Cornwall during the Iron Age and - Cornwall Archaeological Society

CORNISH ARCHAEOLOGY No. 25 (1986) 

50 

IRON AGE and ROMANO-BRITISH 

SITES 

5/0 

Cornwall during the Iron Age and 

the Roman Period 

20 

HENRIETTA QUINNELL 

In later British prehistory change in the archaeological record is currently interpreted as 

the result of contact and exchange between contemporary communities. Detailed studies of 

cross-channel trade and of the development of boats show that frequent contact was feasible 

(eg Cunliffe, 1982). There may have been some migration on a small scale, but the idea that 

major invasions produced social and cultural change has been discredited. This is partly 

because so little material of continental origin has been found in Britain; partly, and more 

importantly, because underlying insular trends in farming, house construction and defence 

have been seen to have deep local roots, and differ substantially from practices along the 

northern littoral of Europe. Cunliffe's survey Iron Age Communities in Britain (1974 and 

1978) explains these trends and simplifies comparison of the Cornish evidence with that from 

the rest of Britain. He describes the British Iron Age in terms of regional tribal areas and 

of chronologies which are increasingly based on radiocarbon dates. Cornwall comes under 

the broad tribal grouping of the Dumnonii (1978, 111) but the study shows how poorly the 

Cornish archaeological record for this period is understood, compared to other areas of 

southern Britain. In particular, the range of radiocarbon dates from Cornwall is not yet 

Fig 1 

Location showing principal sites mentioned in the text. 1 Porthmeor, 2 Cam n Euny, 3 Chysauster, 4 Goldherring, 

5 Mulfra Vean, 6 Maen Castle, 7 Bodrifty, 8 Chun, 9 Gurnard's Head, 10 Bosence, 11 Crane Godrevy, 12 Porth 

Godrevy, 13 Magor, 14 Cam Brea, 15 Trebarveth, 16 Carngoon Bank, 17 Halligye, 18 Grambla, 19 Carwarthen, 

20 Threemilestone, 21 Carvossa, 22 Shorlanesend, 23 Trethurgy, 24 Castle Gotha, 25 Kilhallon, 26 Castle Dore, 

27 Trevisker, 28 St Mawgan in Pydar, 29 Castle-an-Dinas, 30 Penhale, 31 The Rumps, 32 Tregeare Rounds, 33 

Killibury, 34 Trevelgue, 35 nanstallon, 36 Harlyn Bay, 37 Trevone, 38 Trelan Bahow, 39 Stamford Hill, 40 Mount 

Batten, 41 Widemouth Bay. 

111 

30 

40 

5

adequate for the establishment of a detailed local chronology, although those we have present 

a very different outline picture to that perceived 28 years ago, when Dudley (1958) wrote 

the previous review article. Sites not referenced in the present paper are detailed in that 

earlier review. 

The terminology employed to describe the Cornish Iron Age has altered considerably, and 

confusingly. The former ABC system was based on the concept of invasions triggering 

archaeological change, A deriving from the continental Hallstatt Iron Age, B from La Tem 

and C from Belgic groups, but this concept is no longer generally acceptable. In papers on 

Cornish material published up to around 1970 the ABC system was widely used, generally 

with a late chronology which radiocarbon dating now allows us to lengthen. As we cannot 

yet refer to a detailed local chronology, the Cornish Iron Age tends to be divided at present 

into an Earlier Iron Age, to around 400 BC, and a Later Iron Age, and this simple scheme 

will be used in this paper. 

In Britain, iron gradually replaced bronze for all the main metal necessities of life, for the 

tools of fighting and farming, during the 7th century BC. A date around 600 BC may 

therefore be appropriate for the beginning of the Cornish Iron Age. Few iron objects survive 

in the acid local soils and all belong to the Later Iron Age. Iron smelting on a comparatively 

extensive scale has been identified at Trevelgue, and it is possible that iron from the South 

West was reaching Danebury hillfort in Hampshire by the 6th century BC (Salter and 

Ehrenreich, 1984, 151 —2). Traditionally, changes in pottery and the introduction of hillforts 

havebeen used to identify the beginnings of a local Iron Age in which iron objects were sparse. 

Pottery of the Cornish Earlier Iron Age (formerly Iron Age A) is generally undecorated and 

includes a range of shouldered jars and bowls as at Bodrifty (Dudley, 1956) which are 

regarded as copies of metal vessels first introduced during Late Bronze Age II, in the 8th 

century BC (Pearce, 1983, 90). It seems reasonable to assume that the pottery copies started 

somewhere in Britain during the currency of the bronze forms. In Wessex where sufficient 

data exist for a detailed chonological ceramic sequence, it can be shown that some pottery 

formerly regarded as 'Iron Age' belongs to the closing stages of the Bronze Age (Cunliffe, 

1983). In Cornwall, the range of sites producing Earlier Iron Age pottery has not been greatly 

increased in recent years. To the open hut-circle settlements at Bodrifty (Dudley, 1956) and 

at Garrow (Dudley, 1958, 48) have been added Sperris Croft and Wiccca Round in West 

Penwith (Dudley, 1958a), possibly Glendorgal, near Newquay (Dudley, 1962), and Kynance 

Gate (I. Thomas, 1960). Scatters of material on the north coast summarised by Dudley (1958) 

now include material from Gwithian (Thomas, 1964, Fig 21). Maen Castle and Trevelgue 

cliff castles remain the only fortified sites producing Earlier Iron Age pottery. In the present 

state of knowledge these sites may date anywhere between about 800 and 400 BC, before 

or after the introduction of iron. The term Earlier Iron Age, when applied to, or derived 

from, pottery in Cornwall, in fact means Late Bronze Age/Earlier Iron Age. This confusing 

situation will not be resolved until there are further sites with a radiocarbon based 

chronology. An added confusion for readers of local archaeological literature is the use in 

the 1950s and 1960s of the term 'Early Iron Age' for the whole of the pre-Roman Iron Age 

(eg Thomas, 1958, 15). 

The introduction of hillforts is now known to have occurred during the Later Bronze Age 

in some parts of the country, particularly in Wales (Harding, 1976). Some Cornish sites may 

belong to an early stage in the local hillfort sequence, but it is impossible to say whether this 

began in Cornwall before or after the advent of iron. Early sites might be sought among 

simple univallate structures in good defensive positions such as Cadsonbury, near Callington, 

or the stone walled, moorland enclosures, termed 'tor enclosures' by Silvester (1979, 188), 

such as Stowe's Pound, Minions, or Trencrom near St Ives. Recent examination of the 

112

pottery from Trevelgue confirms the amount of Earlier Iron Age material present at this cliff 

castle and raises the possibility that the site may have been occupied from a date in the Late 

Bronze Age. 

The reasons for the scrappy evidence for the period 600-400 BC, the Cornish Earlier Iron 

Age, probably relate to two linked factors. Climatic and soil changes during the first 

millennium BC (Bell, 1984, 53) are thought to have caused major shifts in settlement 

patterns, although Bodmin Moor, at a lower altitude, may have been less affected than 

Dartmoor. If population levels were low and the main pattern of settlement was one of simple 

circular houses scattered among field systems in areas which do not today survive as 

moorland, sites will prove extremely difficult to detect. 

During the Later Iron Age new pottery forms were adopted. Vessels are smaller than 

preceding types and tend to be better made; the most common form is a necked jar, often 

decorated with a zone of incised geometric or curvilinear pattern on the shoulder. This 

pottery has long been related to ceramic trends among La Tene groups in northern France 

and appears in the Cornish literature either as 'South Western Third B' (from an elaboration 

by Hawkes (1960, 15) of his original ABC system) or as 'Glastonbury Ware' from its occurrence 

in large quantities on the so-called Somerset Lake Village. This pottery occurs 

throughout Cornwall and Devon and into Somerset as far north as Mendip, but is hardly 

found in Dorset. Dating has varied, but it was generally assigned to the 2nd and 1st centuries 

BC. Recent excavations with radiocarbon dates have improved our understanding of its 

chronology and the term `(La Tene) South-Western Decorated' is now coming into use for it. 

Detailed study of the pottery excavated at Cam Euny 1964 —72 (Elsdon, 1978) has 

identified sherds with stamp decoration, probably related to material from 5th century 

Brittany (Schwappach, 1969, 272); similar sherds occur at Trevelgue. The similarity with 

the Breton material and an associated radiocarbon date of 420 ± 70 bc (HAR-238) at Cam 

Euny suggest a 5th, or possibly 4th, century date for these decorative beginnings. Avery 

(1972; 1973) has made a detailed study of South-Western Decorated pottery, based for 

Cornwall mainly on the excavations at Trevelgue and Castle Dore. Both at Cam Euny and 

at Castle Dore, there is material with rouletted decoration which appears to represent the next 

ceramic phase after the stamp-decorated sherds. Avery and Elsdon both place this rouletted 

pottery in the 3rd century BC, and the start of the much more common incised geometric 

and curvilinear decorated material in the second century. Excavations at Killibury hillfort, 

Egloshayle, in 1975-6 (Miles et al, 1977) produced a series of radiocarbon dates which 

suggested that South-Western Decorated pottery, including its curvilinear decoration, was 

fully developed by the 3rd century BC. The Killibury dates appear to be corroborated by a 

series from Meare, Somerset (Coles, 1981, 68). It would not be ureasonable to suggest that 

South-Western Decorated pottery was developing during the 4th century BC, relegating the 

sparsely occurring rouletted material to the (earlier?) 4th century BC. South-Western 

Decorated pottery occurs on a wide range of sites and in far greater quantities than all Earlier 

Iron Age styles, and a possible 4th century origin both explains its comparative frequency 

and allows an extended chronology for the sites concerned. 

For convenience, the period from c. 400 BC and the introduction of the decorated styles 

will be referred to as the Later Iron Age. This revision of chronology suggests a period of 

fairly rapid ceramic change in the 5th and 4th centuries BC and then a stable period, in which 

pottery was produced without detectable change, lasting until the 1st century BC. 

This stability may be associated with changes in production centres and distribution. 

Peacock (1968, 1969) demonstrated that virtually all South-Western Decorated pottery from 

Cornwall contains minerals derived from the gabbroic rocks of the St Keverne area of the 

113

Lizard, and distinguished this as Group I, while other decoratively similar pottery had 

petrology which indicated origins in Devon and in Somerset. The Cam Euny material shows 

a gradual change from presumably local granitic fabrics to gabbroic wares, with the majority 

of the South-Western Decorated pottery being gabbroic. South-Western Decorated pottery 

from other sites in West Penwith, such as Bodrifty, appears to be gabbroic, Peacock's Group 

I. Further east in Cornwall, South-Western Decorated pottery is exclusively of gabbroic 

fabric, even at Castle Dore, or at Killibury 50 km from the suggested source. Occasional 

sherds are found in Devon. Despite an intensive programme of field walking by the Cornwall 

Archaeological Society (CAS) between 1978 and 1983, no evidence of pottery manufacture 

in the gabbroic area has been identified (CAS Newsletters, Nos 28-40). However, the area 

produced more gabbroic sherds as surface finds than the Kenwyn valley north of Truro 

which was walked in 1983-4 (CAS Newsletters, Nos 43-44). If the pottery was clampfired, 

manufacture would leave little trace, particularly as overfired or continually reheated 

sherds tend to crumble (pers.comm. on material from Trethurgy). For the Neolithic gabbroic 

pottery found at Cam Brea, it has recently been suggested that appropriate suites of minerals 

could be found in the Camborne area (Mercer, 1981, 179). The location of the production 

centre(s) is unlikely to be finally settled until either the sites themselves are found, or a great 

deal more detailed petrological work is done, both on potential source areas and on possible 

variations within the gabbroic fabric. For the present, however, it appears a reasonable hypothesis 

that, during the Later Iron Age, pottery manufacture gradually became the exclusive 

preserve of groups resident on the Lizard and that gabbroic pottery was distributed through 

exchange networks which became more sophisticated and extensive towards the end of the 

Iron Age. 

The Iron Age and Roman periods in Cornwall have been covered by two general studies 

of the region (Fox 1964, 2nd ed. 1973; Pearce, 1981). A more detailed study by Thomas 

(1966) entitled 'The Character and Origins of Roman Dumnonia' covered settlement patterns 

and emphasised the cultural poverty of Dumnonia compared to other areas of southern 

Britain. Hillforts have been discussed by Forde-Johnson (1976, 30-5). More recently, an 

analysis of enclosed and defended settlements by Johnson and Rose (1982) provided a comprehensive 

summary of the evidence to date, made graphic with 290 plans. Fox had 

demonstrated (1952) that multivallate hillforts with close-spaced ramparts were virtually 

absent from the South West, and that multiple enclosure sites with wide-spaced ramparts 

occurred instead. She discussed the multiple enclosure sites further in 1958 (Fox, 1961), 

emphasising the wide variety both in plan and in geographical location, a variety further 

brought out by Johnson and Rose (1982). Multiple enclosure forts occur throughout Devon 

and Cornwall and in South Wales, with occasional examples elsewhere. Fox (1961, 46) 

suggested that the form was related to the needs of a pastoral, principally cattle rearing, 

economy. No convincing alternative has yet been suggested, and no research design has been 

carried out to test this hypothesis. 

By 1958, excavations at Tregeare Rounds, Castle Dore and Chun Castle, together with 

sites in Devon, had all produced pottery of the South-Western Decorated type. A Later Iron 

Age dating for multiple-enclosure hillforts has been confirmed by excavations at Killibury 

(Miles et al, 1977) and at Castle-an-Dinas, St Columb Major (Wailes, 1963), although the 

final report on the latter site is still awaited. With the exception of Cam Brea (Mercer, 1981), 

no other non-coastal hillfort has been excavated in the last 25 years. At Cam Brea, while 

hut cicles were undoubtedly occupied during the Later Iron Age, the defences were 

tentatively assigned to the Neolithic period. The work at Killibury, although on a small scale, 

was valuable in demonstrating dense occupation in the inner enclosure, with 13 possible 

structural phases between the 4th and 1st centuries BC/AD, while occupation in the outer 

114

enclosure was apparently scantier. The 1936-37 excavations at Castle Dore (Radford, 1951) 

had shown permanent timber structures, perhaps with some degree of planned layout, but the 

features were sparse compared to the density at Killibury. This comparative sparsity, together 

with other evidence, suggests that Castle Dore had phases of disuse (Quinnell and Harris, 

1985). Both Killibury and Castle Dore have four-post structures, probably grain storage 

units, of a type that was not previously thought to occur in Cornwall. There is another of 

Roman date at Trethurgy (Quinnell, forthcoming). Hillforts as permanently occupied 

centres — albeit deserted for periods as local patterns of power changed — and hillforts with 

a degree of interior planning, incorporating both round timber houses and rectangular storage 

units, are concepts consistent with our present understanding of these monuments in Britain 

as a whole (Cunliffe, 1984). 

Cliff castles — fortified promontories — have tended to be regarded as a special Cornish 

feature, perhaps initiated by Breton immigrants (Cotton, 1959; Thomas, 1966, 78). A recent 

study by Lamb (1980, Fig 1) has emphasised that this hillfort variant occurs both in the 

British Isles and in northern France wherever the geography of the coast line is suitable. 

Maen Castle and Trevelgue have been mentioned above as occupied during the Earlier Iron 

Age. Trevelgue certainly continued in the Later Iron Age as well, and has produced a large 

collection of both Earlier Iron Age and South-Western Decorated pottery but very little 

Cordoned Ware and none that need be of pre-Roman date. It has the most complex defences 

of any Cornish cliff castle, yet its floruit appears to have been before c. 50 BC. CAS excavations 

at The Rumps, St Minver (Brooks, 1974) between 1963 and 67 showed occupation 

associated with South-Western Decorated pottery (including rouletted sherds) and Cordoned 

Ware, with a potential date range from the 4th century BC to the 1st century AD. The 

defences at The Rumps had undergone substantial alteration during this period, and round 

timber houses similar to those in inland hillforts and settlements were identified. In 1983, 

work at Penhale cliff castle, Perranzabuloe (Smith, 1984) yielded a single round house and 

South-Western Decorated pottery within a large excavated area; radiocarbon dates are 

awaited. Inconvenient though these fortified promontory sites may appear to modern eyes, 

it is now well established that they were permanent settlements of varying duration, which in 

excavated detail may not differ from inland hillforts. Misty ideas of temporary refuges have 

to be discounted, while their wide date-range shows that they cannot be explained as the 

product of one historical event, such as immigrant Veneti fleeing the Roman conquest of 

Gaul. This long chronological range and the wide variation of their defences suggest that the 

relationship of their inhabitants to those of nearby settlements may have taken different forms 

during the Iron Age. 

Small univallate enclosures, known in Cornwall as 'rounds', were scarcely recognized or 

understood in 1958 (Dudley, 1958). They may lie on hillslopes, or other not obviously 

defensible positions; their ditches tend to be shallow, 1.5 to 2m deep (though there are 

exceptions) and their entrances are simple without inturns. The parish check-list programme of 

the CAS has greatly increased the numbers known. Thomas, in the first general published 

discussion on rounds (1966, 88), suggested a potential total of 750 for Devon and Cornwall. 

This figure is likely to be an underestimate as aerial photography is now producing new sites 

(Johnson and Rose 1982; Griffith, 1985). Johnson and Rose have shown that rounds have 

considerable complexity and variation in plan and can exist in close proximity, for example 

the pair on the slopes of Tregonning Hill below Castle Pencaire. About' fifteen Cornish 

rounds have now been excavated to a greater or lesser (minimal) extent and their occupation 

ranges in date from the Later Iron Age to c. AD 600. The interior of Threemilestone, near 

Truro, was about three quarters excavated and is the only round so far demonstrated to have 

115

St.Mawgan V 

....',.. - • 

•'' Orti° 

/ 49.adij 

i .01.Pr° 

Cain Euny Al 

4c0°•,; NI) 

• 0 ID 

• itt%•Pl°e' q■ 

, 

e • HP • ge(5.70 

, 

• 

,v,, 

*te:013 Actro?6,'s 

&WA c;,,; ', 

• a • ',, o •,g,'5-41 

vs . e • ge°4'rejl 

:141.4'64v 

s - - - - . 4(0E' 

- - - - -11q5;40.4S0e4 • ' 

St.Mawg-an W 

Ft 

10 20 30 

0 5 10 

Fig 2 

Bodrifty House E, Earlier Iron Age (after Dudley); Castle Dore Hut 4, 4th or 3rd century BC (?) (after Radford); 

Trevisker House 1, 1st centuries BC/AD (after Greenfield); Cam Euny House Al, 4th century BC or later (after 

Christie); St Mawgan-in-Pydar Houses W and V (latest phase) 1st century AD (after 77zreipland). 

116

een occupied entirely within the Later Iron Age (Dudley, 1960; Schweiso, 1976). The plan 

shows dense occupation with timber houses in the centre as well as around the perimeter. 

The round at Trevisker, St Eval (ApSimon and Greenfield, 1972) was initially small, 

enclosing c. 0.1 ha with a timber round house 13 m across. Material on the floor of this house 

produced a radiocarbon date of 180 ± 90 bc (NPL —135). This phase was associated with 

South-Western Decorated pottery but was replaced, within the Later Iron Age, by a far larger 

enclosure of c. 1.2 ha and at least one large timber round house (Fig 2). Some occupation 

at Trevisker continued until the 2nd century AD. Castle Gotha round, near St Austell, also 

had a date range from the Later Iron Age until the 2nd, or possibly 3rd, century AD 

(Saunders and Harris, 1982). 

The Iron Age house plans so far recorded in Cornwall are circular, with an inner ring of 

posts taking the main weight of the roof, and an outer wall of granite blocks, shale, turf or 

timber depending on the locality. Stone-walled houses such as those at Bodrifty (Fig 2) are 

identical in plan to those found in the local Bronze Age, while the size of the houses at 

Trevisker is comparable with the largest found elsewhere in Southern Britain. Reconstruction 

work by Reynolds on round houses of this size, using straw thatch, has shown in detail how 

they might have looked (Reynolds, 1978; 1982). 

The relationship of a round to a field system was demonstrated at Goldherring, Sancreed 

(Guthrie, 1969). This round, built initially in the first century AD and containing smallish 

round houses, was set in a block of small rectangular fields. Fieldwork and the study of air 

photographs carried out since the late 1970s by the Cornwall Archaeological Unit (formerly 

Cornwall Committee for Rescue Archaeology) show traces of field systems regularly 

appearing around such sites, wherever conditions permit their survival (Johnson and Rose, 

1982; Cornwall Committee for Rescue Archaeology Reports 1977 onward). The full implications 

of this fieldwork, which is still continuing, have yet to be assessed. Fields indicate 

arable farming. Evidence for specific cereal production is sparse, because no proper wet 

sieving and analysis programme has yet been carried out. Hordeum spp. (barley), Avena spp. 

(wild or cultivated oats) and Secale cereale (rye) were recorded at Goldherring (Guthrie, 

1969, 38), Triticum dicoccum (emmer), T. spelta (spelt) and Avena sp. at Killibury 

(Hillman, G. in Miles et al, 1977). Pollen analysis suggested arable farming around Cam 

Euny, where traces of fields survive around the unenclosed settlement (Christie, 1978, 311, 

426). Querns are found on all types of settlement site. Rotary querns were introduced into 

Cornwall in the Later Iron Age, as in to the remainder of Britain (eg Trevisker, ApSimon 

and Greenfield, 1972, 349) but .,addle querns continued to be used as well, and both types 

are found throughout the Roman period. Detailed understanding of settlement patterns and 

farming practices awaits further work on macroscopic plant remains and pollen analysis. It 

must also depend on more detailed knowledge of the relationship of rounds and hillforts to 

the unenclosed settlements, which remain elusive throughout the Iron Age. That at Cam Euny 

was located because it underlay a later settlement of courtyard houses. 

The variations in enclosed and fortified sites presented by Johnson and Rose (1982) 

indicate complex social patterns varying through time. The rounds have been seen as the 

settlements of landowning kindred groups (Fox, 1964, 125). Multiple enclosure hillforts 

should represent an upper social stratum of chiefs. The extent of defences on some larger 

sites indicates control of considerable man power. Cattle have been seen as the basis of wealth 

for these chiefs, but sheep, better suited to South Western upland grazing, should also be considered 

as an important local resource. Grazing for either implies considerable tracts of land 

controlled from multiple enclosure hillforts. Another possibility would be areas of arable 

worked by bondmen, a system foreshadowing the landholding system of Early Medieval 

Wales. The occupants of rounds may have paid dues in kind or labour. It is probable that 

117

a combination of all these resources contributed to the position of hillfort-building chiefs. 

Material wealth, represented by archaeological finds, still seems to have been less than that 

evidenced elsewhere in Southern England, but perhaps has been underestimated. No new 

pieces of decorated metalwork have been found since 1958 although a study of the Trenoweth 

collar (Megaw, 1967) relocates its findspot to St Stephen-in-Brannel. 

The univallate hillslope fort of St Mawgan-in-Pydar (c. 1.2 ha) near Newquay (Threipland 

1956) hints at a different kind of social unit to those represented by rounds or multiple 

enclosure hillforts. Despite its small size, its entrance had an elaborate unturn, a feature 

lacking in rounds, and it produced a rich array of finds, including a decorated shield 

mounting. The date of occupation may have been from c. 100 BC to AD 100. Thus, unlike 

the multiple-enclosure hillforts, it continued after the Roman occupation. 

The dead of Iron Age Cornwall are less well represented than the living, yet compared 

to most of Britain the evidence is good. Cornwall is one of the few areas to have produced 

cemeteries. Those at Harlyn Bay and at nearby Trevone (Dudley, 1965) have long been 

known, as have the smaller sites at Trelan Bahow in the Lizard, Stamford Hill outside 

Plymouth and related sites in the Isles of Scilly. These sites are all dateable as graves have 

produced decorated metal artefacts, especially brooches and mirrors, ranging in date from 

the 3rd century BC to the 1st century AD. Other cemeteries such as Crantock without 

dateable artefacts may also be of Iron Age date (Whimster, 1977, 81). Whimster's excavation 

at Harlyn Bay in 1975 showed that there may have been a circular stone temple in the vicinity 

of the cemetery (1977, 69). Whimster's work at Harlyn Bay formed part of a major study 

of burial practices in Iron Age Britain (1977a, 181). He was able to demonstrate the regular 

pratice of crouched inhumation with head to the north in Cornwall, as in scattered burials 

across Wessex, and in the Yorkshire cemeteries which include 'chariot' burials. Until 

Whimster's study, the south-western cemeteries had been linked to those in Brittany because 

of the presence of long stone cists in both areas. But in Brittany the burial rite was extended 

inhumation. The Cornish cemeteries seem to be showing us a British funerary tradition made 

archaeologically visible by the use of long stone cists. Whether the stone cists, using local 

slate, reflect Breton influence, is an unanswerable question at present. It must be doubtful 

whether there was any Breton influence on the religious practices of Iron Age Cornwall. The 

Breton evidence, conveniently summarized in Giot (1960), emphasises the importance of 

carved stone pillars (lec'hs). None have ever been noted in Cornwall, despite the present 

author's hopeful examination of standing stones over the past fifteen years. 

Both religious practices and Breton connections enter consideration of the souterrains or 

logous' of West Cornwall. A useful factual summary was published by Clark in 1961. 

Cornish souterrains differ from their Breton counterparts in their method of construction. 

They are constructed of stone in open trenches, whereas Breton soutterains are entirely tunnelled 

out; in this respect Cornish examples are similar to those in Scotland and in Ireland. 

It is becoming increasingly aparent that souterrains in all these areas are closely linked to 

settlements and are not isolated monuments. Thomas (1972) argued that souterrains were 

used for storage, drawing on the similarities with the medieval and later West Cornish 'hulls', 

a point also brought out in a comprehensive survey of hulls by Tangye (1973). A throughdraught 

may have been provided by small gaps in the end opposite the entrance, or in the 

side passage. 

A detailed examination of the Cam Euny fogou and associated 'round chamber' was 

carried out by Christie (1978) during the 1964-72 excavation programme. The 'round 

chamber' with a short entrance passage proved to be the earliest feature of the complex; it 

was thought either to have had a timber roof or to have remained open. A stamped-decorated 

sherd similar to those found in a feature elsewhere on the site with a radiocarbon date of 

118

420±70 be (HAR —238) suggested a possible 5th century BC date. The nature of any 

associated settlement at this early date was unclear. The 'long passage' with a probable 

entrance through the side was added later; a radiocarbon date of 130±80 be (HAR —334) 

from its floor was associated with a local variant of South-Western Decorated pottery. A 

settlement of round turf and timber houses then clustered around the fogou, its starting date 

depending on that assigned to the associated South-Western Decorated pottery. Elsdon (1978, 

404) suggested the 3rd or 2nd centuries BC, but an origin in the 4th century BC would be 

possible if the back-dating of this pottery type suggested above is allowed. Finally, at Camn 

Euny an 'east entrance' was added to the 'long passage' when the first courtyard houses were 

built on the site, in the 1st century BC according to Elsdon (1975, 404) but more probably 

later (see below). The settlement and fogou continued in use until the 4th century AD. More 

recently Startin (1982) re-excavated Halligye fogou on the Lizard and also found evidence for 

multi-phase construction associated with a round, and a date range from the Later Iron Age 

until a blocking in the Roman period. The dating evidence from Cam Euny and Halligye is 

consistent with that from earlier investigations in other fogous. It is not possible to show that 

any fogou was constructed before the Later Iron Age. The end date for fogou construction 

is difficult; use of some like that at Porthmeor continued into the 4th century AD. The only 

cases in which a construction date within the Roman period is probable are those, as at 

Bossullow Trehyllys, where the fogou appears as an integral part of a courtyard house. 

Christie (1978, 332; 1979) reconsidered the purpose of Cornish souterrains, and favoured 

a religious function because of their impracticality of access, the narrow side passages or 

'creeps' and the tantalizing implications of possible fragments of cremated human bone associated 

with the construction of the Cam Euny 'round chamber'. The modern mind tends to 

divide activities into modern categories. Is it not possible that storage of special commodities, 

food reserves (which would need to be immune from damp!), could have been linked to 

religious practices? The amount of Breton influence is difficult to assess, beyond the broader 

similarities. The Irish souterrains seem mainly to date to the Early Christian period, which 

weakens the argument for a general spread of the practice of souterrain construction from 

a single Breton source. 

During the 1st century BC a new form of pottery, 'Cordoned Ware', appears in Cornwall 

(sometimes referred to by various Iron Age C labels). The fullest published range of 

Cordoned Ware is still that from St Mawgan-in-Pydar (Threipland, 1956), though its study 

can be supplemented by the material from Carvossa, Probus excavated 1968 —70 (Douch and 

Beard, 1970). Cordoned Ware has features which are also found in the pottery of northern 

France during the century preceding the Roman conquest of Gaul. A direct Breton origin was 

invoked by Thomas (1966, 79) but similarities are general rather than close and no imported 

Breton material has been found in Cornwall. Imported north French wares have been found 

in Dorset (principally at Hengistbury Head) and in Devon at Mount Batten (Cunliffe, 1983, 

125) and date to the years immediately preceding Caesar's conquest of Gaul (Cunliffe, 1982, 

50). Cordoned Ware may have originated as copies of this north French imported pottery, 

knowledge of which spread westwards with coastal trade, and if so may have started either 

before Caesar's campaigns, or a little after while the imported forms were still current. 

Cornish Cordoned Ware is invariably of gabbroic fabric, suggesting that the new forms were 

adopted by the potting communities of the Lizard. The ware tends to be thinner and better 

made than South-Western Decorated pottery, improvements which have been attributed to 

the introduction of the potter's wheel. The present author is not convinced that any of this 

ware, or of the subsequent gabbroic forms of the Roman period, was wheelmade, and, until 

microscopic studies of particle alignment are available, suggests that 'wheel made' should 

be read as 'well made' for Cornish pottery. 

119

On most Cornish sites, even with modern excavations such as Killibury and The Rumps, 

it is not possible to separate stratigraphically South-Western Decorated pottery from 

Cordoned Wares (pace Threipland, 1956, 53). There was some fusion — cordons appearing 

on South-Western Decorated pottery (eg Killibury P17, Miles et al, 1977) and the two styles 

were probably manufactured together for a while. The actual terminal date for the manufacture 

of South-Western Decorated pottery is not yet established and the style may have 

continued well into the 1st century AD. Goldherring (Guthrie, 1969) is the only site with a 

clearly defined stratified phase associated only with Cordoned Ware, and no South-Western 

Decorated or Roman forms. Threipland emphasised in 1956 that Cordoned Ware was current 

until the 2nd century AD. Detailed analysis of the pottery from Trethurgy (Quinnell, 

forthcoming) makes it quite evident that large cordoned jars, with either rolled (Type H) or 

everted (Type J) rims, were being made in the 3rd and 4th centuries AD. In fact is is probably 

only the bowls, Threipland's Type F and G, which can be restricted to the 1st centuries 

BC/AD: interestingly she notes (1956, 53) that these are concentrated early in the sequence 

at St Mawgan-in-Pydar. Cordoned Ware should be regarded as a long lasting ceramic tradition 

and not a chronological horizon. Many sites have been dated in the past on the basis of 

a few sherds with cordons and need re-assessment. The courtyard house at Mulfra Vean 

(Thomas, 1963) has a pottery assemblage which contains large cordoned jars but also forms 

which would best belong in the 3rd century AD. 

If a starting date in the 1st century BC is accepted for Cordoned Ware, this still provides 

a useful terminus post quem, particularly in regard to the courtyard houses of West Penwith. 

These distinctive structures have small round and long rooms set within thick walls surrounding 

a central courtyard, generally regarded as unroofed (but see Christie, 1978, 387). 

Courtyard houses may form clusters of eight or so, with a central 'street' indicating an 

element of planning in the layout at Chysauster. Courtyard houses have still not been located 

anywhere but West Penwith and Scilly, despite superficial similarities with settlements in 

areas such as North Wales. They are best regarded as a local development, especially suited 

to local windswept conditions — the extensive contemporary field systems recorded in the 

area by the Cornwall Archaeological Unit bear witness to the opening up of the landscape. 

The date of courtyard houses has been obscured by the presence of South-Western Decorated 

pottery on early excavations at Chysauster and Porthmeor. The Chysauster courtyard house 

settlement occurs in an area of fields with round houses, and the South-Western Decorated 

pottery could relate to earlier use of the site. Porthmeor occupies an earlier enclosure of the 

round class. At Goldherring (Guthrie, 1969) the Cordoned Ware phase is associated with 

circular turf and timber houses within the round. The courtyard houses relate to a later 

occupation, in the 3rd and 4th centuries AD. At Cam Euny (Christie, 1978), courtyard 

houses were thought to have developed about the period when Cordoned Ware was 

introduced. But none of this ware is significantly stratified and all of it belongs to types with 

a long life span. The Cam Euny courtyard house phase could start much later, perhaps in 

the 2nd century AD and its sequence would then be more consistent with that at Goldherring. 

It has already been suggested that the pottery from Mulfra Vean should be redated to the 3rd 

century AD. The bulk of the pottery from Chysauster dates to the 2nd/3rd centuries AD, and 

from Porthmeor to the 3rd and 4th. Courtyard house hamlets or villages do not belong to 

the Iron Age, but to the Roman period. Visible surface remains often suggest they have been 

built after a phase with round houses, as revealed by the recent Cornwall Archaeological Unit 

survey of Nanjulian, St Just in Penwith (CAU Report, January 1986), and may in general 

be less regular in layout than the 'classic' sites such as Chysauster (Weatherhill, 1982). 

Possible influence from Brittany or north France can be seen in cliff castles, burial sites, 

fogous and pottery. While direct immigration (or 'invasion') on any scale is now discounted, 

120

it is only reasonable to accept that there would be some similarities among groups who were 

in contact with each other across the English Channel. General similarities developing 

through contact are compatible with the long date range involved, from the Late Bronze Age 

onwards. The main reason for this cross-channel contact was probably trade, particularly in 

Cornish tin. There have been several recent summaries of the archaeological and historical 

evidence for the pre-Roman Cornish tin trade (Laing, 1968; Cunliffe, 1983; Hawkes, 1984). 

In Cornwall, archaeological evidence continues to be slight. The unstratified and unpublished 

ingot from Castle Dore is probably the best established find of Iron Age date, now that the 

St Mawes ingot has been assigned to the medieval period (Beagrie, 1983). Promising 

evidence is beginning to emerge from pollen studies of levels associated with tin working on 

Bodmin Moor. Those in the Colliford area may date well back into the prehistoric period 

(information, S. Gerrard). Research on material from Mount Batten in Plymouth Sound 

currently favours this location for ktis, the tin trading depot referred to by the Greek 

historian Diodorus Siculus (Cunliffe, 1983). 

The various classical references to Britain, both before and after the Roman conquest, 

were published in full by Rivet and Smith in 1979. These references provide the basis for 

assigning south-western Britain to a tribal group called the Dumnonii. They also explain why 

some at least of the inhabitants of Cornwall were called `Cornovii', providing an origin for 

the present name of Cornwall (1979, 325). Rivet and Smith provide a gazetteer of all classical 

records for place-names, settlements and geographical features throughout Britain and their 

comments provide some fascinating sidelights on contemporary affairs — Belerion (Land's 

End, p 266) incorporates an element meaning 'bright' or 'shining'. This may indicate a 

beacon at Land's End as early as the 4th century BC, which would connect with the local 

importance of seafaring and navigation, whether coastal, to Scilly, or across the Channel. 

Understanding of the Roman conquest of south-west Britain has been dramatically 

advanced by the programme of excavations in Exeter from 1971. The presence of a fortress 

occupied by the Second Legion Augusta between c. AD 55 and AD 75 has been established 

(Bidwell, 1979; 1980). In Devon a programme of aerial survey has both greatly increased 

the number of Military sites known (seven forts, one marching camp, seven fortlets/signal 

stations), and shown that some, notably North Tawton and Okehampton, are multi-period 

(Griffith, 1984). No fort has yet been dated to the Claudian period (AD 43-54), and Bidwell 

(1980, 10) suggests that the military garrisoning of the South West occurred during the 

governorship of Didius Gallus between AD 52 and 7. Presumably the Dumnonii did not pose 

any immediate threat to Rome. It is quite possible that the larger hillforts were out of use 

before the Romans arrived. Castle Dore and Killibury produced little Cordoned Ware 

(Quinnell and Harris, 1985, 128). Hembury, East Devon, was not occupied at the time of 

the Roman Conquest (Todd, 1984) and the end date of other excavated Devon hillforts cannot 

be closely established. Local shifts in power may have been brought about as the focus of 

trade with the continent moved from Wessex to eastern England after the Roman conquest 

of Gaul. The event which caused the Roman military to garrison the South West during the 

50s AD is unknown. 

So far only one Roman fort has been identified in Cornwall, the 1.0 ha site at Nanstallon, 

south of the River Camel, just west of Bodmin. Excavations between 1960 and 1963 by Fox and 

Ravenhill (1972) showed that it had contained permanent timber buildings occupied between 

AD 55 — 60 and AD 75-80. The excavators interpreted the barrack accommodation as 

appropriate for a cohors quingenaria equitata, a mixed infantry and cavalry unit of some 500 

men. Other forts in Cornwall will probably be identified as the current programme of aerial 

survey of the non-moorland areas gets under way. It is possible that some Roman forts were 

sited within Iron Age hillforts, as has been demonstrated at Hembury in Devon (Todd, 1984), 

121

and so may not be detectable without excavation or geophysical survey. Other forts may be 

grouped at present amongst the sub-rectangular enclosures of the county. 

These sub-rectangular earthworks, assumed to be a variant of rounds, were first 

summarised by Dudley (1954); known numbers have been recently increased by study of 

existing air photographs (Johnson and Rose, 1982). These enclosures are scattered throughout 

the county, and occur in Devon, but with concentrations around the Helford River, the upper 

Fal valley and in north Cornwall. Trevinnick, near Wadebridge, was excavated by Fox and 

Ravenhill (1969) to check whether it was a Roman military site, but proved to be an enclosed 

civilian settlement of the Roman period, not closely dateable. Grambla, near Wendron, a 

virtually square earthwork enclosing c. 0.35 ha in the Helford River area, was partly 

excavated in 1972 (Saunders, 1972). This also had been a civilian settlement, occupied from 

perhaps the 2nd century AD until the 6th; it contained distinctive oval houses. Its ditch was 

3m deep and its entrance had a quadrilateral setting of post-holes suggesting a gate tower. 

The 2 ha rectangular site at Carvossa, near Probus in the Fal valley, was excavated from 

1968 to 1970 (Douch and Beard, 1970; information from P.M. Carlyon). It produced a rich 

range of finds, starting with South-Western Decorated pottery. The samian sequence began 

with a single Tiberio-Claudian sherd, increased in the Neronian period, and continued with 

some fluctuation until the mid-second century Antonine period. The coin list and other 

imports reinforce the pattern of the samian, suggesting that occupation still flourished in the 

Hadrianic period and subsequently spanned the 2nd and 3rd centuries AD. The range of finds 

for the Neronian — Early Flavian period would be entirely appropriate for a military site. 

The stratigraphy was not clear and there is some evidence for at least two earthwork phases. 

The main 2 ha earthwork had a ditch 4.5m deep which was not of military type, nor was 

the excavated east entrance. Features found in the interior, such as curvilinear gullies, 

suggest civilian occupation. The ditch had silted by at least 1.5m beneath the lowest samian 

sherd. The rectilinear earthwork may be suggested to be of pre-Roman date, perhaps as early 

as the 1st century BC. The finds indicate that activity continued throughout the suggested 

period of military occupation of Cornwall. A fort may have been established in part of the 

earthwork complex, in which case most of the artefacts would derive from an external vicus. 

Perhaps some trusted chief was encouraged to develop at Carvossa a focus for trade up the 

Fal valley. Alternatively, the finds may indicate a fort in the vicinity, perhaps within the 

substantial earth-work of Golden one mile to the south. If Roman military occupation in 

Cornwall is demonstrated in future to be slight compared to that in Devon, a network of 

trusted local chiefs could be the reason. Another local chief was allowed to continue 

occupation at St Mawgan-in-Pydar, where the range of small finds, imports and elaborate 

gabbroic pottery is second only to Carvossa. Carvossa certainly attracted traded artefacts. It 

also produced a rich range of gabbroic pottery, Cordoned wares and innovative pieces 

copying items such as a Dr 29 samian bowl and a bronze handled patera. There was also 

evidence for metal-working, especially iron smelting in the ditch. 

If Carvossa was at one stage a military post, it will have been handed over to a local group 

c. AD 75. Here would be happening, on a small scale, the regular practice of turning over 

military bases to serve as local centres of administration and trade under Roman supervision. 

At Exeter this process led to the establishment of a cantonal capital. Isca Dumnoniorum, 

which administered both Devon and Cornwall (Bidwell, 1980, 56). There must have been 

a series of small centres throughout the South West. As no formal Roman building, apart 

from the Magor villa known since 1931, has been found in Cornwall, it may be presumed 

that such centres were generally developed in the local tradition. In Cornwall these may be 

looked for amongst the more substantial small enclosures, where the descendants of chiefs — 

122

whose ancestors had occupied hillforts — administered the further parts of Dumnonia. Such 

centres would have been useful as local tax collection points, as the capital of the responsible 

civitas at Exeter was so distant. 

Trethurgy Al 

Trethurgy T2 

Ft 

10 20 310 

5 10 

Chysauster 5 

Castle Gotha 

Fig 3 

Trethurgy Houses Al and 72, late 2nd century AD; Chysauster Courtyard House 5, 3rd century AD (after 

Hencken); Castle Gotha Oval Hut, 2nd century AD (after Harris). 

123

The basic pattern of life does not appear to have changed much under Roman rule, but 

our understanding of it has greatly increased over the last 25 years (see Radford, 1958). The 

round continued as the main (detected) settlement type (Fox, 1964), 148; Fowler, 1976). 

Thomas emphasised this continuity in 1966, and the lack of evidence of much Roman 

influence on basic life styles. The main evidence for the continuance of rounds comes from 

Castle Gotha (Saunders and Harris, 1982), occupied from the Later Iron Age to the 2nd/3rd 

century AD; Trevisker (ApSimon and Greenfield, 1972), with activity until the 2nd century, 

and possibly Merther Euny (Thomas, 1968). During the 2nd century a number of new rounds 

were constructed: Shortlanesend (Harris, 1980), Carwarthen (Opie, 1939; unpublished 

material now in Truro Museum), Crane Godrevy (Thomas, 1968a), possibly Kilhallon 

(Carlyon, 1982), and Trethurgy (Quinnell, forthcoming). Of these, Shortlanesend and 

Kilhallon apparently had a short span of a century or so, and for Crane Godrevy only a small 

sample of pottery has been published (Thomas, 1964, 61). Carwarthen was occupied into the 

4th century, Trethurgy into the 6th. The excavated sample is still small, but it is noteworthy 

that no round seems to have been continuously occupied from the Iron Age throughout the 

Roman period. Any dislocation in settlement patterns seems to have occurred in the 2nd 

century AD rather than the 1st (perhaps related to the ending of occupation of the small 

hillfort at St Mawgan-in-Pydar) and to have been a gradual occurrence. The reasons are 

obscure. Was there some historical event leading to temporary decline in population? Were 

there changes in agricultural practice, brought about by new demands or new techniques? Did 

the establishment of new administration centres and communication patterns (though no 

Roman roads have yet been proved for Cornwall) ultimately cause shifts in settlement? 

It need occasion no surprise that minor fortifications were allowed under the Roman 

administration. Even minor defences were regarded in the Iron Age as a symbol of status. 

It was regular Roman practice to integrate the local upper classes in the administrative, 

military and economic structure of the Empire, and to encourage them to continue to look 

after their own areas. The building of new rounds, and more substantial rectangular 

enclosures, implies the continuance of local traditions of maintaining order, and not 

necessarily a specific threat. However there could be a link between the construction of new 

rounds, after a gap of about one hundred years, and the construction of fortifications around 

towns which occurred at Exeter in the late 2nd century (Bidwell, 1980, 66). Cornish rounds 

have been compared to the raths or enclosed settlements of south-west Wales, but recent work 

in the latter area has shown a significant difference. There, all excavated earth-works were 

constructed before the Roman occupation; some continued in use, but no new works were 

built. Gradually during the Roman period the nature of their occupation changed, from dense 

concentrations of circular houses, to include single rectangular houses, as at Dan-y-Coed 

(Williams, 1985) and Walesland Rath (Wainwright, 1971). This difference must reflect the 

longer presence of the Roman army in Wales and the establishment of an urban centre at 

Carmarthen. South-west Wales was left less to its local traditions than was south-west 

England. In Dyfed there are also several rectilinear enclosures, now proved to start in the 

Iron Age in the case of Pen-y-Coed, Llangynog (Murphy, 1983), though Llangynog II and 

possibly Castle Flemish were built in the Roman period (James and Williams, 1982, 299). 

The population of Roman Britain as a whole is now considered to have expanded rapidly, 

possibly even to around six million (Salway, , 1981, 554). It is reasonable to expect more later 

Roman sites in Cornwall than in the 1st to 2nd centuries. Dates given for sites which differ 

from those in the published reports are based on a current study of pottery from Carvossa 

(information P.M. Carlyon) and Trethurgy (Quinnell forthcoming), sites situated 15 km apart 

but which overlap chronologically to present a continuous sequence from the 1st century BC 

to 6th or 7th centuries AD. The number of rounds may have increased, and this increase may 

124

AL •r, 

.4 •eakitiocto dp° • 

• • kir w..s..- ny o 

bo ,:7 4iVal■pAiii. 

44,4,as• .. 40 

• )1F• a 4 gal I . , ...;-__. 'Z1 

•', so 

443.-0 

; 0 

Trethurgy T4 

1. ..1.1'..1.,, 

, 

. ,...„,„,,, 

...,„ 

n ‘Nip.p ca 

,19. .. 

033 411? -.4?). V Vio8;ks 

... ' •°FliMeLft 

•■ •. 

q • 

0 d'on 

7.1_07L. OVv.elio 

e " ig1610 °,0 

-791 '4;6 

...oz,,;%•,.- , . 4,1,goq 

• C.,,C) CIP 

Trebarveth 3 

rvi 

5 

10 

• 

;',4;p'• 

el 

•• 

.44 

A ■ 

Uob 11-* 

r • 

9 

• 

, 

Porth Godrevy 

Ft 

0 10 20 30 

Fig 4 

Trethurgy Houses Z2 and T3, early 4th century AD; Grambla Building I (simplified,) 2nd century AD or later (after 

Saunders); Trebarveth 3, 3rd century AD (after Peacock); Porth Godrevv, 3rd century AD (after Fowler). 

125 

•

e matched by the development and proliferation of courtyard houses in West Penwith from 

the 2nd century AD. Isolated unenclosed houses are rare, Porth Godrevy (Fowler, 1962) 

occupied between the 2nd and 4th centuries being almost the only example (Fig 4); as with 

the Iron Age, this scarcity may well be due to the difficulty of finding sites. 

Gradual changes can be detected in most aspects of life in Roman Cornwall which are 

archaeologically retrievable. It is difficult to show a true round house of the Cornish Roman 

period. At Trethurgy, Castle Gotha and Grambla (and possibly Crane Godrevy and 

Shortlanesend) the houses were oval, up to 13m in length and providing 80-100 m2 of floor 

space. Posts, where present or detectable, were set in the wall face. Such houses must have 

had a ridged roof, supported probably on a polygonal wall plate. The ridged roof was 

presumably a local adaptation of Roman building styles. The new oval houses left the 

interiors free of supporting posts and allowed a larger area to be roofed with shorter timbers 

than in a round house with similar floor space. The sequence is illustrated in Figs 3-4. 

Strong local tradition is still evident in the oval (as opposed to rectangular) plan which made 

good structural sense in terms of wind resistance. Smaller buildings like Porth Godrevy and 

Trebarveth T3 (Peacock, 1969a) were also oval in plan. The large oval houses were well 

floored, with tamped rab (beaten earth) or paving kept clean. At Trethurgy, quantities of 

small nails suggest interior wooden fittings and furniture. A regular feature was a hearth pit 

or cooking pit, alongside a hearth. Hearth pits (especially at Trethurgy) had heavy burning 

around their tops and may have been used for slow cooking, with pots safely embedded in 

hot ashes below floor level as the fire died down. 

Trethurgy, near St Austell, is the only round of which the interior has been fully excavated 

(Miles and Miles, 1973). Precise chronology is difficult to establish because of the usual 

scarcity of dateable small finds. Occupation appears to start in the late 2nd century AD with 

three large oval houses set around its perimeter and a four post structure (a granary?). An 

area was provided for animal husbandry and another for general activities and storage, one 

on either side of the entrance; both areas continued throughout the three centuries or more 

that the site was occupied. By the early 3rd century, domestic provision had increased to five 

houses, one of them small (Fig 5). This maximum residential capacity continued until the 

early 4th century when there may have been a short gap in the occupation. After this a 

rebuilding phase provided only four houses but a large byre was constructed in the animal 

husbandry area. In the late 4th century, houses may again have increased to five, though their 

size tended to be smaller. During the 5th century accommodation gradually dropped to two 

or three houses. 

Sherds of over 600 vessels were found at Trethurgy, but most represented only small parts 

of the original pots. Presumably domestic refuse was regularly taken out and used as a 

fertiliser. One midden in a disused building was never removed. It dated to the later 4th 

century and produced a substantial proportion of the artefacts from the round. The regular 

removal of debris partly explains the paucity of finds on Cornish sites. Three other rich 

rubbish deposits have been found elsewhere, a 2nd century accumulation in a hollow at 

Carvossa (P.M. Canyon pers.comm.), the 3rd century dump in a possible round ditch at 

Kilhallon (Carlyon, 1982) and a 4th century dump in the ditch at Carwarthen Round 

(unpublished material in Truro Museum). These middens or dumps produce metalwork, glass 

and imported pottery as well as artefacts of local origin, and demonstrate the haphazard 

nature of archaeological recovery. Even a coastal midden at Widemouth Bay produced 

samian ware (Canyon, 1981), showing how widely Roman imported goods penetrated. This 

is not to claim that the average Dumnonian was rich in Roman-style artefacts, only that his 

poverty has been overstressed. The pattern of coinage, especially hoards, reflects that of 

Roman Britain generally, although quantities are small (Pearce, 1970). The scarcity of coins 

126

Fig 5 

Trethurgy Round Period 2 c. AD 200. 

on occupation sites suggests that barter remained the common medium of exchange, and that 

coinage may have been reserved for major transactions and for taxes. In the later Roman 

period a number of tin and pewter vessels have been found; the flanged pewter bowl from 

Hallivick (sometimes referred to as Halviggan, eg Hencken, 1932, 199) represents a 

nationally common type (Penhallurick, 1986, 187 P1 93). At Trethurgy a series of stone 

bowls used as mortars (Fig 6) have been copied from handled metal bowls, hinting at the 

presence of more valuable objects which do not survive. 

127 

0 

20m

• III MI MI Cm 

Fig 6 

El van handled bowl mortaria and weight from Trethurgy Round, 4th or 5th centuries AD. All '4. 

Throughout the Roman period pottery was imported into Cornwall, from as close as 

Devon and as far away as Spain. Continental amphorae representing consignments of wine 

and oil (eg at Castle Gotha; Saunders and Harris, 1982, 135) are being increasingly identified 

as research progresses on their pattern of import into Britain. Local wares are becoming far 

better understood from the work of Bidwell at Exeter (1979; forthcoming). Grey wares and 

black-burnished ware copies made in the Exeter area have been identified at Carvossa, 

Kilhallon and Trethurgy. Black-burnished wares from Dorset (BB I) became the predominant 

pottery at Exeter in the mid-3rd and 4th centuries and it is to this period that the majority 

of Cornish BB1 pottery belongs. South Devon ware, with a distinctive granitic fabric, was 

probably manufactured somewhere in the Dart valley. Again, it is much more common at 

128

Exeter in the 3rd and 4th centuries AD than in the preceding period, and has the same pattern 

of occurrence in Cornwall. Significantly, South Devon ware forms include large jars with 

cordons and their increased import in the later Roman period may have reinforced the local 

Cordoned Ware tradition. Oxford wares have been identified on a number of sites such as 

Trethurgy and Porthmeor and probably reached Cornwall during the 4th century AD (Young, 

1977, 239). 

Imported artefacts of comparatively high value may have had a long life and their use for 

dating the archaeological contexts from which they are recovered needs consideration. It may 

seem obvious on a small excavation to use a piece of Antonine samian to date the site. At 

Trethurgy, sixty samian sherds from a minimum of fourteen vessels were found, despite the 

complete excavation. The largest quantity of samian came from the 4th century midden which 

was the earliest context to produce Oxford Ware on the site. In turn, some of the best pieces 

of Oxford Ware were found in the deliberate infill of a structure which contained post-Roman 

import wares and should date to the 6th century AD. Fine wares could thus be deposited 

perhaps two centuries after their manufacture, which increases the difficulties of producing 

a dateable sequence for the Cornish gabbroic wares of the Roman period. 

About 95% of all pottery from Roman Cornwall is gabbroic. The as yet unlocated Iron 

Age pottery centres, presumably on the Lizard, continued, gradually influenced by Roman 

forms. A brief period of experiment in the late 1st and early 2nd centuries has already been 

mentioned with reference to finds from Carvossa. Thereafter the principal products were 

simple jars, some with cordons, which gradually got larger and more poorly made. Bowls 

and dishes were introduced during the 1st century AD with St Mawgan-in-Pydar Type R 

(Threipland, 1956, 66), and continue with flat, 'pie dish' rims starting in the 2nd century. 

These develop into 'grooved' or 'proto-flanged' forms by the late 2nd century and into 

flanged bowls by the late 3rd. This sequence appears at first sight to copy that of Dorset 

black-burnished bowls, but similarities are general rather than precise. It is more likely to 

be a general reflection of stylistic trends, with the flanged bowls influenced by metal forms 

such as the Hallivick pewter bowl. There is no definite evidence for the use of the potter's 

wheel. Well-made vessels, highly fired and of well-levigated clay, represent careful 

application of traditional techniques. Roughly-made jars and bowls, of a type often labelled 

'hand-made' in contra-distinction to 'wheel-made' occur throughout the period. They are 

certainly present from the earliest contexts at Trethurgy. The proportion of roughly-made 

forms increases in the later Roman period. At Trethurgy and at Grambla (Saunders, 1982), 

occupied throughout the 5th century AD, there were no detectable changes in form, and it 

is only in the 6th century that Roman-type gabbroic pottery becomes scarce. The potting 

centres may have continued during the 5th century because they were less dependent on a 

monetary economy than kilns in other parts of the country. 

Other economic activities are less easy to identify. Most domestic sites produce evidence 

of small-scale metal working. The manufacture of some stone mortaria took place on 

domestic sites (unfinished examples are known from Castle Gotha and Trethurgy), although 

a greisen handled bowl mortarium has been found as far afield as Richborough (Roach Smith, 

1850, 104). The production of salt has been demonstrated on the Lizard, at Trebarveth in 

the 3rd and 4th centuries (Peacock, 1969a), and at Carngoon Bank (from at least the 3rd to 

the 6th centuries AD (McAvoy, 1980). A major and distinctive feature for Cornwall was the 

production of tin. This is generally considered to have been of little importance during the 

1st and 2nd centuries AD, when Iberian mines were supplying the Empire's needs. (Could 

this factor have anything to do with the suggested dislocation of Cornish settlement patterns 

in the 2nd century AD?). In the 3rd century the Iberian mines were in decline, tin was 

increasingly used in coinage, and in Britain the demand for tin went up becaue of the growing 

129

popularity of pewter table ware. It has long been recognized (Fox, 1964, 155) that the main 

British tin and pewter artefacts belong to the 3rd and 4th centuries AD. The possible official 

stamps on the Carnanton tin ingot (Warner, 1967) have been argued as evidence of Imperial 

control, although a recent review is more sceptical (Beagrie, 1985). The five Cornish 

milestones inscribed to Roman Emperors of the 3rd and early 4th centuries have been seen 

as evidence of increased interest in local communications by central authority because of the 

growing importance of Cornish tin (Thomas, 1966, 91). (The milestones have been published 

by Collingwood and Wright, 1965, Nos 2230-2234). 

A number of unstamped plano-convex ingots of varying weights have been found in 

Roman contexts (Tylecote, 1966, 31). An ingot weighing about 12 kg was found in the 4th 

century midden at Trethurgy. The evidence, including a large number of ingots now 

destroyed, has been summarized by Penhallurick (1986, 219-221). Palynological evidence 

from old workings such as those at Colliford (S. Gerrard, pers.comm.) hints that the extent 

of surviving early tin workings may have been under-estimated. But was there an Imperial 

Cornish Tin Enterprise? The practice of letting out mineral rights was common, particularly 

after the initial military occupation, to agents or conductores (Salway, 1981, 633). Local 

Dumnonian leaders may have drawn on local skills, which presumably had never entirely 

ceased being practised, and taken the opportunity to enrich themselves by taking out licences 

as conductores, and sold some at least of the local tin through the official system. An official 

presence for the purchase of tin would account for the Carnanton ingot stamps and an 

improvement, if such there was, in the communication network. Some ensuing local 

prosperity may have been a factor in the general expansion of settlements in the later Roman 

period. Naturally the ingots found on domestic sites do not have stamps, as they never found 

their way, whether by intent or accident, into the official system. 

The success of societies of the Roman period in Cornwall can not be realistically appraised 

if compared to those of Somerset or Sussex. Disadvantaged by geographical isolation and the 

apparent lack of Roman establishments to serve as models, the Dumnonian life style naturally 

seems impoverished — no mosaics, no baths, no theatres. There was sufficient contact for 

gradual changes in life-style — oval houses with more open floor space, bowls and dishes 

indicating new eating habits. But because the Cornish Dumnonii continued in pre-Roman 

ways, slightly adapted, they may have remained socially more stable, escaping the difficulties 

that great differentials in wealth must have brought in the major villa areas. They appear to 

have been largely independent of a monetary economy. Consequently their settlements and 

life-style continued into the 5th and even the 6th century with comparatively little disruption. 

Further Work 

Knowledge of Dumnonia between c. 500 BC and 500 AD has increased considerably since 

1962, with consequent new emphasis in research. This article is wrongly titled — it should 

really be called 'The Iron Age and Roman Periods in Mid- and West Cornwall'. A first 

priority is the completion of the data-base for settlement for the whole of the county, by 

continuance of the parish check-list programme, supplemented by aerial survey. Programmes 

of excavation on settlements both in south-east and north Cornwall are then needed. These 

could start with a series of trial excavations leading to the major investigation of an enclosed 

settlement in each area. No more work on sites west of Bodmin Moor should be sponsored 

until the geogrpahical balance is redressed. An exception is Carvossa, where the richness 

of earlier finds coupled with complexities in the earthwork sequence and possible military 

links, suggests that re-excavation should yield dateable stratified deposits. Exception could 

also be made for any other sites with good possibilities of stratified sequences, such as some 

of the complex earthworks plotted by Johnson and Rose (1982). 

130

The question of Late Bronze Age and Earlier Iron Age settlement is still a county-wide 

problem. A trial excavation programme of hut circle settlements in sheltered or low-lying 

positions is indicated, but not in West Penwith where settlement, although geographically 

atypical, has received more than its share of attention so far. The new programme should 

concentrate on settlements with associated field systems and should be combined with 

detailed environmental studies backed up by radiocarbon dating. It is imperative that any 

excavation programme incorporates a full range of modern techniques, especially wet 

sieving, to retrieve evidence for agricultural practices. Further environmental studies, 

following up the work at Colliford, are necessary finally to pull perceptions of Cornish tin 

out of the realms of semi-myth to firm reality. Programmes requiring specialised knowledge 

but comparatively few resources are needed for the ceramic material. The South-Western 

Decorated pottery sequence needs more detailed analysis to establish whether there are 

detectable stylistic changes through time. Collections from Roman period sites should be 

examined to identify the full variety of non-Cornish products present. The range of future 

work offers something at every level, from the local field work so well done by the Society's 

amateur members, through to major programmes of research excavation. 

Acknowledgements 

I owe a great debt in the preparation of this paper to Miss P.M. Canyon, both for making the Carvossa material 

available and for arguing points relating to the Roman period. I am grateful to Mr P. Donaldson for allowing me 

to examine the finds and records of Trevelgue. The preliminary draft of the article was read by Mrs P.M. Christie, 

Mr P. Ashbee, Lady (Aileen) Fox, Dr V. Maxfield, Miss A. Preston-Jones and Mr P. Rose and I have benefitted 

greatly from their comments. Figs 5 and 6 were drawn by Miss S. Morris and the various drafts typed by Mrs 

J. Hutchings. Finally, Figs 2-4 were drawn by my husband Mr N.V. Quinnell; my thanks to him for these, and 

for general discussion and encouragement. 

Bibliography 

ApSimon, A.M. and Greenfield, E., 1972. 'The excavation of Bronze Age and Iron Age settlements at Trevisker, 

St Eval, Cornwall', Proc Prehist Soc 38, 302-381. 

Avery, M., 1972. Some Aspects of the Iron Age in South West Britain, Oxford University thesis. 

Avery, M., 1973. 'British La 'retie decorated pottery: an outline', Etudes Celtiques 13, 522 —551. 

Beagrie, N., 1983. 'The St Mawes ingot', Cornish Archaeol 22, 107-111. 

Beagrie, N., 1985. 'Some early tin ingots, ores and slags from Western Europe', J Hist Metal Stud 19-2, 

162-81. 

Bell, M., 1984. 'Environmental archaeology in South West England' in H.C.M. Keeley (ed) Environmental 

Archaeology: A Regional Review, 42-54. 

Bidwell, P.T., 1979. The Legionary Bath-house and Basilica and Forum at Exeter, Exeter Archaeological Reports 1. 

Bidwell, P.T., 1980. Roman Exeter: Fortress and Town. 

Bidwell, P.T., forthcoming. Excavations of Roman Sites in Exeter 1971-80, Exeter Achaeological Reports 4. 

Brooks, R.T., 1974. 'The excavation of The Rumps cliff castle, St Minver, Cornwall', Cornish Archaeol 13, 

5-50. 

Carlyon, P.M., 1981. 'Some second century AD finds from Widemouth Bay', Cornish Archaeol 20, 224-6. 

Canyon, P.M., 1982. 'A Romano-British site at Kilhallon, Tywardreath: excavation in 1975', Cornish Archaeol 

21, 155 — 170. 

Christie, P.M., 1978. 'The excavation of an Iron Age souterrain and settlement at Cam Euny, Sancreed, 

Cornwall', Proc Prehist Soc 44, 309 — 434 . 

Christie, P.M., 1979. 'Cornish souterrains in the light of recent research', Bull Inst Archaeol Univ London 16, 

187-213. 

Clark, E., 1961. Cornish Fogous. 

Coles, J.M. (ed), 1981. Somerset Levels Papers 7. 

131

Collingwood, R.G. and Wright, R.P., 1965. The Roman Inscriptions of Britain 1. 

Cotton, M.A., 1959. 'Cornish cliff castles: a survey', Proc West Cornwall Fld Club 2:3, 113-121. 

Cunliffe, B., 1978. Iron Age Communities in Britain 2nd ed, (1st ed 1974). 

Cunliffe, B., 1982. 'Britain, the Veneti and beyond', Oxford J of Archaeol 1:1, 39-68. 

Cunliffe, B., 1983. Ictis: is it here?' Oxford J Archaeol 2:1, 123-126. 

Cunliffe, B., 1984. 'Iron Age Wessex: continuity and change', in B. Cunliffe and D. Miles, 1984. 

Cunliffe, B. and Miles, D. (eds), 1984. Aspects of the Iron Age in Central Southern Britain. 

Douch, H.L. and Beard, S.W., 1970. 'Excavations at Carvossa, Probus, 1968-1970: preliminary report', 

Cornish Archaeol 9, 93-97. 

Dudley, D., 1954. `Sub-rectangular earthworks with rounded corners', Proc West Cornwall Fld Club 1:2, 54-58. 

Dudley, D., 1956. 'An excavation at Bodrifty, Mulfra, near Penzance', Archaeol J 113, 1-32. 

Dudley, D., 1958. 'The Early Iron Age in Cornwall', Proc W Cornwall Fld Club 2:2, 47-54. 

Dudley, D., 1958a. 'Late Bronze and Early Iron Age settlements in Sperris Croft and Wicca Round, Zennor', 

J Roy lnst Cornwall n. ser 3:2, 66-83. 

Dudley, D., 1960. Tendeen earthwork, Threemilestone, Truro', J Roy Inst Cornwall n. ser, supplement. 

Dudley, D., 1962. 'The excavation of a barrow at Glendorgal, Newquay, 1957', Cornish Archaeol 1, 9-17. 

Dudley, D., 1965. 'An Iron Age cist-burial with two brooches from Trevone, North Cornwall', Cornish Archaeol, 

4, 18-23. 

Elsdon, S., 1978. 'The Pottery' in P.M. Christie, 1978, 396-424. 

Forde-Johnston, J., 1976. Hillforts of the Iron Age in England and Wales. 

Fowler, P.J., 1962. 'A native homestead of the Roman period at Porth Godrevy, Gwithian', Cornish Archaeol, 

1, 61-84. 

Fowler, P.J., 1976. 'Farms and fields in the West Country', in K. Branigan and P.J. Fowler (eds) The Roman 

West Country, 162-182. 

Fox, A., 1952. 'Hill-slope forts and related earthworks in South-West England and South Wales', Archaeol J 109, 

1-22. 

Fox, A., 1961. 'South-western hillforts', in S.S. Frere (ed) Problems of the Iron Age in Southern Britain, 35-60. 

Fox, A., 1964. South West England (2nd ed 1973). 

Fox, A. and Ravenhill, W.L.D., 1969. 'Excavation of a rectilinear earthwork at Trevinnick, St Kew, 1968', 

Cornish Archaeol 8, 89-97. 

Fox, A. and Ravenhill, W.L.D., 1972. 'The Roman Fort at Nanstallon, Cornwall', Britannia 3, 56-111. 

Giot, P.R., 1960. Brittany. 

Griffith, F.M., 1984. 'Roman military sites in Devon: some recent discoveries', Proc Devon Archaeol Soc 42, 

11-32. 

Griffith, F.M., 1985. 'Enclosures in the Launceston area', Cornish Archaeol 24, 149-154. 

Guthrie, A., 1969. 'Excavation of a settlement at Goldherring, Sancreed, 1958-1961', Cornish Archaeol 8, 

5-39. 

Harding, D.W. (ed), 1976. Hillforts: Later Prehistoric Earthworks in Britain and Ireland. 

Harris, D., 1980. 'Excavation of a Romano-British round at Shortlanesend, Kenwyn, Truro', Cornish Archaeol 

19, 63-75. 

Hawkes, C.F.C., 1961. 'The A.B.C. of the British Iron Age', in S.S. Frere (ed) Problems of the Iron Age in 

Southern Britain, 1-16. 

Hawkes, C.F.C., 1984. 'Ictis disentangled, and the British tin trade', Oxford J Archaeol 3:2, 211-234. 

Hencken, H.O'N., 1932. The Archaeology of Cornwall and Scilly, London. 

James, H. and Williams, G., 1982. 'Rural settlement in Roman Dyfed', in D. Miles (ed) The Romano-British 

Countryside Brit Archaeol Rep 103, 289-312. 

Johnson, N. and Rose, P., 1982. 'Defended settlement in Cornwall - an illustrated discussion', in D. Miles (ed) 

The Romano-British Countryside, Brit Archaeol Rep 103, 151-208. 

Laing, L.R., 1968. 'A Greek tin trade with Cornwall', Cornish Archaeol 7, 15-23. 

Lamb, R.G., 1980. Iron Age Promontory Forts in the Northern Isles, Brit Archaeol Rep 79. 

132

McAvoy, F., 1980. 'Excavations of a multi-period site at Carngoon Bank, Lizard', Cornish Archaeol 19, 31-62. 

Megaw, J.V.S., 1967. 'The Trenoweth collar', Cornish Archaeol 6, 5-8. 

Mercer, R.J., 1981. 'Excavations at Cam Brea, Illogan, Cornwall', Cornish Archaeol 20, 1-204. 

Miles, H. et al, 1977. 'Excavations at Killibury hillfort, Egloshayle 1975-6', Cornish Archaeol 16, 89-121. 

Miles, H. and T., 1973. 'Excavations at Trethurgy, St Austell: interim report', Cornish Archaeol 12, 25-30. 

Murphy, K., 1983. 'Pen-y-coed enlosure, Llangynog', Archaeol Wales 23, 8-10. 

Opie, S.A., 1939. Excavations in the Roseland Peninsula. 

Peacock, D.P.S., 1968. 'A petrological study of certain Iron Age pottery from western England', Proc Prehist 

Soc 34, 414-427. 

Peacock, D.P.S., 1969. 'A contribution to the study of Glastonbury ware from South-West Britain', Antiq J 49, 

41-61 

Peacock, D.P.S., 1969a. 'A Romano-British salt working site at Trebarveth, St Keverne', Cornish Archaeol 8, 

47-65. 

Pearce, S.M., 1970. 'Late Roman coinage in South-West Britain', Trans Devon Assoc 102, 19-34. 

Pearce, S.M., 1981. The Archaeology of South West Britain. 

Pearce, S.M., 1983. The Bronze Age metalwork of south-western Britain, Brit Archaeol Rep 120. 

Penhallurick, R.D., 1986. Tin in Antiquity, Institute of Metals. 

Quinnell, H. and Harris, D., 1985. 'Castle Dore: the chronology reconsidered', Cornish Archaeol 24, 123-132. 

Quinnell, H., forthcoming. Excavations at Trethurgy Round, St Austell. 

Radford, CAR., 1951. 'Report on the excavations at Castle Dore', J Roy Inst Corn. n. ser. 1. 

Radford, CAR., 1958. 'Roman Cornwall', Proc West Cornwall Fld Club 2:2, 55-58. 

Reynolds, P.J., 1978. Iron Age Farm: The Butser Experiment. 

Reynolds, P.J., 1982. 'Substructure to superstructure', in P.J. Drury (ed) Structural Reconstruction, Brit Archaeol 

Rep 110. 

Rivet, A.L.F. and Smith, C., 1979. The Place-Names of Roman Britain. 

Roach-Smith, C., 1850. Richborough, Reculver and Lymne. 

Salter, C. and Ehrenreich, R., 1984. 'Iron Age iron metallurgy in central southern Britain', in Cunliffe and Miles, 

1984, 146-161. 

Salway, P., 1981. Roman Britain. 

Saunders, A. and Harris, D., 1982. 'Excavation at Castle Gotha, St Austell', Cornish Archaeol 21, 109-153. 

Saunders, C., 1972. 'The excavations at Grambla, Wendron 1972: interim report', Cornish Archaeol 11, 50-52. 

Schwappach, F., 1969. `Stempelverzierte Keramik von Armorica', in O.H. Frey (ed) Marburger Beitrtige zur 

Archdologie der Kelten, 213-87. 

Schwieso, J., 1976. 'Excavations at Threemilestone Round, Kenwyn, Truro', Cornish Archaeol 15, 50-67. 

Silvester, R.J., 1979. 'The relationship of first millennium settlement to the upland areas of the South-West' Proc 

Devon Archaeol Soc 37, 176-190. 

Smith, G., 1984. 'Penhale coastal promontory fort, Perranzabuloe, 1983', Cornish Archaeol, 23, 180. 

Startin, B., 1982. 'Halligye fogou: excavations in 1981' Cornish Archaeol 21, 185-6. 

Tangye, M., 1973. "Hulls' in Cornwall: a survey and discussion', Cornish Archaeol 12, 31-52. 

Thomas, C., 1958. Gwithian. Ten Years' Work 1949-58. 

Thomas, C., 1963. 'Trial excavations at Mulfra Vean, 1954' Cornish Archaeol 2, 23-28. 

Thomas, C., 1964. Minor sites in the Gwithian area (Iron Age to Recent) Cornish Archaeol 3, 37-62. 

Thomas, C., 1966. 'The character and origins of Roman Dumnonia' in C. Thomas (ed) Rural Settlement in Roman 

Britain, C.B.A. Research Report 7, 74-98. 

Thomas, C., 1968. `Merther Euny, Wendron', Cornish Archaeol 7, 81-2. 

Thomas, C., 1968a. 'Excavations at Crane Godrevy, Gwithian, 1969: interim report', Cornish Archaeol 8, 

84-88. 

Thomas, C., 1972. `Souterrains in the Irish Sea Province: a note', in C. Thomas (ed) The Iron Age in the Irish 

Sea Province, C. B . A . Research Report 9, 75-78. 

Thomas, C., 1976. 'The end of the Roman South-West', in K. Branigan and P.J. Fowler (eds) The Roman West 

Country, 183-197. 

133

Thomas, I.. 1960. 'The excavations at Kynance Cove 1953-60', The Lizard 1:4, 5-16. 

Threipland, L.H.. 1956. 'An excavation at St Mawgan-in-Pydar, North Cornwall', Archaeol J 113, 33-81. 

Todd, M.. 1984. 'Hembury (Devon): Roman troops in a hillfort', Antiquity 58, 171-174. 

Tylecote, R.F., 1966. 'The history of the tin industry in Cornwall', Cornish Archaeol 5, 30-33. 

Wailes, B., 1963. 'Excavations at Castle-an-Dinas, St Columb Major: interim report', Cornish Archaeol 2, 

51-55. 

Wainwright, G.J., 1967. Coygan Camp, Cambrian Archaeological Association. 

Wainwright, G.J., 1971. 'The excavation of a fortified settlement at Walesland Rath, Pembrokeshire', Britannia 

2, 48-108. 

Warner, R.B., 1967. 'The Carnanton tin ingot', Cornish Archaeol 6, 29-31. 

Weatherhill, C.. 1982. The Courtyard and Houses of West Penwith: A Survey (Cornwall Committee for Rescue 

Archaeology). 

Whimster, R.P., 1977. `Harlyn Bay reconsidered: the excavations of 1900-1905 in the light of recent work', 

Cornish Archaeol 16, 60 — 88. 

Whimster, R.P., 1977a. 'Iron Age burial in southern Britain', Proc Prehist Soc 43, 317-27. 

Whimster, R.P., 1981. Burial Practices in Iron Age Britain, Brit Archaeol Rep 90. 

Williams, G., 1985. Fighting and Farming in Iron Age West Wales, Dyfed Archaeological Trust. 

Young, C.J., 1977. Oxfordshire Roman Pottery, Brit Archaeol Rep 43. 

134

Cornwall during the Iron Age and - Cornwall Archaeological Society

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?