Marloes Sands combines Silurian fossils with spectacular folded cliffs and the volcanic rocks of an ancient coastline. Corals and brachiopods are the main attraction, with trilobites, crinoids and other marine remains recorded in the fossiliferous beds. The wide sandy bay is also an excellent place to see how shallow seas gave way to river floodplains.
FIND FREQUENCY: ♦♦♦♦♦ – Fossils can be abundant in the right beds, although many neighbouring rocks are barren. Observation and photography are rewarding.
CHILDREN: ♦♦♦♦♦ – The sandy beach offers plenty to explore with close supervision around rocks, cliffs and the incoming tide.
ACCESS: ♦♦♦♦ – National Trust parking followed by an approximately 800 m walk to the beach; the descent and rocky shore are uneven.
TYPE: Cliffs, foreshore and naturally fallen material.
DIRECTIONS
♦ Follow the B4327 from Haverfordwest and signs for Marloes. Turn left in Marloes village for Marloes Sands, rather than continuing to Martin’s Haven.
♦ Use the National Trust Marloes Sands car park, SA62 3BH. Parking charges apply; check the payment machine or current National Trust information rather than relying on a staffed booth.
♦ Follow the signed beach route from the car park down Sandy Lane to the shore, allowing approximately 800 m each way. Follow any diversions or temporary notices on arrival.
♦ The original guide’s reference is 51.73625°N, 5.20926°W. Use the signed parking and beach route for the approach; this reference is not a substitute for the access path.
♦ Plan the shore visit for a falling tide and low water. Marloes Sands can be covered at high tide, so leave ample time to return to the beach access.
FOSSIL HUNTING
The most rewarding fossil-bearing rocks are the grey marine mudstones, siltstones and shelly limestone lenses of the Coralliferous Formation. The beds are tilted and broken by faults, so productive rocks recur in different parts of the bay rather than forming one continuous band beside the path. Recorded exposures include Mathew’s Slade, the area south-east of the Three Chimneys and the shore immediately west of the Sandy Lane access. Look first at weathered surfaces: fossils often stand out more clearly there than on a fresh broken face.
Corals are especially characteristic. Solitary rugose corals may show a cup-shaped or horn-like outline, with fine radiating partitions visible in a broken or weathered cross-section. The small solitary coral Palaeocyclus porpita is documented in the succession; colonial favositid corals are also recorded. Some dark, fragile examples can resemble charcoal, but their organised internal pattern separates them from an irregular stain or mineral coating. Photograph delicate exposed detail before handling any loose specimen.
Brachiopods occur as shells, shell fragments and moulds. Look for repeated ribs, a hinge margin and a symmetrical shell outline. The documented succession includes Eocoelia sulcata and Costistricklandia lirata lirata; these records are useful scientifically because different assemblages track changing water depth and age. A weathered fragment is rarely enough for a confident species identification. The photographs below retain the original specimen labels, which should be checked against the actual shell rather than used as a guarantee that every similar-looking find is the same species.
Other recorded marine groups include trilobites, crinoids, bryozoans, straight-shelled cephalopods, bivalves, gastropods, tentaculitids and cornulitids. Small circular crinoid stem segments and broken shells can be easier to find than a complete trilobite. Shelly lenses commonly represent material concentrated or moved by storms, so several groups may occur together in a small area while the surrounding mudstone is much poorer.
The sedimentary beds above the Marloes basalt flows also contain fossils. Some horizons are dominated by Lingula, whereas others include a more varied marine assemblage with brachiopods, corals, molluscs and crinoid remains. These beds belong to the Skomer Volcanic Group: the fossils occur in its interbedded sediments, not within the solid lava itself. In the younger Gray Sandstone Formation the fauna becomes less diverse towards the top, with records of Lingula, Cornulites and bivalves accompanying the increasingly shallow-water deposits.
Marloes is also important for microscopic fossils, including conodonts and organic-walled microfossils. These scientific records usually depend on specialist preparation and are not a promise of visible specimens on the beach. Likewise, the spores used to date the Red Cliff Formation are evidence about ancient land environments, not the conspicuous corals and shells for which most visitors come.
The established guide identifies two useful search areas south-east of the main descent: around another path reaching the beach, then farther along among the near-vertical flaky beds. Treat these as broad orientation points and check the rock type, because sand cover and fresh erosion change what is visible. Keep clear of cliff-foot debris and inspect only safely accessible surfaces. Ask the landowner about removal of loose fossils before taking material; leave attached specimens and scientifically important exposures undisturbed.
Significant fossil discoveries and geological research at Marloes Sands.
1839 – Silurian geology recognised
Roderick Murchison described Marloes Bay and illustrated its rocks as part of the early recognition of the Silurian succession. The locality also became associated with the classic brachiopod now known as Costistricklandia lirata lirata.
1846 – The marine units distinguished
Henry De la Beche logged the bay and introduced the names Coralliferous Series and Gray Sandstone Series, separating the shelly marine rocks from the sandier succession above.
1916 – A comprehensive Geological Survey account
The Milford memoir brought together detailed descriptions, a map, a sketch section and fossil records, establishing a foundation for later work on the bay’s volcanic and sedimentary rocks.
1969 – Volcanic and sedimentary relationships reassessed
Alfred Ziegler and colleagues placed the volcanic and associated conglomeratic rocks within the Skomer Volcanic Group and described the Marloes succession and its changing marine fossil communities.
1976–1977 – Environments, faunas and deformation
Research on the coastal sediments and their fossil assemblages clarified the changes from restricted shore environments to more open marine conditions. Structural work explained why the bedding and cleavage near the Three Chimneys can give apparently conflicting indications of which way the beds young.
1978 – Red Cliff Formation named
John Allen and Brian Williams formalised the Red Cliff Formation, named from the eastern end of Marloes Sands, as part of their subdivision of the early Old Red Sandstone succession.
1983 – Conodonts sharpened the age evidence
Mabillard and Aldridge studied microfossils around the Llandovery–Wenlock boundary. Marloes is the type locality cited for Apsidognathus ruginosus and Icriodella? sandersi, highlighting its importance beyond visible beach fossils.
2004 – The marine-to-terrestrial transition reinterpreted
Robert Hillier and Brian Williams examined the transition into the basal Old Red Sandstone and its fault-controlled sedimentary setting. Their late-Ludfordian interpretation of the Red Cliff Formation was subsequently retained in the 2015 regional framework.
2024 – Marine sections correlated geochemically
Sarah Veevers and colleagues combined sedimentology and high-resolution geochemistry to compare the Coralliferous Formation at Marloes Sands and Renney Slip, distinguishing eleven chemical subdivisions across the study.
2025 – A revised regional hierarchy
The revised BGS framework adopted the Marloes Group for the Coralliferous and Gray Sandstone formations within a wider regional hierarchy, reflecting their related offshore, shoreface and estuarine depositional systems.
2026 – Red Cliff age reassessed
Hillier and Kenneth Higgs reported a revised age assessment: a basal constraint no older than late Homerian, with spores indicating late Gorstian to early Ludfordian intervals within the formation. These separate age statements should not be collapsed into a single date for every exposed bed.
GEOLOGY
The rocks at Marloes Sands record a changing Silurian coastline on the southern margin of the Welsh Basin. Volcanic activity produced the basalt flows of the Skomer Volcanic Group, while sand, gravel and mud accumulated around them. Later, a break in deposition and erosion was followed by renewed marine flooding. The fossil-rich Coralliferous Formation records this next phase of the sea advancing across the area.
The corals and brachiopods did not live in the same conditions as the lava erupted. Their remains occur mainly in younger marine sediments and in sedimentary intervals within the older volcanic succession. Storms concentrated shells into lenses, and changes in water depth affected which communities could live there. The modern Coralliferous Formation includes a coarse basal Renney Slip Member and the finer marine Deadman’s Bay Member. Older descriptions call the whole unit the Coralliferous Group.
The Gray Sandstone Formation records the gradual dominance of sandier shoreface, tidal and estuarine environments. Together with the Coralliferous Formation it belongs to the Marloes Group in the 2025 framework. These formations are related depositional systems, with lateral equivalents as well as their familiar local upward succession; the whole region should not be reduced to an identical stack at every beach. The Gray Sandstone succession is largely Wenlock in age, but not every individual bed has an exact fossil-based date.
At Red Cliff, the red and purple mudstones and sandstones represent the spread of terrestrial floodplains and river channels. Soil formation produced calcrete, while drying and renewed floods left their own structures. This is an early development of Old Red Sandstone conditions within the Silurian, rather than proof that the rocks must be Devonian. The revised 2026 dating distinguishes a maximum basal age from younger spore-bearing intervals; it does not date all the red beds uniformly.
The near-vertical ribs, folds and faults visible today formed long after much of this sediment had accumulated. Faults repeat or interrupt units around the bay, and later Variscan deformation imposed a cleavage that can cut across original bedding. The Three Chimneys are resistant sedimentary beds, not three volcanic vents. Their striking shape makes a useful landmark, but neither cliff height nor the width of an outcrop on the beach is the true thickness of its formation.
The Marloes basalt flows, marine fossil beds, Three Chimneys and Red Cliff terrestrial succession.
Scope and order. This account follows the exposed Marloes Sands succession from older to younger rocks. Fault repetition means that it is not a single uninterrupted traverse. The adjacent Albion Sands and Gateholm sections are distinct exposures; their measured beds and fossil records are not assigned to Marloes Sands merely because some formations are shared.
SKOMER VOLCANIC GROUP
Only the upper part of this group is exposed in the bay. The approximately 1,000 m regional thickness quoted for the group is not the thickness visible at Marloes. Fossils in the upper sedimentary beds support an Aeronian age. The succession contains lava flows together with sedimentary and ash-rich intervals, recording a volcanic landscape progressively affected by the sea.
Mathew’s Slade basal exposure. At the eastern side of Mathew’s Slade, a faulted outcrop includes approximately 6 m of arkosic sandstone below a basalt flow. It should not be stitched automatically into a continuous thickness with the more complete section immediately eastward.
Lower Marloes Basalt. In the better-exposed, steeply inclined section, grey sandstone is followed by a basalt flow reported as 23 m thick. Its vesicular base and reddened top help identify it as a lava flow, rather than a later intrusive sheet.
Upper Marloes Basalt. The succeeding flow is reported as 19 m thick and also has a vesicular base. These two named flows are the local volcanic landmarks; they do not represent the entire Skomer volcanic succession.
Post-basalt sandstones and conglomerate. Above the lavas, up to 3 m of thin- to medium-bedded sandstone and siltstone is largely removed by faulting. A succeeding conglomerate, approximately 9 m thick, contains quartz and rhyolite pebbles. These coarse sediments record transport and reworking around the volcanic shoreline.
Upper sedimentary succession. Approximately 150 m of sandstone, silty mudstone and conglomerate, with numerous tuffaceous bands, follows in the published section. Scoured bases, ripple-marked tops and burrows are locally visible. Restricted Lingula-dominated assemblages contrast with more varied marine faunas containing other brachiopods, favositid corals, tentaculitids, crinoids, bivalves and gastropods. The more diverse beds represent stronger open-marine influence.
Brecciated tuff horizon. A conspicuous tuff about 6 m thick lies approximately 70 m above the basalt flows. Its upper part is rippled, showing sedimentary reworking of volcanic material. The quoted height is a position within the published section, not the thickness of the whole interval above the flows.
The Three Chimneys. Three near-vertical sandstone beds separated by dark siltstone form the landmark at approximately SM 7860 0732. They lie about 15 m above the tuff and 57 m below the top of the Skomer Volcanic Group in the historical section. These are resistant sedimentary ribs; the name does not describe volcanic chimneys.
Upper fossiliferous beds and erosion surface. Limestone ribs and lenses near the top yield conodonts including Pranognathus tenuis. An erosion surface caps the group, with reddening affecting the uppermost approximately 0.5 m. The contact below the Coralliferous succession has a reported angular discordance of 5–10°, although a small fault complicates it locally. Fossil evidence indicates a break that removes much of the intervening later Aeronian and Telychian record.
HAFREN SUPERGROUP
MARLOES GROUP
The 2025 BGS framework adopts the Marloes Group, which brings together the Coralliferous and Gray Sandstone formations following earlier sedimentological work. Their offshore, shoreface and estuarine facies record repeated changes in shoreline position within the fault-influenced Skomer Sub-basin. Although one succeeds the other in the familiar local succession, lateral equivalence is important regionally. Historical ‘Coralliferous Group’ and ‘Gray Sandstone Group’ labels refer to units now treated at formation rank here.
Coralliferous Formation
The principal fossil-rich unit spans the later Telychian to early Wenlock interval in the traditional Marloes account. Older descriptions give approximately 100 m locally. Fault-bounded exposures occur at Mathew’s Slade, south-east of the Three Chimneys and west of the Sandy Lane descent. A precise Llandovery–Wenlock boundary should not be inferred simply from the first coral or a change in rock colour: the macrofossil and microfossil evidence has been interpreted differently as research developed.
Renney Slip Member
Basal coarse beds. The modern member name applies to the basal transgressive coarse deposits. The older Marloes description records approximately 10 m of sandstone and conglomerate above the unconformity, with prominent asymmetric ripples. Fossils are relatively scarce in these coarse beds. This is the local published description, rather than an import of the separate Renney Slip type-section measurements.
Deadman’s Bay Member
Fine-grained marine beds. Above the basal coarse interval are hard blue-grey silty mudstones, thin sandstones and shelly limestone lenses, with occasional altered volcanic-ash beds. The member represents more offshore deposition following marine flooding. Shell-rich lenses commonly record concentration and transport by storms, rather than an undisturbed reef growing across the entire bay.
Lower shelly assemblages. The published Marloes sequence records Eocoelia sulcata and the solitary coral Palaeocyclus porpita in lower beds, followed by abundant Costistricklandia lirata lirata. The latter becomes less common above about 30 m and disappears by roughly 50 m in the older account. These are assemblage trends and approximate positions, not continuously traceable fossil bands guaranteed at every exposure.
Limestone lenses and microfossils. Conodonts were recovered through the lower approximately 70 m of the historical Coralliferous section. Recorded taxa include Pterospathodus amorphognathoides, Kockelella ranuliformis, Distomodus staurognathoides, Apsidognathus ruginosus and Icriodella? sandersi. The last two are described in the conservation account as having Marloes type localities. This record concerns microscopic elements recovered by specialist methods, not large fossils readily picked out in the field.
Organic-walled microfossil horizons. Siltstone samples in the lower half yielded sparse, darkened and poorly preserved acritarchs including Domasia trispinosa, Micrhystridium stellatum and Veryhachium trispinosum. Scolecodont fragments and possible chitinozoans were also noted. Their poor preservation and the tentative identification of the latter should be retained when interpreting the assemblage.
Upper coarsening beds. In the upper approximately 12 m, siltstones become coarser and pass towards the sandier Gray Sandstone succession. An oncolitic horizon is described in the upper part of the unit and is consistent with shallow-water conditions. An older report of Eocoelia cf. sulcata and Palaeocyclus porpita about 2 m below the top helped frame debate over the series boundary. The separate record of Eocoelia angelini from nearby Deadman’s Bay is not treated as a specimen record from Marloes Sands.
Modern correlation. The 2024 study compares Marloes Sands with Renney Slip using geochemistry and sedimentological trends and recognises eleven chemical subdivisions across the study. Those research correlations do not turn a regional thickness range into a measured Marloes log, nor provide casual collectors with eleven conspicuous field beds.
Gray Sandstone Formation
The older local conservation account gives an approximate thickness of 300 m. The 2025 framework regards the formation as poorly dated but largely Wenlock. It includes a generally upward-coarsening marine-to-shoreline succession; published thickness and age summaries should not imply an uninterrupted, unfaulted exposure at every point in the bay.
Lower mixed sandstone and mudstone interval. Sandstone, siltstone and grey sandy mudstone occur with weathered ‘rottenstone’ bands. The historical Fauna IV is characterised as a post-Eocoelia, Salopina community, consistent with a younger Wenlock marine assemblage. The informal fauna number is a research assemblage label, not a formation or member.
Upper approximately 60 m. Greenish-grey quartzitic sandstones alternate with mudstones that can weather olive-yellow. Cross-bedding, lenticular bedding, bioturbation and channel structures become conspicuous. The impoverished historical Fauna V contains Lingula, Cornulites and indeterminate bivalves; it is environmentally informative but does not independently give a precise age. The sedimentary structures record increasingly shallow, tidal and estuarine conditions.
Transition at Red Cliff. Grey sandstones with inclined heterolithic stratification meet and interdigitate with calcretised red mudstones. This relationship supports a changing coastal-to-terrestrial environment rather than treating colour alone as proof of a major time gap. Older interpretations of the basal Old Red Sandstone boundary differed, so the detailed contact should be considered in its local structural setting.
OLD RED SANDSTONE
DAUGLEDDAU GROUP
MILFORD HAVEN SUBGROUP
Red Cliff Formation
Named from Red Cliff at the eastern end of Marloes Sands, the formation comprises red to purple mudstone and siltstone with red, purple and greyish-pink fine sandstone. Calcrete-bearing palaeosols and desiccation structures occur in the mudrocks, while channel forms and bioturbation are present in some sandstones and coarser mudstones. Floodplains, temporary river channels and sheetfloods developed in a seasonally dry alluvial setting.
Thickness and locality distinction. The 2015 framework records approximately 55 m at Red Cliff, compared with approximately 50 m at Albion Sands and 17 m at Lindsway Bay. These are separate sections, not additive parts of a Marloes log. The same account identifies Albion Sands as the type section and Red Cliff as a reference section, despite inconsistent use of ‘type section’ in its thickness paragraph. The earlier conservation account’s figure of less than 52 m is retained as an older estimate, rather than made falsely identical to the later 55 m value.
Age reassessment. The late-Ludfordian age quoted in the 2015 framework reflects the 2004 interpretation. Hillier and Higgs’ 2026 reassessment instead reports a basal constraint no older than late Homerian and spore evidence for late Gorstian to early Ludfordian intervals within the formation. The basal limit and the sampled spore-bearing intervals are different kinds of evidence. They do not justify assigning every red bed, or every section at Marloes, Albion Sands and Lindsway Bay, one uniform age.
Upper boundary and neighbouring exposures. The regional upper boundary is drawn where pale quartz-rich sandstones of the Albion Sands Formation succeed the Red Cliff red mudstones. The well-known Albion Sands and Gateholm logs continue into younger units, but their conglomerate wedges, fish and plant occurrences and detailed measured beds belong to those named exposures. They are not transferred into the Marloes Sands fossil list.
Structure and exposure. The Marloes structural block lies between the Musselwick and Ritec faults. Faults disrupt and repeat the steep strata around the bay. Near the Three Chimneys, pre-Coralliferous tilting followed by later Variscan cleavage helps explain an apparently anomalous bedding–cleavage relationship. Bedding, sedimentary way-up structures and fault contacts must therefore be considered together when interpreting a particular rock rib.
References
Aldridge, R. J. and colleagues (2000). British Silurian Stratigraphy, Geological Conservation Review 19: Marloes accounts for the Llandovery and Wenlock successions. Detailed local beds, fossil assemblages and historical interpretations.
Barclay, W. J. and colleagues (2015). A lithostratigraphical framework for the Old Red Sandstone of the Anglo-Welsh Basin. British Geological Survey RR/14/02, especially the Red Cliff and Albion Sands formation accounts. Older Red Cliff age interpretation distinguished from the 2026 revision.
Davies, J. R. (2025). A systems-based lithostratigraphy for the Lower Palaeozoic Welsh Basin and adjoining region. British Geological Survey OR/25/018, p. 144 and hierarchy table p. 204: Marloes Group, component formations and member terminology.
Veevers, S. J., Ray, D. C., Ratcliffe, K. T. and Thomas, A. T. (2024). The application of chemostratigraphy and proximality trends to the Silurian Coralliferous Formation of SW Wales: rhythmical sedimentation during the transgression of a palaeo-shoreline. Journal of the Geological Society 181, jgs2024-009. Abstract-level correlation summary used here.
Hillier, R. D. and Higgs, K. T. (2026). Silurian (Gorstian) terrestrialisation records in the Red Cliff Formation of Pembrokeshire, UK: the oldest Old Red Sandstone in England and Wales. Proceedings of the Geologists’ Association 137, 101180. Published abstract used for the bounded age reassessment.
SAFETY
The beach can disappear at high tide. Check tide times and weather before leaving the car park, keep the return route in view and turn back well before rising water reaches headlands or rock ribs. Strong swell can make a low-tide visit unsafe even when the predicted tide looks suitable.
Keep clear of cliff faces, fresh falls and overhangs. Near-vertical layers, loose shale and weed-covered rocks create uneven footing, and the descent can be slippery. Supervise children closely around rock pools, cliffs and the water; the broad sandy beach does not remove the tidal hazards.
EQUIPMENT
Bring sturdy footwear with good grip, weatherproof clothing, drinking water, a hand lens and a camera or phone. A ruler or scale card helps make fossil photographs useful. If the landowner has permitted removal of loose specimens, take soft wrapping and small rigid boxes; fragile coral interiors and shell moulds can be damaged when packed together.
Plan a visual search rather than a hammering trip. Do not break the bedrock or prise fossils from the exposures. A notebook recording the part of the beach and the specimen’s original position is more useful than removing a fossil without context.
CLEANING AND TREATING
For any loose specimen collected with permission, begin with photographs and gentle dry brushing. The flaky mudstones and weathered fossil interiors may be much less robust than they look; avoid soaking, scrubbing or trying to split a specimen simply to expose more of it.
Do not use acid or household chemicals on the corals, shells or their matrix. Harder pieces may need careful work under magnification, but stop if the fossil and surrounding rock fracture together. Keep potentially significant material unprepared and seek specialist advice before applying consolidants or attempting detailed preparation.
IDENTIFY YOUR FINDS
Need help identifying a fossil? Share clear photos, where you found it and its size with the community.
FURTHER READING
Geological Conservation Review: Marloes volcanic and fossil-bearing succession
British Geological Survey: the 2025 Welsh Basin framework
Red Cliff Formation: 2026 research abstract
National Trust: Marloes Sands visitor information
ACCESS RIGHTS
Marloes Sands forms part of a protected coast and the National Trust provides the visitor parking and access information. Follow current notices and ask the landowner before removing fossils or rocks, including loose material. Access to the beach does not itself grant collecting permission.
Protect the fossil-bearing beds and leave attached specimens in place. Do not hammer the bedrock, excavate the cliffs or disturb stabilising vegetation. Any proposed scientific sampling should be arranged in advance with the landowner and, where required, Natural Resources Wales. Observation and photography allow the exposures to remain useful to other visitors and researchers.
It is important to follow our ‘Code of Conduct’ when collecting fossils or visiting any site. Please also read our ‘Terms and Conditions‘
LINKS
♦ Fossil Discussions
♦ Fossil Articles
♦ Buy Fossils, Tools and Equipment
♦ Buy Crystals, Meteorites, and Artefacts
♦ Join Fossil Hunts
♦ UK Fossils Network
