1.

Archaeological summary:

The promontory fort site of Caerfai (also known as Penpleidiau) occupies a very large and visually dominant natural coastal promontory approximately 1.3km to the southeast of the city of St David’s. With its precipitous cliffs the promontory protrudes around 500m into St Bride’s Bay. The archaeological site occupies a 0.5ha sub-rectangular headland at the tip of the natural promontory, enclosed by a combination of formidable ramparts and 20m high cliffs. The 1925 Royal Commission survey also named the site ‘Penpleidiau Camp’. The promontory fort is owned by the National Trust and managed by the Pembrokeshire Coast National Park Authority.

A new programme of work commenced at the site between 2017-2022 under the EU-funded CHERISH Project. This included new ground and drone survey of the coastal promontory fort and a new magnetometry survey by SUMO Survey in March 2019. The detailed magnetometry survey was conducted over approximately 2.5 ha of grassland and scrubland. No definite archaeological responses have been identified, though a large number of anomalies of uncertain origin have been mapped. Areas of enhanced magnetic response have been detected within the fort itself, and it is possible that these are archaeological or natural in origin. Evidence of a ploughing headland is also visible in the data, while one of the uncertain linear trends is possibly related to an historic field boundary.

An evaluation excavation was carried out in September 2021 by DigVentures funded by CHERISH, with a follow-up crowd-funded excavation by DigVentures in September 2022. The excavations were focussed on the eroding isthmus of the promontory fort, threatened by coastal erosion, and the promontory defences which were threatened by scrub and blackthorn encroachment and burrowing animals. Evidence for a possible terrace or rampart was recorded at the isthmus, recovering high concentrations of organic, burned material, a flat-laid stone surface, a possible wall and features cut into the surface level proving to be post holes. The excavation also recovered finds including a spindle whorl, pottery and fired clay fragments, copper and iron slag, and several sling shots, whetstones and rubbing stones, providing evidence for Iron Age or earlier occupation.

Landscape setting and geology:

Caerfai forms a significant component of the late prehistoric coastal landscape of the St David’s peninsula which boasts twelve promontory forts, the most common type of late prehistoric site in this part of Pembrokeshire. The closest promontory fort to Caerfai is Porth y Rhaw 2.3km to the east and within visible range.

The natural promontory is formed of both sedimentary mud and sandstones across the northern area and a southern igneous intrusion. The difference in rock type has noticeably influenced the extent to which the promontory has been shaped by erosion. The most severe erosion is visible at the large gully immediately to the south of the defences where a relatively narrow isthmus has formed at the interface between the two rock types. The exact extent of erosion that has occurred since the site’s initial construction is impossible to recover, however, it is more than likely that at the time of occupation the isthmus would have been considerably wider. The nature of the stacked sedimentary rocks that form the northern side of the gully also suggests land has possibly been lost immediately to the south and west of the defences.

Earthwork description:

The following earthwork description is the result of analytical earthwork and UAV surveys conducted during January 2019 by the Dan Hunt and Louise Barker for the CHERISH Project, led by the Royal Commission.

Phase 1

The earliest phase is represented by the southern-most pair of parallel ramparts that enclose the headland. These visible earthworks appear to represent the original extents of both banks, only broken by two well-aligned gaps that would have previously served as an entranceway on the far eastern side.

The western portion of the inner-most rampart is formed of a 67m long curved bank which straddles the northern side of the vast eroding western gully. It is also the taller of the two inner banks and in places stands around 1m higher than the second rampart. Both the western and eastern terminals appear genuine, with the former seemingly terminating immediately before the cliff edge, indicating limited coastal erosion at this location. To the east of the entrance is a much smaller and more subtle bank which completes the inner rampart circuit. This small segment measures around 11m in length with a maximum height of approximately 0.9m and terminates before a steep drop in the coastal slopes to the east. Immediately to the south is a slight counter scarp which may relate to a possible inner ditch of sorts. Based on previous mapping it is highly likely that the bank would have continued down the steep slope to join the cliff edge. Later soil slippage running down the slope is the likely cause of erosion of the bank.

The second inner rampart runs parallel to bank, both separated by a ditch that would have provided material for the inner bank (now largely silted up). With a total length of around 115m (including bank to the east of the entranceway) this is by far the longest of the four lines of defences which visibly stretch from the west cliff-face across to the eastern cliffs. Like the inner rampart all terminals appear to be genuine, respecting the current clifftop lines. Again, completing the rampart circuit to the east of the entrance is a very shallow earthwork, most of which is only visible on UAV derived DEM visualisations where it visibly continues down the steep coastal slope to meet the cliff edge. By doing so it completes the rampart circuit and fully encloses the promontory. Like bank this part of the rampart has likely suffered from soil slippage down the coastal slopes.

Perhaps most interesting about both ramparts is the fact that they respect the northern edge of the deep eroding gully. This is unlikely to be a coincidence. Whilst there has been significant erosion that has narrowed the isthmus over time, it is likely that there was already a certain amount visible at the time of construction. This poses questions surrounding the reasons as to why the effort was taken to fortify the external side of precipitous cliffs which already served to defend the interior. Understanding Promontory Forts

Phase 2

Immediately to the north of the two inner banks is a third rampart which is possibly later in origin. The total length of this bank is around 88m with a roughly consistent width of around 10m, almost double that of the inner two ramparts. This much larger and broader bank was likely built to further improve the existing bivallate defences. This is demonstrated by the way it respects the curved east/west alignment of for much of its course.

At its eastern end however there are some differences in the bank’s composition and alignment. Here there is a notable dogleg where the rampart changes direction slightly in a northerly direction. It is unclear as to why this is the case, but it is likely that this was done to embellish the eastern entrance way into the promontory fort. It has also been suggested that this doglegged ‘out-turn’ could well be a later modification made to an earlier rampart, similar to what is described in Phase 3 (Barker and Driver pers comms.). Along the top of the bank are also indications of possible alterations to an earlier underlying rampart, notably a possible heightening at the centre of the rampart. It is clear that the earthworks here suggest a series of small modifications made during its use. What is also interesting is that while this bank was seemingly constructed to improve the pre-existing ramparts, anyone approaching the fort directly from the north would now find the two inner ramparts and interior totally blocked from view.

Associated with this large rampart are the clear remains of an external ditch which runs from the western cliff-face and wraps around the rampart’s eastern terminal. Here the ditch is noticeably truncated by a later hollow cut. The ditch is around 1m deep along most of its length but is noticeably deeper in the east than it is in the west.

Phase 3

The addition of further features appears to have significantly modified parts of pre-existing ramparts to further embellish what was at the time still likely to have been the entranceway into the fort interior. The most notable enhancement is the modifications made to the eastern terminal of rampart which saw it significantly heightened to around 2.2m and realigned to respect the doglegged terminal of the large outer rampart. Immediately to the north is evidence of the of a relatively deep hollow cut which truncates earlier ditch. This feature was more than likely created in order to provide material for the improvement and to exaggerate the enhanced terminal. It is plausible to suggest that these enhancements were made at a similar time to phase 2 constructions due to similarities in alignment between the doglegged terminals.

References:

Davies, R. 2019. Geophysical Survey Report 14719: CHERISH Ireland-Wales Project - Caerfai (Penpleidiau), Pembrokeshire. SUMO Services. Unpublished.

Duensing, S.N and Teale, K. 2021. CHERISH Evaluation Excavation 2021. Penplediau / Caerfai Promontory Fort, Post Excavation Assessment. Unpublished report. OASIS ID: digventu1-212051

Hunt, D. 2020. Understanding Promontory Forts: Evaluating the use of combined archaeological survey approaches in researching the coastal promontory forts of Pembrokeshire. Kellogg College, Oxford. Unpublished dissertation.

2.

A 3D model of the site was created following field survey by the CHERISH project which can be accessed with the links below:

Cymraeg: https://skfb.ly/6SCGW

English: https://skfb.ly/6SCHr

3.

In August 2021 CHERISH installed two fixed survey markers (survey nails in an exposed rock and an earth-fast boulder) at Caerfai. The markers and their associated location coordinates (BNG) will enable accurate monitoring and change detection at the site going forward. Details are:

E2 Primary Station Marker - Easting: 176239.9611; Northing: 223928.9810; Height: 21.8663

E6 Secondary Control Point - Easting: 176287.2621; Northing: 224021.3485; Height: 27.5102

See CHERISH Monitoring Network - CF_E2 and CF_E6 control markers Event Report: 02/08/2022 for full details including Witness Diagrams (CHERISH Survey Report No. CH/RCAHMW 44 and Data Archive RCCS31)

Louise Barker, CHERISH - RCAHMW, December 2023

CHERISH (Climate, Heritage and Environments of Reefs, Islands and Headlands) was an EU-funded Wales-Ireland project (2017-2023) led by the Royal Commission on the Ancient and Historical Monuments of Wales, in partnership with the Discovery Programme: Centre for Archaeology and Innovation Ireland, Aberystwyth University: Department of Geography and Earth Sciences and Geological Survey, Ireland. https://cherishproject.eu/en/

Cofnod Safle