Machine-read text (OCR) of 9200671prelim.pdf, 30 pages.
Vale of Glamorgan Council planning file 92/00671 (reserved matters 92/00671/RES, outline 89/01396/OUT), Great House Farm, Llandough.
Supplied by the family from the Vale of Glamorgan Council planning file 92/00671.
Scan: https://wayback.bucklerfamily.estate/pdf/evidence/0000-00-00_wp81_unidentified-scanned-pdf-in-the-wiki-media-library.pdf
Catalogue: https://wayback.bucklerfamily.estate/artifacts/#EV-0209
OCR is machine output; the scan is the record. Drawings and handwriting read poorly.

===== page 1 of 30 =====
EXCAVATIONS AT GREAT HOUSE
FARM, LLANDOUGH, CARDIFF,
SOUTH GLAMORGAN

A PRELIMINARY REPORT

by
Alan Thomas and Neil Holbrook
for

Ideal Homes Wales Ltd.

Cotswold Archaeological Trust

===== page 2 of 30 =====
EXCAVATIONS AT GREAT HOUSE FARM,
LLANDOUGH, CARDIFF, SOUTH GLAMORGAN

Preliminary Report

SENIOR STAFF

SiTE SUPERVISOR: Mike InNGs

Sire PLANNER: JoHN DALTON

SITE PHOTOGRAPHER: DAVE GILBERT

REPORT ILLUSTRATIONS: Rick Morton & GiILLY SHARPE

C.A.T Jos: 0381
C.A.T Report: 94213

DECEMBER 1994

This report has been researched and compiled with all reasonable skill, care, and attention to
detail within the terms of the project as specified by the Client and within the general terms and
conditions of Cotswold Archaeological Trust Ltd. The Trust shall not be liable for any innacuracy,
error or omission in the report or other documents produced as part of the Consultancy and no
liability is accepted for any claim, loss or damage howsoever arising from any opinion stated or
conclusion or other material contained in this report or other documents supplied as part of the
Consultancy.

This report is confidential to the Client. Cotswold Archaeological Trust Ltd accept no
responsibility whatsoever to third parties to whom this report, or any part of it is made known. Any

© Cotswold Archaeological Trust
c/o Corinium Museum, Park Street, Cirencester, GLOS GL7 2BX
Tel (0285) 643625 Fax 0285 644641

===== page 3 of 30 =====
CONTENTS

LIST OF FIGURES

1. INTRODUCTION

2. LOCATION AND GEOLOGY

3. HISTORICAL BACKGROUND AND PREVIOUS EXCAVATION

4. EXCAVATION RESULTS

5. THE FINDS

6. THE RADIOCARBON DATES

7. GENERAL CONCLUSIONS

8. ACKNOWLEDGEMENTS

9. BIBLIOGRAPHY

LIST OF FIGURES

Figure 1 Location of Llandough.

Figure 2 1979 and 1994 excavation areas.

Figure 3 Site plan.

Figure 4 Overall plan of the cemetery defining Areas I-IV.

Figure 5 Distribution of stone-lined graves, wooden coffins and pillow-stone
burials.

Figure 6 _ Distribution of finds.

Figure 7

Distribution of Late Iron Age, Roman and Bii pottery.

18

19

26

27

===== page 4 of 30 =====
EXCAVATIONS AT GREAT HOUSE FARM,
LLANDOUGH, CARDIFF,

SOUTH GLAMORGAN

PRELIMINARY REPORT

1. INTRODUCTION

1.1 In March 1994 Cotswold Archaeological Trust was commissioned by Wyn Thomas &
Partners on behalf of their clients, Ideal Homes Wales Ltd. to undertake the excavation of part
of a site to the north of the church of St. Dochdwy, Llandough, Cardiff in advance of a
residential development. The site had been considered to be of potential archaeological
importance owing to its proximity to a presumed early Christian monastic foundation (eg.
RCAHMW 1988), and when the Vale of Glamorgan Borough Council granted outline
planning permission for residential development in March 1990 a condition required that a
field evaluation survey be undertaken. Accordingly, Glamorgan-Gwent Archaeological Trust
was commissioned to undertake the survey. Eight trenches were dug, suggesting to the
excavators that much of the site had been disturbed by post-medieval quarrying and levelling.
Nearer the church disarticulated human skeletal remains which appeared to have been

disturbed from their original context were located (CBA 1990, 66).

1.2 In March 1992 full planning permission was granted for residential development, with no
conditions requiring further archaeological work. However, in March 1994, Ideal Homes
Wales voluntarily undertook to sponsor an excavation of that part of the site which had been
designated by Glamorgan-Gwent Archaeological Trust (in the light of the evaluation results)
as being of medium-high archaeological potential, and where preservation in-situ was not

possible. The total area to be examined was approximately 0.1ha.

===== page 5 of 30 =====
Great House Farm, Llandough 2

1.3. At an early stage in the excavation it became clear that the site contained numerous
burials which continued beyond the bounds of the excavation area into the zone defined, as a
result of the evaluation, as being of non-existent to low archaeological potential. Many of the
burials lay only a few centimetres below the topsoil. Consequently, Ideal Homes Wales
readily agreed to fund a much enlarged programme of investigation, and the area of
excavation was extended to the north, west and east to cover the full area of surviving
inhumations. Ideal Homes Wales endured a lengthy delay to their building schedule as a result
of the increased scale of works, although CAT sought to assist the developers by excavating
the site in phases which were released in turn for construction to commence. In the event a

0.22ha area was investigated and work finally ceased on site in September 1994.

1.4 Since completion of the excavation, work has concentrated on initial processing of
records, indexing and checking of the site archive, initial scanning of the artefacts, and
obtaining radiocarbon dates to confirm chronology. The results outlined in this report are
therefore necessarily provisional, and may be subject to change as further research is
undertaken. Nevertheless it is hoped that the broad conclusions and overall chronology will

remain valid.

2. LOCATION AND GEOLOGY

2.1 The present day settlement at Llandough lies to the north of Penarth and about 3.5lan.
from the centre of Cardiff (fig. 1). Although now a suburb of Cardiff it was until recently little
more than a village, occupying the gently sloping ground near the crest of an escarpment,
which itself delimits the west side of the low lying ground of the Ely and Taff estuaries. The
uppermost part of this escarpment is made up of Lower Lias, which in this area overlies rocks
of the Rhaetic series and Red Marls (Owen-John 1988, 128-9). The excavation site occupies
an irregularly shaped and relatively flat area, c43m above sea level, to the north of the church
of St. Dochdwy. Here, the Lias took the form of alternating thin horizontal bands of rock and

yellow clay, sloping gently to the south and west.

===== page 6 of 30 =====
Great House Farm, Llandough 3

3. HISTORICAL BACKGROUND AND PREVIOUS EXCAVATION

Introduction

3.1. The church of St Dochdwy at Llandough occupies a site of considerable historical
interest, through its well-known association with the site of an Early Christian monastic
foundation. Its archaeological potential was also established by excavations to the south of
the church by the Barry and Vale Archaeological Group (Beare 1963) and the
Glamorgan-Gwent Archaeological Trust (Owen-John 1988). The latter excavation
established the survival of significant archaeological deposits dating from the late Iron Age
through to the post-medieval period (fig. 2).

Iron Age occupation

3.2 Evidence for a late Iron Age structure was recovered from the 1979 excavation in the form
of a possible roundhouse. This was interpreted as a possible farmstead, perhaps
supplementing agricultural produce with the natural resources of the nearby River Ely and the
sea (Owen-John 1988, 129).

Roman occupation

3.3 The Iron Age structure was superseded by the construction of a well-preserved Roman
villa, probably in the 2nd century AD. A major expansion appears to have taken place in the
early 3rd century after which the site was occupied until the early 4th century. It is also
possible that after the demise of the villa at least part of its buildings were utilised for a

substantial time in the post-Roman period (Owen-John 1988, 144, 152).

Post-Roman monastic settlement

3.4 A number of burials, probably at least 30, were disturbed by building operations outside

of the 1979 main excavation area and it was possible to conduct salvage excavation on two of

===== page 7 of 30 =====
Great House Farm, Llandough 4

these. Radiocarbon dates of 8th-9th century AD were produced from both burials. A further
burial within the excavation area produced an 11th-12th century AD radiocarbon date. The
excavator of the site noted that the burials could belong to the monastic graveyard; the
cemetery of the lay community situated close to the religious settlement; or be the result of a
possible Scandinavian sea-borne attack on the monastery (Owen-John 1988, 145-7). Further
archaeological evidence for post-Roman religious activity is provided by the existence of a
decorated cross shaft, dated to the late 10th or 11th century, within the churchyard. The
presence of early medieval inscribed stones and Christian sculpture is rightly used as a major
indicator of the antiquity of a church site, although it has to borne in mind that they may have
been brought into a churchyard from elsewhere for safe keeping. However, the cross was
in-situ in the 1690's when a sketch was made by or for Edward Llwyd. Descriptions of the
cross can be found in Nash- Williams (1950, 206) and RCAHMW (1976, No. 938). In addition
to its sculptured ornament it is inscribed IRBICI (The stone of Irbic). This is presumably the
name of the person commemorated by the cross or responsible for its erection. The current
location of the cross to the south of the church is not one of antiquity, as a photograph kindly
supplied to us by Diane Brook and Mark Redknap, and probably dated to between 1870-1880,

shows the cross standing to the west of the church.

3.5 These two pieces of archaeological evidence can now be compared with the historical
sources for post-Roman monastic settlement. First amongst these is the place-name of
Llandough itself, first attested in the 7th century, and to be interpreted as “the cemetery or
church of Docco”. Decco (or Docheu) is traditionally believed to be a 5th century British saint
who had links with both Cornwall and south Wales. Later sources record that he became a
bishop and died around AD 473 (although this date only appears much later; Knight 1984b,
378). A monastery which bore his name was founded in Cornwall at St. Kew, near
Wadebridge (monasterium quod Docco vocatur): it may be significant that in the 6th century
Life of St. Samson the community is said to have been in existence for long enough for it to
be noted that it had suffered a decline in the standard of religious practice. In south Wales two
sites bear the name of Docco: Llandough near Cowbridge (where significant pre-Norman
activity has yet to be found) and the present site. It is surely correct therefore to interpret the
frequent references to a monastery at Llandough in the period 650-1075 (according to the

chronology of W. Davies) contained within the Welsh charters as referring to our site. The

===== page 8 of 30 =====
Great House Farm, Llandough 5

monastery may possibly have been transferred from an earlier foundation in Cardiff itself
(Wade-Evans 1943, 125). Of the six monasteries which are likely to have existed in
Glamorgan in the post-Roman period, the sources suggest only Llantwit Major and Llancarfan

may have surpassed Llandough in importance.

3.6 It is not entirely certain when the monastery was abandoned although the cross shaft and
documentary sources certainly point to significant religious activity into the 11th century.
However, it is unlikely that the monastery survived long after the Norman Conquest. Indeed,
the rise of Llandaff, just 3 miles away, to an episcopal centre at this time may have led to the
resources and revenues of Llandough being diverted to its neighbour (Knight 1984b, 379).
Certainly the “village” of Llandough was granted to Tewkesbury Abbey at the close of the
11th century.

Medieval occupation

3.7 A 19th century correspondent noted that “at a short distance from the church, to the
south-east, is a small circular mound, commanding the entrances of the rivers Ely and Taff,
and probably an outpost for the defence of those rivers” (Lewis 1843, 530). This could well
refer to an early castle, possibly a motte or ringwork on a spur of land south-east of the church,
unfortunately destroyed for the construction of a farm in the mid-19th century (Thomas 1988,
177). Otherwise little is known of this earthwork.

3.8 Two large medieval buildings, possibly of late 12th century date, were also found in the
1979 excavation, one possibly part of a barn and the other a dovecote. Pits, crude areas of
paving and sections of walling were also apparent. It is possible that the two structures equate
to a monastic grange belonging to Tewkesbury Abbey (Owen-John 1988, 147-153). At the
Dissolution the grange was confiscated by the Crown and sold (Thomas 1988, 177).

3.9 Part of the medieval village of Llandough was excavated in 1963 when a corner of a
substantial building of 12th or 13th century date was uncovered. A gully containing three
human skeletons of probable 13th century date was also exposed (Beare 1963). Further
structures were revealed during the 1979 watching brief (Owen-John 1988, 155).

===== page 9 of 30 =====
Great House Farm, Llandough 6

3.10 The current church was constructed in the mid-19th century but was built on the site of
a previous church with medieval origins. The current graveyard boundary probably also dates
to the Victorian re-building of the church as a map of 1786, kindly provided by Diane Brook
from her unpublished Ph.D thesis, shows this earlier church with a rectangular boundary.
Both church and boundary were enlarged by the time of a map of 1824, while the current

boundary is shown on a map of 1878.

3.11 North of the church in the area covered by the excavation the land was occupied by the
farmhouse and outbuildings of Great House Farm (demolished 1988). The original date of the
farmhouse is uncertain. However, medieval fabric may have been incorporated into a 17th

century house, reconstructed in the 19th century (RCAHMW 1988).

4. EXCAVATION RESULTS

Preservation.

4.1 Itis clear that the original ground level of the cemetery had been reduced at some point(s)
after it became disused. Over much of the cemetery, particularly to the north and west
articulated skeletons, often in very poor condition, were found only centimetres below the
modern ground level. Although no skeletons were found in the north-eastern corner of the
excavation and only a few in the north-western part of the site, it is almost certain that burials
here had been removed by later truncation. The full extent of the cemetery to the east and west
cannot therefore be certain. Much of the levelling was presumably associated with the
construction of Great House Farm and its outbuildings. Extensive post-medieval quarrying
has also removed large parts of the cemetery. It is therefore certain that the cemetery
contained far more burials immediately after its disuse than survived up to the point of

excavation.

4.2 A total of 858 inhumations were recorded of which 434 were in a good state of
preservation with most of the skeleton intact. A further 153 skeletons were less well

preserved but were on average between 30-70% intact. The remaining 271 were very

===== page 10 of 30 =====
Great House Farm, Llandough 7

fragmentary. The vast majority of the skeletons from the latter two categories were recovered

from the northern part of the site reflecting the intensive use of this area for burial (fig. 3).

4.3 In addition 152 discrete areas of disarticulated human bone groups were identified, mostly
within the fills of later burials and reflecting the disturbance of earlier graves. The vast
majority of these bone groups occurred in the northern part of the site and again reflect the
intensive use of this area for burial. In four instances (B583; B586; B776; and B906), the
disturbance of earlier burials was recognised with the deliberate re-burial of the disturbed

bone in a linear arrangement beside the freshly dug grave.

Spatial Distribution.

4.4 With the lack of general stratigraphy and the non-survival of features such as walls and
ditches it is necessary to analyse the spatial distribution of the burials in order to gain an
understanding of the development of the cemetery. The burials at Llandough appear to occur
in four areas (fig. 4). Area I comprised a possible curvilinear cemetery, bounded to the
north-west by a line of burials on a sw-ne alignment, changing to the east to more of an
east-west alignment. Apart from the orientation of these burials no physical barrier such as a
fence, wall or ditch was detected. However, to the north and west of this alignment a gap of
clm, occasionally cut by later burials, was apparent. It may be that this gap represents the
former existence of a physical barrier, subsequently removed by ground truncation. Certainly
the ground to the north-east of this burial alignment had been reduced after the cemetery had

fallen into disuse.

4.5 Area II lay in the western part of the site and consisted of deliberately placed rows of
occasionally intercutting inhumations. The full western extent of these burials was not
established and ground levelling has probably removed much of its northern extent. However,
a shallow ditch on an almost north-south alignment may delimit this area of burials to the east
as it was not cut by the burials. Unfortunately, much of this feature had been destroyed by

modern disturbance and its full significance is uncertain (fig. 4).

4.6 To the north and west of Area I and on the same general alignment lay further burials

===== page 11 of 30 =====
Great House Farm, Llandough 8

(Area III) which continue to the south-west to cut the original burials in Area II. These
include several graves forming a distinct sw-ne line for c12m lying directly north of and
apparently respecting the postulated northern boundary of the Area I cemetery. This line
included an interspaced, near north-south burial (B381). It is possible that the burials in this
area formed an extension to the burials in Area I. The full northern extent of this expanded
cemetery is unclear as again no feature such as a boundary ditch or wall was detected.
However, further to the north this extension was continued to the northern edge of the site,
close to a steep escarpment and was the area most intensively used for burial (Area IV). Here,
the burials were generally on an east-west alignment and included a distinct row of east-west

aligned graves (fig. 4).

Burial postures.

4.7 The vast majority of the inhumations in all three areas at Llandough were extended and

supine. However, there were at least 8 instances of differing burial postures.

4.8 Flexed burials. Five flexed burials were recovered (B55, B68, B202, B333 and B834).
They are widely spaced over the cemetery, do not form a coherent group and occur in infant,

child and adult burials.

4.9 Crouched, prone and pit burials. Single examples of adjacent crouched and prone burials
were recovered (B474 and B470). In addition, 6m to the west, a further burial (B601)
appeared to have been placed in a pit so that the torso sloped downwards at an angle of 45
degrees. This burial produced a potsherd of late Iron Age date. Unfortunately the legs of this
burial had been removed by modern trenching. All three of these burials occurred towards the
northern edge of Area IV.

4.10 Arm and leg positions. The arm and leg positions have not yet been studied in detail.
However, the arms occur in a variety of positions, usually straight or crossed on the chest,
waist or abdomen. There are also rare instances where one or both arms have been raised by
the head (eg. B80 and B98). The legs are usually both straight although there are rare

examples where the legs have been crossed (eg. B89).

===== page 12 of 30 =====
Great House Farm, Liandough 9

Burial Groups.

4.11 There are very many instances of intercutting graves, especially in Area IV reflecting the
intensive use of this part of the cemetery. There are also several instances where one
inhumation has been placed directly on top of another, although it is uncertain whether these
are contemporary burials. There was also a single instance of a paired burial in Area IT,

where two inhumations appeared to have been placed side by side (B206/212).

4.12 A large amount of infant burials were also apparent in Areas III and IV, probably
representing the final use of at least this part of the cemetery, as they cut and overlie earlier

adult interments.

Age, Sex and Pathology.

4.13 Study of the age, sex, and pathology of the inhumations has yet to commence. However,
infants, juveniles and adults are all represented, along with a single instance of a pregnant
female (B240). Instances of abnormal pathological development were apparent in several

skeletons.

Burial Rites

4.14 Most of the burials in all three areas were plain interments. However, evidence for stone

linings and wooden coffins was recovered from a few graves (fig. 5).

4.15 The evidence for wooden coffins was mostly in the form of black staining around the
edge and bottom of the grave cut. Sixteen instances of such staining were noted in all four
areas of the cemetery. Often these stains were very fragmentary but in a few instances
surrounded much of the skeleton. Several graves also produced very small corroded iron

fragments which may on analysis turn out to be coffin nails.

4.16 A total of 14 burials contained varying degrees of deliberate stone-packing around the

edges of the grave cut, re-using parts of the bedrock removed when the grave was dug. Few

===== page 13 of 30 =====
Great House Farm, Llandough 10

of these graves were totally stone-lined although the vast majority of these graves occurred in

Area I.

4.17 Several graves also contained evidence for burial in shrouds, exhibited normally by a
‘hunching' of the shoulders.

4.18 Four graves containing pillow stones for the head (B18, B61, B74, B81) occurred as a
group within Area I. A few Roman potsherds were recovered from the fills of both B18 and
B61. In addition three burials in Area III (B536, B554 and B635) contained stone supports
either side of the skull.

4.19 Nine burials, mostly in Area I, displayed signs of possible markers. In particular B39
contained a distinct headstone and B110 (and possibly B373) contained a vertical stone
marker at its feet. All three of these graves produced sherds of Roman pottery. The other
burials displayed stones near the top of the skull, possibly supports for some type of grave

marker.

The Post-Cemetery Timber Building.

4.20 In the north-east corner of the excavation the remains of a timber building were
recovered (fig. 3). The full ground plan of this building did not survive due to later levelling
but the remaining post-holes suggest a rectangular structure. Two paired post-holes in its
western side may form an entrance. No floors or other internal deposits survived and it is
therefore difficult to assign a function. Previous excavation suggests that the monastic
grange lay to the south of the church (Owen-John 1988) and it may be that the building was
completely unconnected with any sort of religious usage. The building clearly post-dates the
use of the cemetery as some of the post-holes clearly cut through earlier graves. A total of
122 medieval potsherds were recovered in the excavation, all but five of which were from the
backfill of the post-holes of the building, and presumably therefore relating to its
abandonment. Until further analysis has been undertaken of this material only a broad date

range of the 12-14th century can be offered for the demolition of the structure.

===== page 14 of 30 =====
Great House Farm, Llandough 11

Other Archaeology.

4.21 To the west of the medieval building a ditch ran from near the northern edge of the site,
turning to the south-west. Its date is uncertain although it cut and overlay burials.
Interestingly its alignment respects the curvilinear aspect of the cemetery for much of its
length. It survived as a shallow feature and it could pre-date or post-date the medieval

building.

4.22 At some point much of the site had been the subject of intense quarrying. One large
quarry bisected the excavation and others were apparent to the west of the site during the
construction of two houses. They had been backfilled with what appeared to be a slurry-type
deposit. The date of their excavation and backfill is uncertain, but a large part of the cemetery

must have been removed by these quarrying operations.

4.23 Stone walls and demolition deposits apparent in the excavation all belong to Great
House Farm and its outbuildings. The farm was constructed in the 17th century but may have

had medieval origins RCAHMW 1988).

5. THE FINDS

5.1 Very few of the graves at Llandough contained grave goods (fig. 6). One burial (B796)
contained a complete Roman brooch (Colchester Derivative type; 1st century AD) and three
(B40, B470 and B547) contained glass beads, also of Roman date. Two contained knife
blades (B34 and B898). B449 contained an unidentified iron object, possibly some type of
belt, consisting of bands around the waist of a skeleton. This is currently being examined by
the National Museum of Wales. Five burials (B32, B87, B94, B296, and B326) contained
iron hob-nails, a distinctive late Roman burial tradition. Several graves also contained quartz
and rounded limestone pebbles. Quartz pebbles have been found in association with graves
on other Early Christian cemeteries and in Wales have been found in pre-12th century graves
at Pennant Melangell (Edwards and Lane 1992, 11) and at Capel Maelog (James 1992, 98).
They have also been found within the fill of a holy well at Llandegla (Edwards and Lane 1992,
8).

===== page 15 of 30 =====
Great House Farm, Llandough 12

5.2 A total of 320 potsherds were recovered from the excavation, of which 125 were found in
grave fills (fig. 7). The majority of the latter proved to be of 3rd-4th century date. A
significant sherd amongst the assemblage is a single fragment of ceramique a l'eponge from
an unstratified context. This fabric was produced in south-west France in the late 3rd-4th
centuries, and has a predominately south-western coastal distribution in Britain. Fourteen
sherds of late Iron Age pottery together with small quantities of animal bone, flint, glass,
copper alloy, lead and slag were also recovered from some of the grave fills. Four graves

yielded mid-4th century Roman coins from their fills (B59, B87, B93, B110).

5.3. Of major significance in terms of understanding both the chronology and status of
Llandough is the recognition of five sherds of imported Bii amphora from a variety of grave
fills. There is now a body of evidence for the importation of fine wares and amphorae into
post-Roman western Britain, most probably originating from the Aegean or Constantinople
itself. The main period of importation would appear to be the 75 years between AD 475-550
(Fulford 1989). The discovery of sherds of Bii amphora, a type made mainly along the
southern coast of Asia Minor, especially around the Golfe D'Alexandrette, and generally
considered to have carried olive-oil, is important from a chronological aspect in providing
archaeological evidence of 6th century activity at Llandough. No less significant, however, is
that these imported wares, which were never common, are found on comparatively few sites.
In south Wales by far the most significant site is the defended hilltop enclosure at Dinas
Powys, 2km from Llandough, which has produced fine wares and amphorae in quantity. The
interpretation of Dinas Powys is uncertain, although the high quality metalworking and range
of exotic imports point to a site of high status. By implication therefore the same will be true
for Llandough. The nature of the relationship between Dinas Powys and Llandough, if any, is

one for future discussion.

===== page 16 of 30 =====
Great House Farm, Llandough 13

6. THE RADIOCARBON DATES

6.1. Bones from 4 burials were submitted for radiocarbon dating (fig. 3). Two yielded
insufficient collagen for conventional analysis and were sent for AMS determination. The

results are:

Lab. Ref. Burial Number C14 date (B.P.) 1 Sigma calib.

Beta-76463 2 1570+70 AD 420-590
Beta-76464 65 1170+60 AD 790-970
Beta-76465 228 1070+50 AD 960-1015
Beta-76466 380 1210+60 AD 770-890

6.2. Burial 2 is one of a group of stone-lined graves within Area I. The radiocarbon date from
this burial and the imported amphora fit together well to suggest that the possible curvilinear
cemetery (Area I) was in use at sometime, or throughout, the Sth-6th centuries. Burial 65
also occurs within Area I and suggests that this cemetery was still in use between the late 8th-
mid-10th centuries. Burial 228 occurs in Area III and within a clear linear arrangement of
graves just to the north of the postulated Area I boundary and indicates burial here within the
latter part of the 10th or early 11th century. Burial 380 occurs in a wooden coffin within Area

II and indicates burial here between the late 8th-late 9th centuries.

===== page 17 of 30 =====
Great House Farm, Llandough 14

7. GENERAL CONCLUSIONS

Introduction

7.1 The archaeology of the early Welsh church has received little attention compared with
that of many other parts of Britain and Ireland, partly due to the meagerness of the
documentary sources and more importantly the scantiness of the upstanding remains. Most of
the emphasis has been on sculpture. However, the volume of archaeological excavation in
Wales has greatly increased since the late 1960's and rescue work has been carried out on a
number of minor ecclesiastical sites and cemeteries. Many of these sites have yet to reach
final publication (Edwards and Lane 1992, 1-2). The excavation at Llandough can now be

seen to represent a major addition to the corpus of knowledge on this topic in Wales.

Iron Age and Roman Activity

7.2. Many early Welsh cemeteries had their roots in the prehistoric past (Edwards and Lane
1992, 6) and from this point of view the recovery of a few flints and 14 potsherds of late Iron
Age date is significant. All can be considered as residual in later graves with the possible
exception of a late Iron Age potsherd in pit burial 601. Pit burials are well attested in the late
Iron Age in central southern Britain, and seem to have been particularly popular in the period
prior to the Roman conquest. It has been suggested that the deposition of isolated skeletons in
pits must be considered atypical, and that the bodies were those of criminals, social outcasts,
outsiders and the like, deliberately denied the customary funeral ceremony (Harding 1974,
113). If this burial is of late Iron Age date it is presumably associated with the settlement
beneath the Roman villa 150m to the south.

7.3 A total of 105 sherds of Roman pottery were recovered from stratified contexts. Seventy
of these occur within Area I and 35 in Areas III andIV. This pottery is heavily orientated to
the later 3rd-4th century, and it may be significant that the villa to the south appears to have
been abandoned in the first quarter of the 4th century. Roman grave goods were also
recovered, in the form of the beads and the Colchester-derivative brooch in perfect condition.

The latter is unlikely to date after the 1st century AD. If any of the burials at Llandough are

===== page 18 of 30 =====
Great House Farm, Llandough 15

fully Roman, as opposed to post-Roman, the brooch burial would seem a likely candidate.
Two of the burials containing Roman beads, which are spatially associated in the northern part
of the cemetery, may also be of a similar date. Of further interest is a group of east-west
aligned graves at the southern edge of the site (i.e. nearest the church) which might date to the
late Roman period. Within this group five burials contained iron hob-nails, a well-attested
Roman burial tradition which was widely practised in rural areas of southern England from
the late 2nd into the late 4th century (fig. 6). Their presence is often taken to suggest that the
dead needed a good pair of shoes for the journey into the underworld. One of the hob-nail
burials (B87) also produced a Roman coin of AD 330-45 from the grave fill, suggesting a
mid-late 4th century date for the burial. A further three burials, from the same vicinity as those
containing hob-nails, also produced Roman coins of the period AD 330-50. Without a more
comprehensive programme of radiocarbon dating it is impossible to fully define the scope of
Roman burial on the site. However, at this early stage the possibility of the post-Roman
cemetery having Roman origins cannot easily be dismissed. Indeed, the occurrence of
Christian churches and cemeteries on Roman villa sites is a phenomenon known throughout
most of western Europe and raises the possibility that a Roman villa and/or its estate had been
taken over in working order by the early medieval Church. We may note by way of
comparison that the sub-Roman, and apparently Christian cemetery at Llantwit Major, was
cut into the ruins of a Roman villa (Knight 1984b, 377-8). Indeed, the suggestion that part of
the villa buildings to the south of St. Dochdwy's church were utilised in the post-Roman
period is of great interest (Owen-John 1979, 144 and 152).

The Post-Roman Cemeteries

7.4 The excavation at Llandough represents the largest Early Christian cemetery excavated in
Wales to date. However, it is clear that much of the cemetery within the excavated area has
been destroyed by lowering of the ground level, particularly in the north-western and north
eastern parts of the site. Much of this levelling is presumably associated with Great House
Farm and its outbuildings. Extensive quarrying has also removed large parts of the burial
ground.

===== page 19 of 30 =====
Great House Farm, Llandough 16

7.5 Spatially, the cemetery can be divided into four areas. Areas I and II appear to be the
earliest as the former is respected by and the latter is cut by graves belonging to the Area III
expansion. The radiocarbon dates and the imported Bii amphorae suggest that the possible
curvilinear cemetery (Area I) was in use at sometime, or throughout, the 5th-6th centuries.
Burial 65 also occurs within Area I and suggests that this part of the cemetery was still being

used for burial between the late 8th-mid 10th centuries.

7.6 The radiocarbon date from Area II suggests that this part of the cemetery was in use
during the period of the late 8th-late 9th centuries, and was therefore at least partly
contemporary with Area I. We might perhaps generally note that similar regular burial in
rows is attested at the small cemetery excavated at the Atlantic Trading Estate, Barry, which
could have spanned the period from the 2nd-10th centuries (James 1992, 96-7). Both
cemeteries appear to have been replaced by the expanded, possibly curvilinear cemetery of
Area III. The radiocarbon date from this area suggests that it was in use in the latter part of
the 10th or early 11th centuries. Further to the north the expansion appears to have continued,

possibly until the demise of the monastery, with distinct east-west aligned graves (Area IV).

7.7 Perhaps the most interesting component of the entire cemetery is the identification of the
possible boundary of a curvilinear churchyard (Area I). Charles Thomas first pointed out the
likely antiquity of such curvilinear churchyard enclosures, also hinting that the concept of
circular enclosures might be inherited from the pagan past (Thomas 1971, 38-43, 51-3).
Indeed, rounded churchyards are now generally thought to indicate pre-Norman church
foundations in the Celtic west (Brook 1992, 78). The 5th-6th century radiocarbon date from
Burial 2 in Area I and the Bii amphora sherds confirm this view for Llandough.

7.8 Given the expansion of the cemetery represented by Areas III and IV the cemetery at
Llandough must have been very large in area. This is again of particular interest given the
suggestion by Charles Thomas that the size of ecclesiastical enclosures might be informative,
with the greater the area enclosed the more important the site. Large churchyard areas were
suggested to indicate the sites of early major churches and this has been confirmed by the
work of Diane Brook in the Vale of Glamorgan. Her work indicates that large churchyards

(ie. cl acre or more) seem to be linked with early high-status sites. (Brook 1992, 86). The

===== page 20 of 30 =====
Great House Farm, Llandough 17

relationship with the burials discovered well to the south of the church in the 1979 excavation

is uncertain but in need of re-examination.

7.9 The vast majority of graves at Llandough were simple inhumations laid on their back,
without grave goods. Burial in wooden coffins was rare although this method of burial has
been recorded in other similarly early graveyards in Wales (Edwards and Lane, 1992, 6).
Shroud burials are also noted from other Welsh Early Christian cemeteries. However, the lack
of special graves within, for example, timber structures and ditched enclosures, is very

apparent at Llandough.

7.10 Refining the chronology of the development of the cemeteries at Llandough will
necessarily be speculative until more radiocarbon dates are obtained. Most of the pottery is
residual and artefacts such as the Roman beads could well be heirlooms. The expanded
cemetery at Llandough was established by the latter part of the 10th century to judge from the
radiocarbon date from Burial 228 and burial could well have continued in Area IV until the
demise of the monastery, shortly after the Norman conquest. The marked absence of sherds
of medieval pottery from the graves, in contrast to those associated with the timber building,

strongly suggest that regular burial in graves had ceased by the 12-14th centuries.

7.11 The latest burial activity at Llandough is the cluster of infant burials in Areas ITI and IV.
No direct dating was recovered, although the concentration of infants suggests a special cause
or event. Prof. Charles Thomas has kindly drawn our attention, by way of comparison, with
the burial of children within the churchyard of Tintagel, Cornwall, which might be associated
with a historically recorded plague in the winter of 1546-7 (Nowakowski and Thomas 1992,
27-9). A programme of radiocarbon dates which would include these infant burials would
certainly shed light on the final use of the cemetery and any relationship with the timber

building to the east.

7.12 It cannot be over stressed that the results presented in this report are the result of only
preliminary analysis. In particular the skeletal remains possess huge potential for scientific
study, especially so in the light of the postulated late-Roman/post-Roman continuity of the
site. Full analysis of this material will necessarily take time, but the results will surely be of
cardinal importance for our understanding of the inhabitants of the monastery of St Docco and

their neighbours.

===== page 21 of 30 =====
Great House Farm, Llandough 18

8. ACKNOWLEDGEMENTS

Cotswold Archaeological Trust would like to thank Ray John, Bryn Davies and James Davies
of Ideal Homes for their co-operation both on and off site. Paul Vining of planning consultants
Wyn Thomas & Partners was always on hand to offer much appreciated support and
encouragement throughout the project, as was Royston Clark, the archaeological consultant
from Countryside Planning and Management. Numerous people visited the excavations and
provided knowledgeable advice and comment, notably Charlotte Roberts, our human
osteological consultant from University of Bradford; Jeremy Knight and David Robinson of
Cadw; Mark Redknap, Ken Brassil and Richard Brewer of the National Museum of Wales;
Alan Lane, Bill Manning, and Philip Macdonald of the University of Wales, Cardiff; Charles
Thomas, Mick Aston and Carolyn Heighway. Diane Brook, Jane Stewart and Philip
Macdonald provided several documentary references to the site. The pottery was identified
by Jane Timby and the Bii amphorae examined by David Williams. Pottery from the
evaluation was identified by Joyce Compton of Glamorgan-Gwent Archaeological Trust.
Martin Locock, also of the Glamorgan-Gwent Archaeological Trust, allowed access to the

evaluation archive. Coin identifications were kindly provided by Richard Reece.

The success of the excavation depended on its core team of excavators who worked in the
most difficult of conditions, namely Mark Brett, Jon Matthews, Cate Davies, Stephen Forbes,
Ian Dennis, Jim Turner, Lisa Powell, Rebecca Vaughan, Kim Watkins, Steve O'Rourke and
Barbara Inglis. Several members of CAT staff and students and graduates of the University
of Wales, Cardiff, also worked and volunteered on-site and thanks are in particular due to
Huw Sherlock, Kate Bonner, Brian Jones, Phil Hodges, Brona Langton, Paul Charlton, Andy
Manning, Jane Stewart, Jane Hughes, Liz Pimblett, Gillian Thomas, Susan Fielding,
Kimberley Dore, Louise Barber, Christine Bloor, Duncan Robertson, Richard Ashton, Emma
Roden and Catrin Edwards.

===== page 22 of 30 =====
Great House Farm, Llandough 19

9. BIBLIOGRAPHY

Beare EJ. 1963 Llandough, Penarth. Unpublished manuscript lodged with the Department of
Archaeology, National Museum of Wales.

Brook D. 1992 ‘The Early Christian Church East and West of Offa's Dyke’, in Edwards and
Lane 1992, 77-89.

CBA 1990. 'Llandough’, in Archaeology in Wales, 30, 66.

Edwards N. and Lane A (eds). 1992 The Early Church In Wales and the West. Oxbow
Monograph 16.

Fulford MG. 1989 ‘Byzantium and Britain: a Mediterranean perspective on Post-Roman

mediterranean Imports in Western Britain and Ireland’, Medieval Archaeology, 33, 1-5
Glamorgan-Gwent Archaeological Trust 1990. Land At Great House Farm, Llandough. An
Assessment Of The Archaeological Implications Of Redevelopment At The Site Of Great
House Farm, Llandough.
Harding DW. 1974 The Iron Age in Lowland Britain.

James H. 1992 'Early Medieval Cemeteries in Wales’, in Edwards and Lane 1992, 90-104.

Knight JK. 1984(a) ‘Glamorgan A.D. 400-1100: archaeology and history’, in Savory 1984,
315-364.

Knight JK. 1984(b) 'Sources for the early history of Morgannnwg’, in Savory 1984, 365-409.

Lewis S. 1843 A Topographical Dictionary of Wales.

Nash-Williams VE. 1950 The Early Christian Monuments of Wales, Cardiff.

===== page 23 of 30 =====
Great House Farm, Llandough 20

Nowakowski JA. and Thomas C 1992 Grave News From Tintagel, An Account of a Second
Season of Archaeological Excavation at Tintagel Churchyard, Cornwall, 1991
Cornwall Archaeological Unit.

Owen-John H. 1988 'Llandough: the rescue excavation of a multi-period site near Cardiff,

South Glamorgan’, in Robinson 1988, 123-177.

RCAHMW. 1976 Glamorgan Inventory, I, 3 parts. Royal Commission on Ancient and
Historical Monuments (Wales), Cardiff

RCAHMW. 1988 Notes on sites and finds in Llandough village unpublished notes by Suggett

RF. Royal Commission on Ancient and Historical Monuments (Wales).

Robinson DM. (ed). 1988 Biglis, Caldicot and Llandough: three late Iron Age and Romano-
British sites in south-east Wales. Excavations 1977-79. British Archaeological Report
188, Oxford.

Savory HN. (ed.) 1984 Glamorgan County History, Vol, 2, Early Glamorgan, Pre-history and
early history. Cardiff.

Thomas, C. 1971 The Early Christian archaeology of North Britain, Oxford.

Thomas HJ. 1988 Historical Notes in Owen-John H. 1988, 177.

Wade-Evans AW. 1943 'Sub-Roman Cardiff, Archaeologia Cambrensis, 97, 125-6.

===== page 24 of 30 =====
Great House Farm, Llandough 21

Figure 1. LOCATION OF LLANDOUGH

SOUTH
GLAMORGAN —N-

Cowbridge Llandough

GR: DANS NEC Ne 1 BEN OEN ‘

e/a
ys loctbiiay Grapaetay

u Aspen ee Espate I;

7 “ N ecco
\ Cumalusy tiv Gg

ee! wath i Pa V2
a

ie i
= Fm
.

aK RS
ae of Cae.

~ “geauuite oe
Em

vet
f *Roundbush
Boe [~Tel

Reproduced from the Ordnance Survey's 1: 50000 map
with the permission of the Controller of Her Majesty's

Stationery Office, (c) Crown copyright
CAT c/o Corinium Museum, Park St, Cirencester. ALS0196A

===== page 25 of 30 =====
Great House Farm, Llandough

N
i)

Figure 2. 1979 AND 1994 EXCAVATION AREAS

N
~H_ 1994

excavation

N
\
N
“
XN
XN
N
\ |
\
St. Dochdwy's \
Church \
_—
}
i
Graveyard

wall

1979
ae” excavation

Excavated

areas

Area of 1979
watching brief

50m
—

===== page 26 of 30 =====
Great House Farm, Llandough

23

Figure 3. SITE PLAN

Bl = posthole

(85) = pit

(= radiocarbon dated burials

===== page 27 of 30 =====
Great House Farm, Llandough

Figure 4. OVERALL PLAN OF THE CEMETERY DEFINING AREAS 1-IV

= post hole

Ql = burials excavated against possible
notth boundary (Area |)

disturbance

Bed = pit

(ml = Area 1 burials

boundary of expanded cemetery (Area II)

burials excavated against possible

east-west aligned burial row (Area IV)

===== page 28 of 30 =====
Great House Farm, Llandough

25

Figure 5. DISTRIBUTION OF STONE-LINED GRAVES, WOODEN COFFINS AND
PILLOW-STONE BURIALS

post hole

Unexcavated

= wooden coffi
fed = stone lined grave

bl = pit

"pillow stone” burial

===== page 29 of 30 =====
Great House Farm, Llandough

26
Figure 6. DISTRIBUTION OF FINDS
a
a] a
| es
Bigi2zé
i) ” i 4 i]
EGLiTdaied

===== page 30 of 30 =====
Great House Farm, Llandough

27

Figure 7. DISTRIBUTION OF LATE IRON AGE, ROMAN AND Bii POTTERY

GS = post hole

Late Iron Age

Ee] = pit

1 | = Roman

(9 = Bii Amphorae (imported)