PRN 11 NGR SN41802030
Name BRYN ROMA
Type Bath house
Form Buried Feature
Condition
Summary First revealed in excavations in 1897 (Lloyd 1935) and in 1962. These
revealed a cement lined tank and adjacent room with probable
underfloor heating, presumably the caldarium of a bath house.
Pottery and coin finds dated from the first to fourth centuries (PP
12/5/04).
Long description Excavations for the cellar of No.9 The Esplanade in 1897 revealed
part of a cement lined tank with an adjacent room with flues
entering. This suggested a raised floor and the caldarium of a
bath house. Further remains were found in trenching in the
garden of adjacent Dyffryn House. Finds included a considerable
quantity of box flue tiles and two 3rd century coins. There was 1st
and 2nd century pottery, also 3rd and 4th century BB ware and
mortaria (PP 12/5/04).
PRN 12 NGR SN41712038
Name ST JOHN'S CHURCH
Type Building
Form Documents
Condition
Summary J E Lloyd (1935) writes that there is a plan of the supposed 'Roman
Fort' reproduced in the county history from RCAHM (Carms) which
shows various fields, one of which is marked "Foundations seen
c.1905, now St John's Church". FM 2008
Long description Record of Roman building at St Johns Church. RPS August 2001
St John's Church stands at this site and there are no visible
remains of these foundations. TA James 1976
PRN 15 NGR SN4120
Name RED CROSS
Type Town defences
Form buried feature
Condition Substantial destruction
Summary Relates to a V shaped ditch excavated by GDB Jones in 1968 behind
the offices of the Red Cross on Spilman Street. It produced a quantity
of South Gaulish Samian Ware, the ditch was described as Flavian ,
the 'date would most easily associate it with an early military site'
(from GDB Jones). He believed it represented the southern tip of the
early Roman fort. (PP 12/5/04)
Long description This is the sum-total of published information by GDB Jones on
the siting of the Roman Fort along with site number 38 "Edge of
Fort Paltform" shown on a map published in Carm Ant (1970). TA
James 1976 Behind the offices of the Red Cross, Spilman Street,
an area which is presumed to contain the southern tip of the fort
platform. "A deep section through the late Medieval and Modern
levels located the butt end of a ditch of V-shaped profils (c.5' wide
x 5' deep) belonging to an undoubtedly Flavian context. The ditch
itself produced a quantity of South Gaulish Samian ware .....
while the limited area available restricted examination of the
ditch, its date would most easily associate it with an early military
site controlling the river crossing to the south. Its VICUS will
have grown to the east, eventually to be formalised into the tribal
capital of the Demetae" (Jones 1968).
PRN 21 NGR SN41722024
Name PRIORY STREET INFIRMARY
Type Road
Form Documents
Condition
Summary In 1922 Carmarthenshire Antiquarian Society undertook a salvage
excavation in an area behind Priory Street Infirmary. One feature
discovered was "18ft in length of well-made Roman road....much worn
surface being carefully faced with pudding stone cobbles presumably
brought from the Marros Mountain".
Long description
PRN 30 NGR SN415202
Name BONO REPUBLICAE NATO STONE
Type Inscribed stone
Form Finds
Condition
Summary The discovery of an Roman inscribed stone was recorded by Edward
Donovan when he visted Carmarthen in 1804. The inscription "O RP
NATO" is thought to be of fourth century date. JH Aug 2002
Long description " In the year 1804 the town (of Carmarthen) was visited by
Edward Donovan, who records the discovery of 'an inscribed
stone commemorative of the Roman republic' that was dug up
with some others of a similar nature in the street near the vicar's
house" (Excursion through S.Wales ii,p215). This find was
probably that of a fourth century inscription, the end perhaps of a
dedication rather than of a milestone, mentioning some
fourth-century emperor, with the common fourth-century phrase
bon(o) r(ei)p(ublicae) nato.
PRN 31 NGR SN41952065
Name MORIDUNUM AMPHITHEATRE
Type Amphitheatre
Form Earthwork
Condition Substantially intact
Summary Roman amphitheatre located c250m to the NE of the eastern
boundary of the Roman Town defences. It was excavated in 1968 and
1970, revealing the course of the arena wall and traces of the seating
arrangement, giving an estimated seating capacity of 4500-5000. It
was not closely dated but believed to be mainly 2nd century. (PP
13/5/04)
Long description Half of the amphitheatre is now in the care of the local authority,
and a reconstruction of some of the wooden seating, and of the
line of the arena wall has recently been completed. The other half
of the site is, however, still built over. TA James 1980 A Roman
amphitheatre was excavated by GDB Jones, for DofE, in 1968.
Located 150yds east of the presumed east gate of the Roman
town (SN42SW) and constructed by cutting into the hillside (to N)
and using the excavated soil to make up the south bank of the
cavea. Four sections (three to N and one to S), defined the
position of the arena floor which was under 6-7ft of silt. The arena
wall was located as a 7ft foundation trench with a gutter 1ft wide
on the inner side. Erosion and modern development has reduced
the banking, but the dug-out portion on the north is probably at
the original level. Dimension: arena c.150ft by 90ft. Circ. of
cavea c.300ft by 200ft.
PRN 33 NGR SN4120
Name RICHMOND TERRACE
Type Town defences
Form Earthwork
Condition Substantial destruction
Summary The line of defences of the Roman Town (PRN 69), traced through
small scale excavations, watching briefs and topographical features.
Two phases of defences have been identified, dating is imprecise but
the 1st phase is believed to have been built in the late 2nd century,
replaced by the 2nd phase in the mid to late 3rd century. (PP
13/5/04).
Long description The 1st and 2nd phase Roman defences on the SW side of the
town was well established by excavation on the E side of Parade
road in 1978, the 2nd phase outer ditch probably lay under
present day Parade Road, the western edge of this ditch was
picked up in excavations in 1988. Topographical evidence
suggests the line northwards is fossilised in the alignment of Little
Water Street. Parts of the phase 1 outer ditch was picked up in
1978 excavations. The northern line along Richmond Terrace has
also been traced on topographical and archaeological grounds.
Six trenches were opened up in 1968-9 by GDB Jones, revealing
2 phases of rampart construction.(Event PRN 49644). ^An
evaluation trench in 1996 also picked up parts of the phase 2
rampart. Topographical evidence and recent watching briefs have
picked up the eastern line, running along the western side of Old
Oak Lane and then roughly along the line of a footpath to the rear
of No.23 Priory Road. The southern line has been traced fairly
precisely on topographical grounds and a small amount of
archaeological evidence. The line is expected to run along the
edge of the steep fluvio-glacial terrace above the floodplain. The
1st phase defences cannot have been constructed before the late
2nd century, and there is a suggestion the 2nd phase might date
from the mid to late 3rd century AD. (PP 13/5/04)
PRN 34 NGR SN41452038
Name FOOTBALL GROUND
Type Road
Form Buried Feature
Condition
Summary A Roman-period street in Carmarthen, the construction of which was
dated by a worn coin of Aurelius (AD161-80) recovered from its
primary make-up. M. Page 2008 based on James, T 1980.
Long description
PRN 35 NGR SN41452040
Name ST PETER'S STREET CAR PARK
Type Furnace
Form Buried Feature
Condition Substantial destruction
Summary Excavations by GDB Jones at St Peter's Street carpark in 1969 (event
PRN 49643), revealed an E-W aligned Roman street. Fronting the
street to the north were timber built structures of industrial character
of 2nd to 4th century date. These buildings comprised of 3 small
furnaces, one of which yielded a crucible thought to have been used
in working copper. (PP 13/5/04)
Long description During excavations to the east of St Peter Street car park in 1969
by GDB Jones, a Roman Street or DECUMANUS was revealed
splitting the excavation into two separate parts. Upon further
analysis it was found that the buildings to thenorth of the
Decumanus were of an industrial character, distinct from those to
the south which were larger residential buildings. These industrial
buildings comprised 3 small furnaces, one of which yielded a
crucible thought to have been used in working copper. The
structures were timber built. TA James 1976
PRN 36 NGR SN41472037
Name ST PETER'S STREET CAR PARK
Type Dwelling
Form Buried feature
Condition
Summary The 1967 excavations at St Peters car park revealed a street
(DECUMANUS) running in a W-E direction. To the south of the
Decumanus the buildings were of a residential character, dated to the
mid fourth century.
Long description The 1967 excavations at St Peters car park revealed a street
(DECUMANUS) running in a W-E direction. To the south of the
Decumanus the buildings were of a residential character, whereas
those to the north were more representative of small industry or
trade. Although much of the evidence was badly robbed it
appeared that most of the rooms had been floored with tessarae
of terracotta and Prescelly stone. One room contained a
hypocaust with a furnace. This construction was tightly dated by
the discovery in the foundation level of "two fresh coins of
Magnentius and his brother Decentium (AD 351-3)." Allowing the
structure a normal life (there was no evidence of violent
destruction) this takes urban life in the westernmost cantonal
capital of Roman Britain into the last quarter of the fourth
century, if not later. TA James 1976
PRN 37 NGR SN41652048
Name PRIORY STREET
Type Building
Form Buried feature
Condition
Summary The 1968 trial excavations by GDB Jones in Priory Street car park
revealed traces of timber and half-timbered buildings of at least four
periods.
Long description The 1968 trial excavtions by GDB Jones in Priory Street car park
revealed traces of timber and half-timbered buildings of at least
four periods......"the rise in building levels being created in the
later phases by extensive dumping of river gravel. The buildings
and accompanying roads show a change of alignment between
the early and later periods. At first they appear to reflect a street
grid aligned on the Priory Street axis. Later this seems to have
been abandoned because buildings in the last two phases are
associated with a road running at an acute angle to the Priory St
line towards the presumed site of the east gate."
PRN 39 NGR SN414200
Name MORIDUNUM
Type Fort
Form Documents
Condition
Summary The Roman fort at Carmarthen is thought to have been established
during the conquest of west wales by Govenor Julius Frontinus A.D.
74-77.The exact size, location and layout of the fort is uncertain but
limited evidence from archaeological excavation suggests that an
initial auxilliary fort was later reduced in size, before abandonment in
the second century A.D. There is as yet no clear evidence of a vicus
to the east or west of the fort.DS 27/10/2004.
Long description The possibility that a Roman fort of Flavian context stood to the
west of Moridunum - the cantonal capital - has been raised by
Barri Jones (Roman Frontier in Wales 1969) and a conjectural
plan of the outline of the supposed fort is contained in Delaney &
Soulsby 1975. This is based partially on a ditch (PRN 15) c.5ft
wide by 5ft deep "belonging to an undoubtedly Flavian context"
which produced a quantity of South Gaulish Samian of the
potters: Severus, MeOillus and Cealaus (GDB Jones 1968 see Site
No 15). Furthermore a map by Jones in the Carm Ant (1970)
indicates the existence of the edge of the "Fort Platform". Other
than this there is no further evidence for its existence and it
remains, therefore, highly conjectural. TA James 1976
PRN 69 NGR SN4020
Name MORIDUNUM DEMETARUM
Type Town
Form Documents;Buried Fea
Condition
Summary The Roman Town of Moridunum. The defences (PRN 33) of the town
have been traced on archaeological and topographical grounds,
archaeology has also revealed various internal buildings and streets,
including a bath house and industrial buildings. The original Roman
Fort lay to the west in the King Street/Spilman Street area, the first
phase of the town defences were probably built in the late 2nd
century AD. (PP 17/5/04)
Long description Moridunum = MOR =SEA; DUN = FORTIFIED CITY Caerfyrddin =
CAER.MOR.DUN = FORTIFIED CITY OF THE SEA? (TA James
1976) "The correct form (of the name) is unquestionably
MORIDUNUM with celtic mori "sea" meaning sea fort" (K Jackson
1948)
PRN 218 NGR SN41301940
Name PENSARN
Type Road
Form buried feature
Condition
Summary A possible length of Roman road running south from the west end of
the fort at Carmarthen, and crossing the Afon Tywi at the presumed
location of the Roman crossing point in the same location as the
modern bridge. The road is believed to change to a southeast course
at SN41261928. This part is identified as 'Roman Road' on Ordnance
Survey maps. Recent excavation and observation (PRN 46370)
revealed no evidence of any surviving Roman road. If of Roman
origin this segment would form part of Roman road RR60d (PRN
3401), it may however, be of Turnpike origin. DS 27/10/04
Long description .Possible Roman road running south from Carmarthen and used
into Post Med period PRN 24465. JH 1995
PRN 508 NGR SR8997
Name CASTLEMARTIN BURROWS
Type Findspot
Form Finds
Condition
Summary A group of Roman finds including a blue bead necklace, spindle whorls
and a bronze pendant from an unspecified location in Brownslade or
Linney Burrows. Murphy,K 1993
Long description "An exceptionably well-preserved necklace of blue beads found
scattered in a circle about 12 inches in diameter, were identified
as Roman" Ordnance Survey 1964
PRN 517 NGR SR8997
Name LINNEY BURROWS
Type Findspot
Form Finds
Condition
Summary Roman finds including pottery, a bead, pin, bangle, brooch, ring and
tile from an unspecified location in Linney Burrows indicate a
settlement site. Murphy,K 1993
Long description According to the Ordnance Survey in 1964 the Rev J P
Gordon-Williams recorded discovering the following Roman finds
from Linney, Castlemartin: 1) - frags. of 3 Samian vessels and 2
of white ware jugs 2) - broken bead of green glass 3) - a spindle
whorl 4) - an enamelled pin of bronze, sunflower type, with loops
5) - broken bronze bangle decorated with crosshatch 6) - an
intaglio 7) - hexagonal Roman flooring tile bearing initials VC 8) -
a Roman brooch of 2nd century type found c.1926.
PRN 550 NGR SR9028396629
Name BULLIBER;WARMANS HILL
Type Occupation site
Form Finds
Condition
Summary A find of well-made pottery from Bulliber hillfort 549 indicates Roman
or Dark Age occupation of the site. Murphy, K 1993
Long description
PRN 604 NGR SR9359694326
Name CROCKSYDAM CAMP
Type Occupation site
Form Earthwork
Condition Substantial destruction
Summary Finds including pottery and other occupation debris from excavations
indicate Roman occupation of Crocksydam Camp, 603. Murphy, K
1993
Long description "Crossley refers to remains of a round stone hut to the north of
the outcrop within the defences. Grimes described a floor of burnt
clay 18 inches thick, strewn with Romano-British pottery of the
3rd and 4th century types, burnt and unburnt bones and other
occupation debris..." Ordnance Survey 1964
PRN 610 NGR SR95439340
Name BUCKSPOOL CAMP
Type Occupation site
Form Finds
Condition
Summary Finds, including pottery, indicate Roman occupation of Buckspool
Camp,609. Murphy, K 1993
Long description "Romano-British pottery of 2nd century date was found at this
site........It is in the National Museum of Wales." Ordnance Survey
1964
PRN 872 NGR SN6122
Name TACITUS STONE
Type Milestone
Form Finds
Condition Moved
Summary A milestone bearing an inscription to the Emperor Tacitus
(AD275-276) recorded in 1697 apparantly built into a farmhouse near
Dinefwr and then subsequently built into a surrounding wall although
the object is now lost. (EGH 2003).
Long description
PRN 989 NGR SN10753850
Name DYFFRYN QUARRY
Type Findspot
Form Finds
Condition
Summary The remains of a black Roman urn of 3rd to 4th century date found in
Dyffryn Quarry in 1925.
Long description "A broken, inverted, black Roman urn of 3rd/4th century date
was found in Dyffryn Quarry, Meline, in February 1925. The finder
and precise site are unknown, and the urn...is now in the County
Museum, Carmarthen" Ordnance Survey 1966
PRN 1011 NGR SN15573360
Name MOEL TRIGARN
Type Occupation site
Form Finds
Condition
Summary Occupation of this hillfort site in the Romano-British period is
suggested by the presence of artefacts of that date at the site. MM
May 2003
Long description "Occupation in the Romano-British period was suggested by
fragments of a mortarium, spindle whorls, beads and a coin of
Faustina II. Most of the finds are in Tenby Museum." Ordnance
Survey 1965
PRN 1278 NGR SN00712679
Name CASTELL FLEMISH;CASTLE
Type Farmstead,Enclosed settlement
Form Earthwork
Condition Damaged
Summary
Long description
PRN 1417 NGR SN07072365
Name FFYNNON-GAIN
Type Enclosure
Form Cropmark
Condition
Summary In 1975 the Ordnance Survey identified a roughly square cropmark
enclosure on aerial photographs. It measured approximately 50m
square with rounded corners. They suggested that the shape, size and
situation of the vegetation mark, on ground falling gently to the SE,
indicated that it is a Romano-British homestead. They noted that there
are no extant remains, but that changes in soil texture had been
noted. This feature is not visible on 1955 meridian aerial photographs.
F Murphy 2008
Long description
PRN 1654 NGR SN428170
Name ABERCYFOR
Type Homestead,Findspot
Form Finds
Condition
Summary Allegedly in the late 18th century there was evidence of a Roman
building at Abercyfor, nr Caramarthen. The site location is unknown.
"Abercyfor" is now incorporated in the names of three separate farms
in the area centred at SN428170, at none of which is there any
visible evidence of a possible earthwork or building. (F Murphy 2009
based on Ordnance Survey 1967)
Long description The site location is unknown. "Abercyfor" is now incorporated in
the names of three separate farms in the area centred at
SN428170, at none of which is there any visible evidence of a
possible earthwork or building. (Ordnance Survey 1967) A Roman
building possessing "a remarkable ....tessellated pavement with
a prodigious quantity of silver and copper coins of the Lower
Empire", was found in the late 18th century at Abercyfor, about 2
miles south of Carmarthen. The character of the finds , in
conjunction with the isolated site, are suggestive of a villa or
Romanised farmstead. "The hoard of late coins suggests the
decay or destruction of the site in the late 3rd or early 4th
century". (History of Carm 1935)
PRN 1887 NGR SN653397
Name CWRT-Y-CILION
Type Cremation burial
Form Finds
Condition Unknown
Summary An urn, thought to be a Romano-British cremation urn, was
discovered on land at Cwrt y Cilion in 1909. It was described as being
'on the left bank of the River Cothi, and about 25 yards from the
water (RCAHM:1917:p.37).This site is in an area which is well known
for Roman archaeology. It lies in close proximity to the mines at
Dolaucothi, and to the Roman road and fort at Llandovery. PG June
2003.
Long description The urn was found in the side of an alluvial terrace in a field
adjoing Dolaucothi estate about 200ft south on th left side of the
River Cothi looking downstream. (TCASFC 1909-10) From the
above description the find-spot would appear to be SN653397.
Urn was given to Carmarthen Museum in 1921.
PRN 1945 NGR SN65624024
Name YNYSAU-UCHAF
Type Bath house
Form Documents
Condition
Summary In 1831 two hypocausted room with plain tessellated pavements were
excavated by Mr John Johnes. Thought to be the remains of
bath-buildings for a mining settlement
Long description Approximately 40m south of the location of PRN 1945, a number
of trial excavations undertaken in 1997 (PRN 42780) on behalf of
the National Trust revealed the presence of a substantial road,
with traces of related structures and features along its eastern
side. A related geophysical survey demonstrated that the road
and the associated settlement also extended across the field to
the south. CN based on BB 1997 The OS siting corresponds with
an artificial hollow set into a wooded slope above the Afon Cothi
flood plain. Inspection produced numerous fragments of Roman
brick and a scatter of crushed quartz, but it is likely that these
finds repesent excavation spoil, and that the actual site of the
building falls on the flat ground in the field a few metres to the
south above the hollow at SN65624023 (Ordnance Survey 1975)
The late Sir Ian Richmond had suggested to Jarrett that the
bath-house might be military. There is a long gap between the
forts of Llanio (SN6456) and Llandovery (SN7735) and a site
south of the Cothi and close to the bath house would be likely.
Daniel (writing the section on Bath Houses) points out that the use
of brick tesserae to decorate one of the floors is difficult to
parallel in a military bath-house. The building, therefore , is still
best interpreted as part of the mining settlement. (Roman Frontier
in Wales 1969).
PRN 1946 NGR SN66454025
Name DOLAUCOTHI;OGOFAU;CAIO
Type Gold mine
Form Complex
Condition Substantially intact
Summary The site of the historic gold mines at Dolaucothi, which has been
shown to have been exploited from at least Roman times until the
20th century. RPS 24.09.02
Long description The Dolaucothi goldmines. These goldmines were utilised over a
long period of time, and a large part of the mining complex has
been designated as a SAM (CM-208).Dolaucothi was probably first
used for gold extraction in the pre-Roman period. A small group
of hut circles south of the main complex seem to be associated
with small-scale mining of an exposed rock-face (Lewis and
Jones:1969:p.253), indicating that the site was known prior to its
development by the Romans. The Roman phase of workings is
the one for which Dolaucothi is most famous. The site is
immensely complex and the precise nature and chronology of the
workings is still not well understood. However, an outline
understanding of the mining process can be gained by
interpretation of the surface remains (principally earthworks).
Water was conveyed into the site by means of two main
aqueducts - the Cothi leat (PRN 1947, SAM Cm200N) and the
Annell leat (PRN 8882, SAM Cm210) - and a number of smaller
leats. The Cothi leat is notable for bringing water into the site
from a distance of over 7 miles! Both of the leats are carefully
engineered, running along gentle gradients and charging a
number of tanks along their routes. Most of the mining features
themselves - adits and levels as well as opencast workings - are
in the area just north of Gwarnoethle farm, and a number of them
are open to the public. It is supposed that the earlier mining on
the site was opencast, and that as the easily accessible gold was
exhausted the underground workings were put in place. Many of
the Roman features at Dolaucothi have been obscured or
modified during later phases of use and it has been suggested
that a crushing mill was in use at the site somewhere between
1200 and 1700 A.D (Austin and Burnham:1984:p.304 - 313).
Although a Roman date could also be applied to this mill
(Burnham:1997: p.325 - 336) it does seem that sporadic,
small-scale extraction took place at intervals until the late 19th
/early 20th century. At this stage the mines were re-opened more
fully, and several of the Roman levels re-used (Annels and
Williams:1983:p.3). During 1938-9 the site was taken on by
British Goldfields, and this phase of use has left very visible
remains (PRN 20429 and 49154 - 49162). Sadly, Dolaucothi
proved financially unviable and was abandoned by British
Goldfields only a year after work had started in earnest. The
visible mining remains across the whole site form a complex,
multi-period landscape. The importance of the Roman elements
has been recognised by the designation of the aqueducts and the
central mining area as Scheduled Ancient Monuments. However,
the importance of the later mining activities should not be
underestimated.
description
PRN 1947 NGR SN6883542996
Name DOLAU COTHI AQUEDUCT;COTHI
Type Leat,Aqueduct
Form Earthwork
Condition Damaged
Summary The remains of a leat, parts of which are a scheduled ancient
monument, (CM 200, section K & J). Much of the leat is visible as a
linear feature running across several fields. Parts of it are quite well
preserved as a broadly level shelf in the hillside c. 2m across, in other
sections however, it is reduced to a vague break in slope which
closely follows the contours. A survey by the National Trust points out
that it is most easily appreciated from the road below Llwynceiliog
farm, where it can be seen as a darker green than the surrounding
grass where it crosses pasture and a clear line of differential growth
where it crosses bracken. Recent reports from the Cadw Field
Monument Warden indicate that the leat is in a stable condition.
During a farm visit to Llwynceiliog, the length of the leat within the
farm holding was observed. The line of the leat can be seen with
difficulty within the triangular woodland parcel between the road and
the track leading to Glan-meddyg, due the undergrowth of bramble,
however it seems to survive reasonably intact. In the fields
immediately to the east the leat is barely a levelling off of the natural
hillslope, but its course is still traceable. Its profile appears to be best
preserved in the hedgebanks. In the field directly behind and to the
southeast of Llwynceiliog farm the leat is well preserved and survives
as an almost level terrace over 2 metres wide. Where it continues
northeast into the valley behind the farm it becomes less clear, in part
due to the rough terrain and bracken. To the north, the course of the
leat becomes increasingly difficult to trace in improved fields where
reseeding and cultivation has taken place. AMP 2005.
Long description The leat brings water along the slopes of Allt Cwmhenog into the
Dolaucothi mines (PRN 1946). The engineering is complex, and
the leat charged several tanks along its route before entering the
largest, last tank (PRN 49148). It has been suggested that a
sluice controlled water flow into the tank (PRN 49148) and
washing tables (PRN 49150) (Lewis and Jones, 1969,p.258) but
this is hard to verify on the ground. The line of the leat is visible
as a break in the northeast corner of the tank and can be traced
faintly running northwest towards the fenceline.;This Scheduled
Ancient Monument lies just east of the Dolaucothi woodlands. The
Cothi leat brought water along the west side of the slopes of Allt
Cwmhenog into the Dolaucothi mines and eventually charged the
largest tank in the complex (Tank E, PRN 49148). It probably also
provided water for the washing tables (PRN 49150). The
best-preserved sections of the leat have been scheduled and
therefore have statutory protection. PG 2003. A short section of
the leat passes within Gwarnoethle farm boundaries (PRN 47200).
The leat brings water along the slopes of Allt Cwmhenog into the
Dolaucothi mines (PRN 1946). The engineering is complex, and
the leat charged several tanks along its route before entering the
largest, last tank (PRN 49148). It has been suggested that a
sluice controlled water flow into the tank (PRN 49148) and
washing tables (PRN 49150) (Lewis and Jones, 1969,p.258) but
this is hard to verify on the ground. The line of the leat is visible
PRN 1948 NGR SN668404
Name ALLT CWMHENOG
Type Aqueduct
Form Earthwork
Condition Damaged
Summary Part of the Roman aqueduct system that supplied water to the nearby
gold mines at Dolaucothi. RPS 24.9.02
Long description
PRN 1949 NGR SN67614139
description Name ALLT Y BRUNANT as a break in teh northeast corner of the tank and can be traced
Type Reservoir faintly running northwest towards the fenceline. PG 2003. The
Form O.Struct Cothi leat is partly scheduled as CM-200. The leat is, in general,
clearly visible as a break in slope which in places is up to 2m wide
Condition Substantially intact and in part has been used as a trackway. In places the leat is
Summary Small rectilinear earthwork in the corner of a field which has been rock cut. LA 1999
identified as a water tank associated with the Roman aqueduct system
that supplied water to the Dolaucothi gold mines. RPS 24.9.02
Long description A Roman water tank, discovered during the field investigation and
confirmed by Dr Lewis, is situated on the north facing slope of
the Cothi valley above the Cothi aqueduct. Measuring 14m NE-SW
by 9m transversely it is fed by a fast flowing spring in its SE
corner. This water is now allowed to flow out of the tank through
a breach in the retaining bank on the NW side. It must have been
constructed to supplement the water supply flowing in the Cothi
aqueduct though there is now no trace of an effluent channel
leading from it. (Ordnance Survey 1975)
PRN 1952 NGR SN66334028
Name CARREG PUMSAINT
Type Anvil stone
Form Finds
Condition Substantially intact
Summary This stone slab has been set upright in the fashion of a standing
stone. It is likely that it is an early anvil stone,used in the stone
crushing process at the nearby gold mine and that it was placed in its
current position long after its use had ceased. The stone is notable for
five depressions in its surface (worn by the hammers used in the
crushing process, but the basis for the legend which has it that the
five saints of Pumsaint used the stone as shelter from a storm and
that the depressions are the impressions left where they leant against
the stone. RPS 24.09.02
Long description The stone is 1.0m high and the four faces are each 0.6m across
and show numerous depressions presumably the result of stone
crushing. (Ordnance Survey 1969). Carreg Pumsaint is a weather
beaten boulder, 3ft 6in long, roughly quadrangular and on each
of its sides which are about 2ft wide there are four or five
basin-like cavities overlapping each other. It was used for
crushing quartz and now stands on a small mound as the
entrance to the gold workings. A drawing of c.1835 shows it
prostrate and probably not in its present position. (Ordnance
PRN 1958 NGR SN660410
Name CAE GARREG LWYD
Type Hoard,Findspot
Form Finds
Condition
Summary An assemblage of Roman gold ornaments found in the late
1700s/early 1800s in a field called 'Cae garreg llwyd' on the
Dolaucothi estate.
Long description A number of gold ornaments of 2nd-3rd century AD date were
ploughed up in a field called 'Cae garreg llwyd' on the Dolaucothi
Estate in 1796 or 1797. According to the RCAHM (1917) the gold
ornaments comprise a necklace, a small eight spoked wheel,
fragment of an armlet and a rod both of serpentine form, a chain
with an attcahed crescent pendant., and reputedly three or four
rings found at the same time. In 1819 another similar wheel and
chain (PRN 1669) were recovered, possibly from the same
location. In 1975 the Ordnance Survey failed to establish the
position of 'Cae garreg llwyd' field, reporting that the area of the
finds as originally reported (centred approximately SN660410)
falls in arable land below 700ft, on the SE side of Allt Dolaucothi,
where many of the late 18th century field boundaries have since
been destroyed. F Murphy 2009 Findspot of a number of Roman
gold jewellery objects during the early 19th century. RPS
24.09.02
PRN 1961 NGR SN66144030
Name BRYN GLAS
Type Industrial site
Form Documents;Find
Condition
Summary An area of the Dolaucothi gold mine complex which has been dated to
the Roman period. It includes a scatter of quartz fragments of a clay
floor, the result of rock crushing associated with basic gold extraction
processes. RPS 24.09.02
Long description There are no surface indications of the quartz processing area,
which remains under pasture. (Ordnance Survey 1975) An
exploratory excavation in the field 175m W of Ogofau Lodge,
close to the line of the minor road leading to the main upcasts of
Ogofau Gold Mines, revealed a spread of crushed quartz
fragments reduced to a width of 0.6cm, lying over an artificial
clay floor. These fragments represent remains of ground quartz
from the basic gold extraction process, and the artificial clay floor
may be interpreted as a working or washing surface. (Jones &
Little 1972)
PRN 2097 NGR SN28905137
Name PARC CARREG Y LLUNIAU
Type Cremation burial
Form finds only
Condition Intact
Summary An urn containing ashes was discovered together with some coins of
Vespasian when a cairn (PRN 2098) was demolished. The urn, a small
grey-ware olla of 1st or 2nd century type, is now in the National
Museum of Wales.
Long description Site of barrow PRN 2098 (now levelled), possibly bronze age in
origin, (re-)used for a c.2nd century Romano-British cremation
burial PRN 2097. It was also the findspot of a Group I
Latin-inscribed ECM (PRN 2096) which is a memorial to one
Corbalengus, bearing the inscription `of Corbalengus, he lies, an
Ordovician', suggesting that the barrow was again (re-)used in
the 5th - early 6th century for the burial of a post-Roman, ?high
status individual. The stone has been re-erected in the centre of
the natural rise (forming a slight plateau) in the field, close to the
site of the ?barrow. It is a Scheduled Ancient Monument, but has
not been fenced off and the field has been recently ploughed or
harrowed close to the stone. NDL 2004 "In the 18th century a
windmill stood on the top of a small cairn on the spot now
occupied by the stone (PRN 2096) which at that time lay to one
side. When the windmill and cairn were demolished (shortly after
1800 according to Meyrick) an urn (PRN 2097) containing ashes
was discovered, together with some coins of Vespasian. The
stone was then set upright on the site of the cairn from which it
was supposed to originally have come, although Walrond, noting
the early date of the finds, suggests that there may have been
more than one burial mound on the site. The urn and an aureus
of Titus said to have come from the mound, although from some
accounts it may have been discovered a short distance away,
were presented to the National Museum of Wales. The stone
stands in a field still called 'Parc Carreg y Lluniau'. " Ordnance
Survey 1972-3
PRN 2243 NGR SN3423
Name PANTYKENDY
Type Cremation burial
Form Finds
Condition
Summary A Roman cremation urn discovered in c.1833 during repairs in a road
near Pantycendy (SN347232). It was found "a few feet below the
surface .....containing ashes, with which also the place where it was
found was discoloured to a considerable extent; the urn immediately
on its being removed fell in pieces" (Topog Dict Wales 1833).
Long description According to the landowner no further finds had been made ( E
Evans 1984) Roman coins are also said to have been found but it
is not stated if they were associated with the urn. (Ordnance
Survey 1966)
PRN 2286 NGR SN37613712
Name CASTELL WAUN WITHAN
Type Fortlet?,Defended enclosure
Form Earthwork
Condition Good earthwork
Summary A 30m square, univallate earthwork with rounded corners, that lies in
marshy ground on Waun-Withan Farm. It has two centrally placed
entrances on the north and south sides. There is little trace of an
outer ditch. The earthwork is possibly a Roman fortlet or signal
station.
Long description Situated on low lying marshy ground at 160m above sea level, on
a gentle northwest-facing slope, the date and purpose of this
substantial square earthwork remain uncertain. In 1917 The
RCAHM recorded this earthwork as a 96ft (c.30m) square with
rounded angles lying in marshy ground on Waun-Withan Farm.
The bank, 6ft (1.8m) in height and of great thickness, had
entrances centrally placed on the north and south sides both 10ft
(3m) wide. The ditch had almost entirely disappeared. They noted
that locally it is known as ‘Castel,’ ‘Castell waun Withan’ or Yr
amddiffynfa,’ ‘the post of defence’ and that an adjacent field is
called ‘Parc sarn’ (causeway field), and that traces of a paved
way were found when ploughing. A stone basin was reported as
having been found within the enclosure. By 1977 the Ordnance
Survey site visit records that the earthwork is as described by the
RCAHM, except that the corners are not of Roman type. They
report that of the two entrances the one on the north is only 1.5m
wide and may be the result of mutilation, as a small spoil heap
lies outside it. The other entrance is 3.0m wide and overgrown.
The bank of the enclosure is 1.5m high internally and 1.9m high
externally. It is noted that the interior is raised above the
surrounding marsh and is dry. The southern side for a length of
12.4m is further raised in two stages of about 0.1m each, and on
the higher stage, abutting the south bank of the enclosure, there
are indications of buried walling. A drainage ditch mutilates the
southwest corner; otherwise the earthwork is well preserved. No
trace of the causeway was found. The 2008 site visit found that
the earthwork lies in very wet and marshy ground. It was
extremely overgrown with small trees, scrub, reeds and
brambles, so much so that access was not gained onto the site.
Just outside of the earthwork on its west side the owner of
Waun-Withan Farm had recently excavated a wide and deep
drainage ditch that ran very close to the bank. The ditch follows
for a time the line of the public footpath from Blaenwthan Farm to
Waun-Withan Farm, where it passes the earthwork. The owner
was aware of the site and said that he had looked out for
anything unusual during the excavation of the ditch, in the way of
finds for instance, but had seen nothing. The land on which the
site lies is now owned by Llwyn-pur Farm, and is currently fenced
off to stop cattle from entering the wet area. It appears that the
area is so marginal and wet, that it has been left alone over the
years and it is probable that the earthwork survives much as it
was recorded in 1977. The shape and size suggest the possibility
that this is a Roman fortlet or signal station, but its situation does
not. (F Murphy 2008)
PRN 2384 NGR SM94972648
Name UPPER NEWTON
Type Villa,Homestead
Form buried feature
description Condition Substantial destruction
Summary The remains of a 'viila' or Romano-British farmstead. Located on a
gentle north-facing slope approximately 800m west of Wolfscastle.
Long description In 2003 Dr Mark Merrony undertook fieldwork in the area to
ascertain the location of the 'villa' site. The results of a
gradiometer survey showed evidence of a rectangular building
approximately 18.5 by 7.5m, running in a SE-NW direction. It was
very close to the location as proposed by Fenton. The survey
also showed a possible trackway heading west towards the
building, as well as other rectilinear lines to the east. Partial
excavation revealed the floor surface of the building, which was
paved with large stone slabs. 2nd and 4th century pottery, as well
as hexagonal roofing slates were also discovered. F Murphy 2008
140m to the north air photographs show a large rectangular
crop-mark (PRN 2017). The Ordnance Survey report that in April
1960 a scatter of roofing slates, two pieces of hypocaust, and a
fragment of a tegula were found at SM948265 approximately
130m east of Ford Camp Iron Age defended enclosure (PRN
2385). At SM94972648 Fenton records the finding of a large
amount of Roman brick and a structure with two flues, which in
his opinion was a Roman 'bath', by a labourer who was digging up
a hedge in a field near Ford in Haycastle in 1806.
PRN 2632 NGR SM73452703
Name HEN EGLWYS YR;MENAPIA
Type stone foundation
Form Documents
Condition
Summary Historical documents report that vague fragments of stone
foundations located within sand dunes above Whitesands Bay are the
remains of a possible Roman settlement.
Long description Set within sand dunes, not identified during field walking. Murphy
1996. There are vague fragments of possible walls in the area,
but that they are without apparent significance. Ordnance Survey
1966 Foundations, said to have been found when building the sea
wall, are thought by local inhabitants to have been a church
founded by St David. Coins are reported to have been found but
are now lost. There is not the slightest evidence that a Roman city
ever stood here. Ordnance Survey 1964 Fenton appears to have
originated the beleif that Menapia was in the vicinity of
Porthmawr. Fenton 1903
PRN 3086 NGR SM8900
Name KILPAISON BURROWS
Type Occupation site,Findspot
Form Finds
Condition
Summary A 'third brass'of the late third century, together with Samian ware and
Roman glass, were found on Kilpaison Burrows, Rhoscrowther.
(Ordnance Survey 1964).
Long description
PRN 3284 NGR SM98170160
Name ST MARY
Type Finds
Form Finds
Condition
Summary A number of Roman coins found during excavations of Pembroke
Castle in 1881.
Long description The RCAHM (1925) report that while excavating at Pembroke
Castle in 1881, Mr J R Cobb found a coin of Carausius which had
been used to fix a blade of some kind to a handle. With it were
one coin each of Constantine, Constantine II, and Constans, also
two uncertain issues. (F Muphy 2009).
PRN 3382 NGR SN4020
Name SARN HELEN (S);SARN ELEN
Type Road
Form buried feature
Condition Damaged
Summary This PRN was allotted to the line of the Roman road RR69d between
Carmarthen and Llanio Fort. The southern end of this route between
SN43282353 and Carmarthen is doubtful in the light of cropmark PRN
14308 which appears to indicate line of the Roman road heading on a
south westerly course directly towards Carmarthen. North of here the
course is assumed top follow the modern road. There is little evidence
to dispute this until SN47333696, when cropmark evidence indicates
that the Roman road follows a slightly different course in places to the
proposed route. The Roman road appears to cross the Afon Teifi
floodplain on either side of Lampeter at SN55634629 and
SN61985051, perhaps suggesting the possibility that there was a fort
at Lampeter. The continuation of this proposed route northwards from
Llanio to Pennal Fort is PRN 5222 DS.Sept.2004.
Long description The line of the road crosses the SE corner of the field just W of
the junction of the A485 and the B4578. The modern road layout
has a dogleg, but the original Roman line continued straight
across the field. There is no surface indication of the road,
although the present owner has noted stones in the area when
ploughing. NAP 2000
PRN 3437 NGR SN80052596
Name RHOS LECHACH
Type Dam
Form Earthwork
Condition Damaged
Summary The site of a possible dam which has been postulated to be of Roman
date. RPS Nov.2002
Long description South of the large Roman temporary camp at Arosfa Gareg-Llwyd
is the site of a possible dam across the Afon Llechach. However,
in 1968 the Ordnance Survey described it as "an insignificant
earthwork some ten metres long, irregular in profile and nowhere
more than two feet high. It would have created a very small pool
of water in an area where there is an abundant supply. Any
suggestion of its being associated with the Roman Camp must be
a very tentative suggestion."
PRN 3449 NGR SS13809685
Name DAYLIGHT ROCK
Type Occupation site
Form Finds
Condition
Summary A number of sherds of pottery identified as Romano-British were
found on Caldey Island.
Long description The finds were discovered in 1916 and 1919 from the rainwash
(under the blown sand) forming a low cliff a few yards below the
path between the Guest House and Rubbishy Corner, where a
small valley runs down to Priory Bay. One piece of
Romano-British pottery is probably 1st century or early 2nd
century, and the other fragments are 3rd or 4th century.
Ordnance Survey 1964
PRN 3664 NGR SN17500785
Name TRELISSEY
Type Farmstead
Form Earthwork
Condition Substantial destruction
Summary A Romano-British stone-built rectangular structure situated within an
Iron Age defended enclosure, approximately a quarter of a mile south
of Trelissey Farm. The building was partially excavated in 1950-1.
The site appears to be an Iron Age defended enclosure whose use
continues well into the the 2nd century as a Romano-British
homestead. F Murphy 2008
Long description Although shown as "Roman Building" on the 1975 Ordnance
Survey map, this site has been interpreted by the Royal
Commission as a native Iron Age settlement under Roman
influence. MM March 2003. Circular enclosure on a southwest
facing slope formed by a double bank with central ditch which
becomes a single bank on the northwest side, where there are
the remains of what appears to be an annexe. The site strewn
with stones, has been badly ploughed down on the northeast. The
entrance on the west is weak and the site of the Roman building
is marked by a weak, fragmentary bank. Finds indicated that the
building was occupied from about the early 2nd century to at least
the 3rd century. No part of the ringwork, which consisited of a
higher inner and lower outer bank, with a ditch between them,
was excavated and while ringwork and building may be
contempoary, it is possible that the latter was erected within an
existing earthwork. Finds included pottery, samian pottery, tile,
spindle whorl, bronze stud, scrap of sheet lead and slag.
Ordnance Survey 1964-8
PRN 3815 NGR SN18601336
Name YET WEN
Type Fortlet
Form Earthwork
Condition Substantial destruction
Summary
Long description In 1965 the Ordnance Survey reported that aerial photographs
taken by the RAF in 1946 clearly show a small, square single
banked enclosure, with an entrance on the western side. The site
is situated on level high ground at 160m above sea level. By 1975
a site visit by the Ordnance Survey records that it is an almost
square enclosure whose sides measure approximately 30m, with
rounded corners. It stands on high ground that affords extensive
views especially to the south and west. It has a bank on the
higher east side but only an outward facing slope up to 0.6m high
around the remainder, although lush vegetation on the north and
south suggests that there may well have been a ditch. They
suggest that a lowering of the scarp in the centre of the west side
almost certainly indicates an entrance, and that the shape, size
and situation suggest the possibility that this is a Roman fortlet or
signal station. When visited by Dyfed Archaeological Trust in 1985
the land had obviously been ploughed since 1975 and the bank
on the east was less definite. It was noted that the site was
overlooked by slightly higher ground top the north, so perhaps
not in such a commanding position as outlined by the Ordnance
Survey. The reporter suggested that it was a Roman fortlet. (F
Murphy 2008)
PRN 3847 NGR SN296076
Name LAUGHARNE BURROWS
Type Settlement,Findspot
Form Finds
Condition
Summary A collection of Roman finds from a number of shell mounds that occur
in an area about three-quarters of a mile long on the sands of
Laugharne Burrows. "Shell mound" is a loose term applied to surface
deposits covering a wide area centred at SN296076.
Long description The mounds, eleven of which are described in detail by Cantril,
have yielded misc. finds of several periods. The Roman finds
(which this PRN covers) included a Roman coin, brooch , pottery,
and 2 querrns. The whole area is now part of a strict security
area.
PRN 3855 NGR SN28420914
Name COYGAN CAMP
Type Occupation site
Form Earthwork
Condition Substantial destruction
Summary A number of finds and settlement evidence belonging to the
Romano-British period found within the Iron Age hillfort of Coygan
Camp. The finds included pottery, coins and bone. The settlement
evidence included a hut circle, rectangular enclosure, midden and
possible inhumations. The site has been destroyed by quarrying and
is now descheduled. F Murphy 2008
Long description
PRN 3884 NGR SN26541543
Name DOL GARN;DOL GAM
Type Moat,Defended enclosure
Form Earthwork
Condition Damaged
Summary In my opinion this is a moated site;the road has not destroyed all of
the second half of the monument as seen in my APs of
12-2-88.(T.A.J.22-04-88)
Long description
PRN 3900 NGR SN25371213
Name CWMBRWYN
Type Farmstead
Form Earthwork
Condition Substantial destruction
Summary The later Roman occupation of the Iron Age univallate enclosure (PRN
7437). It was, according to the Ordnance Survey, “a most interesting
combination of Roman farm with native walled homestead”.
Long description A sub-rectangular, univallate enclosure with an external ditch and
a single entrance in the centre of the east side. Situated on a
hillslope at 120m above sea level. The bank is c.1.0m high
externally and 0.70m high internally. There is only a slight trace
of an outer ditch. The entrance is approximately 6.0m wide.
According to the Ordnance Survey partial excavation in 1906
showed that the original ditch was probably V-shaped, 17ft (5.2m)
wide and 8ft (2.4m) deep. The gateway had stonewalls but the
rampart was built largely of earth. Foundations of a stone building
were uncovered at the western end of the enclosure, and there
were remains of a small bath building as well as indications of
timber buildings. No traces of these buildings can now be seen on
the ground. It is probable that the enclosure was occupied
throughout both the Iron Age and Roman periods and that it was,
according to the Ordnance Survey, “a most interesting
combination of Roman farm with native walled homestead”. The
whole site is under pasture, and is regularly grazed by livestock.
F Murphy July 2007 – compiled from several sources
PRN 4047 NGR SN70472545
Name LLYS BRYCHAN
Type Villa,Homestead
Form Building
Condition Substantial destruction
Summary The site of a Roman Villa, first reported by Fenton in c1800. Appears
to have been excavated in the mid 19th century, and again in 1961-2.
Excavations revealed masonry walls, wall plaster, hypocaust tiles,
roof tiles, pottery and coins suggesting a villa of considerable size,
built around a courtyard, with an associated bathhouse. Possibly
occupied from c200AD to the late 4th century AD. (PP 2004)
Long description In 1961 The site was re-discovered prior to the erection of a silo.,
these operations having destroyed everything except the lowest
courses and the floors of the area excavated, but walls standing
to a height of about 3ft were visible in the area untouched. The
foundations exposed formed 2 rooms (possibly 3) part of a
NE-SW range, the walls of dressed stone banded with lime mortar
and the floors of opus signinum. On one floor were two pillars of
soft red tiles - likely to be supports for a raised floor. A room of
later date abutted on the south side, possibly indicating that the
buildings had surrounded a courtyard. Finds included hypocaust
tiles, painted wall paster, roofing slates and two fragments of
mortarium. The 1962 excavations showed that the east range
extended northwards beyond the heated suite found in 1961, and
then turned west, giving two sides at least of a courtyard
structure. In 1968 the Ordnance Survey recorded the site of the
villa at SN70492546, where, although there was no trace of the
villa, an area of disturbed ground clearly marked the excavation.
PRN 4081 NGR SN71613455
Name LLANWRDA
Type Moat,Fortlet
Form Earthwork
Condition Substantial destruction
Summary A small rectangular earthwork enclosure 27m by 33m, interpreted as
a possible Roman fortlet. If so, it is located over 2.5km from the
nearest known Roman Road RR623 between Llandeilo and
Llandovery. Three of the banks and ditches were buldozed in the
1960s.DS.27.10.2004
Long description A small rectangular enclosure, measuring approximately 27m by
33m, visible on aerial photographs. It is situated on high ground,
at 260m above sea level, with good views to the east and south,
but it is overlooked by slightly higher ground on the west. In 1968
the Ordnance Survey record that on three sides a ploughed down
bank and a filled in ditch are visible, but that on the remaining
southwest side the bank survives as an earthwork. It rises 1.1m
above the inner level of the enclosure, and then drops 1.4m
forming an outer ditch. On this southwest side the outer edge of
the ditch once had a modern hedge running along it, but this has
been removed. They also report that the then landowner stated
that six years previously he had bulldozed three of the banks into
their outer ditches, but that prior to that the banks had been 4ft
(c.1.2m) high on all sides and that the ground inside of the
enclosure was dry. By 1977 the Ordnance Survey report that the
southwest bank and ditch have been ploughed out and are now
only discernable as a change in soil colouration. The landowner
informed them that there is a local legend that the enclosure once
held buildings, but that when he had recently drained the interior
no features or remains of buildings were seen. The Ordnance
Survey suggest that the shape, size and position of the enclosure
point to it being a Roman fortlet. (F Murphy 2008)
PRN 4087 NGR SN77003473
Name CAEFELIN CIE
Type Cemetery
Form Finds
Condition
Summary An extensive Roman cemetery was discovered in a field called
'Caefelin cie' (SN770347) to the southeast of Llandovery Roman Fort,
during development of the Maesglas building estate.
Long description In 1968 the Ordnance Survey reported that the Maesglas housing
estate now covered the site of the Roman cemetery and that
there was nothing significant to be seen. Some of the urns were
given to the County Museum, Carmarthen in 1961.
PRN 4253 NGR SS14589697
Name NANNA'S CAVE
Type Cave occupation
Form Finds
Condition
Summary A collection of finds of Roman-British date from Nanna's Cave. The
rectangular cave is located at SS 14589697, and is 3m wide and 3.5m
high and 5m deep. It is situated near the top of coastal cliffs and the
entrance faces east. There is a level turf covered platform in front of
the cave.
Long description The 1977 excavation consisted of opening a 1m wide trench
inside the cave. It revealed back-fill, to a depth of 64cm,
containing an undisturbed midden of animal bones,
Romano-British potsherds and a human bone. (Ordnance Survey
1979) In the winter 1971-72 a denarius of Severus Alexander (AD
222-235) was washed out of spoil; possibly from Leach's
excavation of 1913. (Ordnance Survey 1978) The Romano-British
finds came from a kitchen midden at the mouth of the cave and
indicated occupation in the 3rd and 4th centuries. Pottery included
a hammer-head mortarium, flanged-and-beaded rimmed bowls
and imitation Samian ware of late Roman date. The midden also
yielded a blue glass bead, a flint scaper, and marine and land
shells, and animal bones. (Ordnance Survey 1974) Excavated in
1911, 1915 and 1950-51. Later finds included Romano-British
pottery of the 3rd-4th centuries AD.
PRN 4324 NGR SM82572293
Name DINAS FACH
Type Occupation site
Form Finds
Condition
Summary An Iron-Age promontory fort (PRN 2790) consisting of two defensive
banks with a ditch between them. Within the interior a number of
hut-circles were constructed. Use of the site during the
Romano-British period is attested by the presence of fragments of
Samian ware and other occupation debris.
Long description
PRN 5999 NGR SN64475627
Name CAE'R CASTELL
Type Bath house
Form Building
Condition Substantial destruction
Summary The remains of a Roman bath house situated southeast of 'Bremia'
Roman auxiliary fort (PRN 5998). The bath house was discovered in
1887, re-opened in 1910 and excavated in 1969-71. The area is now
a Scheduled Ancient Monument SAM-CD129 (CER).
Long description In 1976 the Ordnance Survey recorded that the "excavations at
the bath house have been left open and the remains of a 6.0m
square building with walls up to 1.0m high can be seen. the floor
lies 1.5m below the current ground level. A large quantity of
broken tiles forming part of the Hypocaust are scattered about
the site. The building is surrounded on three sides by the remains
of a small embanked enclosure measuring about 40.0m N-S by
20.0m tranversely." "The much robbed bath building found in
1887 and re-opened 1910 was re-examined, the caldarium,
praefurnium and frigidarium being identified." Ordnance Survey
1974.
PRN 6271 NGR SN809398
Name TRE GOCH
Type Station,Findspot
Form Finds
Condition
Summary Reputed site of a Roman station, known traditionally as Tre Goch. A
number of finds of Roman pottery and a stone seal here suggest that
there was indeed a Roman presence. RPS Nov 2002
Long description In 1976 the Ordnance Survey report that the no ground or further
documentary evidence has been gained. The graveyard to the
NE of the chuch has lain undisturbed since the church was rebuilt
in the mid-19th century and no further finds are known of locally.
Llanfair-ar-y-bryn church is supposedly built on the site of a
Roman station, traditionally known as Tre-Goch. (Lewis Topog
Dict Wales 1842)
PRN 7367 NGR SN62192221
Name ST TYFEI'S CHURCHYARD
Type Temple,Homestead
Form Documents
Condition
Summary The parish church of Llandyfeisant is allegedly built on the stone
foundations of a Roman building - either a villa or a temple. No
indication of this earlier building was identified during the rebuilding of
the church in 1875. However, reported finds of tesserae during work
in the churchyard have been made in the latter part of the 20th
century. (EGH 2003).
Long description The church is said to be built on the site of a Roman temple "upon
the foundation of which its northern angle is said to rest. In
levelling the churchyard (c.1800) ..... the walls of the Roman
edifice were discovered". (Topo Dict Wales 1833)
PRN 7374 NGR SN35141014
Name LLANSTEPHAN CASTLE
Type Occupation site
Form Finds
Condition
Summary A Romano-British pottery sherd found during excavations at
Llanstephan Castle in the 1960-70's
Long description
PRN 7414 NGR SN00731740
Name MERRYBOROUGH CAMP
Type Occupation site
Form Finds
Condition
Summary A find of a Samian sherd from a Dr 45 bowl from Merryborough Camp
Iron Age defended enclosure (PRN 3554) suggests Roman period
occupation of the site.
Long description A rescue excavation sectioned the single inner bank in three
places and indicated a decayed turf or timber revetment with a
ditch 7ft max. depth. Light structures were indicated in the interior
but the only stratified find was a Samian sherd from a DR 45
bowl. (Ordnance Survey 1964)
PRN 7567 NGR SM896388
Name GARN FAWR
Type Occupation site
Form Finds
Condition
Summary A single sherd of Roman pottery has been interpreted as evidence
that Garn Fawr Iron-Age hillfort was occupied in the Romano-British
period. MM March 2003.
Long description
PRN 7626 NGR SN19280834
Name PARC YR EGLWYS
Type Occupation site,Findspot
Form Finds
Condition
Summary A piece of black Romano-British pottery of 3rd-4th century found
during excavations in 1906 of the Iron Age enclosure 'Parc yr Eglwys'
(PRN 7625)
Long description
PRN 7635 NGR SN24100934
Name CASTLE LLOYD
Type Occupation site,Findspot
Form Find
Condition
Summary One sherd of Roman samian pottery that reputedly came from Castle
Lloyd Iron Age inland promontory fort.
Long description
PRN 7694 NGR SM91521730
Name WALESLAND RATH;EAST HOOK
Type Occupation site
Form Documents
Condition
Summary Evidence of Roman occupation was discovered within the Iron Age
defended enclosure of Walesland Rath.
Long description The earthwork listed as East Hook Rath at Lambston is a small
scarped enclosure on a spur between two streams. Preliminary
excavation has shown this site to have been occupied in the Iron
Age and Roman periods, with a considerable period of desertion
in between . Large post holes were revealed under the bank to
the south and west and traces of iron and bronze working were in
evidence in the Iron Age levels (Ordnance Survey 1969)
PRN 7761 NGR SN424021
Name COURT WOOD
Type Occupation site
Form Buried Feature
Condition
Summary A buried feature of possible Roman/Post-Roman date discovered
during trial excavation in 1977 of Court Wood Iron Age defended
enclosure (PRN 7648). It may have been a post setting. No dating
evidence was recovered from this feature, only its relationship with
other features indicated it may be of Roman date. (F Murphy 2009)
Long description
PRN 7778 NGR SR94719377
Name OGOF MORFRAN
Type Cave occupation
Form Finds
Condition
Summary A group of Romano-British finds from the excavations of Ogof Morfran
Cave.
Long description This south facing cave is situated in the coastal carboniferous
limestone cliffs, about halfway between Stack Rocks and St.
Govan's Head. It is roughly 50ft above the present sea-level and
forms a rcok shelter 30ft long and 18ft deep. Finds of Roman
pottery, a bronze brooch and three Roman coins indicate Roman
occupation. See also PRNs 7395, 7776-7780.
PRN 8337 NGR SN674828
Name GREAT DARREN;DARREN
Type Zinc mine,Silver mine,Lead
Form Complex
Condition Various
Summary The Darren mine was one of the most important in Ceredigion. There
was probably prehistoric activity here and the mine flourished from
the 17th until the early 20th century. The industrial character of the
area is enhanced by the old miners cottages in Daren and
Penybontrhydybeddau villages. RPS 2003
Long description "Lewis Morris stated that the 'Romans or Britains' wrought the
lode 'in some places 40 or 50 yards deep without shafts' (ie
opencast)" (Bick 1988) Copper ore output - 1000 tons lead ore
output - 20000 tons zinc ore output - 200 tons. ^The Daren mine
was one of the most important in Ceredigion. There was probable
prehistoric activity here and the mine flourished from the 17th
until the early 20th century. The industrial character of the area is
enhanced by the old miners cottages in Daren and
Penybontrhydybeddau villages.^Evidence of mining activity is
now best preserved in an area stretching from west of Darren
Farm northeast over the ridge and down to the stream opposite
Cwmdarren. This includes a large opencut working that cuts
across the ridge-top on a northeast-southwest axis, and notable
spoil tips that cascade down the slopes above Coed y Darren at
the northeastern end of the opencut, forming a significant
landscape feature. A series of drainage levels cut into the slope at
this end are now blocked by collapsed spoil material and a
number of fold shafts have been capped by concrete plugs and
steel grills in modern times. There has been significant
disturbance of the lower spoilt tips above Cwmdarren, where a
County Council or private storage depot has been built. ^17th
century activity was focused at the southwestern end of the
opencut, west of Darren Farm. Most of the features here are
slight earthworks, with some collapsed airshafts and vegetated
spoil tips. Bushell's Level, which was probably associated with
these features, is no longer visible. There are also early 19th
century shafts and features to the north and west of Darren Farm.
These too appear to be very degraded, but this whole area is
assessed as having significant potential in terms of the
preservation of sub-surface archaeological features. This area
already has SAM status.^Some 800m west of the workings at the
northeastern end of the opencut was the late 19th century
ore-processing mill, which forms a detached block of land with
archaeological interest. Surface remains here are ephemeral, but
it is still possible to trace some features in the field that are
shown by the late 19th century OS map survey. This area was
connected to the mine by a tramway, which can still be traced in
part running through Coed y Darren, although for most of its
course it appears to have been widened and levelled by modern
forestry operations. Adjacent to the mine is PRN 2023 Pen y
Darren Iron Age Hillfort, also a SAM Cd 28. RPS 2003
PRN 8508 NGR SN41582034
Name PRIORY STREET NO.136
Type Building
Form buried feature
Condition
Summary The foundation of a north-south wall and mosaic surface were found
in 1871 when a Captain Davies was sinking a shaft to find water
below the garden of 136, Priory Street. JH Aug 2002
description
Long description
PRN 8864 NGR SR97159480
Name BOSHERSTON
Type Occupation site
Form Finds
Condition
Summary Evidence that Bosherston Pormontory Fort was occupied during the
Romano-British period is provided by the discovery of three pieces of
Roman pottery, including one Samian sherd from the site (Gardner
1964)
Long description
PRN 8882 NGR SN67504055
Name ANNELL AQUEDUCT (SOUTH)
Type Aqueduct
Form Earthwork
Condition Unknown
Summary This site (PRN 8882) falls into two separate areas. The east end of the
site is a Scheduled Ancient Monument and is the known and visible
line of the aqueduct. The leat then continues along a projected line to
the south and west and in places is visible as an earthwork, but it has
not been scheduled. PG 2003.
Long description PRN 8882 (east) - SAM CM210 This Scheduled Ancient Monument
is a section of the Annell Leat. The leat would have carried water
into the Dolaucothi mines (PRN 1946) in order to wash the ore
which was being extracted from the mines. It is not known
whether the leat took water from the Annell itself, or whether it
picked up small tributaries and surface water. Only a short
section of the Annell leat is scheduled. This part survives as a
clearly visible, broad track-like cut in the hillside. For the most
part, the leat stays very level and follows the natural contour of
the hillside. However, at its northernmost end, the leat seems to
curve quite sharply upslope and then disappears into a boggy
area at the head of a small valley. In general this feature is very
visible and well preserved and in places it appears that another
'track' is carved just upslope from the leat. In one place, it
appears that there are three parallel sections of the leat. Sections
of these upper paths may simply be sheep tracks, but Lewis
(1976) argues for a triple channel system within the Annell leat,
with wooden shuttering supporting the banks between the
channels. The leat passes just in front of a rock-cut spring (PRN
49146) which probably fed into it. The relationship between these
two features is not entirely clear, and the line of the water
channel is harder to pick up at this point. The southwest end of
the Scheduled area is quite overgrown with bracken and scrub,
and the line of the leat is harder to see. There are also a few
sheep scrapes along the length of this part of the leat. PRN 8882
(west) - The projected line of the Roman Annell leat that is a
continuation of the Scheduled section to the north. The southern
part of this feature is not visible on the ground, partly due to
scrub and bracken growth. Recent bracken clearance has
revealed clearly the earthwork of the leat further north, two fields
south of the scheduled area. The earthwork here is in a good,
stable condition, and it is likely that further physical remains may
survive. PG 2003. A section of the Roman aqueduct system that
carried water to the nearby gold mines at Dolaucothi. RPS
24.09.2002
PRN 8883 NGR SN69904275
Name ANNELL AQUEDUCT
Type Leat,Aqueduct
Form Earthwork
Condition Unknown
Summary Originally thought to be an aqueduct taking water from the River
Annell to the Dolaucothi gold mines, this scheduled earthwork
(SAM-CM211) is possibly rather part of a system of mineral extraction
and hushing.
Long description SAM Carm 211. This feature was originally considered to be the
line of a Roman aqueduct taking water from the River Annell to
the Dolaucothi gold mines. Recent survey and research has
recognised that a channel running a considerable distance across
unenclosed land supplied water to the scheduled section of the
feature and it could not have been fed from the Annell. The
multiple channels have been explained as a result of the severe
change of gradient needed to bring water around this route and
the problem of managing this, either the channel had to be rebuilt
a number of times or the flow was split between channels.
However, recent investigation suggests that one or more of the
channels may have been cart tracks. It has been postulated that
the upper channel might be an ancient excavation along the back
of a mineral load, the leat could then have supplied water for
hushing. It has also been suggested that the leat, rather than
taking its load all the way to the mine, may have followed the
contours of Cefn-y-Bryn north and east almost to Rhiw Garegog,
before dumping its load in a tributary of the Cothi just below this
point. A maximum of four channels are visible within the
Scheduled section of the feature. The results of the 1999 field
visit indicated that all four continued into the area visited. At the
north-western end of the multiple channels they have been cut
across by a much later boundary bank.LA 1999
PRN 8884 NGR SN68764285
Name COTHI AQUEDUCT;UPPER COTHI
Type Leat,Aqueduct
Form Earthwork
Condition Damaged
Summary The Upper Cothi Leat runs approximately 30m above the Cothi leat.
There are four sections of the leat which have been scheduled (CM
213; a,b,c & d). However, the leat must continue along the contour
between these sections, but these areas have lost the definition of the
scheduled areas. Clearly visible on the north-west facing slope of
Banc Llwyn-y-ceiliog, it is partly incorporated into field boundaries. In
places it survives c.2m across and 0.5m deep. AP 2005.
Long description During a farm visit to Llwynceiliog, the length of the leat within
the farm holding was observed. At the southwestern limit of the
leat within the farm holding, the leat is not a visible feature and
the earthwork appears to have been completely eroded. It
becomes visible in the field boundary which crosses it and then
becomes a break in the profile of the hillslope when it enters the
unimproved cwm, this is where most of the scheduled section lies.
The section to the east of the stream and east of the farmstead
does not survive as an earthwork and again appears to have
largely been eroded. AP 2005 SAM Carm 213. The Upper Cothi
Leat runs approximately 30m above the Cothi leat. Clearly visible
on the north-west facing slope of Banc Llwyn-y-ceiliog, it is partly
incorporated into field boundaries. In places it survives c.2m
across and 0.5m deep. To the north-east it is possible that the
leat partly follows the northern edge of Coed Maes-yr-Haidd. LA
1999
PRN 8921 NGR SN425163
Name DAN Y GRAIG
Type Non antiquity,Building
Form Cropmark
Condition
Summary Cropmark site of possible Roman building suggested by Terry James
in 1979. No traces of any earthworks were visible on the ground
during site visits in 1983 and 2008. Probably a non antiquity. F Murphy
2008
Long description Dan y graig is a cropmark site that lies on a northwest-facing
gentle hillslope at 80m above sea level, 300m east of the
destroyed, possibly Roman, rectangular enclosure of Gelligaeros
(PRN 1655). It has been suggested that this cropmark shows the
site of buildings of possible Roman date. These were identified
from aerial photographs taken in 1979, by the photographer
Terry James. He recorded that there appeared to be traces of
earthworks visible on the photographs. In 2008 no traces of any
earthworks could be found on the ground, as was the case in
1983 when Dyfed Archaeological Trust made an earlier visit. One
possible linear earthwork on the photograph is along the line of
an old hedge boundary, and another is where there is a natural
shelf in the topography. With only the aerial photograph and its
near proximity to a possible Roman site as evidence, this site
cannot be classified as a definite earthwork building of Roman
date. The site lies within a field of improved pasture. F Murphy
2008
PRN 8923 NGR SN699252
Name PANT-MEREDITH
Type Enclosure
Form Cropmark
Condition
Summary A small square shaped enclosure identified from a 1979 aerial
photograph. Unfortunately the grid reference is not precise enough to
pinpoint its position on the ground, and there are no features on the
photograph to relate the enclosure too in order to estimate its size or
position. It appears to be lying within scrubby heathland - Carn Goch
area? Possibly small Medieval/Post Med enclosure? F Murphy 2008
Long description Cropmark of unknown significance. RPS October 2001
PRN 9180 NGR SN7040
Name GWENLAS
Type Leat
Form Earthwork
Condition Damaged
Summary The Gwenlas leat is part of a leat system that carried water over
significant distances to the Roman gold mines at Dolaucothi. RPS Nov.
2002
Long description
PRN 9624 NGR SN6351
Name HEOLDDWR
Type Leat
Form Earthwork
Condition Substantial destruction
Summary A long, wide watercourse or leat perhaps of Roman origins,
contouring the hills above the ancient silver-lead mine (PRN 8290) of
Graig Ddu Isaf. The leat begins at SN 655520.
Long description
PRN 9811 NGR SN32751397
Name CASTELL COGAN; OLD CASTLE
Type Findspot
Form Buried Feature
Condition
Summary During the excavations in 1971 of Castle Cogan Iron Age defended
enclosure (PRN 2154) a possible Roman spearhead was discovered on
the burnt floor of Hut II.
Long description
PRN 9869 NGR SN4221
Name CARMARTHEN
Type Aqueduct
Form Earthwork
Condition Substantial destruction
Summary There is no clear evidence of the course of an aqueduct feeding the
Roman settlement. Professor Jones (1971) has suggested the
reservoirs at Cwmoernant may have Roman predecessors but James
(1980) suggests the route (PRN 41) used to feed the Medieval Priory
mills is more likely. (PP 23/9/04 from T.James 1980).
Long description The Roman aqueduct could well have followed the course of the
Medieval priory (SN42 SW 2) mill stream as it terminated not far
short of the eastern boundary of Moridunum (SN 425223 to SN
422209). A low bank sometimes visible in a field above Tanerdy,
SN 42152090, is a less likely candidate as it appears to be sloping
away from the area of Roman occupation. Ordnance Survey
1981
PRN 1056 NGR SN768351
Name LLANDOVERY
Type Findspot,Bath house
Form Documents
Condition
Summary This PRN applies to a group of Roman finds discovered during the
18th and 19th centuries by successive vicars of Llandovery. One, in
the 18th century, was said to have detected baths. The location given
to this site is the grid reference for Llandovery vicarage. However, the
RCAHM in 1917 thought the baths were more likely to be on the east
side of the fort near the River Bran. F Murphy 2008
Long description In 1917 the RCAHM recorded, as part of their inventory of
Carmarthenshire, information on the Roman fort just north of
Llandovery. Included in this report was a reference to an 18th
century vicar of Llandovery who it was said "collected some
relics, now lost - a broken altar, lamps, potsherds, coins of
Constantine - and to have detected baths, while a successor in
the early 19th century had a copper coin of prehaps republican
date, some Samian and other small objects. A few remains of
building have been noted outside the fort on the east side, where
the ground slopes to the River Bran: here we may perhaps seek
the bath-house". RCAHM 1917
PRN 1109 NGR SN41822048
Name PRIORY STREET ALLOTMENTS
Type Hoard
Form Finds
Condition
Summary Seven Roman coins, dated to between 253 and 278AD, were found at
the northern end of the allotments outside the SE boundary of the
Roman town of Carmarthen. JH Aug 2002
Long description NGR appears to be outside the area of the allotments but have
made no attempt to revise the NGR due to lack of information. JH
Aug 2002
PRN 1194 NGR SN66454016
Name MELIN-Y-MILWYR
Type Reservoir
Form O.Struct
Condition Unknown
Summary The unusual name of this site (Soldiers Mill) has been said by some to
be a direct reference to the presence of Roman military personnel at
the mine complex. It is possible that the ponds here are of early
origin, possibly Roman, and associated with the water system of the
early gold mine complex. Roman pottery, dating from the 2nd - 3rd
centuries AD, has been found in the vicinity. RPS 24.9.02
Long description
PRN 1194 NGR SN66454016
Name MELIN MILWYR
Type Mill
Form Documents
Condition
Summary It is thought that an early mill was located at, or near, this site,
possibly Roman in date. RPS 24.09.02
Long description
PRN 1284 NGR SN66375531
Name MIBUS TAST STONE;LLANDDEWI
Type Inscribed stone
Form stone slab
Condition Moved
Summary Inscribed stone, built into the outside wall of Llanddewi Brefi parish
church tower PRN 5135 and 5138, filling the arch of the south
transept. Roman? Not early medieval and not included in Dr N.
Edwards recent study (Edwards forthcoming). NDL 2003
Long description
PRN 1318 NGR SN61931458
Name MAES DEWI
Type Inhumation
Form Finds
Condition
Summary In 1912 an inhumation and 3 coins were discovered during the
building of a farmhouse near Llandybie. It appeared that the skeleton
had a wound mark across the skull.
Long description The owner of Maesdewi (formerly Piode Fach), Llandybie
remembers as a child in 1912 that when the foundations of the
new farmhouse were being excavated, that 3 Roman coins - one
large bronze and two small silver - and a human skeleton were
discovered. It appeared that the skeleton had a wound mark
across the skull. The finds were supposed to have gone to
Carmarthen Museum, but there is no trace of them. K Murphy
1985
PRN 1383 NGR SN117391
Name CASTELL HENLLYS
Type Settlement
Form None
Condition
Summary Romano-British settlement immediately north of the Iron Age hillfort
PRN 952. Mytum H AW 1992
Long description
PRN 1419 NGR SM59800930
Name GRASSHOLM
Type Well
Form Documents
Condition
Summary Well site partially excavated in 1946. During the excavtion several
sherds of Roman pottery including Samian Ware are reported to have
been found. ZSBR 2006.
Long description
PRN 1428 NGR SM86272395
Name BRAWDY CASTLE
Type Occupation site
Form finds
Condition
Summary Excavations directed by Ken Dark in 1989 found numerous fragments
of Roman glass and pottery at this promontory fort site, suggesting
occupation of the site during the Romano-British period. The site is
defended on the eastern and southern sides by steep natural slopes,
and by a multivallate system of banks and ditches on the northern
and western sides. Finds from other periods sugest that this is a
multi-phase site. MM May 2003.
Long description
PRN 1950 NGR SN65129200
Name LLANGYNFELYN;LLANCYNFELYN;TAL
Type Lead mine,Copper mine
Form Earthwork
Condition Unknown
Summary A large complex of mine shafts and worked lodes. Visible above
ground is a fine example of a Cornish stack.
Long description Evidence of bronze age mining has been identified at three
locations, PRNs 51835, 51832 & 51839. In the 1740's the mine
was worked by a Flintshire company. Larger scale workings took
place again a century later and a large 'fire engine' was erected
in 1846. Flat rods were used to pump the engine shaft which was
28 fathoms deep. However, although the mine seemed to be a
promising venture, it closed in a short space of time. It was
resurrected some years later as Taliesin mine and was drained
and worked from 1854. Although this venture was more
successful in its output the costs of re-equipping the mine meant
that returns were not sufficent to cover these costs. After a
further proposal to restart mine using water power the venture
was finally abandoned. AMP 2005 Lead ore output-100 tons
PRN 2584 NGR SN682834
Name CWMDAREN;LEFEL
Type Lead mine,Copper mine
Form Earthwork
Condition Unknown
Summary A large mine, producing copper and lead, worked from at least the
Roman period, if not earlier, until as recently as 1920. It is situated
east of the village of Penbontrhydybeddau, Ceredigion. The name
'Cwm Darren' originally applied to the large workings adjacent to
Cwm Darren house but it later embraced 'Twll y Mwyn' mine, a short
distance away. F Murphy 2008
Long description A large mine, producing copper and lead, worked from at least
the Roman period, if not earlier, until as recently as 1920. It is
situated east of the village of Penbontrhydybeddau, Ceredigion.
The name 'Cwm Darren' originally applied to the large workings
adjacent to Cwm Darren house but it later embraced 'Twll y
Mwyn' mine, a short distance away. It is believed that the
workings were wrought in prehistoric and Roman times. "In 1852
....it was reported that in the deepest parts 'many Roman
hammers have been found, very perfect," (Bick 1988) Copper
ore output-110 tons; Lead ore output-200 tons;
PRN 3017 NGR SM822052
Name DALE POINT;DALE PROMONTORY
Type Occupation site
Form Finds
Condition
Summary Excavations directed by W.F. Grimes between 1966 and 1983, of Dale
Iron Age promontory fort, found Romano-British pottery and a
brooch, suggesting occupation in this period.
Long description Dale Iron-Age promontory fort was formed by the construction of
a fairly complex sequence of defences in the late Bronze-Age and
the Iron-Age.
PRN 3103 NGR SM77512582
Name TREPEWET
Type Aerial photograph site
Form Cropmark
Condition
Summary L-shaped cropmark with rounded corner, characteristic of playing card
shape of Roman Fort identified on 1955 Meridian aerial photographs.
Nothing is visible on more recent aerial photographs. The field
appears heavily ploughed. F Murphy 2008
Long description L-shaped cropmark with rounded corner, characteristic of playing
card shape of Roman Fort identified on 1955 Meridian aerial
photographs. The feature bears no obvious relation to the
contemporary field enclosure pattern which is of irregular form in
this area. Interestingly, the field names to the south of this
feature contain the name element 'Caerwen' = 'White Fort'. The
site is located at a point where the Ffos-y-mynach deviates
sharply in zig-zag fashion away from its usual north-south
alignment to an east-west one. The via Julia passes close by this
site to the south. DE Morgan 1998
PRN 3392 NGR SN07071880
Name BROADWAY
Type Quarry
Form Cropmark
Condition
Summary A line of quarries identified as cropmarks from aerial photographs,
running along the north side of the Roman road west of Carmarthen
(PRN 14277). DS.09.2004.
Long description
PRN 3392 NGR SN08451851
Name TRE-WYNT
Type Quarry
Form Cropmark
Condition
Summary Two areas of quarrying to the south of a short segment of Roman
road (PRN 28102) part of the Roman road west of Carmarthen,
identified as cropmarks from aerial photographs.DS.09.2004.
Long description
PRN 3392 NGR SN10061848
Name GREAT VAYNOR
Type Quarry
Form Cropmark
Condition
Summary A segment of the Roman road west of Carmarthen (PRN 14277) with
a group of quarries along the south side, identified as parchmarks
from aerial photographs. DS.09.2004.
Long description
PRN 3393 NGR SN29891824
Name CLUNGWYN
Type Quarry
Form Cropmark
Condition
Summary A cluster of probable quarries associated with the construction of the
Roman road west of Carmarthen (PRN 14277), identified as
cropmarks from aerial photographs.DS.09.2004.
Long description
PRN 3393 NGR SN32581891
Name WENALLT
Type Quarry
Form Cropmark
Condition
Summary A group of quarries associated with the construction of the Roman
road west of Carmarthen (PRN 14277) identified as cropmarks from
aerial photographs.DS.09.2004.
Long description
PRN 3395 NGR SN55822235
Name COURT HENRY
Type Quarry
Form Cropmark
Condition
Summary A cluster of quarries to the south of the probable line of the Roman
road RR623 between Carmarthen and Llandeilo (PRN 11089).
Identified as cropmarks from aerial photographs. DS.09.2004.
Long description
PRN 3400 NGR SN64534660
Name ESGAIR-CORN
Type Quarry
Form Cropmark
Condition
Summary A cluster of quarry cropmarks identified from aerial photographs
located to the east of the Roman road RR62c from Pumsaint to
Llanfair Clydogau, and possibly associated with its construction.
DS.09.2004.
Long description
PRN 3590 NGR SN66678438
Name SALEM
Type Hoard
Form Finds
Condition
Summary A disturbed hoard of Roman coins found after the reinstatement of
ground above a Welsh Water pipeline. JH May 1998
Long description
PRN 3917 NGR SN16904585
Name PANTEG HOARD
Type Hoard
Form Documents
Condition Unknown
Summary A hoard of some 200 unspecified Roman coins, in a vessel or 'crock',
were reported in the local newspaper, The Cardigan & Tivyside
Advertiser, as being found by a farmhand whilst ploughing a field on
Panteg land. No further information is known about the find or what
became of the coins.
Long description
PRN 3941 NGR SN699428
Name MYNYDD MALLAEN
Type Hush
Form Earthwork
Condition Substantially intact
Summary It is possible that these features are the result of intentional
prospecting for mineral bearing rock during Roman or prehistoric
times.
Long description This west facing hill slope has a number of what appear to be
roughly parallel water worn gullies giving an appearance of
irregular ridge and furrow running down the hillside. These
appear to be cut through by the Annell Leat. It is possible that
these features are the result of intentional prospecting for mineral
bearing rock. By using collected water, possibly using a system
of small tanks or dams and washing away the soil and or mineral
within gullies down the hillside. LA 1999
PRN 4039 NGR SN07793992
Name ALLT JANE
Type Fort,Aerial photograph site
Form Cropmark
Condition
Summary Identified on aerial photograph taken 26/07/1999. Possible
double-ditched square enclosure, c.50mx50m with internal parching
possibly indicating the site of masonry buildings. Internal ditches also
present. The whole has a very rectilear appearance. KM 1999
Long description
PRN 4647 NGR SM93953080
Name ROMAN ROAD
Type Road
Form Placename
Condition
Summary A section of east-west road to the north of Letterston marked as
Roman Road on 1st and 2nd edition Ordnance Survey 1:10560 maps
(Pembrokeshire sheet XVI.NE) published in 1891 and 1908. The name
presumably originated through local tradition as there is no evidence
of a Roman Road in this area.
Long description
PRN 5194 NGR SN64454856
Name SARN HELEN
Type Quarry
Form Earthwork
Condition Damaged
Summary A group of roadside quarries between SN64624836 and SN64114921,
associated with the construction of the Roman road RR62c from
Pumsaint to Llanfair Clydogau which runs immediately to the west. At
the southern end a short section of the road may also be visible
where the modern road diverges slightly from the Roman route.
Several of the quarries are clearly visible on the ground.DS.09.2004.
Long description
PRN 5195 NGR SN63615291
Name PEN-DDOL
Type Quarry
Form Cropmark
Condition
Summary A cluster of quarries located on the west side of the Roman road
RR69c from Llanfair Clydogau to Llanio (PRN 5222) at SN63615291,
identified from aerial photographs.The Roman and modern roads
follow the same course in this location. DS.09.2004.
Long description
PRN 5196 NGR SN74363853
Name TY ISAF
Type Quarry
Form Cropmark
Condition
Summary Several clusters of probable quarries located between SN74123851
and SN74623850 associated with the construction of the Roman road
RR62c (PRN51972) which runs E-W to the south of these quarries.
Identified as cropmarks from aerial photographs. DS.09.2004.
Long description
PRN 5196 NGR SN72843834
Name PEN-Y-BANC-UCHAF
Type Quarry
Form Cropmark
Condition
Summary A cluster of quarries centred on SN72833853 to the west of a segment
(PRN 33994) of the Roman road RR62c, identifed as cropmarks and
plotted from aerial photographs.DS.09.2004.
Long description