Druidston Haven combines graptolite-bearing Ordovician shale, younger Carboniferous rocks and a remarkable record of Ice Age deposition. The broad beach offers opportunities to observe loose fossil-bearing material and see how faulting and glaciation have brought very different chapters of geological history together.
FIND FREQUENCY – Graptolites occur in suitable shale; plant fragments and reworked shells belong to different rocks and deposits.
CHILDREN – A sandy beach with steep access, tidal limits and unstable cliff margins; supervise closely.
ACCESS – Signed paths and a steep approach with steps; use authorised parking and keep narrow roads clear.
TYPE – Loose beach material, a fault-bounded Ordovician bedrock section and much younger glacial deposits.
DIRECTIONS
♦ Approach Druidston on the local coast road and use the signed parking and walking arrangements. The National Park short-walk information gives parking at grid reference SM866169; follow current signs and do not obstruct the narrow road or private entrances.
♦ Use the signed path around the hotel and the established descent to the beach. The coast-path information describes steep access including 45 steps.
♦ On the beach, observe safe natural surfaces and loose shale while keeping a clear return to the access path. The caves, arches and cliff sections discussed in geological accounts are not collecting routes.
♦ Original locality reference: 51.81064°N, 5.10208°W.
FOSSIL HUNTING
Graptolites are the characteristic fossils of Druidston’s Ordovician shale. These colonial marine animals are flattened into narrow films, and some show two conspicuous rows of small cups. The original photographs illustrate the range of appearances, but individual species cannot be assigned simply from a dark line or a partly exposed fragment. A hand lens, scale and oblique light are more useful initially than splitting the rock.
Look at already exposed natural surfaces of loose shale in safe areas where inspection is permitted. Bedding is the original sedimentary layering; cleavage is a later pressure-related fabric. They may intersect here, explaining why a graptolite can appear to cross the surface along which the rock most readily separates. Repeated splitting can cut through a fossil rather than expose it. Photographs of in-situ examples retain useful evidence without disturbing the section.
The 1916 Geological Survey memoir recorded separate graptolite assemblages near Priest’s Vault and at the small stream mouth south of the drift-covered gap. Historical names include Dicranograptus brevicaulis and Diplograptus (Mesograptus) multidens; they are genuine section records rather than identifications newly made from the visitor photographs. The expanded geology retains the original locality and taxonomic qualifications instead of treating the whole beach as one species list.
Other fossils need different labels. Poorly preserved plant impressions in sandstone blocks relate to Carboniferous rocks, while marine shell fragments in glacial deposits were transported and reworked during the Ice Age. They are neither companions of the Ordovician graptolites nor proof of a marine community living at Druidston when the till accumulated. Keep shale, sandstone and reworked shell records separate and record the actual find spot and loose-block status.
Do not import the detailed plant or bivalve lists of Rickets Head, Black Cliff or North Maidenhall Point into an unidentified Druidston block. Those are separately documented localities in the broader coastal succession. Permission and preservation matter more than enlarging a collection: leave fragile or potentially significant material undisturbed until it can be assessed.
Significant discoveries, geological research and conservation milestones at Druidston Haven.
1826 – Early observations of transported blocks
De la Beche described abundant exotic blocks between Druidston and Broad Haven. A later survey corrected his suggested local igneous source, illustrating the developing understanding of glacial transport.
1909–1910 – Shell-bearing drift recorded
Survey work in 1909, reported in the following year’s Summary of Progress, documented shell-bearing drift at Druidston, then a new record for this part of Pembrokeshire.
1916 – Detailed bedrock and drift account
The Geological Survey memoir published the separate graptolite lists, locality-associated specimen registers, faulted contacts and shelly drift, providing an enduring reference for the beach’s different geological records.
1965–1970 – The Quaternary section investigated
Detailed work developed the interpretation of the channel-filling sequence and the evidence for in-situ till, including sediment fabric rather than fossil content alone.
1967 – Carbonised wood
Carbonised wood including Pinus was recorded from the till, adding transported organic material to the evidence preserved in the glacial deposits.
1970 – A radiocarbon lower limit
Reworked marine shells gave a result greater than 36,300 BP. This lower limit concerned the sampled shells and was not a direct date for till deposition.
1973–1984 – Revised beach and till interpretation
Later work favoured an Ipswichian raised beach with a Late Devensian glacial sequence, revising earlier interpretations of the deposits’ ages.
1989 – Conservation-review synthesis
The detailed Druidston account established a six-unit reference succession and explained the outstanding debate over the origin of the shelly glacial deposits.
2002 – Combined conservation interests documented
The statutory coastal citation recorded the area’s Carboniferous, structural and glacial interests, including Druidston’s faulted older rocks and Quaternary record.
2003 – Deformation-based interpretation
A study of sediment deformation refined the depositional interpretation through evidence of sheared basal tills and subsequent collapse or flow deformation, challenging a marine origin for the studied sequences.
2025 – Revised stratigraphic hierarchy
The preliminary regional framework placed Mydrim Shales in the Drefach Subgroup of the Montgomery Group, while the historically recorded Druidston specimens retained their separate locality provenance.
GEOLOGY
The older rocks at Druidston are exposed in a fault-bounded block beside much younger Carboniferous strata. The saved locality point maps to the Mydrim Shales Formation, traditionally described using Caradoc and ‘Mesograptus Beds’ terminology. Folding, cleavage and faulting record several episodes of deformation; the structure is not a simple depositional sequence in which the Ordovician shale passes upwards into all the neighbouring sandstones.
The distinction between the two sides of the bay is important. To the north, grey pennant-like sandstones extend towards Nolton and belong to the Carboniferous context. Immediately south of the Ordovician boundary, the older survey described dark shales with thin ironstone bands under its historical ‘Middle Shales of Millstone Grit’ terminology. These should not all be labelled Pennant Sandstone merely because they lie next to the graptolite-bearing block.
The 2025 BGS preliminary framework places Mydrim Shales in the Drefach Subgroup of the Montgomery Group. Its formation-scale graptolite zones provide regional context, not a modern reidentification of every specimen collected from the separate Druidston sections in the early twentieth century. No complete local Mydrim thickness is established here, and old fault-throw estimates based on uncertain coal correlations are not used as measured modern quantities.
Much younger sediments fill a deeply cut bedrock channel. The published Quaternary section includes a reported raised-beach gravel, slope debris, a thick shelly till, fluvioglacial sand and upper slope deposits. The till contains northern erratics, broken marine shells and carbonised wood, recording transport and reworking as well as deposition. The cliff is an archive of changing environments, not a source from which visitors should excavate samples.
Interpretations of the glacial sequence have changed as new evidence became available. A radiocarbon lower limit from reworked shells did not directly date the till, and later deformation studies refined earlier debates over its origin. The detailed account separates the six units of the older published log from later study schemes rather than combining their similarly numbered units into a new, unsupported section.
Explore Druidston’s faulted bedrock, separately recorded graptolite faunas and six-unit Quaternary succession.
Separate geological records. Ordovician bedrock, neighbouring Carboniferous rocks and channel-filling Quaternary deposits are distinct parts of Druidston’s history. They are not one continuously exposed vertical succession. Historical observations below retain their section positions and uncertainty; they do not establish present safe access to caves or cliff faces.
MONTGOMERY GROUP
DREFACH SUBGROUP
Mydrim Shales Formation
Terminology and age. The original locality maps to Ordovician Mydrim Shales. The group and subgroup assignment follows the BGS preliminary 2025 framework; it is not Hengae Group. Historical ‘Mesograptus Beds’ and Caradoc terminology are retained as context. Formation-scale foliaceus and clingani biozones do not assign both zones to every loose Druidston block or modernise every 1916 identification.
Northern boundary and Priest’s Vault. The 1916 memoir records black shale faulted against grey Coal Measure sandstone at the northern end, with a subsidiary fracture a few yards south. North-dipping beds flatten south of the promontory pierced by Priest’s Vault. Cleavage was measured dipping towards N20°W at 60°; this is a historical cleavage observation, not a bedding measurement or a current access instruction.
Stream-mouth and southern traverse. Bedrock reappears at the small stream mouth south of a drift-covered gap. A little farther south, beds were recorded dipping north-west at 40°, steepening towards the cliff top. Low-angle south-dipping thrusts and veining occur farther south, where bedding and cleavage become difficult to distinguish; eventually a high southerly dip leads towards the southern boundary fault. No complete local Mydrim thickness is derived from this traverse.
Priest’s Vault historical fauna. The memoir’s first list records Climacograptus antiquus, Dicranograptus brevicaulis, D. rectus, Diplograptus (Orthograptus) calcaratus var. acutus, Diplograptus (Mesograptus) multidens and its var. compactus, Diplograptus (Amplexograptus) perexcavatus, Diplograptus (Glyptograptus) teretiusculus?, the same species’ var. euglyphus, and Lasiograptus (Hallograptus) mucronatus var. bimucronatus. The question mark and historical combinations are preserved. Appendix locality 21 associates the section with Pem. 26 NE, Eλ7, register Pl.107–127; that range is not a one-to-one allocation of individual taxa to specimens.
Stream-mouth historical fauna. The separate second list contains Climacograptus scharenbergi, Dicranograptus brevicaulis, Diplograptus (Orthograptus) calcaratus var. acutus, Diplograptus (Mesograptus) multidens and its var. compactus, and Diplograptus (Glyptograptus) teretiusculus var. euglyphus. Appendix locality 22 is Pem. 26 NE, Eλ6, with Pl.128–146 and T.C.C.1999–2000. These are historical locality records, not a new taxonomic revision or identifications of all illustrated visitor specimens.
Faulted southern Carboniferous neighbour. Immediately south of the Ordovician boundary, the 1916 account describes dark shale with lean ironstone bands roughly an inch thick. Near the fault the dip was SSE at 85°, flattening farther south. These were called ‘Middle Shales of Millstone Grit’, not Pennant Sandstone. The farther Druidston Chins–Haroldston Chins–Settling Nose traverse is a separate broader section and is not a thickness for the Ordovician target.
WARWICKSHIRE GROUP: NORTHERN CARBONIFEROUS CONTEXT
Pennant Sandstone Formation
Northern sandstone and plants. Grey pennant-like sandstone was traced from Nolton past Madoc’s Haven to the Druidston fault. Modern regional Pennant Sandstone belongs to Warwickshire Group, not within South Wales Coal Measures Group. Poorly preserved plant impressions in the original guide’s sandstone blocks are retained without importing named plant assemblages or a precise Bolsovian/Asturian assignment from neighbouring collecting sites.
Structure and correlation limits. The protected-site account recognises an Ordovician horst, pre-Variscan extension and Caledonian structures overprinted by Variscan deformation. ‘Older rocks exposed in a fault-bounded block’ avoids implying a simple uplifted stratigraphic stack. Historical resultant fault throws of about 4,000 ft or at least 6,000 ft depended on uncertain coal correlations and are not adopted as exact modern measurements.
Quaternary channel fill: published units 1–6, bottom to top. The conservation account describes a sequence extending laterally for about 150 m in a deeply cut bedrock channel, interpreted as an older meltwater incision. Lateral extent is not stratigraphic thickness. The numbered units below belong to that account and are not merged with similarly numbered units from later deformation studies.
1. Raised-beach gravel. Stained and cemented gravel was historically described at the base. It had not been relocated in the years before the conservation account was published, so present visibility is not promised. No thickness was stated.
2. Lower head, up to 2 m. Blocky quartz-sandstone debris with iron staining represents periglacial slope material. ‘Up to’ is retained; the deposit is not assigned a uniform two-metre thickness throughout the channel.
3. Irish Sea till, approximately 15 m. The calcareous till includes northern erratics, marine shell fragments and carbonised wood, including Pinus sp. Its upper approximately 1.3 m is decalcified, with a thin decalcified layer also reported at the base. These are alteration zones within the till, not extra beds to add to its thickness.
4. Fluvioglacial sands. Sand is interbedded with the upper till, with increasing sand and gravel. No separate thickness is established. The transition should not be drawn as an everywhere-sharp boundary between uniform units.
5. Upper head. Solifluction material records later slope movement. No thickness is given in the source, and its geometry is not inferred from another locality’s log.
6. Sandy loam, 0.3 m. This thin upper unit completes the published six-part succession. The numeric value belongs to that documented section.
Historical reworked-shell record. The 1916 Druidston drift and stream-bank list comprises Astarte (Nicania) compressa?, Astarte sulcata, Mytilus edulis, Pectunculus glycimeris, Trophon (Boreotrophon) clathratus, Trophon (Boreotrophon) scalariformis, Turritella sp. and ?Tellina (Macoma) balthica. Historical names and doubts are retained. The locality register is Pem.26 NE, Eλ4, T.C.C.2001. Later work reported sixteen species, but the complete later table has not been reconstructed by adding guessed names. These are reworked shells, not Ordovician fossils or an in-situ living assemblage.
Dating and changing interpretations. A bulk shell sample gave a radiocarbon lower limit greater than 36,300 BP (I-1687), reported in 1970. Because the shells were reworked, it does not directly date till deposition or supply a calibrated calendar age. Bowen’s 1966 interpretation favoured older or reworked till above a Hoxnian beach; John’s 1965–70 work argued for in-situ till using fabric evidence. Bowen’s later 1973–84 interpretation accepted an Ipswichian beach and Late Devensian glacial sequence. These are stages in interpretation, not simultaneously adopted ages.
Later deformation evidence. The 1989 conservation synthesis allowed a glaciomarine alternative. McCarroll and Rijsdijk’s 2003 study instead reported sheared basal till and rejected a marine origin for the studied successions, refining the earlier debate. Its full original figure and numerical datasets were not available for this revision, so the later unit geometry and fabric plots are not redrawn. In particular, the later six-unit scheme is not silently substituted into the older six-unit log.
References. Cantrill and colleagues (1916), The geology of the South Wales Coalfield, Part XII: the country around Milford, pp.46–48,107,111–113,134–137,140–141,155–157 and168; Campbell and Bowen (1989), Quaternary of Wales, Druidston, pp.75–76; Natural Resources Wales, Newgale–Little Haven Coast citation (2002); McCarroll and Rijsdijk (2003), Journal of Quaternary Science 18,473–489; BGS Mydrim Lexicon and mapping; Davies (2025), BGS OR/25/018,pp.89–90.
SAFETY
Plan around local tide, swell and weather conditions and retain a clear return to the beach access. Avoid rounding points on a rising tide; do not rely on the old assurance that the sea reaches the cliffs only near high water.
Keep away from cliff bases, overhangs, caves and unstable glacial deposits. Loose blocks, wet shale and uneven rocks add hazards, and the steep steps need care. Scientific section descriptions are not invitations to stand beneath or climb the exposures.
EQUIPMENT
Take tide information, sturdy footwear, a hand lens, camera, scale and notebook. For authorised removal of loose material, bring individual padding and a rigid box. Natural fossil surfaces can be examined without hammering or repeated splitting.
CLEANING AND TREATING
Protect the thin fossil film and fragile shale by using only gentle dry brushing of loose dirt. Avoid acids, varnish, aggressive scraping, prolonged soaking and universal consolidant treatments. Leave important or delicate specimens unprepared for specialist assessment.
Retain counterparts and keep Ordovician shale, Carboniferous sandstone and reworked shells in separately labelled records. Record the precise spot, date and whether the material was loose; do not infer its original bed solely from where it lay on the beach.
IDENTIFY YOUR FINDS
Need help identifying a fossil? Share clear photos, where you found it and its size with the community.
FURTHER READING
Geological Survey memoir: the country around Milford (1916)
Geological Conservation Review: Druidston’s Quaternary sequence
BGS Lexicon: Mydrim Shales Formation
BGS preliminary revised Lower Palaeozoic framework (2025)
Natural Resources Wales: Newgale–Little Haven Coast SSSI citation
National Park: Druidston walking route
ACCESS RIGHTS
Druidston lies within the Newgale–Little Haven Coast / Arfordir Niwgwl–Aber Bach SSSI, whose interests include the structure, Carboniferous succession and glacial deposits. Ask the landowner before collecting and follow current conservation notices. Do not hammer the bedrock, excavate the cliffs or disturb large blocks.
Beach access does not authorise fossil removal or scientific sampling of designated sections. Use the signed paths and authorised parking; keep private access and the narrow roads unobstructed.
It is important to follow our ‘Code of Conduct’ when collecting fossils or visiting any site. Please also read our ‘Terms and Conditions‘
Joe Shimmin. Last inspected 2007, Updated 2017, Edited by Jon Trevelyan 2016
LINKS
♦ Fossil Discussions
♦ Fossil Articles
♦ Buy Fossils, Tools and Equipment
♦ Buy Crystals, Meteorites, and Artefacts
♦ Join Fossil Hunts
♦ UK Fossils Network
