
F4
F4
F4
F4
F4
F4
F4
F12
F9
F102
F102
F102
F102
F106
F102
F102
F104
APPENDICES
497
E No.
No.
Type
Description
Area
Location
Context
E338:
21
Pottery
3 sherds transfer-printed saucer
SOT
Cutting 1
E338:
E338:
E338:
22
23
24
Pottery
Coin
Lignite ring
Large oratory
Large oratory
Large oratory
SE quadrant
SE quadrant
SE quadrant
E338:
E338:
E338:
E338:
E338:
25
26
27
28
29
Flint
Flint
Stone cross fragment
Lignite bead
Leather shoe
F302
F219
F219
F306
F1
30
Knife
Large oratory
SOT
SOT
Large oratory
S entrance 2, inner
enclosure
Large oratory
SE quadrant
Cutting 2
Cutting 2
SE quadrant
Found in wall
E338:
SE quadrant
F306
E338:
E338:
E338:
31
32
33
Flint
Flint
Flint
9 sherds shell-edged (blue), plates
Silver long-cross penny, 1247–50
Half of polished ring,
2.6cm internal diam.
Possible end scraper
Tiny waste flake
Rhyolite cross fragment
Disc bead
Probably mans shoe, late 19th/
early 20th-century
Worn iron blade with remains of
wooden handle
Waste flake
Small nodule
Small nodule
F201/F202
interface
F302
F302
F302
SE quadrant
SE quadrant
Cutting 1
F304
F306
E338:
E338:
E338:
E338:
E338:
34
35
36
37
38
Clay pipe
Clay pipe
Slate
Cord
Crucifix
E338:
39
Pottery
E338:
40
Plastic inset
Large oratory
Large oratory
S entrance 2,
inner enclosure
Stem, early 19th-century
SOT?
Stem, early 19th-century
SOT?
Ovoid with remains of perforation SOT
Length of modern cordage
SOT ledge
Fragments of modern wooden
Large oratory
crucifix
Sherd of shell-edged ware,
Large oratory
1790–1820
Circular inset with inscription
Unprovenanced
Unprovenanced
Fill of drain F226
Cutting 2
Found within make-up of altar
F224
F102
Entrance cutting
F321
Unprovenanced
Surface find
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
498
E No.
APPENDIX II—CONCORDANCE OF CONTEXT/FEATURE
NUMBERS USED ON SITE AND IN THIS REPORT
499
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
Feature no.
(report)
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
520
521
530
531
532
533
534
535
536
537
538
539
540
541
542
542
543
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
500
Context no.
(excavation)
No number
No number
No number
00:34
00:29
00:30
00:31
00:32
00:33
00:35
00:36
No number
No number
No number
No number
No number
No number
No number
00:28
No number
No number
98.2
98.1
98.18 = F1008
98.21A
98.21B
98.22
98.23
98.24
98.25
98.27
98.28
98.29
98.3
98.3 = 98:3
98.4
98.31 = 98:4
98.5
98.1
98.6
98.7
98.8
98.11
98.12
98.13
98.14
98.2
98.15
98.19
98.9
98.16
Area
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
Monks graveyard
Monks graveyard
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
South entrance, inner enclosure
APPENDICES
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
98.17
No number
0
99.1
99.2
99.3
99.4
99.5
99.6
99.7
99.8
99.9
No number
00:01
00:02
00:03
00:04
= 568
00:05
00:06
00:07
00:08
00:09
00:10
00:11
00:12
00:13
00:14
00:15
00:16
00:17
00:18
00:19
00:20
00:21
00:22
00:23
00:24
00:25
00:26
00:27
No number
No number
No number
96.1
96.2
96.3
96.4
96.5
96.6
96.7
96.8
96.9
96.10
South entrance, inner enclosure
South entrance, inner enclosure
Leacht
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
Inner enclosure near leacht S of St Michaels Church
East of St Michaels Church
West of lower monks garden
West of lower monks garden
West of lower monks garden
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
South of St Michaels Church. Last cut for new revetment
South of St Michaels Church. Last cut for new revetment
South of St Michaels Church. Last cut for new revetment
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
East of St Michaels Church
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
501
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
611
612
613
614
615
616
617
618
619
620
621
622
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
502
96.11
96.12
96.13
96.14
96.15
96.16
96.17
96.18
96.19
96:20
96:22
96:23
No number
No number
No number
A
B
C
D
E
F
No number
No number
No number
No number
No number
No number
G
G?
No number
= F633(1)
= F628
= F657
No number
No number
No number
No number
H
I
No number
= F633(4)
= F648
= F1029
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
Cistern
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
APPENDICES
669
670
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
700
701
702
703
704
705
706
707
708
709
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
No number
No number
03:01
03:01a
03:02
03:03
03:03a
03:04
03:05
03:06
03:07
03:08
03:09
03:10
03:13
03:14
03:15
LMG
LMG
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
Structure at base of east steps
No number
No number
No number
90:05
90:04
90:03
90:02
90:01
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
No number
Structure at base of east steps
UMG
UMG
UMG
UMG
UMG
UMG
UMG
UMG
UMG
UMG
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
East entrance
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
503
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
1025
1026
1027
1028
1029
1030
1031
1050
1051
504
No number
No number
No number
No number
No number
No number
No number
No number
No number
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
LMG
APPENDIX III—CATALOGUE OF DISARTICULATED HUMAN
SKELETAL REMAINS
Codes used in catalogue of disarticulated human skeletal remains
(based on Chamberlain and Witkin 2000):
??
GB
GC
GT
KK
WW
AI
AM
AS
CC
CE
CF
CH
CL
CN
CO
CP
CS
CT
CV
CX
CZ
QH
D1
D2
D3
D4
D5
E1
E2
E3
E4
E5
D?
DD
DR
DX
M1
M2
M3
M4
M5
M6
M7
M8
Unknown
Burnt bone
Calcified soft tissue
Soft tissue
Skeleton
Unknown (faunal)
Auditory: Incus
Auditory: Malleus
Auditory: Stapes
Cranium
Endocast
Frontal
Ethmoid
Lacrimal
Nasal
Occipital
Parietal
Sphenoid
Temporal
Calvaria
Vault fragment
Zygomatic
Hyoid
Upper dI1
Upper dI2
Upper dC
Upper dM1
Upper dM2
Lower dI1
Lower dI2
Lower dC
Lower dM1
Lower dM2
? Deciduous tooth
Deciduous tooth
Deciduous tooth root
Deciduous crown fragment
Lower I1
Lower I2
Lower C
Lower P1
Lower P2
Lower M1
Lower M2
Lower M3
X1
X2
X3
X4
X5
X6
X7
X8
MC
MM
MR
MS
MY
XD
XP
XX
PP
PR
PX
QM
QS
QX
QC
S?
SA
SB
SC
SG
SS
SX
QR
VC
VT
VL
VS
VY
VV
VX
I?
IA
II
IL
IP
IS
IX
Upper I1
Upper I2
Upper C
Upper P1
Upper P2
Upper M1
Upper M2
Upper M3
Mandibular body
Mandible
Mandibular ramus
Mandibular symphysis
Mandibular condyle
Demimaxilla
Premaxilla
Maxilla
Permanent tooth
Permanent tooth root
Tooth crown fragment
Manubrium
Sternum
Sternum fragment
Clavicle
? Scapula
Acromion
Scapula blade
Coracoid
Scap. glenoid cavity
Scapula
Scapula fragment
Rib
Cervical vertebra
Thoracic vertebra
Lumbar vertebra
Sacrum
Coccyx
Vertebra
Vertebra fragment
? Hip bone
Acetabulum
Hip bone
Ilium
Pubis
Ischium
Hip bone fragment
505
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
H?
HH
HP
HM
HD
R?
RR
RP
RM
RD
U?
UU
UP
UM
UD
YC
YY
YS
YL
YQ
YI
YZ
YD
YA
YH
YM
YP
LS
506
? Humerus
Humerus
Humerus, proximal
Humerus, midshaft
Humerus, distal
? Radius
Radius
Radius, proximal
Radius, midshaft
Radius, distal
? Ulna
Ulna
Ulna, proximal
Ulna, midshaft
Ulna, distal
Carpal
Hand bone
Scaphoid
Lunate
Triquetral
Pisiform
Trapezium
Trapezoid
Capitate
Hamate
Metacarpal
Phalanx (hand)
Sesamoid
F?
FF
FP
FM
FD
LL
T?
TT
TP
TM
TD
B?
BB
BP
BM
BD
ZT
ZZ
ZA
ZC
ZN
ZE
ZI
ZL
ZU
ZM
ZP
? Femur
Femur
Femur, proximal
Femur, midshaft
Femur, distal
Patella
? Tibia
Tibia
Tibia, proximal
Tibia, midshaft
Tibia, distal
? Fibula
Fibula
Fibula, proximal
Fibula, midshaft
Fibula, distal
Tarsal bone
Foot bone
Talus
Calcaneus
Navicular
Medial cuneiform
Intermedial cuneiform
Lateral cuneiform
Cuboid
Metatarsal
Foot phalanx
ID
1
2
3
Area
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
Detail
Puffin burrow
Puffin burrow
Puffin burrow
4
5
6
7
8
9
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
10
11
Context
Code
MR
MM
M5
Portion
Ascending ramus & condyle
Body, ascending ramus, condyle, gonial angle
Lower second premolar
Side Sex Age1 Age2
L
JUV
L
JUV
L
JUV 10–11 yrs
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Tooth
Tooth
Mandible
Tooth
Tooth
Mandible
M6
M7
MC
M6
M7
MS
L
L
L
L
L
JUV
JUV
JUV
JUV
JUV
AA
E of St Ms Ch.
E of St Ms Ch.
Puffin burrow
Puffin burrow
Tooth
Mandible
M2
MC
Lower first molar
Lower second molar
Body frag.
Lower first molar
Lower second molar
Frag. of anterior symphysis
canine; assoc. with 21?
Lower second incisor
Body frag.
R
L
AA
AA
12
13
14
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
Puffin burrow
Puffin burrow
Puffin burrow
Tooth
Tooth
Mandible
M7
M8
MC
Lower second molar
Lower third molar
x 2 body frags
L
L
R
YA
YA
AA
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Tooth
Tooth
Tooth
Tooth
Mandible
Mandible
Tooth
Maxilla
Tooth
Tooth
Tooth
Tooth
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
M5
M6
M7
M8
MY
MR
M3
XX
X3
X4
X1
X6
CT
CT
CT
CT
CF
CF
CF
CF
CO
CO
CP
CP
CX
Lower second premolar
Lower first molar
Lower second molar
Lower third molar
Frag. of condyle
Frag. of gonial angle
Lower canine
Small frag. only
Upper canine
Upper first premolar
Upper first incisor
Upper first molar
Temporal frag. with occipital groove
Temporal—mastoid & petrous portion
Temporal—mastoid & petrous portion
Temporal—frag. of petrous portion
Frontal—eye orbit
Frontal—eye orbit
Frontal—vault frag.
Frontal—vault frag.
Occipital—posterior of base
Occipital—pars bas., lat. wings, sphenoid frag.
Parietal frag.
Parietal frag.
x 14 vault frags
R
R
R
R
R
R
L
R
R
R
R
R
L
R
R
R
L
R
R
AA
AA
YA
YA
JUV
JUV
AA
AA
AA
AA
JUV
YA
AA?
AA?
AA?
AA?
M? AA
M? AA
AA
AA
AA
AA
AA?
AA?
Pathology
In two frags; contains 3, 4, 5
Root forming; contained within 2
10–11 yrs
10–11 yrs
10.5–11.5 yrs
10.5–11.5 yrs
Moderate calculus
Slight calculus
17–25 yrs
17–25 yrs
Severe calculus
First molar lost
ante-mortem
Slight calculus
Slight calculus
Slight calculus
17–25 yrs
17–25 yrs
Additional comments
Slight calculus
Complete; contained within 2
Root forming; contained within 2
Contains 7, 8
Contained within 6
Contained within 6
Contained roots of all incisors and left
Contains 12 and 13; same individual as 14
Contained within 11
Contained within 11
Contains 15, 16, 17, and 18; same
individual as 11
Enamel broken p-m; contained within 14
Contained within 14
Contained within 14
Contained within 14
Hypoplastic defects May be assoc. with 9
Contains 23 and 24
Hypoplastic defects Contained within 22
Slight calculus
Contained within 22
Moderate calculus No wear evident
Slight calculus
Joins with 32, 33 and 34
Joins with 31, 33 and 34
Joins with 31, 32 and 34
Joins with 31, 32 and 33
Joins with 36
Joins with 35
APPENDICES
507
Bone
Mandible
Mandible
Tooth
Area
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
Detail
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
60
61
62
63
64
65
66
67
68
69
70
71
72
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
73
74
75
76
77
78
79
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
Context
Bone
Cranial
Sternum
Clavicle
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Vertebra
Code
CS
QS
QC
VC
VC
VC
VC
VC
VC
VC
VT
VT
VT
VT
VT
VT
VT
VL
VX
VX
Portion
Sphenoid—frag. of greater wing
Complete
Frag. of midshaft
C1—complete
Cervical—body only, C3?
Cervical—body only, C4?
C5?—complete
Cervical—frag. of arch and body
Cervical—frag. of posterior neural arch
Cervical—frag. of neural arch
Thoracic—x 6 body frags
Thoracic—body and arch frag.
Thoracic—neural arch frag.
Thoracic—frag. of arch and body
Thoracic—neural arch frag.
Thoracic—frag. of arch and body
Thoracic—frag. of arch and body
Lumbar—neural arch frag.
Small vertebral frags x 6 frags
Thoracic/lumbar—inferior body frag.
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Puffin burrow
Vertebra
Rib
Rib
Rib
Rib
Rib
Humerus
Humerus
Metatarsal
Unidentified
Unidentified
Scapula
Unidentified
VC
QR
QR
QR
QR
QR
HM
HP
ZM
??
??
SB
??
Cervical—complete
x 10 medial frags
x 3 medial frags
x 6 shaft frags
x 1 medial frag.
x 2 shaft frags
Shaft frag. c. 15cm long
Frag. of proximal neck
Unidentified
x 4 long bone shaft frags
x 20 small frags
Frag. of blade
Long bone frag.—hum/fem/tib?
Puffin burrow
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Vertebra
Femur
Patella
Cranial
Cranial
Tooth
Vertebra
VT
FD
LL
VX
VX
D5
VX
Thoracic—neural arch frag.
Distal condyle of femur
Complete
Vomer in two frags
x 3 vault frags
Upper 2nd decid. molar
Body frag.
Side Sex Age1 Age2
AA
AA
L
AA
AA
AA
AA
AA
L
AA
AA
L
AA
AA
R
AA
R
AA
R
AA
L
AA
R
AA
R
AA
L
AA
AA
R
L
R
JUV
AA
AA
AA
JUV
JUV
AA
AA
AA
AA
L
AA
AA
AA
AA
AA
JUV
AA
Additional comments
Articulates with 45
Articulates with 44 and 46
Articulates with 45
Herniation &
osteophytes
Fused except for vertebral plates
AA
AA
?
L
Pathology
Bony process—may Possible non-metric trait
be supraconhyloid
process
<10 yrs
Roots damaged p-m
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
508
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
97
98
99
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Assoc. w/Skel. 1
Cervical—frag. of artic. facets
Body only
x 15 long bone frags
x 2 shaft frags
x 1 lateral frag.
Inter. phalanx—complete
x 2 vault frags
x 2 iliac body frags
x 12 small frags
x 1 frag. of palatine
Frontal—orbit fused to palatine
Temporal—zygo. arch, EAM & TMJ
Temporal—EAM, TMJ, & pet. portion
Temporal—frag. of zygo. arch
Zygomatic—fused to maxilla
Small body frag. w/2 sockets
Lower molar—unidentified
L
R
R
R
?
?
AA
OA?
AA?
AA?
AA
OA?
OA?
OA?
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
7
7
7
Tooth
Tooth
Tooth
PP
M7
X7
Lower 1st/2nd perm. molar
Lower 2nd perm. molar
Upper 2nd? perm. molar
?
?
R
AA
AA
AA
100
101
102
103
104
105
106
107
108
109
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
E of St Ms Ch.
7
7
7
7
6
6
6
6
Tooth
Mandible
Unidentified
Cranial
Cranial
Cranial
Cranial
Cranial
Scapula
Vertebra
X8
MC
??
CX
CT
CT
CO
CS
SC
VX
Upper 3rd? perm. molar
Frag. w/sockets
x 14 long bone shaft frags
x 3 small frags
Temporal—petrous portion
Temporal—petrous portion
Occipital—squamous frag.
Sphenoid—body frag.
Frag. of corocoid
x 2 body frags
R
JUV
AA?
AA?
L
R
AA?
AA?
AA?
AA?
AA
AA
Bleached and weathered
Refers to Skel. 1, this report
110
E of St Ms Ch.
Femur
FM
Shaft frag. c. 20cm long
R
JUV
Refers to Skel. 1, this report; joins w/111
111
E of St Ms Ch.
Femur
FP
Prox. metaphysis
R
JUV
Refers to Skel. 1, this report; joins w/110,
112
E of St Ms Ch.
Femur
FM
Unfused greater trochanter
R
JUV
113
E of St Ms Ch.
Unidentified
??
x 23 long bone shaft frags
AA?
114
E of St Ms Ch.
Hand phalanx YP
Inter. phalanx, prox. end unfused
JUV
115
E of St Ms Ch.
Metacarpal
Unidentified—prox. end only
JUV
Refers to Skel. 1, this report; joins w/111
112
Refers to Skel. 1, this report; femur/tibia
frags
Refers to Skel. 1, this report; same ind. as
115, 116, 117
Refers to Skel. 1, this report; same ind. as
114, 116, 117
From clearance
Skeleton 2
(1998), bone 5
Skeleton 2
(1998), bone 5
Skeleton 2
(1998), bone 5
Skeleton 2
(1998), bone 5
Skeleton 2
(1998), bone 10
Skeleton 2
(1998), bone 6
Skeleton 2
(1998), bone 6
YM
AA
AA
Unfused
AA
JUV
AA
?
R
Porosity on prox. end—probably p-m
Matches w/94
Moderate calculus
Severe calculus
Large caries on
mesial side
13–14 yrs?
Matches w/90
Contains 96
Extremely worn almost to roots, contained
w/in 95
Roots damaged p-m
Roots damaged p-m
Roots damaged p-m
Roots forming
APPENDICES
509
VC
QH
??
QR
QR
YP
CX
IL
??
CX
CF
CT
CT
CT
CZ
MC
PP
R
7
7
7
7
7
7
7
Vertebra
Hyoid
Unidentified
Rib
Rib
Hand phalanx
Cranial
Pelvis
Unidentified
Cranial
Cranial
Cranial
Cranial
Cranial
Cranial
Mandible
Tooth
Area
E of St Ms Ch.
117
E of St Ms Ch.
118
E of St Ms Ch.
119
E of St Ms Ch.
120
E of St Ms Ch.
121
E of St Ms Ch.
122
E of St Ms Ch.
123
E of St Ms Ch.
124
E of St Ms Ch.
125
E of St Ms Ch.
126
E of St Ms Ch.
127
E of St Ms Ch.
128
E of St Ms Ch.
129
E of St Ms Ch.
130
E of St Ms Ch.
131
E of St Ms Ch.
132
E of St Ms Ch.
133
E of St Ms Ch.
134
E of St Ms Ch.
135
E of St Ms Ch.
136
E of St Ms Ch.
137
E of St Ms Ch.
Detail
Context
Skeleton 2
(1998), bone 6
Skeleton 2
(1998), bone 6
Assoc. w/Skel. 2
(1998)
Assoc. w/Skel. 2
(1998)
Assoc. w/Skel. 2
(1998)
Assoc. w/Skel. 2
(1998)
Assoc. w/Skel. 2
(1998)
Assoc. w/Skel. 2
(1998)
Assoc. w/Skel. 2
(1998)
Assoc. w/Skel. 2
(1998)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Bone
Code
Hand phalanx YP
Portion
Unidentified—distal end only
Side Sex Age1 Age2
JUV
Metacarpal
YM
Unidentified—distal end
Tibia
TP
Anterior prox. metaphysis
L
AA
Additional comments
Refers to Skel. 1, this report; same ind. as
114, 115, 117
Refers to Skel. 1, this report; unfused epip.
absent; same end[¿ind.?] as 114, 115, 116
Refers to Skel. 1, this report; joins with 119
Tibia
TP
Posterior proximal metaphysis
L
AA
Refers to Skel. 1, this report; joins with 118
Femur
FD
Frag. of distal femur
AA
Refers to Skel. 1, this report
Femur
FD
Frag. of distal metaphysis
JUV
Tibia
TM
x 4 shaft frags
AA?
Femur
FM
x 5 shaft frags
AA
Unidentified
??
x 49 long bone shaft frags
AA?
Refers to Skel. 1, this report; unfused epip.
absent
Refers to Skel. 1, this report; anterior spine
identified
Refers to Skel. 1, this report; linear aspera
identified
Refers to Skel. 1, this report
Unidentified
??
x 1 long bone shaft frag.
?
Refers to Skel. 1, this report
Humerus
HM
Shaft frag.
AA
Refers to Skel. 1, this report; joins with 127
Humerus
HD
Distal third of humerus
R
M? AA
Refers to Skel. 1, this report; joins with 126
Ulna
UP
Frag. of olecranon process
?
AA
Refers to Skel. 1, this report
Radius
RP
Prox. head
?
AA
Radius
RP
Frag. of prox. metaphysis
Ulna
UM
Shaft frag.
R
JUV
Ulna
UP
Frag. of prox. head
R
JUV
Scapula
SS
Glenoid + frag. of blade
R
JUV
Refers to Skel. 1, this report; head—
27.7mm diameter
Refers to Skel. 1, this report; unfused prox.
epip. absent
Refers to Skel. 1, this report; joins with
132; not part of Skel. 1
Refers to Skel. 1, this report; joins with
131; not part of Skel. 1
Refers to Skel. 1, this report
Scapula
SB
Blade frag.
L
AA
Refers to Skel. 1, this report
Rib
QR
x 8 shaft frags
AA?
Refers to Skel. 1, this report
Unidentified
??
x 11 long bone shaft frags
Carpal
YA
Capitate—complete
JUV
M
JUV
Pathology
Refers to Skel. 1, this report
L
AA
Refers to Skel. 1, this report
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
510
ID
116
138
141
E of St Ms Ch.
142
E of St Ms Ch.
143
E of St Ms Ch.
144
E of St Ms Ch.
145
E of St Ms Ch.
146
E of St Ms Ch.
147
E of St Ms Ch.
148
E of St Ms Ch.
149
E of St Ms Ch.
150
E of St Ms Ch.
151
E of St Ms Ch.
152
E of St Ms Ch.
153
E of St Ms Ch.
154
E of St Ms Ch.
155
E of St Ms Ch.
156
Small oratory
F229
Tibia
TP
Prox. half of diaphysis
157
158
159
160
161
Small oratory
Small oratory
Small oratory
Small oratory
Small oratory
F229
F229
F229
F229
F229
Humerus
Unidentified
Unidentified
Tooth
Mandible
HD
??
??
PX
MC
Distal diaphysis
x 2 large long bone shaft frags
Small quantity of small long bone shaft frags
x c. 65 small broken enamel frags
Body frag.
139
Misc. assoc.
Carpal
YZ
Trapezium—almost complete
R
AA
Refers to Skel. 1, this report
Misc. assoc.
Metatarsal
ZM
4th metat., distal head absent
R
AA
Refers to Skel. 1, this report
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Misc. assoc.
w/Skel. 2 (1997)
Metacarpal
YM
5th metac., distal head absent
R
AA
Refers to Skel. 1, this report
Unidentified
??
Shaft of metat. or metac.
AA
Refers to Skel. 1, this report
Unidentified
??
Metat. or metac. prox. end
JUV
Refers to Skel. 1, this report
Unidentified
??
Metac. or metat.—distal end
JUV
Pelvis
IL
Ilium—frag. of aricular surface
AA
Refers to Skel. 1, this report; unfused distal
end absent
Refers to Skel. 1, this report
Pelvis
IL
Ilium—frag. w/sciatic notch
AA
Refers to Skel. 1, this report
Pelvis
IL
x 1 iliac body frag.
AA
Refers to Skel. 1, this report
Vertebra
VL
Lumbar?—x 6 body frags
AA
Vertebra
VL
Lumbar?—frag. of sup. facet
Vertebra
VL
Lumbar x 2 inferior facets
AA
Refers to Skel. 1, this report
Sacrum
VS
Frag. of sup. sacrum
AA
Refers to Skel. 1, this report
Femur
FP
Prox. head
?
AA
Tibia
TD
Poss. frag. of distal epip.
?
Refers to Skel. 1, this report; 51.2mm in
iameter
Refers to Skel. 1, this report
Unidentified
??
x 2 unfused ends of long bones
JUV
Humerus
HP
Unfused prox. epip. of humerus
JUV
Unidentified
??
x 21 small frags
R
AA
M
Some possible
degeneration
Osteophytes on
facet
AA
Refers to Skel. 1, this report
Refers to Skel. 1, this report
Refers to Skel. 1, this report;
unidentified—epiphyses absent
Refers to Skel. 1, this report; frag. only
Refers to Skel. 1, this report
511
R
JUV
R
JUV?
?
?
?
JUV 10.5–12.5 yrs?
?
Healing perio. on
medial side of
ant. shaft
Unfused epip. absent; lat. half of
metaphysis absent; c. 9–11 rs
Superior to olecranon fossa; c. 9–11 rs
Humerus/femur/tibia?
Broken p-m
Frag. w/ mental foramen; contains 162,
163
APPENDICES
140
E of St Ms Ch.
w/Skel. 2 (1997)
E of St Ms Ch.
w/Skel. 2 (1997)
E of St Ms Ch.
Area
162
Detail
Context
Bone
Code
Portion
Side Sex Age1 Age2
Small oratory
F229
Tooth
M5
Lower 2nd premolar
R
?
JUV
10.5–12.5 yrs?
163
Small oratory
F229
Tooth
M6
Lower 1st perm. molar
R
?
JUV
10.5–12.5 yrs?
164
165
166
167
168
169
170
171
Small oratory
Small oratory
Small oratory
Small oratory
Small oratory
Small oratory
Small oratory
Small oratory
F229
F229
F229
F229
F229
F229
F229
F229
Tooth
Tooth
Tooth
Tooth
Tooth
Tooth
Tooth
Tooth
M2
M7
X5
M5
X6
PP
M4
M7
Lower 2nd perm. incisor
Lower 2nd perm. molar
Upper 2nd premolar
Lower 2nd? premolar
Upper 1st perm. molar
Upper 2nd/3rd perm. molar
Lower 1st premolar?
Lower 2nd perm. molar?
R
?
?
?
R
R
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
172
173
Small oratory
Monks Graveyard
F229
Tooth
Humerus
M7
HD
Lower 2nd perm. molar?
distal 4/5, capit. absent, troch. ant. 1/2 only
?
R
?
?
?
AA
174
175
176
177
178
179
Monks graveyard
Monks graveyard
Monks graveyard
Monks graveyard
Monks graveyard
Lower monks garden
Femur
Scapula
Tibia
Mandible
Unidentified
Femur
FD
SS
TM
MR
??
FD
Distal 2/3; med. con. ab.; lat. cond. very incomp. R
Glenoid, frags of acromion & spine
R
Shaft frag. 65mm long w/frag. med. cond.
R
Frag. of sup. ascending ramus, no condyle
L
x 12 small frags
Distal 1/6 of femur
L
?
?
?
?
AA
AA
AA
AA?
180
E of St Ms Ch.
Tooth
X1
1st upper incisor
R
AA
Significant wear
Within 182, fragmenting
181
E of St Ms Ch.
585/95?
Tooth
X2
2nd upper incisor
R
AA
Significant wear
Within 182, fragmenting
182
E of St Ms Ch.
585/95?
Maxilla
XX
Anterior frag. containing 180 and 181
R
AA
With pelvis of
Skel. 1 (1998)
With pelvis of
Skel. 1 (1998)
With pelvis of
Skel. 1 (1998)
Cutting 5,
Layer C,
E ext.
585/95?
Pathology
Additional comments
Contained w/in 161; enamel extensively
broken; Rc or A1/2
Contained w/in 161; enamel extensively
broken
Root broken off p-m
Enamel only; broken p-m
Enamel only; broken p-m
Enamel only; broken p-m
Distal half of crown only; broken p-m
Enamel only; broken p-m
Enamel only; broken p-m
Frag. of enamel and root; extensively
damaged
Small frag. of enamel; broken p-m
P-m damage—min. 22cm; min. cir. 67mm;
max. dia. 23.1mm; min. dia. 20.6mm
P-m damage—min. 25cm long
P-m damage; glenoid length min. 35.6mm
Small frag. only
Severe p-m damage; bone is like paper
AA
SKELLIG MICHAEL, CO. KERRY: THE MONASTERY AND SOUTH PEAK
512
ID