The Archaeological
Landscape of Bute
George Geddes and Alex Hale
The Archaeological
Landscape of Bute
George Geddes and Alex Hale
The Archaeological
Landscape of Bute
George Geddes and Alex Hale
British Library cataloguing in Publication Data: A CIP
catalogue record for this book is available on request
from the British Library.
Crown Copyright © RCAHMS 2010
ISBN: 978 1 902419 74 9
Book layout by Mitch Cosgrove
The maps are based on Ordnance Survey material with
the permission of the Ordnance Survey on behalf of
Her Majesty’s Stationery Office © Crown Copyright. All
rights reserved , OS Licence Number OS 1000 25406.
Unauthorised reproduction infringes Crown Copyright
and may lead to prosecution or civil proceedings. Except
where otherwise stated , all the illustrations and images
in this publication are Crown Copyright RCAHMS.
All rights reserved. No part of this publication may be
reproduced, stored in or introduced into a network
retrieval system, or transmitted, in any form, or by any
means (electronic, mechanical, photocopying, recording
or otherwise) other than for initial download without
the prior written permission of RCAHMS. Any person
who does any unauthorised act in relation to this
publication may be liable to criminal prosecution.
Front cover: An oblique aerial view of the vitrified fort of
Dunagoil (at right, 40291) and the fort of Little Dunagoil (at
left, 40280). See page 22. DP062659
Front cover inset: St Colmac cross, standing stone with later
inscribed cross (40317). See page 15. SC408104
Back cover: A detailed plan from 1780 of Dunagoil vitrified fort
(40921) and its surrounding landscape. See page 4. DP075180
© Bute Archive at Mount Stuart
Page iv: A view from Glencallum bay at the SE corner of Bute
towards the Dunagoil archaeological complex in the distance.
Much of the dramatic intervening landscape was cultivated
until the 18th century, when the south of the island became a
sheep farm. DP062664
Published in 2010 by The Royal Commission on the
Ancient and Historical Monuments of Scotland.
The Royal Commission on the Ancient and
Historical Monuments of Scotland (RCAHMS)
John Sinclair House
16 Bernard Terrace
Edinburgh EH8 9NX
tel 0131 662 1456
fax 0131 247 4163
www.rcahms.gov.uk
Registered Charity SC010240
For over 100 years, the Royal Commission on the
Ancient and Historical Monuments of Scotland
has been collecting, recording and interpreting
information on the archaeological, architectural,
industrial and maritime heritage of the nation,
creating a unique archive that offers a remarkable
insight into the special nature of Scotland’s places.
Many millions of items, including photographs, maps,
drawings and documents are made widely available
to the public via the web, through exhibitions and
publications, and at the RCAHMS search room in
Edinburgh.
Commissioners of the Royal Commission on the Ancient and Historical Monuments of Scotland, 2010
Professor John R Hume OBE BSc ARCST Hon FRIAS FSA Scot (Chairman)
Dr Gordon G T Masterton OBE BA MSc PhD DIC CEng FICE FIStructE FIES FRSE MCIWEM FConsE (Vice Chairman)
Dr Kate Byrne MA MSc Phd CEng MBCS CITP FRSA
Mr Tom Dawson FSA Scot
Mr Mark Hopton MBA Dip Arch BArch RIBA ARIAS FSA Scot
Dr Jeremy Huggett BA Phd MIFA FSA Scot
Professor John Hunter BA PhD MIFA FSA FSA Scot
Mr Paul Jardine BA MBA ACMA
Ms Jude Quartson-Mochrie BA (Hons) BArch (Hons)
Ms Elspeth Reid MA MA DAA
Secretary, Chief Executive, Mrs Diana Murray MA FSA FSA Scot MIFA
iii
iv
Contents
Foreword
vi
Site References and More Information
vii
Introduction
1
The Mesolithic and Neolithic
7
The Early and Middle Bronze Age
13
The Late Bronze Age and Iron Age
21
The Early Historic and Medieval Periods
29
The Pre-Improvement Period
37
The Improvement Period
41
Conclusion
47
Further Reading
48
Index
51
v
Foreword
The partnership between the Discover Bute
Landscape Partnership Scheme (DBLPS) and the
Royal Commission on the Ancient and Historical
Monuments of Scotland (RCAHMS), from which
this volume emanates, has been a truly rewarding
example of successful partnership working
between Scottish Government agencies and local
people. With the assistance of funding from
the Heritage Lottery Fund , the long wished for
reassessment of the archaeological record of Bute
and the enhancement of that record into a modern,
easily accessible resource has become a reality.
As significant, the symbiotic knowledge transfer
between the RCAHMS Survey and Recording
staff and the local Bute residents has been a
process that has been educational, inspirational
and , most importantly, a whole lot of fun! We
are delighted that this volume will stand as a
long term legacy for the project, and one that
will encourage involvement in, learning about
and enthusiasm for the archaeology of Bute for
generations to come.
Paul Duffy
Deputy Project Manager (Archaeology)
Discover Bute Landscape Partnership Scheme
September 2010
Acknowledgements
This booklet was written by G Geddes and A Hale
and edited by J Sherriff, JB Stevenson, R Bailey
and R Turner. The maps and the survey drawings
have been prepared by G Brown and I Parker and
the aerial photographs were taken by R Adam and
D Cowley and other members of the RCAHMS
aerial survey team. Additional photography was
provided by D Smart, S Wallace and by RCAHMS
field investigators. In addition to those noted above,
fieldwork was undertaken principally by P Dixon,
A Dutton, A Gannon, L Fisher, J Hepher, A Leith,
L McCafferty, P McKeague and M Middleton.
Thanks also to O Brookes, M Sutherland and
W Toole for production, proof reading and indexing.
The Commission is grateful to many individuals
who provided assistance throughout the project.
Those to be thanked at the Discover Bute Landscape
Partnership Scheme (DBLPS) include B Paterson
(Scheme Co-ordinator), and the members of
the Archaeology Research Project, including
P Duffy. Of the members of the local community
who participated in the fieldwork (too many to
individually mention here), special thanks are
extended to A Hannah and J Herriot. Thanks also
to S Hothersall from the Association of Certificated
Field Archaeologists (ACFA), S MacDonald
(Administrator), I McArthur (Buteshire Natural
History Society), A Speirs (Curator, Bute Museum),
A McLean, archivist at Mount Stuart, and
J Turner, who facilitated access to Inchmarnock.
For providing other specialist advice, thanks are
due to T Cowie and A Sheridan (both National
Museums of Scotland) and N Finlay (Dept of
Archaeology, University of Glasgow). RCAHMS
is also especially grateful to all the landowners,
including the Marquess of Bute, Lord Attenborough
and Sir Robert Smith, and to all the tenants who
willingly allowed access to the sites.
vi
Site References and More Information
Sites and monuments are consistently labelled
throughout this publication with their RCAHMS
number. This is a unique identifier that will allow the
reader to make the connection between sites which
are mentioned in the main text or in captions, and
sites which are highlighted on maps. The number
will also enable readers to obtain further information
about a particular site, including bibliographic and
archive details, by accessing the RCAHMS sites and
monuments database Canmore. This can be found at
rcahms.gov.uk. Within the advanced search page, the
reader should enter the number in the box marked
‘Canmore ID’ and click on the search button.
Readers can also discover more information
about sites and monuments on Bute by searching the
ScotlandsPlaces website: scotlandsplaces.gov.uk.
This offers an alternative way of accessing the
site-records held within the RCAHMS database as
well as other source material.
Each illustration in this publication has a caption
which includes, when appropriate, the name of the
site, its main classification and its number. Further,
a code number at the end of each caption identifies
the digital file held in the RCAHMS archive and
this should be cited if the illustration is required
for reproduction.
Plan Conventions
vii
viii
Introduction
Introduction
This booklet is a result of a partnership project
undertaken in 2008–10 between the Royal
Commission on the Ancient and Historical
Monuments of Scotland (RCAHMS) and the
Discover Bute Landscape Partnership Scheme
(DBLPS), the co-ordinators of a four-year
programme of improvements to Bute’s rural
landscape funded largely, but not solely, by the
Heritage Lottery Fund. The short-term aims of
the project were to revise the existing RCAHMS
archaeological records for Bute, working closely
with the local community, and to produce a booklet
that summarises the archaeology of Bute. The
principal long-term aim is to provide the local
community with the information that will allow
them to determine priorities and make decisions
about the archaeological work they are likely to
initiate in the future.
This volume contains a commentary on the
character of the archaeology of the island , one
that has been informed principally by the evidence
contained within the revised archaeological record.
Although there has never been a systematic
archaeological survey of Bute a great deal of work
has been undertaken. In recent times the locations
of several early prehistoric sites have been identified
through the discovery of surface scatters of stone
artefacts; documentary research has provided clues
regarding pre-Improvement period landscapes;
fieldwork has resulted in many previously
unrecorded upstanding monuments being reported;
and , since 1977, aerial survey has yielded evidence
The Isle of Bute, showing Rothesay, other settlements and
geographical features. GV004571
RCAHMS staff and local volunteers discuss the archaeological
landscape of Scalpsie Bay, where prehistoric forts, burial cairns,
an improvement farmstead and what may be a medieval building
lie in close proximity. DP084694
of a broad range of cropmarks. Nevertheless,
comparatively large areas of the island still have
secrets to reveal, and further work in all these
disciplines, plus others, such as geophysical survey
and excavation, will undoubtedly reveal more
evidence in the years to come.
The revision of the Bute archaeology records
has made some use of documentary evidence. First,
second and third editions of the Ordnance Survey
6-inch-to-a-mile map, respectively published in
1
The Archaeological Landscape of Bute
1869, 1897 and 1924, have been referred to, but
two 18th century estate surveys, now held within
the Bute Archive at Mount Stuart, have been
particularly useful. One, by John Foulis in about
1758–9, is roughly contemporary with Roy’s
Military Map of 1747–55, but it provides a more
accurate and detailed record. The other, compiled
about 1780 –2 by Peter May, the Bute Estate factor,
records precise details of the land-use within
the estate, including the location of farmsteads,
smallholdings and single buildings, with notes on
acreages and tenancies. It is supported by a ledger
that contains an astonishing amount of detail.
Extract from the 3rd edition of the 6-inch Ordnance Survey map
showing Dunagoil (Buteshire 1924, Sheet CCCXXVII). This
depiction of the landscape, which was clearly influenced by Dr
J N Marshall’s contemporary survey and excavation, includes
a number of archaeological sites that had not been previously
shown, some of which have subsequently been lost or reclassified.
In contrast, there are only two sites depicted on the earliest
edition of the survey of Bute (published in 1869) where their
interpretation has been subsequently doubted – forts at Achamor
Wood (40393) and Suidhe Chatain (40286). SC1208177
2
Scope of survey
The partnership agreement between RCAHMS
and DBLPS required the archaeological site
records for Bute to be thoroughly revised , with
the necessary fieldwork providing opportunities
for local volunteers to become involved. The
evaluation of the record started as a desk-based
exercise late in 2008, and a programme of field
survey, which was undertaken in the spring and
summer of 2009, incorporated the informal
Introduction
training of members of the local community in
archaeological survey techniques. The overall
contribution made by the volunteers was
invaluable, not only saving RCAHMS staff a
considerable amount of time in an unfamiliar
landscape by identifying the locations of sites
that would have otherwise proved difficult
to find , but in making RCAHMS aware of a
significant number of sites that had not previously
been reported.
Although prospective fieldwork – proactively
searching for previously unreported monuments
– did not constitute a formal part of the project,
the simple act of visiting known sites with local
people who had already undertaken a considerable
amount of fieldwork resulted in about 120 ‘new’
sites being recorded. This was in addition to the
525 previously recorded Bute archaeological
records in the RCAHMS database that were
revised and the 260 or so other records that
have been created through the accessioning of
recent work.
One aspect of the existing RCAHMS database
that the revision did not cover was the broad range
of artefacts that have been found; these form
a significant proportion of the archaeological
record. However, this topic is currently being
addressed by another project within the Landscape
Partnership Scheme.
Although the project specification stipulated
the revision of the archaeological records for
Bute, it should be recognised that the distinction
between archaeology and architecture, especially
when ruined buildings are concerned , can be
a fine one. A number of relatively recently
abandoned farmsteads were visited – not because
of any intrinsic architectural merit (though this
may be present), but because map or documentary
evidence suggested that earlier remains might
have been evident. As in the case of those working
farmsteadings that were either visited on the
ground or photographed during aerial survey,
a note about the place was recorded for the
RCAHMS database.
Extract from the Survey of the Isle of Bute undertaken in 1758–9
by John Foulis. His survey, which is accompanied by brief texts,
recorded both the position and orientation of each building, and
the size and land-use of each farm. DP077260 © Bute Archive at
Mount Stuart
John Patrick Crichton-Stuart, the 3rd Marquess of Bute
(1847–1900). Along with his predecessors and descendants, the
3rd Marquess sponsored a number of archaeological projects
and funded the restoration of buildings that included St Blane’s
church. © Mount Stuart Archive
Previous work
The present archaeological record owes much
to the Earls of Bute, who, since the late 18th
century, have either initiated work or readily given
permission for others to do so. Perhaps the earliest
noted archaeological investigation on Bute was that
undertaken at Dunagoil, planned in 1780, where
several cairns (40070) were ‘excavated’ soon after
by John Norton, chief gardener at Mount Stuart,
under the direction of John Blain. The 19th century
saw several small excavations being documented
and probably a lot more not, and culminated with
the publication in the 1890s of The Isle of Bute
in the Olden Time by James Hewison, Church of
Scotland minister of Rothesay parish. Although
Hewison’s volumes reproduced a lot of the
rather vague information that appeared in earlier
publications, including the Statistical Account, they
provide an excellent summary of the archaeology
of Bute at that time, some of their illustrations
containing information that is now lost.
The leading figure in Bute archaeology in
the early part of the 20th century was Dr John
Marshall, a founder member of the Buteshire
Natural History Society (BNHS) and of the Bute
Museum. Dr Marshall excavated at Dunagoil from
3
The Archaeological Landscape of Bute
A detailed plan and sketch section of Dunagoil vitrified fort
(40291) and its surrounding landscape. The plan, dated 1780,
and probably produced by either Peter or Alexander May,
depicts and describes archaeological features at Dunagoil. It
is a remarkable symbol of the early interest and investment of
the 3rd Earl and 1st Marquess of Bute in aspects of the island’s
heritage. DP075180 © Bute Archive at Mount Stuart
1913, first in the cave (40278) and then in the
fort (40291), and undertook what would be called
today a watching brief at Townhead gravel quarry
(40377), where he recorded details of a Neolithic
settlement site.
Dr Marshall’s contribution to Bute
archaeology was eventually eclipsed by his
youngest daughter, Dorothy, who studied
under Sir Mortimer Wheeler at the Institute of
Archaeology in London and , after her mother’s
4
death in 1949, gained considerable experience
working on excavations both at home and
abroad. Much of what is now understood about
the archaeology of Bute is a consequence of
work that she instigated , work that has been
warmly appraised by Jack Scott (1992). Dorothy
Marshall’s legacy, however, is not just the records
of the important excavations she undertook or
her reorganisation of Bute Museum, it is also
the enthusiasm she engendered among the local
community. That passion has persisted since
her death in 1992 and to this day others have
continued to investigate aspects of Bute’s heritage
through research, excavation and survey.
Initially encouraged by Dorothy Marshall,
and more recently led by Edwina Proudfoot,
members of the Society have been recording rural
Introduction
Hawk’s Nib, cave and midden (40675). A group of volunteers
accompanied by RCAHMS staff ascend to explore this cave
site. Excavators in the 1930s discovered midden material
and evidence for hearths, but no categorical dating evidence.
DP073220
Excavations at Eilean Buidhe island dun (40458) in August
1936. One of a number of important archive photographs held
by Bute Museum, this image shows Dr J N Marshall (1860 –
1945) in the trench flanked by his daughter Dorothy (1900 –1992)
and J Harrison Maxwell, who directed the excavations. The
photograph was taken from the south-west showing one of
the trenches cut through the rampart; the building in the
background is the boat house at Caol Ruadh School. SC1210134
© Bute Museum
settlement remains for many years. Since 1991, the
Bute Settlement Survey project has systematically
added to the corpus of information through deskbased research, walk-over surveys and detailed
recording. As a result, the local community has
made and continues to make a very valuable
contribution to Scottish rural settlement studies.
At the start of the revision of the RCAHMS
records for Bute a large proportion (about 200
out of 525) of the archaeological records related
to settlement sites, mainly farmsteads, which had
been reported by members of the Society.
The abundant legacy outlined here has been
built upon by the partnership project between
RCAHMS and DBLPS, which has seen such a
significant involvement by the local community,
and by the RCAHMS Scotland’s Rural Past
project. The revision of the archaeological records
will provide a solid platform on which to base
future research, and the fieldwork undertaken by
the local community in tandem with RCAHMS in
2009–10 will better equip them to undertake that
work. The authors of this volume hope that it may
help in directing and encouraging their efforts.
5
6
The Mesolithic and Neolithic
c12,000BC – c2200BC
Mesolithic c12,000BC – c4500BC
Evidence from Howburn Farm in South
Lanarkshire has recently demonstrated that
Scotland was visited as early as 14,000 years ago.
To date, no evidence for such early settlement has
been found on Bute, where, as far as is known,
the first visitors were itinerant Mesolithic foragers
(fisher-hunter-gatherers) who populated the west
coast of Scotland between the end of the last ice
age (about 11,000 years ago) and the beginning of
the Neolithic period.
Three locations on Bute, all close to the south
end of the island , have so far yielded Mesolithic
artefacts. In the early 1970s an assemblage of agate
tools and waste material was discovered on Plan
Farm (40287). Most of the material is of Neolithic
or later date, but the collection includes two earlier
very small worked blades – microliths. A larger
assemblage of about 200 flints (40645) from a
raised beach at the north end of Kilchattan Bay
also contained microliths as well as late Mesolithic
cores and waste. Again, most of this material is of
later date. In 2003 fieldwalking on Garrochty Farm,
just 1km south-west of the source of the agateworking site, resulted in the discovery of another
flint and pitchstone assemblage (268254) with a
small Mesolithic component, this time including
two platform cores and a core rejuvenation blade.
Distribution of Mesolithic and Neolithic sites and artefacts.
This map probably reflects to a large degree an original bias in
the overall distribution of rock carvings and chambered tombs
in the northern third of the island. However, that distribution
has certainly been affected by other factors, including intensive
agriculture since the medieval period, which may have resulted in
the removal of even the most substantial monuments. GV004572
A shell-midden below the chambered cairn at
Glecknabae (40371) was partly examined when the
tomb was excavated at the beginning of the 20th
century. No artefacts were found at that time, but a
suggestion that the midden was Mesolithic has not
necessarily been disproved by the identification of
the flint core found more recently in the eroding
edge of the midden as probably Neolithic.
The small number of Mesolithic sites so far
identified on Bute precludes any in-depth comment
on their distribution. However, there is considerable
potential for identifying further sites of this date on
the island , as fieldwalking exercises in other areas
of Scotland have demonstrated that Mesolithic
material will invariably be found on prolonged
and extensive searches of cultivated ground. The
fact that the fieldwalking at Black Park Plantation
(123472, 123473) at the south end of Bute yielded
only material of Neolithic and Bronze Age date
need only be a reflection of the comparatively small
area examined.
Although arable fields provide the most
obvious targets for fieldwalking, other locations
on Bute ought to be considered, and work
elsewhere in Scotland has shown the potential of
examining moorland landscapes. However, the
best opportunity for locating relatively undisturbed
Mesolithic evidence on Bute and Inchmarnock may
be in the numerous caves and rock-shelters that
have been identified. Several of these were looked
at by Dorothy Marshall, who found no evidence of
human activity in most of them, the one exception
being at Hawk’s Nib (40675) at the south end of
Bute, which contained an undated shell-midden.
Evidence for the more recent use of caves and
7
The Archaeological Landscape of Bute
8
The Mesolithic and Neolithic
Bicker’s Houses, chambered cairn (40381). Sketch of the
chambered long cairn viewed from the south by J K Hewison
(1893, 65). In addition to the main burial chamber at the north
(far) end of the cairn, this sketch shows what may have been a
cist that is no longer visible inserted in the southern part of the
mound. The ‘fort’ in the background is probably Barone Hill
(40424). SC1163929
rock shelters has been found close by at Dunagoil
(40278, 40279) and on Inchmarnock (300171,
300178) see page 20.
Neolithic c4500BC – c2200BC
Our understanding of how and where the first
groups of farmers on Bute lived and worked
is hampered by the dearth of identified and
accurately dated settlement sites. Consequently,
the information gained from the distribution of
Kilchattan Bay. Excavations at the quarry in the distance
uncovered evidence for a Neolithic settlement (212266). The
nearby hamlet once included a thriving brick and tile works
(158613) and a parish church (81886), while prehistoric burial
cairns stood sentinel on the bay’s edge (40639). The pattern of
rectangular fields, established in the Improvement period, rolls
over the edge of the old shoreline cliffs. DP057287
funerary, ceremonial and ritual sites, rock art (which
cannot easily be included in any of those categories)
and a thin scatter of diagnostic artefacts, paints an
incomplete picture of events that were spread over
at least two millennia.
Between 1914 and 1919, and again in 1929,
the working of a gravel-pit at Townhead (40377)
on the southern edge of Rothesay revealed a range
of artefacts, including a polished stone axe and
pottery, which indicated the presence of some
form of Neolithic settlement. However, little detail
is known of this settlement and there is scant
evidence of any structure that might have existed
here in association with the artefacts and what was
probably a hearth. A number of shallow trenches
that were revealed in the surface of the gravel were
interpreted as probable ‘sleeper beams’, but they
are more likely to have been field drains or the
bottoms of cultivation furrows. A radiocarbon date
for a burnt shell of a hazelnut from the probable
hearth produced a date around the middle of the 4th
millennium BC.
More recently, excavation at another gravel
quarry – at Kingarth (212266) towards the south
end of Bute – revealed what may be a round-ended
9
The Archaeological Landscape of Bute
Michael’s Grave, chambered cairn (39905). The position of
even the most significant monuments has been carefully checked
with survey grade GPS. This particular cairn was excavated
by T H Bryce in 1903, and the burial chamber now sits proud
of the surrounding field, divorced from its protective round
cairn which has been heavily robbed and largely ploughed out.
DP082856
Neolithic building. This structure has been
compared to broadly similar buildings elsewhere,
particularly in Argyll and Ireland , but it yielded no
absolute dating evidence and could not be shown to
have had any direct association with any of the other
excavated features or with the scatter of flint, quartz
and pitchstone artefacts recovered in the vicinity.
If the paucity of identified habitation sites
limits discussion about the character of Neolithic
settlement on Bute, then the situation is significantly
improved with regard to the funerary, ritual and
ceremonial sites, and , of course, the rock art.
There are six chambered cairns on Bute, five in
the northern third of the island and one, Bicker’s
Houses (40381), in the central part, some 4km
south-west of Rothesay. A chambered cairn is a
large mound of stones that contains a stone burial
chamber which could be opened whenever the
10
remains of the dead needed to be interred. All of
the known chambered cairns on Bute have either
been ‘explored’ by antiquaries or excavated in
more recent times, and they have been thoroughly
discussed elsewhere. Audrey Henshall’s work
makes it clear that each cairn probably had a long
and complex history, in some cases extending into
the Early Bronze Age. Three of the chambered
cairns, Glecknabae (40371), Hilton (40306) and
Michael’s Grave (39905), appear to have been built
as round cairns and remained that way throughout
their use. In its final form, Glenvoidean (39897)
is a trapezoidal cairn that covers two earlier round
cairns, and Bicker’s Houses and Carnbaan (40370)
are long mounds that may well have developed from
earlier round cairns. There is only one radiocarbon
assessment for any of the Bute tombs, a burnt
deposit from beneath one of the side-slabs of the
chamber at Glenvoidean producing a date around
the middle of the 4th millennium BC.
The bias in the distribution of the tombs in the
northern third of the island probably reflects their
original range even allowing for the undocumented
destruction of some tombs and the possibility that
previously unrecorded examples will be identified
The Mesolithic and Neolithic
in the future. The most recently recognised
chambered cairn is situated at Hilton, overlooking
Kames Bay. First reported in the early 1970s, it
exhibited no evidence of its Neolithic origin until
it was excavated. Similar monuments could easily
have been swept away in the tide of agricultural
improvements; others may yet be recognised.
There is also the possibility that other forms of
Neolithic burial may also be present on Bute. For
instance, there may be tombs in which the mortuary
chambers were constructed in timber rather than
stone. If this is the case, their identification can only
be confirmed by excavation. In this context, mention
should be made of a number of apparently manmade mounds that have been examined without
yielding any evidence of burial. Investigation of
two such mounds (40639), which formerly stood
about 40m apart at Kerrytonlia on the north side
of Kilchattan Bay, established that each comprised
a boulder cairn under a cover of beach sand. A
thin layer of charcoal was detected under the cairn
within the western mound , but no artefacts were
recovered. Whether these cairns represented an
episode of early prehistory that the excavators
simply failed to recognise will now never be known
since both mounds have been destroyed.
The most ubiquitous, yet least understood ,
manifestation of Neolithic settlement is rock
art – the carving of symbols on bedrock and
on stones ranging in size from small portable
boulders to large glacial erratics. There are over
80 recorded examples of rock art on Bute, all of
them comprising either simple cups or cups within
rings. What stands out about this corpus is that it
contains only simple carvings. None is elaborate
or complex, and this contrasts markedly with
neighbouring Argyll.
Although widespread throughout Bute, the
distribution of the rock art appears to favour the
west side of the island and the main concentration is
Scarrel, cup-marked rock (40326). Typical of the rock art on
Bute, this example is located in improved pasture, and bears a
number of cup marks, highlighted by recent rain. The island of
Inchmarnock is visible in the background. DP066517
found in the north. This distribution is probably in
part influenced by the likelihood that this category
of monument will survive (and be detected) in
unimproved land , but it may not be a coincidence
that this is also where most of the chambered cairns
are found. On this note, none of the Neolithic
monuments should be considered in isolation,
with each forming an integral part of the wider
landscape. The pottery and artefacts that were
deposited with the dead in the chambered cairns
were the same wares and tools that were being used
on the settlements.
11
40070
12
The Early and Middle Bronze Age
c 2200BC – c1000BC
Three general features that mark the transition
from the Neolithic to the Early Bronze Age are the
introduction of bronze metalwork, the adoption of
different ceramic forms, and the move away from
the use of megalithic chambered cairns that were
essentially charnel houses used over a long period
of time. New rites included the inhumation and
cremation of single and multiple burials in pits,
cists and pottery vessels, often under specially
constructed mounds (cairns and barrows), but
also sometimes within earlier monuments such
as the chambered cairns, or under conspicuous
natural knolls. No doubt some burials were interred
without anything to mark the grave and there may
well have been other methods of disposing of the
dead that have left no trace in the archaeological
record. It ought to be emphasised that no Bronze
Age houses have been positively identified on Bute
and that any discussion of the pattern of settlement
is dependent on what is known about other
contemporary monuments, burials and artefacts.
Excluding for the moment those features
directly associated with funerary practices, other
monuments within the Bronze Age landscape
of Bute include stone circles, stone settings and
standing stones. The stone circle at St Colmac
Cottages (40338) – the only surviving stone circle
on the island – originally comprised nine stones,
of which four still stand erect, three survive as
stumps and an outlier stands close-by to the southThe distribution of Early Bronze Age sites and artefacts
demonstrates that activity was widespread across the whole of
Bute at this time. However, the main foci of monuments such as
standing stones and stone circles appear to have been in the two
low-lying and narrowest points of the island. GV004573
St Colmac Cottages, stone circle (40338). This stone circle is one
of two that once stood in what is now a cultivated field. The other,
which was situated 25m to the west, is depicted on an early 19th
century plan of Kames Estate (RHP 14262), but it was removed
before 1814, some of the stones having to be blown up by
gunpowder first. This circle survives because by that time it had
already been enclosed within a small plantation. GV004721
west. The circle appears to have stood adjacent to
another about 25m to the west, of which nothing
is now visible. However, it was described by Aiton
13
The Archaeological Landscape of Bute
St Colmac Cottages, stone circle. A number of monuments in the
valley between Kames and Ettrick Bay can be seen to represent
an early prehistoric ritual landscape. The most significant of
these, this circle of eight stones, was set within a protective
plantation bank around 1800, though very few of the trees now
survive. DP064299
in 1816 as ‘a circle of stones, of great size and
height; but they are all thrown down and most
of them removed. Some of them were so large
as to require to be blown with powder’. A late
18th or early 19th century estate plan appears to
show an arc of five stones on the west side of the
current circle.
One of the most interesting features of the
site is its position within the local topography.
Overlooking Ettrick Bay to the south-west, the
surviving stone circle stands within a natural
amphitheatre in which there are also three
standing stones (40316, 40317, 40360), one of
them (40317) bearing a later early Christian cross.
Also, the stone circle is overlooked by what is
likely to be the roughly contemporary Watch Hill
14
burial cairn (40442) at the south-east end of the
bay and a possible cist cemetery (40071, 239544)
above Kildavannan at the north-west end. Even
if the Bronze Age ritual landscape here has been
reduced to a series of individual monuments
isolated from each other within modern farmland ,
the range and density of those sites provides
enough evidence to suggest that this location
was somehow special, apparently more so than
anywhere else on Bute. Further, these monuments
appear to represent a major change in focus in
ceremony, ritual and burial activity away from
what are today the marginal lands in the northern
third of the island favoured by the Neolithic tomb
builders, to a much more fertile farming landscape
only a relatively short distance to the south-east.
An earthwork, Cnoc an Rath (40312), which
stands on the watershed on the north-east edge of
the amphitheatre, has been proposed as a henge.
If this is the case then the monument may have
served as the main focus for ritual, ceremony
and funerary practices throughout the Late
The Early and Middle Bronze Age
St Colmac cross, standing stone with later inscribed cross
(40317). Standing near the site of Colmac’s chapel (40349), now
removed, this stone may have originally formed part of a Bronze
Age ritual landscape. Recorded in the 1990s as part of a long
running RCAHMS survey of medieval carved stones, it forms
part of an important corpus from Bute, much of which is now
housed in Bute Museum. SC408104
Cnoc an Rath, earthwork (40312). This site, situated on the
watershed between the bays of Ettrick and Kames, is defined by a
ditch and an inner bank, with a causewayed entrance on the eastsouth-east. First surveyed by RCAHMS in 1982, the date and
function of the monument remains elusive. SC1208314
Neolithic and Early Bronze Age. However, there
is no dating evidence from the site, which others
have suggested is an Iron Age fort, a medieval
ringwork, and , most recently, a Norse ‘ting’ site.
Elsewhere on the island there is a scatter of
standing stones and stone settings which are
presumed to be of Early Bronze Age date. Of
the latter, the three-stone alignment at Largizean
(40257) at the south end of the island is similar
to many other examples elsewhere in Britain, and
it may be significant that the stones stand only
about 350m south-east of the find spot of a hoard
of bronze halberds found in the 19th century
(see below). Another stone setting, at Blackpark
Plantation (40258), stands just 850m east of the
hoard site. First referred to as ‘the remains of one
of the circles of the Druids’, the evidence provided
by various descriptions of the site is at best
ambiguous, and there is little compelling evidence
to show whether the stones belong to a circle that
once comprised more than the three stones that are
visible today or some other form of setting. There
must also be some doubt about a setting of two
stones at St Ninian’s Bay (40426). Measuring only
up to 1.4m in height and 1.3m apart, the stones
are slighter than might be expected of a Bronze
Age monument.
15
The Archaeological Landscape of Bute
16
The Early and Middle Bronze Age
James Skene’s sketch and plan of the ‘druidical monument’
at Blackpark (40258) captures the unusual character of the
individual stones. The original sketch, dating to 1818, is part of a
collection from the the Society of Antiquaries of Scotland, gifted
originally by James’ son. We are also fortunate to hold a larger
version, reproduced here, redrawn in a sketchbook and probably
dating to around 1832. DP084695
There are at least 30 Early Bronze Age burial
sites on Bute that have produced evidence of either
single burials or cemeteries. There are also seven
cairns, the characteristics of which suggest they
are probably also of this date. The whole corpus
of Bronze Age burials cannot be reported here in
detail, but a few sites deserve particular mention.
At four sites graves have been inserted into earlier,
Neolithic, cairns. At Glecknabae (40371) one of
the two chambers contained sherds of a Beaker,
possibly indicating re-use of the tomb in the Late
Neolithic/Early Bronze Age. Later still, a cist
containing an inhumation was inserted into the
west edge of the cairn. A vessel that was described
as ‘looking like a flowerpot’, which was found
and then lost before excavations in the early 20th
century, was probably another secondary burial.
At Glenvoidean (39897), an Enlarged Food Vessel
and a cremation were found in an unusual twotiered cist, and at Hilton (40306) Marshall found an
apparently empty cist that had been sunk into the
body of the chambered cairn but which had , in turn,
been covered by its own discrete mound. At Bicker’s
Houses (40381), the possible cist inserted into the
mound close to its south end , which was noted as
‘rifled’ by Hewison, can no longer be seen.
Fine spacer-plate necklaces of jet or jetlike material have been found accompanying
inhumations in cists at Kerrycroy (40722) and
at Northpark on Inchmarnock (40427). As high
prestige artefacts with a broad distribution across
northern and western Britain, they provide
striking evidence of how rich some sections of the
Early Bronze Age population on Bute were. The
Kerrycroy example was found along with a Food
Vessel and a corroded bronze object in 1887; the
Northpark necklace, which comprised 139 beads
Scalpsie, barrow (40232). Centred on the Bronze Age barrow,
this view captures the character of Bute’s Improvement
landscape. The regular fields, the timber plantations, and the
well spaced farmsteads are all products of this period, and
the water from the distant hill slopes once helped to drive
Rothesay’s mills. DP083916
Recently relocated, the Bronze Age cist at Mount Stuart was
first opened in 1887 (40722). Careful, given the standards of the
time, the excavations yielded a crouched burial, a Food Vessel, a
bronze awl, and a jet necklace, all of which were donated to the
National Museum in 1888. © Trustees of the National Museums
of Scotland. Licensor www.scran.ac.uk. 000-100-035-704-R
recycled from at least four other necklaces, was
found in one of three cists discovered in 1960 and
originally excavated by Marshall. One of the other
cists yielded an old and worn spacer plate from
another jet necklace.
Recent research has seen the re-excavation of
the cist that contained the necklace. A sample of
bone yielded a radiocarbon date around 2000BC, in
the middle of the Early Bronze Age. Bone isotope
analysis, undertaken as part of the Beaker People
Project, revealed that the bones were those of a
woman who was almost certainly local and who
did not eat fish – a surprise considering that she
was buried (and therefore presumably lived) on a
comparatively small island.
In addition to the Northpark cist, only one other,
at Cnoc-na-mhanan (239544), has been scientifically
dated. Here, one of what may be a cemetery of cists
contained a miniature Food Vessel and a cremation
that has been dated to between 2000 and 1700BC.
Another cemetery, which comprised no less than
17
The Archaeological Landscape of Bute
eighteen cists and was found beneath a cairn at
Ardbeg (40315), provides a striking contrast to
the single burials or those in small groups. It
was dug into before 1840 and again in 1858, but
nothing now survives either of the cairn itself or
of the several ‘sepulchral urns’ that were found in
the graves.
Edwina Proudfoot has recently reviewed the
evidence for Early Bronze Age burials on Bute
and , notwithstanding those sites (mainly noted
18
by antiquaries) from which the evidence is either
dubious or, at best, ambiguous, the picture presented
suggests that probably the whole of the island
(and Inchmarnock) was exploited by a farming
population whose material culture was no different
to that on the mainland to the east and Argyll to
the north and west. This distribution is heavily
influenced by the discovery of burials mainly
through the agencies of arable cultivation, quarrying
and construction, but it should be recognised that
The Early and Middle Bronze Age
A group of three cists were found in 1960 by a farmer at
Northpark Inchmarnock (40427). One of the cists contained
the remains of a single woman, accompanied by a flint knife
and a spacer plate necklace of 139 jet beads, re-strung and
photographed by Alison Sheridan, NMS. Re-excavation in 2006
formed part of a significant re-assessment of the discovery by by
Bute Museum and NMS Archaeology Department. © Trustees of
the National Museums of Scotland.
weapons, the precise function of those of Early
Bronze Age date is not known. The circumstances
surrounding the deposition of the hoard are also
unknown, but the find-spot is located in an area
of poor natural drainage and it is possible that the
halberds were a votive offering within an originally
wetland location.
activities such as these occur in the very areas that
are likely to have most attracted early farming
populations. In mitigation, though, an additional
pointer to the use of the landscape at this time is
provided by those Bronze Age burials in Neolithic
cairns that are situated on or just outwith the limit
of recent intense arable cultivation.
A full consideration of the Bronze Age pottery
from Bute is beyond the scope of this booklet and
commentary is restricted here to noting that there
is a paucity of Beaker pottery – especially outwith
the chambered tombs, there is only one confirmed
discovery of a cinerary urn (40237), but Food
Vessels are comparatively common, with no fewer
than eight examples being found.
In addition to the badly corroded metal object
found in the Kerrycroy cist noted above, there has
been only one recorded discovery of Early Bronze
Age metalwork on Bute – that of five halberds
(40255) found in the 19th century when the
farmer at Langalbuinoch was digging a drain. A
halberd comprises a long, axe-like, blade attached
at right-angles to a wooden or metal shaft. While
medieval versions were designed and used as
Five Bronze Age halberds (40255) were found in the mid 19th
century by a farmer digging a drain on Langalbuinoch farm.
Three survive in the National Museum of Scotland, where they
have been dated stylistically to between 2300BC and 2000BC.
© Trustees of the National Museums of Scotland. Licensor
www.scran.ac.uk. 000-190-004-241-R
19
20
The Late Bronze Age and Iron Age
c1000BC – cAD600
If the Early Bronze Age on Bute is categorised by
a lack of settlement sites in favour of ceremonial,
ritual and funerary monuments and the Middle
Bronze Age is marked by a total dearth of identified
monuments of any description and no artefacts, then
the Late Bronze Age (c1000BC – 700BC) at least
offers some tangible evidence for settlement. Any of
the prehistoric round-houses or burnt mounds so far
recorded could date from this time and there is also
evidence for metalworking in this period.
Prehistoric round-houses take two forms on
Bute: hut-circles, where the building is preserved as
a low, roughly circular stony bank, and platforms,
where all that is now visible of the structure is
the artificial terrace, dug into a hillside, on which
it was built. Perhaps the best example of one of
these platforms is that built within the earthworks
at Dun Scalpsie (40254). It is impossible to date a
prehistoric round-house from visible evidence alone
and any of them could date to a period spanning
over a thousand years. None of the ten examples so
far recorded on Bute has been excavated.
A burnt mound is a large pile of fire-shattered
stones, charcoal and soil that is thought to have
accumulated as a result of an open-air cooking
process. Three burnt mounds have been recorded
on Bute and there is one possible example on
Inchmarnock. The burnt mound at Mecknock
Distribution of sites and artefacts of the Late Bronze Age and
the Iron Age. Many of the settlements identified here could have
been built and/or occupied within a period lasting up to 1,800
years. The bias in the distribution of forts and duns on the west
coast may reflect contemporary politics, though the small fort at
Kerrytonlia (40643), visible today as a cropmark, indicates that
similar sites existed on the east coast. GV004574
Barr Hill, hut-circle and cultivation remains (124782). This
hut circle, with its thick stone wall and east-facing entrance,
is typical of the small prehistoric round-houses found all over
Scotland. The cultivation adjacent to it is probably medieval
or later in date. GV004729
(40267), up to 10m across and 1.4m in height, has
been excavated but has not been dated. Elsewhere
in Britain and Ireland , while similar mounds have
generally been found to date to the Middle and Late
Bronze Age, others have been found to be Early
Historic or even medieval in date.
The Late Bronze Age and the first part of the
Iron Age – the centuries spanning the end of the
2nd millennium BC and the beginning of the 1st
millennium BC – was a period of great transition
across Scotland. It was a time of social upheaval,
which saw the widespread abandonment of
permanent occupation in upland locations affected
by increased rainfall, cooler temperatures and the
21
The Archaeological Landscape of Bute
An oblique aerial view of the vitrified fort of Dunagoil (at centre
right, 40291) and the fort of Little Dunagoil (at centre left,
40280), taken from the north-west. There is no evidence
of fortification on the intervening ridge, but the landscape
between the two forts is dotted with the remains of burial
cairns, clearance cairns and cultivation remains. DP062659
formation of peat. The resulting pressure on the
most fertile and productive land is thought to be
one of the reasons why forts and other forms of
defensive enclosure began to be constructed.
However, whilst there are a comparatively
large number of sites that can fairly confidently be
assigned to the 1st millennium BC, the settlement
pattern on Bute at this time is not at all well
understood. None of these sites has produced
good dating evidence from reliable contexts
and it is likely that the picture is complicated by
several sites being re-used in the Early Historic or
medieval periods.
The vitrified fort on Dunagoil (40291), at the
south end of Bute, and the fort on the summit of
Barone Hill (40424), overlooking Rothesay, are the
22
only two large forts on the island. Barone Hill is
undated , but the partial excavation of Dunagoil
nearly a hundred years ago recovered artefacts
dating from the Neolithic to the Iron Age, and
included evidence for bronze working. A vitrified
fort is one that had a high and thick stone wall,
internally braced by timbers, which has been burnt
to such a high temperature that large portions of its
rubble core have been fused into a solid mass. As a
technique, the use of timber-lacing cannot be used
to date forts, but those which have been dated by
other means have been shown to lie within a long
period spanning the first millennium BC and the
first millennium AD.
The excavation account of Dunagoil is vague
about the exact nature of the metalworking, but
given the close proximity of the site to Little
Dunagoil (40280), where socketed axes were being
produced in the Late Bronze Age, it seems fair to
infer that it, too, was probably occupied in some
form at that time. Most of the artefacts found
during the excavation of Dunagoil are undiagnostic
The Late Bronze Age and Iron Age
Map of the archaeological landscape at Dunagoil, including: (A)
the vitrified fort (40291); (B) a probable round-house platform;
(C) the lengths of wall that partly define the enclosure north-east
of and below the fort; (D) the fort on Little Dunagoil (40280);
(E) the later buildings and building-platforms (90299), and
burial cairns (40070). GV004726
in terms of date, but they include a brooch (dating
to the 3rd century BC) that probably provides some
indication of when the actual fort was occupied.
The results of the excavation at Little Dunagoil fort
were inconclusive.
Dunagoil and its environs, including the
fort on Little Dunagoil, have recently been
reconsidered by D Harding of Edinburgh
University. His conclusion – that the excavations
and artefacts demonstrate that the Dunagoil
‘complex’ attracted settlement from the Neolithic
to the Early Historic periods is sound enough, but
he has also radically reinterpreted the fort and
its relationship to the immediate topography and
other features within the landscape. Foremost in
this reinterpretation is his proposal that adjacent
23
The Archaeological Landscape of Bute
to and contemporary with the timber-laced fort was
what he calls a ‘bailey’. This comprised a terrace
immediately north-east of and below the fort that
was defined largely by topographical features,
Little Kilchattan, cropmark enclosure (40644). The transcription
of oblique aerial photography has allowed cropmarks across
Bute to be accurately mapped and their features measured. In
the case of this enclosure, which was first recorded in 1977, its
purpose is unknown, but its 4m broad ditch suggests that it may
have been for something more significant than stock control.
SC1203299
24
which offered excellent natural protection, linked
by lengths of timber-laced wall (C on the plan
on page 23). The claim that this enclosure was
contemporary with the summit fort is based on an
observation that there is in situ vitrified bedrock
visible in one of these lengths of wall at the northwest end.
The presence of vitrified material, wall core
rather than bedrock, was confirmed in the present
survey. However, the function and date of this wall
has yet to be established , and it is difficult to see
The Late Bronze Age and Iron Age
how a ‘bailey’ may have worked: there is a gap of
some 100m between the south-east ‘bailey’ wall
and the summit fort, a gap that would have allowed
relatively unimpeded access to the lower enclosure
from the south.
Whether these large forts were built as tribal
centres, community gathering places or simply
for the purpose of making ostentatious political
statements may never be known, but, whatever
the reason, large forts probably served a very
different purpose to smaller sites, which were
most likely built by small farming communities
as a response to some perceived threat. The
majority of these sites on Bute are relatively
small, thick-walled enclosures that have variously
been classified as forts and duns. However, the
distinction between the two is sometimes a fine
one and the classification used by RCAHMS in its
recent revision of the Bute archaeological records
is based on a definition of forts and duns used in
the first Argyll Inventory. This simply states that
the internal area of forts was large enough to serve
the needs of a small community while duns were
capable of accommodating only a single family.
This fairly loose definition is almost certainly
an over-simplification, but using it means that the
enclosures at Kerrytonlia (40643), Little Dunagoil
(40280), Clachan Ard (40269) and probably
Ardnahoe (40242) – the smallest fort with an
internal area measuring 25m by 23m – have been
added to the list of forts. The enclosure at Balilone
(40378), with an internal area of only 24m by 19m,
is on the cusp between fort and dun. Previously
described as a fort, the enclosure at Aultmore Burn
(39906), which has an internal area of only 19m
by 16m, has been reclassified as a dun and takes
its place with four others – Dunstrone (40276),
Dun Scalpsie (40254), Dun Burgidale (40300)
and Castle Cree (40441). The last mentioned site,
however, is unusually narrow and rectangular and it
may actually be medieval.
Few of the forts or duns display any notable
architectural features, despite several of them
having been investigated by excavation. Exceptions
are Dun Scalpsie, where a door-check is still visible
within the entrance passage, and Dun Burgidale,
where there are almost certainly passages between
the walls.
As far as the distribution of the forts and
duns is concerned , there appears to be a definite
The cropmark fort at Kerrytonlia (40643) is one of a number of
archaeological sites that were first noted during an aerial sortie
undertaken by RCAHMS on 7 July 1977. SC1203301
emphasis on the west coast, though some
allowance should be made for sites that have
either been destroyed or have yet to be recognised.
Late 18th and 19th century local historians
made unsupported references to all manner of
archaeological monuments, including forts which
had already been destroyed. In part, that lost record
has been retrieved by aerial survey, which, since
1977, has revealed the cropmarks of a number of
sites, in particular broad-ditched enclosures that
may represent the earthwork equivalent of the
stone-built sites.
One of these sites is at Kerrytonlia (40643),
on the north side of Kilchattan Bay. Here, a
promontory extending from the old sea-cliff has
been defended by two broad arcing ditches drawn
across its neck, the inner one accompanied by
a narrow palisade trench along its inner edge.
Whether this last feature was an integral part of a
system of defence with the ditches or represents
an independent period of enclosure is not known.
Its presence, however, is a useful reminder that,
irrespective of how well preserved a site may
appear on the surface, more ephemeral features
will seldom be seen and the visible remains
will only tell a fraction of the full story of any
particular site.
That some sites enjoyed extended periods
of use has already been noted at Dunagoil, but
25
The Archaeological Landscape of Bute
the same is also true at Little Dunagoil (40280)
and probably also at Dun Scalpsie (40254).
Little Dunagoil is noted in the next chapter, but
Dun Scalpsie, which occupies the summit of a
rocky knoll on the north side of Scalpsie Bay, is
of considerable interest because it is one of the
few sites on Bute where more than one phase of
construction can be demonstrated from the visible
evidence alone. Until recently, the site was thought
to comprise simply a dun with a narrow entrance in
its north-west side (A on plan). However, survey by
RCAHMS in 2010 has demonstrated the presence
of a very substantial stone wall (B) lower down the
slope on the south and east. This wall follows the
Dun Scalpsie was partly excavated in the 1950s (40254). This
plan shows the dun (A) that sits on the summit; the wall of the
possible fort (B) below; the timber round-house (C), and the foxhole type shelters (D) that were probably constructed by local
defence volunteers (LDV), later the Home Guard, during the
Second World War. GV004725
26
crest of a steep slope on the north-east and appears
to underlie the wall of the dun on the north. On
the west flank there is little trace of the wall, but it
may have been destroyed by a ditch that was dug
to protect the later dun. The visible evidence is
ambiguous and it is not possible to state whether
the outer wall represents a fort that occupied the
rocky knoll before the dun was placed on the
summit or is a contemporary outwork of the dun,
providing additional defence on its vulnerable
flanks. The results of a trench dug inside the dun in
1959, which revealed two periods of occupation,
is not helpful in this respect. Also unhelpful is
the disturbance to the summit area caused by the
construction during the Second World War of
several dug-outs (D).
Of particular interest at Dun Scalpsie, however,
is the recognition of a stance for a timber roundhouse (C) immediately outside the wall of the dun
and within the area of the outer enclosure. The
exact context of this building, which is one of only
a small number of timber round-houses recorded on
the island , is not known. It may be contemporary
with either the earliest or later enclosures on
the knoll; equally it may represent a period of
settlement that has no association with either phase
of enclosure.
Something similar may be said of what is
probably the stance for a round-house on the southwest-facing flank of Dunagoil, some 50m south of
the fort (B on plan on page 23). Measuring little
more than 5m in overall diameter, the undated
structure sits on a very steep, craggy area of
ground characterised by short lengths of wall, some
incorporating blocks of vitrified stone, which define
a series of discrete enclosures.
Other forms of settlement that deserve mention
are crannogs, and caves or rock-shelters. A crannog
is an artificial island built to support a building,
usually of timber. There are three crannogs on
Bute: one in Dhu Loch (40380) and two in Loch
Quien (40233, 40234), all undated. The crannog in
Dhu Loch was discovered in 1812, when the water
level was low enough for a plan to be drawn of the
remains, which included horizontal timber beams
and numerous piles. Subsequently the water level in
the loch was raised and the site now lies completely
submerged. The two crannogs in Loch Quien,
which are visible today as wooded islets close to the
shore, were both surveyed in 2003.
The Late Bronze Age and Iron Age
Dhu Loch, crannog (40380). First surveyed in the early 19th
century, this crannog was subsequently completely submerged
when the loch became a reservoir for Rothesay’s water supply.
There are three crannogs on Bute, and a survey in 2003 resulted
in the discovery of a rotary quern stone at Quien north (40234),
now at Bute Museum. SC1163938
Caves and rock-shelters (shallow cave-like
openings at the base of a cliff, often with a wall
built across the mouth), have been used in a variety
of ways by humans from earliest prehistory to
recent times. Two rock-shelters at the south end of
Inchmarnock (300171, 300178) have been shown
to contain evidence of some form of use in the
second half of the 1st millennium BC. Dorothy
Marshall considered the cave in the cliff below
the fort on Dunagoil (40278) to have been used
in the Iron Age as well as the Neolithic. As noted
above, several caves and rock-shelters contain
no obvious evidence for any human activity, but
only a relatively small number on Bute have been
examined. With more sites of this type yet to be
recorded , caves and rock-shelters may represent
one of the best opportunities for recording in situ
later prehistoric deposits.
27
28
The Early Historic and Medieval Periods
cAD600 – cAD1600
For Bute, it is convenient to consider the Iron
Age as ending at the start of the Early Historic
period , which for the purposes of this text has been
assumed to commence about the beginning of the
7th century AD and end around the 10th century.
Vikings are only briefly referred to below, and
actual Norse settlement in the 11th to 13th centuries
is considered to be medieval, a period accepted here
as running to the end of the 16th century.
In the Early Historic period Bute found itself
sandwiched between the power-bases of Dal Riata
in Argyll to the west and Strathclyde to the east. In
what must have been extremely turbulent times, a
dynamic landscape was first disturbed by the Viking
incursions in the 7th and 8th centuries and later,
during the 12th century in particular, became heavily
influenced by the Kingdom of Man and the Isles.
Finally, it was included in the Stewart kingdom.
Currently, our level of knowledge about the
archaeology of this time is limited by a paucity of
securely dated sites and a comparatively modest
amount of documentary evidence. Consequently,
almost nothing is known of the settlement pattern
at this period and the little knowledge available is
necessarily biased towards three major sites – the
monastic settlements at ‘Cenn Garah’ (Kingarth;
40292), at the south end of Bute, and that dedicated
to St Ernan on Inchmarnock (40268), both probably
established around the beginning of the 7th century
AD, and at Little Dunagoil (40280), where Dorothy
Marshall’s excavations in the late 1950s and early
Distribution of Early Historic and Medieval sites and artefacts.
As with the preceding periods, the record is dominated by high
status religious and secular establishments. GV004575
St Blane’s Church (40292). An oblique aerial view of the 12th
century church and graveyard, taken from the north-west.
The upper and lower burial-grounds, both to the right of the
church, are reputed to have been reserved for men and women
respectively. What may be a medieval chapel is visible in the
lower graveyard. DP066101
1960s yielded artefacts indicating activity in the
second half of the 1st millennium AD.
The evidence from the excavation of the
monastery on Inchmarnock has yielded a wealth of
information, detailed in an excellent book (Lowe
2008), but little is known about contemporary
secular settlement there. At Kingarth we know the
names of two of the 7th century bishops but, despite
the late 19th century exploration there at the request
of the 3rd Marquess of Bute, we know very little
about that establishment’s relationship with either
its lay neighbours or the wider landscape.
The documentary evidence for the Early Historic
period in the upper Firth of Clyde implies that Bute
contained a flourishing population, the spiritual
29
The Archaeological Landscape of Bute
St Blane’s Church (40292). This early photograph, taken
around 1895, of the medieval church shows the nave prior to its
restoration, which was undertaken by the architect RW Schulz,
and sponsored by the 3rd Marquess of Bute. The photograph
was presented to the National Buildings Record (now part of
RCAHMS) by Mr A Curtis, whose given address at 14 Grays Inn
Square, London, was the former office of Schultz. SC1161664
needs of which were serviced through a network
of chapels. The existence of these chapels is not
documented , although their presence can surely
be assumed to coincide with the occurrence of
cross-slabs at Kilmachalmaig (St Colmac), where
the carved stone (40317) still stands; at St Mary’s
Church, Rothesay (40373); and at an unidentified
site near Rothesay Castle (301961), where a broken
slab (now in Bute Museum) was found in the early
19th century. The MacAlister Stone (40406), also
Rothesay Castle, crossslab (301961). This
rectangular cross-slab
of sandstone was found
in Rothesay Castle in
1816, re-used in two
fragments at the foot of
the stair in the northwest tower. It remained
there until some time
after 1903 when it was
moved to Bute Museum.
SC403493
30
now in Bute Museum, probably also came from
a chapel, the site of which is not known. Apart
from St Ninian’s Chapel (40437), which has been
excavated , there is little evidence that any of the
other surviving chapels date earlier than about the
12th century AD.
The assumption must be that there was a
relatively thriving population that could sustain the
infrastructure implied by the presence of a network
of religious establishments. Elsewhere in the north
and west of Scotland there is evidence that some
duns were reoccupied in the Early Historic period
and in this context we can note that at least one
Norse place-name, Dunburgidale, may be linked
to a specific structure of earlier date. However,
the majority of Bute place-names are later, many
being derived from Scottish Gaelic, and the best
evidence for a continuity of settlement is provided
by the fort on Little Dunagoil (40280) and its
immediate environs.
Rothesay High Kirk, cross (40406). This cross-shaft, known
as ‘MacAlister’s Stone’, was re-erected in 1886 after lying
recumbent for many years on a burial-plot belonging to a family
of that name. Tradition associated it with the Ascog area, about
2km south-east of Rothesay, but there were conflicting accounts
of its earlier provenance. Hewison tentatively identified it with a
stone which formerly stood at Kilwhinleck (NS c.0580 6230). The
stone was removed to Bute Museum in 1996. SC403494
The Early Historic and Medieval Periods
The fort (40280) and farmstead (90299) on Little Dunagoil,
showing the location of: (A) the summit fort; (B) the later
addition to its east side; (C) the additional wall on the north
side; (D) the cluster of subrectangular building platforms,
and (E, F) the buildings (farmstead) partly excavated by
Dorothy Marshall around 1960. GV004724
Excavations demonstrated that the fort had a
long history stretching from prehistory through
into the Early Historic period. However, while the
excavations demonstrated the depth of chronology
of the site in general, it failed to reach an adequate
understanding of the defences and the structural
remains within the fort. The resurvey of the site
in 1994, part of a reassessment of the Dunagoil
complex, had a fresh look at the earthworks without
radically altering the previous interpretation of
the remains.
The most recent survey, undertaken by
RCAHMS in 2010, shows that the small thickwalled enclosure on the summit of the rocky knoll
(A on plan) was added to on the east by another
enclosure (B) that effectively doubled the size of the
defendable area. It was probably at this time that a
short length of additional wall (C) was added on the
north. Perhaps more important is the recognition
of a group of subrectangular building platforms
(D), which lie below the east end of the fort. These
platforms are very different in character from the
large 12th/13th century buildings (E, F) previously
investigated and are demonstrably earlier than those
buildings (90299).
The significance of these platforms cannot
be overstated as they represent part of the Early
Historic occupation of the hill that probably
dates to a period after the fort went out of use but
before the 12th or 13th century buildings were
constructed. This small group of platforms therefore
represents the most tangible link between the
monastic settlement of Kingarth, if it survived into
the 9th and 10th centuries, and its nearest secular
neighbour. That link may have continued into the
late Norse period in the form of the two large halllike structures which are quite unlike any other
buildings so far recorded on Bute.
31
The Archaeological Landscape of Bute
The hall-like buildings at Little Dunagoil
excepted , the only structures of the medieval period
(13th to 16th centuries) that are readily identifiable
in the landscape are churches, chapels and castles.
The churches and chapels are predominantly
found on the west coast, perhaps an indication of
the direction from which the inspiration for their
foundation derived. At Kingarth, a parish church
(St Blane’s) succeeded the monastic settlement
(40292) and there was also a parish church (St
Mary’s) in Rothesay (40373). Inchmarnock may
have had medieval parish status, but it is more likely
that Rothesay and Kingarth continued as the only
parishes until the creation of North Bute in the
19th century.
St Michael’s Chapel and Burial Ground (39902). This aerial view
taken from the north-east shows the ruinous medieval chapel set
within its circular burial-ground. DP067130
Kilchousland Chapel (40436), which lies within an enclosure
formed by a thick bank on the south-west and the face of a
relic shoreline cliff on the north-east. The original entrance to
the enclosure, which may have served as a burial ground, was
probably situated in the broad gap in the wall on the south.
GV004723
St Blane’s, cross
no. 10 (301996).
Free-standing cross
of white sandstone,
re-used as a grave
slab south of the
nave of the church.
SC403036
32
The Early Historic and Medieval Periods
In addition to the medieval parish churches there
was a network of chapels, one in Rothesay Castle
(301959) and a series of others – St Michael’s
(39902), St Ninian’s (40437), Kilchousland
(40436), Kilmory (40260) and East St Colmac
(40349) – dotted at fairly regular intervals along the
western seaboard. A chapel may also have existed
at Chapelton (40388), where an alleged font was
Rothesay Castle (40395) is the most powerful symbol of Bute’s
medieval period. Sitting in a prominent location that dominates
the town, it is surrounded by a moat that was redesigned as a
garden in the late 18th century. SC800098
once seen, and the possible locations of others,
including Acholter (40314), Cnoc-na-Mhanan
(40369), Kilwhinleck (302265), Kilbride (40372)
and Kilmory Hill (294600), are hinted at either in
tradition or in place-name.
Medieval castles existed at Rothesay (40395),
Kames (40309) and possibly also at Kelspoke
(40674), Ascog (40724) and Wester Kames (40298).
Buildings associated with the Stewart dynasty
include Meikle Kilmory (40374), where there is
evidence for what may have been the house of the
‘Sheriff Crowner’ of Bute in the 16th century. In
contrast to the majority of the churches and chapels,
33
The Archaeological Landscape of Bute
Kames Castle (40309). This tower, one of only a handful of
medieval strongholds on Bute, is surrounded by a suite of
19th century buildings, and set within extensive later gardens.
DP010303
these secular foci are usually found on the east side
of the island , and this may be an indication that by
that time the people of Bute were focusing more on
the Scottish mainland than anywhere else.
Other than castles, towers and religious
buildings, evidence for rural settlement on Bute
in the medieval period is mainly restricted to
the mention of farms and estates in documents.
With the exception of two excavated kilns (see
below), no rural farm buildings earlier than 18th
century in date have been identified on Bute,
though recent excavation at Quien (88671) may
have uncovered a small medieval building and
enclosure. It is clear, however, from the detailed
histories published in the 19th century that the
majority of the modern farms on Bute have roots in
land-holdings that were already in existence in the
late medieval period. Exchequer rolls from 1450
include references to many names that are still in
use today, but it is difficult to tie these references to
34
particular locations and impossible to link them to
specific structures.
Nevertheless, there are several locations where
there are groups of rectangular buildings, albeit
surviving only as grassed-over wall-footings, which
are not depicted on detailed estate maps of the
mid to late 18th century, the inference being that
they were already abandoned by that time and as
such could belong to a much earlier period. The
problem here, though, is a 13th century building
that has been reduced to footings will probably
look very similar to one dating to the 15th, 16th or
17th century. These potentially early sites have been
recognised in the north of the island in particular –
including on the lower flanks of Fly Hill (124772,
124773, 302308) and Edinbeg Hill (124763) and
around Split Craig Hill (139140, 294622) in the
comparatively remote north-west corner. One of
them (124793), which is tucked into a corner of
a pasture field well above Stuck farmsteading,
comprises two buildings beside a stream. It is often
difficult to establish from field evidence alone what
the function of buildings that have been reduced to
footings has been. Here, however, it is possible to
state with some confidence that the building which
The Early Historic and Medieval Periods
that the building which has a drain running out of it
is a byre and that the other structure was a dwelling.
That medieval buildings can and do survive
within a heavily modified landscape has been
demonstrated at St Blane’s (302028) and more
recently on Inchmarnock (300170), where corndrying kilns have been excavated. However,
medieval kilns take much the sam