1. Introduction

The narrow isthmus on Ynys Enlli/Bardsey Island (see main island record: NPRN 402783) separates the settled and farmed `mainland' to the north from the low-lying rocky southern peninsula which terminates in Pen Diban, upon which the lighthouse (NPRN 422200) was constructed in 1821. The isthmus is defined by the sweeping bay of Henllwyn on the south-east side, now used for modern landings with the concrete slipway and boathouse located at the eastern end, and the smaller sandy bay of Porth Solfach on the north side together with the rocky stretch of lower-lying coast on the west, Porth Hadog, these two inlets separated by the angular headland of Trwyn Dihiryn.

Coastal erosion is aggressively affecting both sides of this geologically weaker section of Bardsey Island, with Porth Hadog and Porth Solfach bearing the brunt of the most severe westerlies and north-westerly storms respectively, as evidenced by storm-thrown deposits upcast from the beaches, with the soft sediments of the more sheltered bay of Henllwyn still susceptible to southerly gales and storm surges.

The exposed section at Henllwyn changes annually, and has been intermittently recorded for its eroding archaeology since 2003. Basic comparison of the mapped high water mark with First Edition Ordnance Survey mapping from c. 1890 suggest as little as 8m has been lost from the eastern, slipway, end of the bay while up to 26m has been lost from its south-western end. In February 2017 the EU-funded CHERISH Project commissioned a 25cm LiDAR survey for the entire island from BlueSky International LTD, in order to precisely map erosion baselines for future generations (Archive Number: CHR_04 (6540709). The project also carried out drone photogrammetry surveys of the eroding coast in May 2018 (Archive Number: RCCS12 (2021-09-28_654), and in May 2018 and May 2019 carried out section cleaning to record eroding archaeology (see below).

A major synthesis of the CHERISH Project work on Enlli/Bardsey was published by Kirsty Ackland in 2023 using LiDAR and other survey datasets to assess the impacts of coastal erosion on the heritage assets (Ackland et al. 2023). The paper is available to read on Open Access.

Note: Mesolithic finds also identified at the west end of Henllywn (NPRN 422858).

2. BARDSEY ISLAND, Henllwyn Eroding Isthmus, Mesolithic Finds and Occupation Layers. (SH 1151 2109)

The CHERISH – Climate Change and Coastal Heritage project undertook fieldwork on Bardsey Island during May 2018 and May 2019. A priority of the work was survey and monitoring of the Henllwyn eroding isthmus.

In May 2018, on the north-west side of Henllwyn Bay at SH 11299 20929, close to one end of a post-medieval field wall, monitoring of the eroding section revealed prehistoric flints and remains of a limpet shell midden (NPRN 422858). Eroding from the south facing section at the eastern end of Henllwyn deposits of burnt bones were identified, which included an inital interpretation of 2 possible cremation burials. All deposits were located with GNSS and the sections were drawn. Samples of charcoal and bone were removed for analysis and C14 dating with permission of Bardsey Island Trust and Coroner’s Office at the Ministry of Justice.

Label	Location
Bone 2018/1	SH 11536 21093, 5.157m OD
Bone 2018/2	SH 11534 21093, 5.163m OD
Cremation 2018/1	SH 11529 21093, 5.021m OD
Bone 2018/3	SH 11526 21093, 5.196m OD
Bone 2018/4	SH 11521 21093, 5.040m OD
Bone 2018/5	SH 11514 21093, 5.098m OD
Bone 2018/6	SH 11514 21093, 5.056m OD
Cremation 2018/2	SH 11513 21093, 5.010m OD

Deposits of burnt bone were first noted and recorded here in 2003 by archaeologists Tom Dawson, Joanna Hambly, Robert Johnston and Esther Elin Roberts. This included 8 deposits of bones and what at the time were interpreted as 4 cremation burials, with charcoal and bone samples removed from two of them, labeled as Cremation 1 (context 110) at SH 11495 21092 (PRN 59966) and Cremation 2 (context 113) at SH 11514 21099 (PRN 59967). In the deposit (context 107) below Cremation 2 and directly on top of natural a flint was recovered. The flint was analysed and drawn by Ian Dennis for the CHERISH project in 2019. Examination of the flint revealed fine worn down serrates on one edge which classifies it as a microdenticulate flake. This flint tool type is usually associated with Mesolithic activity, although similar types can be found in early and late Neolithic contexts. A translucent deposit was also observed on the dorsal surface of the tool for which scanning electron microscopy was used to image and determine composition. The deposits were found to be largely siliceous with a high tin level. Plant processing has been shown to leave a siliceous deposit on flint and the high tin siliceous composition of the deposits are consistent with the use of the flint for plant processing in the local area (Dennis, 2019). Information, section drawing and samples from the 2003 work was transferred to the CHERISH team in 2018, for analysis and deposition within the NMRW.

In 2018, soil samples from six of the deposits were sent for analysis to Wessex Archaeology to facilitate the recovery of bone and charcoal for C14 dating. These comprised three samples from the 2003 work (two from Cremation 1 - PRN 59966 and one from Cremation 2 – PRN 59967) and three from the 2018 work (bone 2018/1 and 2018/6, and cremation 2018/1). Following environmental processing by floatation and wet-sieving to 1 mm sieve fraction, samples of charcoal suitable for radiocarbon dating were selected by the archaeobotantist.

Analysis of the very small quantities of bone recovered indicate none were human and the small size of the fragments, slightly eroded condition of the bone and lack of clearly identifiable morphological features rendered it difficult to identify species, but fragments representative of medium and small-sized mammals were noted. The majority of bone had been burnt to some degree but fragments of unburnt bone were also present, including a fragment of worked bone. None of the burnt bone was well oxidised, much of it being only charred and suggests the bone was burnt at relatively low temperatures and/or in reducing conditions rather than the sustained high temperature/oxidising conditions commonly associated with cremation pyres (McKinley, 2019).

This has led to the conclusion that the remains relate to domestic debris associated with an occupation on the isthmus, rather than a burial ground.

Three small samples of the charcoal recovered by Wessex Archaeology, all from the 2003 samples, have subsequently been analysed for potential C14 dating, and have been identified as heather, alder and gorse charcoal. In terms of material suitable for radiocarbon dating, all of the samples are below the size recommended, although some of the material is above the acceptable minimum weight. Of the three samples only one (alder), from context (110) - PRN 59966, has been recommended for dating (Elliot L, 2019) [see below]. This alongside animal bone fragment 2018/6 was sent to SUERC for dating.

In May 2019 the CHERISH team returned to Bardsey for further fieldwork which included continued monitoring of Henllwyn. Work involved drawing and context recording of a 37 metre length of the eroding section. Further burnt and calcined bone deposits were recorded at 4 locations during this work and a burnt flint deposit in the layer above natural.

3. Radiocarbon dating November 2020 yielding an early medieval date

A Charocal Identification and C14 Preparation report by Archaeological Services, Durham University (Archaeological Services, December 2019) for the CHERISH Project was prepared following flotation by Wessex Archaeology. All of the charcoal samples were below the recommended size (50mg) for radiocarbon dating, although some of the material was above the acceptable minimum weight, which is 10mg. The 2003 Context (110) sample 1B was the only sample that offered the possibility of suitable material. The alder stemwood was thought to provide enough carbon, by bulking up the material by adding the fragments of ash and hazel.

Radiocarbon dating was carried out by SUERC, Glasgow, on the 2003 context (110) sample 1B and reported on 24 November 2020 as SUERC-95520 (GU56126). The radiocarbon age was determined as 1157 +/- 24 BP (before 1950 AD), thus AD 793, early medieval for the domestic activity and small mammals represented.

Louise Barker, Toby Driver and Daniel Hunt, RCAHMW, for the EU-funded CHERISH Project.

Record updated: January 2025.

Survey for the CHERISH Project. Crown: CHERISH PROJECT 2018. Produced with EU funds through the Ireland Wales Co-operation Programme 2014-2020. All material made freely available through the Open Government Licence.

References:

Ackland, K., Griffiths, H., Barker, L., Davies, S., Driver, T., & Hunt, D. (2023). Mapping the impacts of coastal erosion on the heritage assets of Ynys Enlli (Bardsey Island), North Wales, UK. The Journal of Island and Coastal Archaeology, 19 (4), 786–813. https://doi.org/10.1080/15564894.2023.2227944

Archaeological Services, Durham University, 2019. Yny Enlli – Bardsey Island, Gwynedd, Wales. Report No. 5230, December 2019. Unpublished (NPRN 424528).

Barker, L., Driver, T. and Hunt, D. 2019. Bardsey Island, Henllwyn Eroding Isthmus (SH 1151 2109). Archaeology in Wales 59, 119-122

Dennis, I. 2019. Report on the lithic finds from Bardsey Island (unpublished)

Elliot, L. 2019 Ynys Enlli - Bardsey Island, Gwynedd, Wales. Charcoal identification and C14 preparation. Archaeological Services, Durham University. Report No. 5230.

Hunt, D., Barker, L., and Driver, T. 2019. CHERISH Project Lidar Capture and Analysis for the Islands of Wales. Archaeology in Wales 59, 91-96.

McKinley, J. I. 2019 Bardsey Island, Gwynedd (HEN 03 and 18). Report on the Burnt Bone. Wessex Archaeology. Ref: 104062.6

SUERC, Radiocarbon Dating Certificate 24 November 2020 (Laboratory Code: SUERV-95520 (GU56126)

Cofnod Safle