Upper Gilwern Quarry, near Howey in Powys, is known for its well-preserved Ordovician trilobites, from tiny growth stages to occasional complete specimens. Its fine-grained rocks record an ancient marine environment. This privately owned quarry is available to booked accommodation guests under the operator’s current access and collecting conditions.
FIND FREQUENCY: ♦♦♦♦♦ – Trilobite remains are abundant, particularly isolated heads and tails; complete specimens are less frequent.
CHILDREN: ♦♦♦ – Children need close adult supervision around loose rock, quarry faces and hammering. Confirm suitability with the operator.
ACCESS: ♦♦♦ – Private accommodation-guest access by arrangement; follow the allowance and conditions attached to your booking.
TYPE: – Inland quarry; Middle Ordovician mudstone, shale and fine-grained sandstone/siltstone.
DIRECTIONS
♦ Upper Gilwern Quarry is on the hills east of Howey, near Llandrindod Wells, Powys. The accommodation address is Upper Gilwern Quarry Hut, Howey, Llandrindod Wells, LD1 5PA. The postcode covers the accommodation locality; use the operator’s arrival instructions for the final approach.
♦ Approach Howey from Builth Wells to the south or Llandrindod Wells to the north, then take the road east towards Hundred House. The older approach description follows the steep climb and cattle grid onto open ground, then a left-hand single-track turning signed for Upper Gilwern. Confirm this final route against the directions supplied with your booking.
♦ The operator provides directions after booking and states that parking is available at the hut. Use the allocated parking and access route; do not drive through quarry gates or park on private tracks simply on the strength of an old photograph or map.
♦ Quarry reference: 52.214417, -3.3313867. This is a locality reference, not permission to enter. Research has used SO09095819 for the modern quarry.
FOSSIL HUNTING
Trilobites are the main attraction. The most characteristic is Ogyginus corndensis, a broad-bodied asaphid represented by large smooth tail shields, head shields, detached cheeks and occasional more complete bodies. The spelling corndensis is important: older visitor material sometimes gives “cordensis”. Tiny examples and separated body parts are worth keeping with good labels, since this locality’s growth series has been scientifically important.
Begin by looking carefully over loose material in the collecting area approved for your visit. Examine the surfaces and edges of grey-green to rusty-weathered slabs before splitting anything. Moulds may be subtle: a gently curved tail outline, paired eyes, repeated ribs or a fine pitted fringe can be more useful than a conspicuous dark fossil. If splitting is permitted, work gradually along natural bedding surfaces, leaving supporting rock around any fossil rather than trying to prise it out. Wear eye protection and keep other visitors beyond the range of flying chips.
Many finds are internal or external moulds, sometimes with a matching counterpart. A negative impression is part of the evidence, so inspect both halves and keep them together. The assemblage contains numerous isolated parts and juveniles; these can reflect moulting, breakage, transport and preservation as well as animals dying. A slab crowded with remains does not by itself demonstrate a single catastrophic event or an annual breeding colony.
Smaller trinucleid trilobites provide a useful contrast. Bettonolithus chamberlaini, formerly called Bettonia chamberlaini, has a distinctive pitted head fringe and a relatively small tail. Hughes illustrated an enrolled quarry specimen, BMNH It.2816, and a complete internal mould, It.2811. The type material of this species comes from nearby Howey Brook, so the quarry records should not be described as its type specimens.
Trinucleus abruptus is also securely recorded. Its head-fringe details need careful comparison, especially on worn or distorted moulds. The 1971 description includes the Upper Gilwern paratype BMNH It.2736. The holotype, In36920, had an old label naming another locality; Hughes inferred Upper Gilwern from its rock and comparison with known quarry specimens. That inferred origin is a useful distinction when discussing the scientific importance of the site.
The rarer Anebolithus record needs particular care. Hughes described quarry specimen It.2744 as an unnamed species and compared it with Anebolithus simplicior from Shelve; he did not identify the quarry specimen as that species, and even the generic assignment was qualified because of its preservation. A fragment with a pitted fringe is therefore not enough to label a find A. simplicior. Preserve fine fringe and spine details and seek specialist comparison for unusual trinucleids.
Other documented quarry fossils include brachiopods and graptolites, with rare sponges, asterozoans, the crinoid Iocrinus pauli and fragments of large arthropods reported by researchers. These are exceptional finds rather than a routine checklist for a visit. The naming of I. pauli in 1989 did not establish this quarry as its type locality: the original type was probably from near Hundred House.
Record which permitted area produced each find, and whether it was loose or found in place. Keep different horizons and loose scree finds separately. The lower quarry’s grey shale was described in a 2019 study as containing two fossiliferous beds, but that does not mean every shale layer is equally productive or that a loose fossil can be assigned to either bed. Photograph unusual articulated fossils before attempting more splitting or preparation, and retain the surrounding matrix.
Significant discoveries, geological research and conservation milestones at Upper Gilwern Quarry.
1971 – Published description of the quarry’s trinucleids
On 22 June, C. P. Hughes’s second monograph on the inlier described Trinucleus abruptus, including the securely attributed Upper Gilwern paratype BMNH It.2736. The holotype’s Upper Gilwern provenance was inferred from its rock and comparison with known material, correcting an older locality label. The same work figured quarry Bettonia chamberlaini, now Bettonolithus chamberlaini, and recorded an unnamed Anebolithus from specimen It.2744.
1979 – A detailed record of trilobite growth
Published on 27 September, Hughes’s third monograph used Upper Gilwern quarry-complex material to document the development and variation of Ogyginus corndensis. Its juvenile series, head and tail moulds and hypostomata made the small and incomplete fossils scientifically important alongside complete specimens. This was a developmental study, not the naming of a new quarry species.
1991–1993 – Commercial extraction and conservation assessment
A private company approached Powys County Council in December 1991, followed by a full planning application in April 1992 for commercial trilobite extraction. Kathy Kennedy’s 1993 account described the conservation assessment and proposed provisions for access, recording, curation and a final exposed face. Those proposals document the historical debate; they do not establish today’s access rights or prove that every proposed condition was implemented.
2002 – Upper Gilwern material in a trinucleid revision
Bowdler-Hicks’s revision included small Upper Gilwern morphometric samples in work on the trinucleids. The study supports the modern use of Bettonolithus for the genus previously called Bettonia. Its regional and Howey Brook specimens must be distinguished from the Upper Gilwern sample; the revision did not turn the quarry into the type locality of B. chamberlaini.
2019–2020 – Modern quarry logging and collecting study
Sam Parry began fieldwork at Upper Gilwern on 23 June 2019 and produced an undergraduate dissertation in 2020. It described lower, middle and upper exposures, two separate partial logs and a trilobite sample dominated by Ogyginus. The study adds useful modern observations while leaving detailed bed correlation and ecological interpretation subject to its sampling and preservation limits.
GEOLOGY
The quarry exposes fine-grained marine sedimentary rocks of the Camnant Mudstone Formation. The current researcher-led fieldtrip account places it in the artus graptolite Biozone, within the Darriwilian Stage of the Middle Ordovician. The British regional term Abereiddian, historically part of the lower Llanvirn, also occurs in geological accounts. These terms describe the age of the sedimentary succession; they do not make the six volcanic and graptolite units formerly listed for the wider hill a measured quarry section.
Mud, silt and fine sand accumulated in the Ordovician sea of the Welsh Borderland basin. The rocks now appear as shale or mudstone, uneven rusty-weathering siltstone and locally fine green sandstone. The BGS definition of the Camnant Mudstone Formation includes sandstone and conglomeratic sandstone interbeds becoming more frequent upwards, but this is the formation’s regional description, not a claim that all those rock types or its boundaries are exposed in the visitor quarry.
Volcanic activity was important in the surrounding district, and a modern quarry study described broken ashy, tuff-like material towards the top of one exposure. That observation does not prove that hot ash or poisonous volcanic fumes killed the trilobites. Nor do the abundance of juveniles and occasional complete specimens establish a fixed shallow-water breeding ground. The strongest evidence is the fossil preservation itself: abundant moulds, many separate body parts and a range of growth stages within fine-grained marine deposits.
The exposure is incomplete and has changed through working, weathering and slumping. Two partial sections logged in 2019 were reported as 2.53 m and 3.68 m, but their relationship is not resolved well enough to add them into a single thickness. Historical Upper Gilwern specimens came from small quarries in the same complex, close to the modern site; their old locality labels do not locate them in a particular bed of today’s face. The expandable account below separates the direct exposure descriptions from broader stratigraphic context.
Detailed geology, partial quarry sections and fossil evidence at Upper Gilwern.
Locality and historical provenance. The modern quarry is near SO09095819, east of Howey. Hughes’s 1971 and 1979 collections were attributed to small quarries east of Upper Gilwern at SO092582. Those records belong to the historical quarry complex, roughly 110 m east of the modern study coordinate; their precise modern face and bed cannot be recovered from the locality name alone. Howey Brook, Little Wern and other Gilwern Hill exposures are separate sections.
Camnant Mudstone Formation
Age, rank and limits. The preferred assignment is the Camnant Mudstone Formation, Middle Ordovician, Darriwilian, with the quarry placed in the artus Biozone by the current IGCP735 fieldtrip account. The BGS Lexicon gives the British regional age Abereiddian and specifies no parent unit. No formal member is established here. Its formation definition describes a lower boundary that is not seen and a gradational upper boundary; neither boundary is demonstrated in the visitor quarry. The Lexicon does not define a formation thickness.
Lithology and regional context. The formation consists of mudstone with sandstone and conglomeratic sandstone interbeds becoming more frequent upwards in its regional development. At Upper Gilwern, modern descriptions emphasise shale, rusty-weathering siltstone and fine green sandstone. Botting’s account describes a laterally restricted sandstone body with a tentative thickness of “perhaps 20 m”; that estimate must not be treated as a measured total for this quarry. The older district-wide Camnant thickness estimates likewise do not describe the exposed collecting section.
Lower quarry exposure: grey shale and two fossiliferous beds. Parry’s observations in 2019 describe grey shale containing two fossiliferous beds, followed upwards by uneven, rusty-weathering siltstone and then soil or slumped material. Fossils were also found in loose siltstone. Figure 13 is captioned as a 2.53 m partial lower section. The two fossiliferous beds are not formally named, and separate bed thicknesses and precise fossil positions cannot be established from the accessible description. A loose nodule from this area, specimen 168, contains incomplete biserial graptolites tentatively identified as ?Diplograptus ellesi; it cannot securely date a particular bed.
Middle quarry exposure: siltstone, shale and ashy material. A thick basal siltstone was described as continuing the lower exposure at a higher elevation. Shale occurs above it, with broken ashy, tuff-like material towards the soil cover. Fossiliferous horizons were inspected where exposure and access permitted. Figure 14 is captioned as a 3.68 m partial middle section. These lithologies are field descriptions, not formal members or demonstrated correlations with named regional tuffs. Their individual thicknesses and contacts are not resolved here.
Upper hillside pits and surface cover. The upper collecting pits in the same study yielded mostly loose material, with at most 0.5 m of bed exposed at the top. Soil, slumped shale and broken surface rock obscure the succession and can mix fossils from different positions. They should not be drawn as an additional named bed or used to place an unlabelled loose fossil in a stratigraphic sequence.
What the measured figures do and do not establish. The lower 2.53 m and middle 3.68 m values are two partial logs, not a verified composite; they must not be added to make 6.21 m. The full log images and their appendix key were not available for checking individual boundaries, so the chart deliberately summarises the separate exposures without a bed-by-bed scale. The study’s lateral quarry span and elevation range describe geometry rather than bed-normal thickness. The 10 m final face mentioned in the 1993 conservation account was a proposed outcome of quarry working, not a measured present-day section.
Historical terminology and unresolved correlation. Older papers used lower Llanvirn and Didymograptus speciosus terminology for Upper Gilwern material, while Bowdler-Hicks’s 2002 revision placed its samples in the Builth Volcanic Formation and a mid-murchisoni interval. Modern descriptions instead use Camnant and artus. Accounts also differ over whether the quarry lies low or high within Camnant. These schemes should not be silently combined into a precise position for historical specimens. The present guide uses the modern formation and biozone assignment without claiming that the old samples can be correlated to an individual modern bed.
Trilobite growth and preservation. Hughes’s 1979 study reconstructed an almost complete developmental sequence of Ogyginus corndensis, principally from this quarry complex. The figured material includes very small transitory tail shields, juvenile meraspid stages and larger holaspid individuals, together with detached cheeks and hypostomata, the plates beneath the head. For example, BMNH It.2873 represents meraspid degree 4, It.2871 degree 7, and It.2847 the smallest certain holaspis in the account. The historical BMNH collections are now at the Natural History Museum. These are research specimens, not the species’ type material: its holotype came from near Middleton in Shropshire.
Variation, deformation and sample limits. The 1979 work tabulated ten explicitly quarry-marked juvenile specimens, compared successive growth stages and measured hypostomata including NMW 68.376.G208 and G209. Its bivariate comparisons use different, overlapping samples; their counts are not separate populations to add together. Seven-, eight- and nine-rib tail shields occur in the regional species, and Hughes did not recognise formal subspecies simply from rib number. Growth stage and tectonic distortion both affect shape, so a differently proportioned mould need not represent a new species.
The 2019 collecting sample. Parry’s undergraduate study recorded 202 trilobites from 168 fossil-bearing Upper Gilwern rock specimens: 193 assigned to Ogyginus, eight to Bettonolithus and one indeterminate trinucleid. It combined selected, identifiable in-situ and loose material from several exposures. Those numbers describe that collecting sample, not the living community or the probability of a visitor finding each species. Inconsistent denominators in its preservation figures are another reason not to turn the reported percentages into a precise ecological census. The poorly preserved trinucleid GL45 was compared tentatively with ?Protolloydolithus ramsayi, not securely identified as that species.
Deposition and interpretation. Fine sediment accumulated in a marine basin affected by regional volcanism. Mould preservation, disarticulated parts and juveniles provide evidence to investigate burial, moulting, breakage and habitat, but they do not demonstrate annual breeding returns or death from burning ash and poisonous fumes. A tentative hydrothermal interpretation on a researcher’s website depended partly on ferruginous bodies no longer available for re-examination. It remains a hypothesis, rather than an established origin for the fossil concentrations.
Rare fossils and provenance. Modern research accounts record Iocrinus pauli, asterozoans, sponges and large arthropod fragments at Gilwern Hill Quarry, without published accession numbers for all recent material. These occurrences add to the locality’s interest but do not make it the type locality of I. pauli. Its 1989 type specimen BMNH E71413 was probably from near Hundred House, while E71424 came from a different recorded locality. The diverse soft-bodied fossils of Castle Bank and the type-locality faunas of Howey Brook must likewise remain separate from the quarry record.
References. British Geological Survey, Lexicon entry for the Camnant Mudstone Formation; Hughes, C. P. (1971), The Ordovician trilobite faunas of the Builth–Llandrindod Inlier, central Wales, Part II, Bulletin of the British Museum (Natural History), Geology 20(4), 115–182; Hughes, C. P. (1979), Part III, Geology 32(3), 109–181; Bowdler-Hicks, A. (2002), The taxonomy, phylogeny and biogeography of the Marrolithinae (Trinucleidae, Trilobita), PhD thesis, University of Glasgow; Parry, S. (2020), A study of the trilobite assemblages of the Middle Ordovician Quarries Upper Gilwern and Little Wern in central Wales, University of Portsmouth undergraduate dissertation; Kennedy, K. (1993), “Fossils for profit – the Gilwern Hill controversy”, Earth Science Conservation 32, 27; Donovan and Gale (1989), description of Iocrinus pauli; IGCP735 Builth–Llandrindod fieldtrip account; Botting and Muir, regional stratigraphy research pages. Links to the principal sources are listed under Further Reading.
SAFETY
Quarry conditions change. Loose blocks, steep worked faces, uneven scree and unstable edges require care even where collecting is permitted. Stay within the area currently allowed by the operator, obey closures, keep back from faces and edges, and never undercut a face or work directly beneath another person. Do not climb a face or assume that an old photograph shows its present condition.
Wear sturdy footwear and safety goggles whenever splitting rock; keep hands clear of the striking area and other people away from flying chips. Children require close supervision by their own adult. Stop if safe spacing cannot be maintained or rock movement is apparent.
This is an exposed, remote hill setting. Bring suitable layers and waterproofs, drinking water, sun protection when needed and a charged phone. Confirm current arrival and site instructions before travelling. Do not rely on the old guide’s description of sloping faces as an assurance that any particular area is safe.
EQUIPMENT
A hand lens, camera, notebook, labels, soft wrapping and a rigid box are as useful as a hammer. Bring safety goggles and sturdy boots. Use a geological hammer and a bolster or splitting chisel only where the operator permits those tools and methods; work on loose approved material rather than attacking the face.
Carry small finds and delicate spines in separate padded containers. Number the two halves of a split fossil together, adding “part” and “counterpart”, and keep any detached fragments with the same record. A bucket alone gives little protection against other rocks rubbing across a fine mould.
CLEANING AND TREATING
Keep preparation minimal at first. Photograph the specimen and both halves of a split slab, record its collecting area and retain a generous margin of matrix. A thin trilobite mould belongs with its supporting slab; attempting to free it can destroy the very surface needed for identification.
Under magnification, remove only loose dust with a soft dry brush. Do not assume that every shale will withstand water: try any proposed cleaning on an unimportant piece of the same matrix first. Avoid prolonged soaking, acids, bleach, wire brushes and rotary polishing. These conservative precautions follow the fine-grained matrix and mould preservation here; they are not a tested treatment formula for every Upper Gilwern specimen.
Stop if the matrix starts flaking, a crack crosses the fossil or a spine disappears beneath the rock. Fine mechanical preparation and any consolidation should be assessed by an experienced preparator with appropriate magnification and dust control. Do not cover fossils indiscriminately with household glue, varnish or oil, and do not assume that a recipe used on material from another quarry is suitable here.
Keep a permanent locality label, field number and treatment record with each specimen. An unusual articulated echinoderm, sponge or large arthropod fragment may be much more informative before extensive preparation. Seek museum or specialist advice and retain photographs and provenance; accurate records make even small or incomplete finds useful.
IDENTIFY YOUR FINDS
Need help identifying a fossil? Share clear photos, where you found it and its size with the community.
FURTHER READING
Deposits: The disparids, weird and weedy crinoids of the Palaeozoic (general crinoid background)
Hughes (1971): The Ordovician trilobite faunas of the Builth–Llandrindod Inlier, Part II
Hughes (1979): The Ordovician trilobite faunas of the Builth–Llandrindod Inlier, Part III
Parry (2020): Upper Gilwern and Little Wern trilobite assemblages (undergraduate dissertation)
BGS Lexicon: Camnant Mudstone Formation
IGCP735: Builth–Llandrindod Inlier fieldtrip, including Gilwern Hill Quarry
Kennedy (1993): Fossils for profit, the Gilwern Hill controversy (page 27)
Botting and Muir: Stratigraphy of the Builth–Llandrindod Inlier
ACCESS RIGHTS
This is private land. The current operator offers quarry access as part of booked accommodation arrangements. Quarry Hut guests and Rhiw Wood Treehouse guests have different access allowances, so confirm the hours, collecting areas, tools and methods that apply to your own booking. An accommodation booking does not imply permission for unrestricted excavation, entry into closed areas or visits by additional unbooked people.
Check the official Upper Gilwern Quarry Hut website and its booking information before arranging a visit. The original contact address is uppergilwernquarryhut@hotmail.com. Follow the operator’s latest instructions rather than historical prices or access descriptions. A nearby public route does not provide permission to collect across private ground.
It is important to follow our ‘Code of Conduct’ when collecting fossils or visiting any site. Please also read our ‘Terms and Conditions‘
Craig Chapman. Last inspected 2016, Updated 2017, Edited by Jon Trevelyan 2016. Our gratitude also goes to Pete Lawrance for providing much of the information about this site.
LINKS
♦ Fossil Discussions
♦ Fossil Articles
♦ Buy Fossils, Tools and Equipment
♦ Buy Crystals, Meteorites, and Artefacts
♦ Join Fossil Hunts
♦ UK Fossils Network
