Content and licensingview original scan buy a printed copy

Geology of the South Wales Coalfield, Part III, the country around Cardiff Memoir for 1:50 000 geological sheet 263 (England and Wales)

By R. A. Waters and D. J. D. Lawrence

Bibliographic reference: Waters, R. A. and Lawrence, D. J. D. 1987. Geology of the South Wales Coalfield, Part III, the country around Cardiff. 3rd edition. Mem. Br. Geol. Surv. , Sheet 263 (England and Wales).

British Geological Survey Natural Environment Research Council

Geology of the South Wales Coalfield, Part III, the country around Cardiff Memoir for 1:50 000 geological sheet 263 (England and Wales)

• R. A. Waters and D. J. D. Lawrence
• Third Edition Contributors:
• Palaeontology H. C. Ivimey-Cook M. Mitchell G. Warrington D. E. White
• Water supply M. A. Lewis
• Authors
• R. A. Waters, BSc, PhD British Geological Survey, Aberystwyth
• D. J. D. Lawrence, BSc British Geological Survey, Newcastle
• Contributors
• H. C. Ivimey-Cook, BSc, PhD, G. Warrington, BSc, PhD and D. E. White, MSc, PhD British Geological Survey, Keyworth
• M. Mitchell, MA Formerly BGS
• M. A. Lewis, BA British Geological Survey, Wallingford

London: Her Majesty's Stationery Office 1987 © Crown copyright 1987

First published 1902 Second edition 1912 Third edition 1987 Printed in the United Kingdom for HMSO. ISBN 0 11 884402 4

Other publications of the Survey dealing with this district and adjoining districts

Books

• British Regional Geology
• South Wales, 3rd edition, 1970
• Mineral Assessment Reports
• No. 141: The hard rock resources of the country around Caerphilly, South Wales: description of parts of 1:50 000 Sheets 249 and 263, 1984
• One-inch and 1:50 000 memoirs
• Pontypridd (248), 3rd edition, 1964
• Newport (249), 3rd edition, 1969
• Monmouth and Chepstow (233 and 250), 1961
• Bridgend (262), 1904
• Bristol (Bristol special sheet), The Lower Jurassic rocks,
• 1984
• Weston-super-Mare (279), 1983
• Wells and Cheddar (280), 1965

Maps

• 1:625 000
• Geological map of the United Kingdom (solid geology), South sheet, 3rd edition, 1979
• Quaternary map of the United Kingdom, South sheet, 1977
• Aeromagnetic map of Great Britain, South sheet, 1965
• 1:250 000
• Sea bed sediments and Quaternary geology, Bristol Channel, 1986
• Solid geology, Bristol Channel (in press)
• Aeromagnetic anomaly, Bristol Channel, 1980 Bouguer gravity anomaly, Bristol Channel, 1986
• 1:50 000 or 1:63 360
• Pontypridd (248) Solid 1963; Drift 1960, reprinted 1975
• Newport (249) Solid 1968, reprinted 1975; Drift 1969
• Chepstow (250) Solid and Drift 1958, reprinted 1981
• Bridgend (262) Solid and Drift 1901, reprinted 1974
• Cardiff (263) Solid 1986; Drift 1897, reprinted 1969
• Bristol (264) Solid and Drift 1963, reprinted 1974
• Weston-super-Mare (279) Solid and Drift, 1980
• Wells (280) Solid and Drift 1963, reprinted 1984

Notes

• In this memoir the word 'district' means the Welsh part of the area, excluding the island of Flat Holm, included in 1:50 000 geological sheet 263 (Cardiff).
• Figures in square brackets are National Grid references within 100 km square ST.
• Numbers preceded by the letter A refer to the Survey's collection of photographs.
• The authorship of fossil species is given in the index of fossils.

List of six-inch and 1:10000 maps

The following is a list of the six-inch and 1:10 000 geological maps included, wholly or in part, in the Cardiff Sheet with the date of survey of each map. The surveying officers are: Mr G. W. Green, Dr B. Kelk, Mr D. J. D. Lawrence, Mr K. Taylor and Dr R. A. Waters.

Preface to the First edition

This volume, which forms the third part of the Memoir descriptive of the geology of the South Wales Coalfield, illustrates the country around the principal town connected with the coalfield and the chief outlet for its products. The map (Sheet 263) of which it forms the explanation was originally surveyed about the year 1840 by Sir H. T. De la Beche with the assistance of W. T. Aveline; but additions, chiefly in the Secondary Rocks, were made in 1872 by H. W. Bristow and H. B. Woodward. The re-survey on the six-inch scale was carried out during the years 1892–6 under the superintendence of Mr Strahan, who himself surveyed the greater part of the sheet, while Mr Cantrill was engaged upon the western part of it. The following pages have been written mainly by Mr Strahan, but Mr Cantrill has supplied the descriptions of the area surveyed by himself.

The oldest strata, consisting of Ludlow and Wenlock rocks, come to the surface near Cardiff in the axis of a great anticline of pre-Triassic age. Their existence was first detected by the Rev. Norman Glass in 1861, but it was not until 1879 that their extent and age were placed beyond doubt by Professor W. J. Sollas. Much information concerning them was obtained at a later date also by Mr John Storrie. The Old Red Sandstone presents the same general features as in Monmouthshire, and the precise position of the boundary between Upper and Lower Old Red Sandstone still remains doubtful.

The Carboniferous Limestone is concealed for the most part by later rocks, but it is evidently far thicker on the southern than on the northern side of the coalfield, as is the case with all the sub-divisions of the Carboniferous system.

In the examination of the Rhaetic Beds and Lias the surveyors have availed themselves of the detailed work of H. W. Bristow, R. Etheridge and H. B. Woodward, and of the additions made to it by Mr John Storrie and Mr F. T. Howard.

Superficial Geology is well illustrated in the neighbourhood described in these pages. The raised beach so well known around the shores of the Bristol Channel, exists near Weston-super-Mare, and the suggestive observations of Mr E. C. H. Day as to the age of the bones contained in it have received confirmation from recent observations in Cower, where the beach and its associated 'head' are seen to be of earlier date than the local Glacial drift. Recent work on the Glacial deposits confirms Professor Edgeworth David's conclusions made in 1883.

Post-Glacial deposits were admirably displayed in the excavations of the Barry Docks, where conclusive evidence was obtained of a subsidence of the land upwards of 50 feet during and since Neolithic times. In this investigation also, Mr Storrie gave valuable assistance.

Chapter X contained a list of well-sections, largely made up from Mr Storrie's information. To him also are due some corrections made in the published account of wells sunk under his advice. Chapter XI deals with the more important economic products.

Cardiff being the principal town in South Wales, the Geological Bibliography of South Wales has been inserted in this part of the memoir. It is founded on a list drawn up by Mr W. Whitaker and on other sources, but it has been revised and completed by Mr Cantrill.

The Map is issued in two editions. On the Edition for Solid Geology the Glacial Drift is omitted, while on the Drift Edition, the areas occupied by Drift are coloured as well as those portions of Solid Geology which are not concealed by Drift. MS. six-inch Maps geologically coloured, are deposited in the Office where they can be consulted. Copies of these Maps can be obtained at cost-price.

J. J. H. Teall Director, Geological Survey Office 28 Jermyn Street, London 21 April 1902

Preface to the Second edition

The exhaustion of the First Edition of this Memoir gives the opportunity of making the additions and corrections rendered necessary by the work of several observers. Moreover, the six-inch survey of the western part of South Wales having been completed since the issue of the First Edition, it is now possible to speak definitely on some general questions which were then left open. Thus, the unfossiliferous Brownstone sub-division is now included in the Upper Old Red Sandstone on the strength of evidence obtained near Carmarthen.

The Carboniferous Limestone of the Cardiff district has not yet been zoned, but some observations by Professor G. Delepine in the ground further west suggest that the complete Avonian sequence is there represented. On the other hand, Mr Dixon, in the ground further east, found reason to believe that the upper zones are missing. In Somerset the palaeontological sequence has been investigated by Professors Lloyd Morgan and Reynolds, and in greater detail by Dr Sibly, with the result of proving the existence of a great repeating fault along the Worlebury ridge. The last-named has also shown that the volcanic episodes mainifested in the two headlands of Worlebury and Middle Hope respectively were not strictly simultaneous. The characters of the igneous rocks have been described by Professor W. S. Boulton.

In the Rhaetic Beds and the strata associated with them much has been added to our knowledge of the palaeontology by the detailed work of Mr Richardson. Some precursors of the Rhaetic fauna have been found by him in the uppermost bands of the Keuper Marl.

Several new wells near Cardiff have been described in Chapter XI, and the Geological Bibliography relating to that part of South Wales which has been surveyed on the six-inch scale has been brought up to date.

Such re-examination of the ground as was necessary for the purposes of this Edition was made by Dr Strahan and Mr Cantrill.

J. ,J. H. Teall Director, Geological Surrey Office 28 Jermyn Street, London 8 December 1911

Preface to the Third edition

The area covered by the Cardiff (263) Sheet of the 1:50 000 geological map of England and Wales was originally surveyed on the six-inch scale by Sir Aubrey Strahan and T. C. Cantrill and was published on the one-inch scale as Sheet 263 in two editions, solid and drift, in 1897. An explanatory memoir written by the two surveyors was published in 1902, with a second edition incorporating some revision, following in 1912. Since the publication of the second edition there have been considerable advances in the understanding of South Wales stratigraphy. Because of this and the fact that the district includes several expanding urban areas a complete resurvey has been undertaken. It was mostly resurveyed in 1976–80 by Mr D. J. D. Lawrence, Mr K. Taylor and Dr R. A. Waters under the supervision of Dr D. B. Smith as District Geologist; the Avon district and the island of Flat Holm were resurveyed in 1967, 1972 and 1975 by Mr G. W. Green and Dr B. Kelk as part of the survey of the Weston-super-Mare (279) Sheet, published in 1980 with an accompanying memoir in 1983.

The third edition of this memoir has been written by Dr R. A. Waters and Mr D. J. D. Lawrence with contributions on the palaeontology by Dr H. C. Ivimey-Cook, Mr M. Mitchell, Dr G. Warrington and Dr D. E. White and on the water supply by Mrs M. A. Lewis; it was compiled by Dr R. A. Waters and edited by Dr R. A. B. Bazley, Regional Geologist.

The Silurian Shelly macrofaunas were identified by Dr D. E. White; Devonian palynomorphs by Dr B. Owens; Lower Carboniferous macro-faunas, and foraminifera and algae by Mr M. Mitchell and Dr A. R. E. Strank respectively; Namurian faunas by Dr W. H. C. Ramsbottom; Triassic and Jurassic macrofaunas and palynomorphs by Drs H. C. Ivimey-Cook and G. Warrington respectively. We are grateful to Dr K. Dorning of Pallab Research and Dr R. B. Rickards of Cambridge University for the identification of the Silurian palynomorphs and graptolites, respectively, Miss S. V. Young of the British Museum (Natural History) for those of the Devonian fish, Professor G. D. Sevastopulo of Trinity College Dublin for those of the Lower Carboniferous conodonts and Professor D. T. Donovan of University College, London for those of Lower Jurassic ammonites. Petrological descriptions were provided by Messrs R. K. Harrison and G. E. Strong; clay mineralogy determinations were carried out by Mr R. J. Merriman. The photographs were taken by Mr C. J. Jeffrey.

We thank Dr M. G. Bassett and the National Museum of Wales for unpublished data on the Ludlow and Wenlock Series and for the provision of facilities fin- logging the BGS Rurnney Borehole, and Professor J. G. C. Anderson for unpublished borehole logs. We also acknowledge the information and assistance generously provided by officials of South Glamorgan County Council and many private site investigation contractors. The willing co-operation of local landowners and quarry operators at the time of the resurvey is gratefully appreciated.

G. Innes Lumsden, FRSE, Director, British Geological Survey Keyworth, Nottingham NG12 5GG 17th August 1987

Summary of geology

This memoir describes the district around the City of Cardiff and Barry situated beyond the southern margin of the South Wales Coalfield. The variable geology encompasses gently folded Silurian, Devonian and Carboniferous rocks, now largely seen in inliers exposed through an unconformable cover of late Triassic and Lower Jurassic strata. These younger rocks and the basal unconformity are spectacularly displayed in classic coastal sections between Barry and Penarth.

It presents a description of the geological history of the district, which can be traced back some 420 million years, through an examination of the stratigraphy, palaeontology and depositional environments of the various rock sequences. Much new information from several boreholes drilled during the resurvey is provided. The deposits of the last glaciation, well developed in the north of the district, are also described together with the clays and peats flanking the Severn estuary north of Penarth, the products of the subsequent rise in sea level.

Further chapters deal with the tectonic evolution and economic geology, including the water supply, of the district.

(P945681) Geological sequence

Geological sequence

Chapter 1 Introduction

Area and location

This memoir describes the geology of the Welsh part of the Cardiff (263) Sheet of the 1:50 000 Geological New Series map of England and Wales; the English part of the sheet and the island of Flat Holm are described in the memoir for the Weston-super-Mare (279) Sheet (Whittaker and Green, 1983).

The district occupies the eastern part of the Vale of Glamorgan, which comprises strongly dissected lowland, reaching above 130 m only in the north where it rises towards the upland area of the South Wales Coalfield. Much of the coastline bordering the Bristol Channel is dominated by high cliffs which provide excellent sections of the Mesozoic succession, but north-east of Penarth there is a wide coastal plain. The city of Cardiff is situated on the widest part of this plain where the rivers Taff, Ely and Rhymney enter the sea. Inland the best exposures in the varied rock sequence are mostly in old or working quarries. Several boreholes were drilled during the course of the survey to clarify poorly exposed parts of the stratigraphic sequence.

Geological history

Throughout the geological history of the Vale of Glamorgan two major basement structures repeatedly played a major role. One is the north–south Usk Axis, situated east of the district, and the other trends east–west roughly coincident with the axial trace of the Cardiff–Cowbridge Anticline. The nature of this axis, here termed the Vale of Glamorgan Axis, is as yet uncertain.

The oldest exposed rocks show that in the mid-Silurian (Wenlock) the district was situated on the southern margin of the Lower Palaeozoic Welsh Basin in a shallow marine shelf setting with land probably not far to the south. Wenlock mid-shelf mudstones and subordinate thin sublittoral sheet sandstones, siltstones and limestones (Pen-y-Lan Mudstone) are followed by a major shoreline regression which led to the establishment of an inner shelf facies of thin sandstones and mudstones with a restricted marine fauna (Cae Castell Formation). A major subtidal sand bar (Rhymney Grit) is developed at the base. At the beginning of Ludlow times a widespread transgression saw a rapid return to mid-shelf conditions and the deposition of further sublittoral sheet sandstones, siltstones and limestones (Cardiff Group) in a mud-dominated environment. Towards the top of the Group, shoreline progradation produced an upward transition from mid- to proximal-shelf sediments, and eventually into Old Red Sandstone facies at the end of Ludlow times.

In the late Silurian (Přídolí), red mudstones with calcretes and dominantly fluvial sandstones (Raglan Mudstone Formation) were deposited on alluvial and coastal mudflats.

During the Lower Devonian southward flowing rivers deposited alluvial mudstones and sandstones in an upward-coarsening sequence. This sequence, beginning with mudstones with associated calcretes and sandstones (St Maughans Formation), and ending with sandstones (Brownstones), reflects increasing Caledonian late orogenic uplift to the north, and southward migration of the more proximal alluvial facies. However, localised tectonic uplift, probably in the Bristol Channel area, interrupted this southward progradation to give rise to a wedge of coarse southerly-derived alluvial fan conglomerates, mudstones and sandstones (Llanishen Conglomerate).

Late Caledonian regional uplift is responsible for the lack of Middle Devonian strata, whilst local faulting and tilting along the Vale of Glamorgan Axis followed by erosion is responsible for the Upper Devonian, Upper Old Red Sandstone overstep in the district. During the Upper Devonian, southward flowing rivers were at first characterised by high sinuosity channels depositing pebbly sandstones and mudstones with associated calcretes (Cwrt-yr-ala Formation), and later by low sinuosity channels depositing quartz conglomerates and sandstones (Quartz Conglomerate Group).

By earliest Dinantian (Courceyan) times a marine transgression began to invade the coastal plain. Cyclical siliciclastic and carbonate, offshore, barrier, lagoonal and restricted bay sediments at the base of the Tongwynlais Formation demonstrate that the transgression was pulsed. However, a shallow marine shelf was eventually established on which mudstones with storm-generated carbonate sheet sands were deposited (Tongwynlais Formation and Cwmyniscoy Mudstone). During this period a regressive interlude saw the brief appearance of an oolitic and skeletal barrier complex (Castell Coch Limestone). The establishment of a carbonate-dominated shelf took place as a result of the next shoreline regression which led to the deposition of proximal shelf, storm-generated skeletal limestones (Barry Harbour Limestone). The acme of this regression is manifested in the north by the Brofiscin Oolite shoal. Renewed transgression saw a rapid transition to mid-shelf, storm-generated, argillaccous, bioclastic limestones (Friars Point Limestone). A major regression at the top of the Friars Point Limestone led to emergence in the north of the district and the establishment of a prograding oolite shoal complex (Gully Oolite) in the south. The shoals later migrated northwards during a further transgression. However, due to pro-gradation they eventually became emergent over the northern half of the district, leading to the establishment of a widespread karst and palaeosol. During the deposition of the Friars Point Limestone and the Gully Oolite the Vale of Glamorgan Axis acted as a hinge, giving rise to much thicker sequences to the south of it. Renewed transgression saw first the establishment of a peritidal environment (Caswell Bay Mudstone), and then proximal and mid-shelf bioclastic limestones (High Tor Limestone). The latter eventually coarsened upwards into a 'regressive' oolite shoal (Cefnyrhendy Oolite) that became emergent, and is capped by a palaeokarst. Further transgression saw the re-establishment of oolite shoals forming a major barrier complex interfingering with peritidal deposits (Hunts Bay Oolite), for by this time the shelf had graduated to a carbonate platform. Sedimentation ceased in the late Visean due to uplift along the Usk Axis, and did not recommence until a late-Narnurian transgression when marine shales were deposited. Although Westphalian sediments are not exposed in the district they were undoubtedly deposited but have been removed by post-Variscan erosion.

Variscan (end Carboniferous) compressional deformation resulted in open east–west-trending folds, of which the Cardiff– Cowbridge Anticline is the largest in the district. It developed along the line of the Vale of Glamorgan Axis. Other Variscan structures include NW–SE cross-faults and local thrusts. A later tensional phase related to the opening of the Atlantic led to the initiation of the Bristol Channel basin offshore.

During much of the Permian and Triassic periods the district was deeply dissected by erosion and only in late Triassic time did sediments begin to accumulate against the Palaeozoic uplands. Red mudstones with evaporites, and later green and grey mudstones with evaporites and thin carbonates (Mercia Mudstonc Group), both of lacustrine origin, gradually successively onlapped onto an irregular topography. Marginal to the lake, coarse shore-zone sediments accumulated, whilst coarse scree and alluvial fan sediments (marginal facies) were deposited in subaerial situations.

A major transgression in Rhaetian times established marine conditions over most of the area depositing dark grey and grey shallow marine shales with thin sandstones and limestones (Penarth Group). A lagoonal interlude occurs within the upper part of the Penarth Group. Continuing transgression into the early Jurassic saw the establishment of an epeiric sea over much of southern Britain. In the district calcilutites and mudstones (Blue Lias) accumulated in an offshore environment; a mudstone-dominated formation (Lavernock Shales) occurs in the lower part of the sequence. The upper part coarsens upwards and oolitic, peloidal and bioclastic limestones (marginal facies) developed over the Vale of Glamorgan Axis as a response to fault movements; renewed transgression followed. There are no post-Lower Jurassic solid rocks in the district, but the presence of later Mesozoic and Tertiary strata in the Bristol Channel basin suggests that such deposits were probably deposited in the district but have since been stripped off. Post-Lower Jurassic faulting is widespread and may be Alpine in age.

Evidence for pre-Devensian Quaternary glaciation in the district is enigmatic. However, in the Devensian an ice-sheet, generated in the upland areas to the north of the district, flowed southwards out of the South Wales Coalfield into the Vale of Glamorgan, depositing widespread hummocky till and sand and gravel. Beyond the southern edge of the ice-sheet fluvioglacial gravels were deposited, especially in the river valleys. Following the melting of the ice and the resultant sea-level rise, clays, (estuarine alluvium) were deposited in the major river estuaries and, along the low-lying coastal plain bordering the Bristol Channel.

Chapter 2 Silurian

Silurian rocks are the oldest present at outcrop in the district and occupy much of the higher ground north of central Cardiff. They comprise about 455 m of mudstones, sandstones and limestones, deposited on a shallow marine, muddy shelf, and belonging to the Wenlock and Ludlow Series, overlain by about 300 m of red mudstones with subordinate sandstones and calcrctes, deposited on tidal and alluvial flats, that belong to the Lower Old Red Sandstone magnafacies and are referable largely to the Přídolí Series (formerly Downton Series). The Silurian/Devonian boundary in South Wales and the Borders is generally regarded as lying somewhere within the middle or upper part of the Ledbury Formation (Holland and Richardson, 1977; Allen and Williams, 1981), which is equivalent to the Raglan Mudstone Formation, the basal Old Red Sandstone unit in the district (House and others, 1977).

The Wenlock and Ludlow Series crop out in an inlier centred on the suburbs of Rumney and Pen-y-Lan and are poorly exposed. The oldest rocks were shown to be late Wenlock by Bassett (1969). Reference to the Wenlock faunal communities represented within the inlier is made by Hurst and others (1978), and plant material has been described by Harris (1884), Storrie (1892) and Barber (1892). Shallow drilling, including that of Anderson and Blundell (1965), has shown that the inlier continues in a south-eastward direction at least as far as the coast as an exhumed Triassic ridge flanked on either side by the Mercia Mudstone Group. A reexamination of Anderson's (1968, 1974) shallow borehole material, curated in the National Museum of Wales, has resulted in the discovery of two further isolated inliers; beneath the Taff flood-plain at Pontcanna [ST 165 777], and beneath estuarine alluvium on the Wentlooge Level at Sluice Farm [ST 2538 7917]. Samples processed for palynomorphs proved barren and these rocks are, therefore, depicted on the map as 'Ludlow and Wenlock Series, undivided'. The strata are also known beneath thick Triassic cover to the south of the main inlier in several water bores in the industrial and clock area of the city (Strahan and Cantrill, 1912; North, 1915a and b).

The main inlier was first discovered by Glass (1861), and the first detailed account of the succession was by Sollas (1879); Strahan and Cantrill (1902, 1912) largely followed Sollas's account. Additional information on the succession has come from temporary sections and boreholes (Storrie, 1879, 1908; North 1915a and b; Anderson and Blundell, 1965). Both Storrie (1879) and Sollas (1879) considered the Ludlow Series/Old Red Sandstone junction to be conformable due to the presence of a bone bed at the top of the Ludlow Series, whilst North (1915b) and Heard and Davies (1924) suggested that there was a non-sequence. The lack of evidence for 'Leintwardinian' or 'Whitcliffian' faunas in the main inlier (Allender in discussion of Warmsley, 1959) led Walmsley (1962) to suggest that either there was an unconformity or that there was an early development of red beds

(Old Red Sandstone) in the Ludlow Series. However, the conformable nature of the junction was demonstrated by Rumney Borehole (Waters and White, 1980) drilled during the resurvey, which provided the first complete succession from the Wenlock Series through to the Old Red Sandstone.

The lithological classifications used in the past and in this memoir are illustrated in (Figure 1). The Wenlock Series comprises two newly named formations, the Pen-y-Lan Mudstone and the Cae Castell Formation. The Pen-y-Lan Mudstone is the age-equivalent of the upper part of the 'Wenlock Shale' of the Welsh Borderland, and is characterised by the presence of thin sheet-sandstones. The Cae Castell Formation is a sandstone-dominant age-equivalent of the 'Wenlock Limestone' of the Welsh Borderland. The Ludlow Series comprises the Cardiff Group which is divided into two formations, namely Hill Gardens Formation and Llanedeyrn Formation; the latter is further divided into three members (Eastern Avenue Member, Chapel Wood Member and Roath Park lake Member), on the basis of the relative amounts of mudstone, sandstone and limestone, and of the style of the bedding. The classification is strictly lithostratigraphical, but in both the Wenlock and Ludlow Series, brachiopods, graptolites and palynomorphs have been used to recognise the standard chronostratigraphical divisions (Bassett and others, 1975; Holland, 1980) (Figure 2) and also to correlate with the lithostratigraphic divisions (Holland and others, 1980) of the type-area of each series.

In the Přídolí Series diagnostic fossils are absent in the district, and regional correlations are based on lithology.

Pen-y-Lan Mudstone

The Pen-y-Lan Mudstone is a new term for the richly fossiliferous mudstones with scattered thin beds of sandstone and impure limestone that occur beneath the Rhymney Grit. The base of the formation has not been proved. The designated type-locality is a cutting [ST 1906 7875]–[ST 1923 7876] on the northern side of Eastern Avenue at Pen-y-Lan. The formation crops out as an east–west strip between Roath Park and Rumney; it is poorly exposed. Assuming that there are no structural complications, the total thickness exposed is 225 m at maximum.

Mudstones account for 80–90 per cent of the formation. They are grey when fresh, but weather grey-green, olive-green and buff, and are locally affected by Triassic staining. They are variably calcareous and silty, in parts grading into siltstone. The mica content is variable.

Two mudstone types can be differentiated but there are all gradations between. The first and most prevalent occurs in massive, poorly defined but commonly blocky beds, 0.1 to 0.45 m thick. It is characterised by intense burrowing, mainly dominated by Chondrites; bedding within the mudstones is locally preserved as thin wisps and streaks of siltstone, whilst shelly fossils, commonly broken, are scattered randomly throughout in wisps and clots. The second occurs in thinner blocky beds 2 to 7 cm thick. The fauna on the whole is sparser; there is little bioturbation and, unlike the other mudstones, hand specimens can usually be split parallel to the bedding.

The sandstones within the formation occur as distinct, laterally persistent, parallel-sided beds 20 mm to 0.2 m thick, and exceptionally up to 0.6 m. They are variably calcareous, grey to greenish grey when fresh, but weathering buff, and susceptible to Triassic staining. They are fine to very fine grained, some grading to siltstones. Their bases are flat, while the tops are either gradational, diffuse or flat to wave-rippled. Internally they are mainly parallel-laminated, though some exceedingly weakly so. Broken shelly debris is common in the basal 10 mm or so, but only rare above. Burrows occur, but are uncommon in the thicker beds except at the tops. The scattered exposures suggest that sandstones are distributed unequally throughout the succession with an increase in the uppermost part.

Scattered within the bioturbated mudstones are irregular ball-like masses of argillaceous bioclastic limestone, on average 0.1 m in diameter, developed along the bedding. The masses have diffuse margins and contain variable amounts of bioclastic material, predominantly crinoid and broken shelly debris. They appear to be the result of both mechanical and biological soft-sediment disruption of original thin parallel-sided, silty limestone beds.

Bentonites were recorded (Dr M. G. Bassett, personal communication) during the excavations for Eastern Avenue [ST 1906 7875]–[ST 2019 7880]; they consist of pale blue-green clay beds up to 0.85 m thick.

The junction with the Rhymney Grit is abrupt. Rumney Borehole [ST 2108 7925] (Figure 2) showed fine- to medium-grained sandstones of the latter resting on interbedded thin sandstones and siltstones of the Pen-y-Lan Mudstone. A temporary section (Storrie, 1908) along Lake Road East [ST 1862 7915]–[ST 1863 7929] showed a similar contact.

The formation contains a rich, high diversity fauna, predominantly of brachiopods and trilobites but also including corals, gastropods, bivalves, cephalopods and bryozoa; graptolites are represented by only one species, Monograptus flemingii , from localities low in the exposed formation. The richest faunas (listed p.11) have come from Pen-y-Lan Quarry [ST 1981 7873]. Within the uppermost part of the formation, seen in Rumney Borehole, this high diversity fauna is replaced by a restricted assemblage of cf. Microsphaeridiorhynchus nucula, Atrypa reticularis, crinoid columnals and the alga Pachytheca sp.; other plant-like material also makes its first appearance. sp.

Shelly fossils are characteristic of the late Wenlock. Bassett (1969) has commented that they are best correlated with the upper part of the 'Wenlock Shale' or the 'Wenlock Limestone'. M. flemingii elsewhere ranges from the Cyrtograptus rigidus Zone to the C. lundgreni Zone. On this basis it appears that the formation correlates with the upper part of the Coalbrookdale Formation (Homerian Stage) of the type area for the Wenlock Series, Much Wenlock, Shropshire.

The fauna of the silty mudstoncs indicates accumulation in a shallow marine, mid-shelf, low-stress environment below normal wave-base. The sandstones exhibit many of the features of sublittoral sheet-sandstones (Goldring and Bridges, 1973). Each sandstone represents a distinct 'event', probably due to storm-wave activity accompanied by tidal or storm-surge ebb when sand was brought onto the shelf from nearshore environments (Goldring and Bridges, 1973; Dott and Bourgeoise, 1982). The limestone beds are also 'event deposits' and are carbonate analogs of the sublittoral sheet-sandstones. They may represent storm events when there was insufficient energy to transport sand but sufficient to rework autochthonous bioclastic debris into carbonate sheetsan ds

The transition into the restricted, predominantly shallow water inner shelf facies of the overlying Cae Castell Formation is manifested by the less diverse fauna and increased sandstone content in the uppermost part of the formation.

Cae Castell Formation

The Cae Castell Formation comprises 70–80 m of sandstones, siltstones and subordinate mudstones. Scattered thin limestones, some sandy and conglomeratic, are also present as well as a thin ironstone and one bentonite. The two type localities are the south-east bank of the River Rhymney [ST 2096 7892]–[ST 2101 7898] below the earthwork known as Cae Castell, and the adjacent Rumney Borehole (Figure 2), in which the formation was 70 m thick (adjusted for dip). A fuller account of the formation in the borehole is in preparation.

At the base is the Rhymney Grit (Sollas, 1879), 6–20 m thick. It comprises locally pebbly, partly cross-bedded, fine-to coarse-grained sandstones with scattered mudstone partings. Above, the remainder of the formation predominantly comprises interbedded very thin sandstones, siltstones and subordinate sand- and silt-streaked grey mudstones. Thicker beds of sandstone up to 2 m thick, commonly with mudstone partings, occur scattered throughout, as do thin beds of intraformational conglomeratic sandstone and limestone. This part of the formation contains a low diversity restricted marine fauna of molluscs and brachiopods. Plant-like material and phosphate are common. Burrowing is variably developed and in places intense.

A distinctive 10 m thick unit, the Newport Road Member, has been recognised in the middle of the sequence in Rumney Borehole. It comprises mudstones with subordinate interbedded siltstones, sandstones, and limestones, and contains a diverse open marine fauna in its lower part. A thin ironstone occurs near the base.

Rhymney Grit

The term Rhymney Grit was used by Sollas (1879) for the thick- to well-bedded, cross-bedded sandstones best seen in Rumney Quarry [ST 2148 7880]. The member is laterally persistent across the main inlier and varies in thickness from 6 m (Rumney Borehole, corrected for dip) to 11 m at Rumney Quarry (Sollas, 1879). It is probably locally thicker in the west, for Storrie (1908), reported 19.2 m of 'grits' from south of the dam at Roath Park Lake.

The strata consist of grey, buff-weathering, variably calcareous, cross-bedded and variably laminated medium-and in part fine- and coarse-grained or locally pebbly sandstones. A few scattered mudstone partings occur. The only surface exposure is at Rumney Quarry which exposes the uppermost 4.3 m of the member. This is trough cross-bedded. Individual sets commonly are graded, with intraformational lags at the base that pass up into micaceous 'plant'-rich shaly partings. Lenticular units, up to 0.17 m thick, of flaser- and wavy-bedded sandstones also occur interbedded within the trough-cosets. The trough-set bases locally exhibit irregular load casts.

The only complete section is that of Rumney Borehole. Here two types of laminated or cross-bedded sandstone were interbedded. The first was medium- to fine-grained sandstone, in beds on average 0.2 m thick, becoming fine-grained and silty in their upper part; some beds have cross-lamination or a mudstone drape at the top, while mud- and silt-filled 3 mm diameter burrows are common, especially in the upper part of beds. The second type, in beds 0.26 m to 0.4 m thick, consists of coarse sandstone, with scattered granules and small pebbles of quartz and medium- and fine-grained sandstone. Some beds are graded, but burrows and silts are rare. Large mudflake intraclasts are common in the lower parts of beds, whilst scattered micrite pebbles also occur; calcareous ooids have been noted in one bed. The base of the member is sharp in the borehole. The top is rather gradational both in the borehole and in Rumney Quarry.

Scattered crinoid columnals, bryozoans and indeterminate shell fragments are present and plant-like debris is common. Evidence of a Wenlock age is provided by acritarchs from Rumney Borehole samples.

The initiation of the Rhymney Grit represents a rapid shallowing event as demonstrated by the appearance of high energy structures such as trough cross-bedding. The member is best interpreted as a shallow marine, subtidal sand bar.

Strata above the Rhymney Grit

With the exception of the Newport Road Member, this part of the formation dominantly comprises very thin beds of fine- to medium-grained sandstone and siltstone interbedded with mudstone together with scattered thicker beds of sandstone. The mudstones are grey, slightly silty and micaceous, and contain sparse lenticular phosphate nodules. The thin sandstones and siltstones are greenish grey, but weather buff, and are variably calcareous and micaceous. They occur as beds and laminae up to 4 cm thick, but mainly less than 2 cm thick. The laminae are streaky whilst the thin beds exhibit graded, lenticular, or wavy bedding. The percentage of sandstone and siltstone in this lithology varies from 10 per cent in some packets to 75 per cent in others. Burrowing is common but varies from sporadic, to places where it is so intense that packets up to a metre thick have been homogenised to muddy sandstone. 'U'-shaped spreite burrows are abundant. Thicker beds of sandstone include units of wavy-bedded, flaser-bedded and cross-laminated, fine- to medium-grained sandstone up to 2 m thick; units of cross-bedded or planar-laminated, fine- to coarse-grained sandstone up to 0.6 m thick; beds ranging from intraformational conglomeratic calcareous sandstone to sandy limestone, up to 0.26 m thick and containing shell debris as well as intraclasts of phosphate, mudstone and micrite; and distinctive sheet-like graded beds, 0.07 to 0.51 m thick, of fine-grained sandstone which are planar laminated, some with rippled tops.

The Newport Road Member comprises mudstones with subordinate interbedded thin siltstones, sandstones and limestones. It can be divided into two. The lower part sits sharply on a thick unit of flaser-bedded and cross-laminated sandstone and contains a thin, burrowed, bioclastic oolitic ironstone in the lowest metre. It comprises grey, very silty Chondrites-burrowed mudstones with thin beds (up to 0.16 m thick) of siltstone, very fine-grained sandstone and bioclastic limestone, a general lithology identical to that of the Hill Gardens Formation (see p.7). The upper part of the member consists of greenish grey, slightly silty, mudstones, with scattered lenticular phosphate nodules and numerous interbedded, graded, or rippled siltstones and fine-grained sandstones 1 to 3 cm thick. The member is overlain sharply by a packet of several thick sheet sandstones

Although fossils are fairly common at some levels within the formation above the Rhymney Grit, with the exception of the Newport Road Member, low diversity is a pronounced feature, with articulate brachiopods represented almost exclusively by generally small and fragmented specimens of the rhynchonelloid cf. M. nucula; Lingula sp.and Orbiculoidea sp.also occur. The remainder of this fauna consists principally of molluscs, including a large, thick-shelled bivalve common in a 14-cm friable, yellow-weathered sandstone (the Ctenodonta Bed of Sollas, 1879), which occurs just above the top of the Rhymney Grit in Rumney Quarry. In contrast, the Newport Road Member contains a diverse fauna, mainly of articulated brachiopods and bivalves, together with crinoid, hyolithid and tribolite fragments, which is replaced in the highest beds of the member by an unusual assemblage of hyolithid fragments and bivalves. The macrofauna does not provide a precise age for the formation but together with the lithology indicates a regressive event, for which evidence is widespread in late Wenlock sequences in South Wales and the Welsh Borderland (Hurst and others, 1978). However, acritarchs in samples from the upper part of the formation in Rumney Borehole confirm that it is of late Wenlock (Homerian) age, and indicate that the Wenlock–Ludlow Series boundary can be taken at the top of the formation.

The succession above the Rhymney Grit, excluding the Newport Road Member, was deposited in a very shallow, storm-dominated, inner shelf environment dominantly above storm wave base but occasionally above normal wave base.

The thin sandstones, siltstones and limestones of the lower part of the Newport Road Member are identical to storm-generated sublittoral sheet-sandstones (Goldring and Bridges, 1973) and this, together with their fauna, confirms that they were deposited in a relatively deeper open marine shelf environment but still above storm wave base. This lower part of the member is interpreted as a short lived, previously unreported, transgressive interlude in the regional late Wenlock regression. The upper part of the member with its very restricted fauna is enigmatic. It may represent a relative stillstand prior to the return to the inner shelf conditions at the top of the member.

Cardiff Group

The Cardiff Group consists of about 155 m of interbedded mudstones and thin siltstones with subordinate thin sandstones and bioclastic limestones that sharply overlie the sandstone-dominated Cae Castell Formation and are in turn themselves sharply overlain by the red beds of the Lower Old Red Sandstone. The type localities are the east bank of the River Rhymney [ST 2101 7898]–[ST 2101 7909] and Rumney Borehole [ST 2108 7925] (Figure 2). The group crops out between Heath and Rumney, and has been proved in shallow boreholes beneath the estuarine alluvium on the Wentlooge Level near Maerdy Farm. As the outcrop is largely urban, exposure is very poor. In the Cardiff city centre two deep boreholes, Helen Street [ST 1992 7704] (North, 1915b) and Pellet Street [ST 1881 7611], proved the group beneath the Trias.

The rather monotonous sequence has been subdivided on the relative percentages of the component lithologies, bed thickness and sedimentary structures. Two formations have been recognised in Rumney Borehole, and the upper one has been divided into three members; these have not been mapped at the surface due to the lack of exposure; indeed only the Hill Gardens Formation and Eastern Avenue Member can be seen commonly in surface exposures. A full description of the group in Rumney Borehole is in preparation.

Hill Gardens Formation

The Hill Gardens Formation (Plate 2) is an interbedded sequence of siltstones with subordinate mudstones and sandstones, sparse limestones and a few bentonites, which sharply overlies the Cae Castell Formation and is gradationally overlain by the Llanedeyrn Formation. It is characterised by abundant prominent (more than 9 cm thick) sheet-sandstones and siltstones, the former being most numerous and thickest in the lowest 13 m. The formation is 96 m thick, and has a thin conglomeratic bioclastic limestone at its base containing scattered micrite and phosphate pebbles derived from the underlying formation. Some 9 m above the base a 0.6 to 1 m ferruginous oolitic limestone—the Ty Mawr Ironstone—is present. The type localities of the formation are the east bank of the River Rhymney [ST 2101 7898]–[ST 2102 7909], immediately south-south-west of Rumney Hill Gardens, and Rumney Borehole.

The siltstones and sandstones form distinct, thin, laterally persistent, sheet-like beds. The siltstones are medium to pale grey; the sandstones are usually pale grey. They are calcareous, and have sharp bases with scattered bioclastic debris in the basal portions of some beds. The siltstones occur in beds 2–30 cm thick. Internally they are weakly laminated or apparently structureless. The tops of the beds are mainly burrowed or graded and a few are ripple-laminated. The sandstones are fine grained and occur in beds 2–52 cm thick. Internally they are low-angle to planar-laminated; their tops are either wave-rippled or diffuse due to bioturbation.

The mudstones are medium to dark grey, sparsely micaceous and very silty, locally verging to siltstones. Chondrites mottling and rare streaky lamination are present.

Two types of limestone beds are present. Brachiopod coquinas, with a silty mudstone matrix, form parallel-sided beds 3–20 mm thick; coarse, crinoidal, shelly packstones form beds 1–10 cm thick with sharp bases. The tops of the latter beds are either graded, wave-rippled or extensively burrowed. The limestones contain brachiopod, crinoid, bryozoan and coral debris in a matrix of similar, but finer-grained, material with a variable silt content (5–30 per cent). At some levels, where thin limestones are abundant in the mudstones, intense bioturbation has led to complete destratification, the resultant shelly mudstone containing clots and wisps of bioclastic material.

Six bentonites, varying from 1 to 260 mm thick, were noted in Rumney Borehole. The Ty Mawr Ironstone was first recognised by Sollas (1879) who described its petrology and suggested that it was equivalent to part of the Wenlock Limestone. It is the 'red limestone' referred to by Bassett (1974), and Cocks and others (1971). The type locality of the member is the east bank [ST 2101 7898] of the River Rhymney where it is 0.6 m thick. It is also present in Rumney Borehole where it is 1 m thick. It comprises red, strongly bioturbated, muddy hematitic 'oolitic' limestone, with a rich shelly fauna of brachiopods, crinoid columnals, bryozoa, corals and orthocones. The ferruginous ooids are generally oblate with cores of biotic fragments. A few thin beds of siltstone and argillaceous limestone, identical to those in the rest of the formation, are present within it. A conglomeratic lag of limestone pebbles, some bored, occurs at its base. A shallow borehole [ST 2242 7832] at Pwll-mawr proved three separate thin red ferruginous mudstone beds, up to 0.5 m thick, with green ferruginous ooids, interbedded in 2.2 m of grey mudstones. Both above and below these red mudstones, green ferruginous ooids occur either scattered or in wispy nests in grey mudstones with thin siltstones and sandstones. In total, 7.2 m of strata contained varying amounts of ferruginous ooids. These beds may be the lateral equivalent of the Ty Mawr Ironstone.

A high diversity fauna consisting of brachiopods and molluscs with bryozoans, trilobites and crinoid columnals, at the base of the Hill Gardens Formation, abruptly replaces restricted assemblages of the underlying Cae Castell Formation. It persists through the Ty Mawr Ironstone into higher beds where Atrypa reticularis and Gypidula galeata are particularly common. The latter becomes rare in the upper half of the formation, but A. reticularis is abundant to the top, with bryozoans and corals, including favositoids and the solitary rugosan Phaulactis sp., gradually becoming more prominent in the upper part of the sequence. In the lower part of the formation in Rumncy Borehole, the monograptids Colonograptus cf. colonus and C. cf. varians were recorded and, from the middle part, Saetograptus chimaera salweyi and the dendroid graptolite Koremagraptus sp. These are the first records of Ludlow graptolites from the district.

Bassett (1974) placed the base of the Ludlow Series tentatively at the top of the Ty Mawr Ironstone, in the belief that this bed possibly equated with part of the 'Wenlock Limestone' of the Welsh Borderland. This view was provisionally followed in the published summary log of Rumney Borehole (Waters and White, 1980), but acritarchs from the borehole have since demonstrated that the Hill Gardens Formation below the ironstone is early Gorstian in age, and the base of the Ludlow Series should therefore be taken at the base of the Cardiff Group.

The graptolites mentioned above and acritarch assemblages from the lower and middle parts of the Hill Gardens Formation indicate either the Neodiversograptus nilssoni or Lobograptus scanicus Zone of early Gorstian age. Although no graptolites were obtained from the upper part of the formation, the presence of numerous corals suggests a late Gorstian age, an interpretation confirmed by the acritarch assemblages.

Llanedeyrn Formation

This comprises about 58 m of interbedded mudstones and thin siltstones with subordinate thin beds of sandstone and limestone at the top of the Cardiff Group. The type locality for the formation and its three constituent members is Rumney Borehole. The formation crops out between Roath Park and Rumney, but the sparse exposures are limited to the Eastern Avenue Member. The boundary with the Hill Gardens Formation is gradational.

Eastern Avenue Member

This member comprises 41 m of interbedded blocky mudstones and thin siltstones with subordinate thin limestones and a few thin sandstones. It shows a marked increase in the percentage of limestone beds compared to the underlying formation, and has no prominent (more than 9 cm thick) sheet-sandstone beds. The percentage of mudstoncs is the highest in the whole of the Cardiff Group.

The mudstones are medium to dark grey, but greenish weathering, sparsely micaceous and very silty, in parts verging to siltstones. They are variably calcareous, and become tough and rather massive with increasing carbonate content. Chondrites mottling is common, and streaky lamination occurs in places. Brachiopods and crinoid debris are common in the mudstoncs. Corals, absent in the remainder of the formation, are even more common than in the underlying Hill Gardens Formation; they occur both as rolled specimens and in life position.

The beds of siltstone and fine-grained sandstone are grey, slightly calcareous, and form thin sheet-like beds mainly less than 5 cm thick but up to 8 cm thick in the lower part of the member. The siltstones are either structureless or laminated, whilst the sandstones exhibit both planar- and ripple-lamination.

The limestones range from silty coquinoid packstones, 2–60 mm thick, to coarse bioclastic packstones in sheet-like beds up to 25 mm thick.

Four bentonites up to 0.34 m thick were noted in Rumney Borehole.

Burrowing in the member is commonly intense, and many levels with thin limestones have been destratified. The bioclastic material then occurs in the mudstones as clots and wisps of limestone that weather out as nodules with diffuse margins concordant with the bedding.

The increase in the number of corals in the upper part of the Hill Gardens Formation becomes even more pronounced in the lower part of the Eastern Avenue Member but, apart from this, there is no significant faunal change. However, in the upper part of the member a sharp decrease in the number of corals occurs and the upper stratigraphical limits of several strophomenacean brachiopods are reached. Saetograptus sp.was recorded from these upper beds in which Isorthis orbicularis is the dominant brachiopod and the stratigraphically important Shaleria ornatella also occurs. The Gorstian–Ludfordian boundary is indicated by these faunal changes. Its position in the upper part of the Eastern Avenue Member is compatible with the acritarch evidence, the record of S. ornatella and the occurrence of the specimens of Saetograptus sp. These latter could be examples of a rare semispinose form which occurs with S. leintwardinensis at some localities (Dr R.B. Rickards, personal communication).

Chapel Wood Member

The Chapel Wood Member comprises 13 m of very thinly interbedded mudstones, siltstoncs and sandstones, with a few limestones. The mudstones are greyish green to grey, and are variably silty with vague streaky laminations and burrow mottlings mainly referable to Chondrites. Brachiopods and crinoid debris occur locally. The siltstones and sandstones are greenish grey and variably calcareous; they occur in sharp-based beds most 3–30 trim but up to 60 mm thick. The siltstones are mainly weakly laminated or structureless, whilst the very fine- to fine-grained sandstones are parallel-and cross-laminated. Wavy and lenticular-bedding is present in places. The few limestones are up to 3 cm thick; they are either coquinoid or silty coarse bioclastic packstones. Two bentonites were noted in Rumney Borehole. Burrowing occurs throughout and is locally intense.

The fauna in the upper part of the Eastern Avenue Member persists into the lower half of the Chapel Wood Member up to a thick bentonite. Above this bed there is a pronounced decrease in faunal diversity, particularly affecting the brachiopods which are represented mainly by Microsphaeridiorhynchus nucula and Protochonetes ludloviensis. The remainder of the fauna consists principally of bivalves. Acritarch assemblages indicate an early Ludfordian age and this is compatible with the recorded macrofauna.

Roath Park Lake Member

This comprises about 4.2 m of very thinly interbedded mudstones, siltstones and sandstones that make up much of the sequence, together with thicker, commonly amalgamated, beds of sandstone and limestone. It is further characterised by the presence of reworked phosphate pellets.

The mudstones are dark grey to greenish grey, micaceous and variable silty. Chondrites burrows are ubiquitous. The thin beds of sandstone and siltstone are generally less than 3 cm thick. Both lenticular and wavy-bedding are present, and parallel-lamination is common. Shell hash is in places present at the bases of sandstones, whilst some bedding planes have scattered fish scales. Load structures and burrows are common.

The thicker beds of sandstone are 0.12 to 0.75 m thick. Some beds are amalgamated. 'They are reddish pink to pale green mottled, very calcareous, fine grained, and contain up to 25 per cent of bioclastic material (brachiopods, crinoids, bryozoans, gastropods, orthocones and fish scales), mainly concentrated in thin coarse bands. Mudstone intraclasts are locally present, as are phosphate pellets in the form of phosphatised internal moulds of gastropods. Some beds show parallel-lamination; and cross-lamination occurs in the upper parts of beds.

The limestones occur in beds, 5 to 44 cm thick; the thicker ones are composite with mudsione partings. They are pale greenish grey to pinkish green mottled, fine- to coarse-grained packstoncs, and variably sandy. The bioclastic content is the same as in the sandstones and they also contain similar phosphate pellets. The beds have sharp bases, and the tops are either sharp or rippled or are gradational into fine-grained calcareous sandstone.

The fauna is essentially the same as that recorded from the upper part of the Chapel Wood Member, with the important addition of numerous phosphatic, internal mould fragments of the gastropod Loxonema sp. and also abundant fish fragments which are mainly thelodont scales.

The member is lithologically, faunally and stratigraphically comparable with the so-called 'Holopella' Beds of the Usk Inlier (the Upper Llangibby Beds of Walmsley, 1959), and constitutes the uppermost beds of the Ludfordian Stage.

The conditions of deposition of the Cardiff Group strata can be summarised as follows:

The sheet-sandstones, siltstones and limestones of the Hill Gardens Formation are identical to sublittoral sheet-sandstones (Goldring and Bridges, 1973), and therefore reflect storm-generated 'event' deposits on a muddy open shelf above storm-wave base. This interpretation is supported by the presence of a rich and varied fauna. The sudden appearance of this formation above the inner shelf deposits of the Cae Castell Formation reflects the rapid earliest Ludlow transgression known elsewhere in Wales and the Welsh Borderland (Hurst, 1975; Hurst and others, 1978). The bored pebbles at the base of the Ty Mawr Ironstone are locally derived from a hardground that reflects a pause in deposition.

The upward disappearance of the thick sheet-sandstones and the establishment of the much muddier Eastern Avenue Member are probably due to the baffle effect of an offshore shoal rather than to further deepening, particularly since the member equates with the 'Aymestry Limestone' facies of the Welsh Borderland (Watkins and Aithie, 1980).

The upward succession from the Chapel Wood Member into the Roath Park Lake Member reflects upward shoaling, as evidenced by the appearance of lenticular and wavy-bedding, the increasing sandstone content, and the greater bed thicknesses. The presence of amalgamated sheet-sandstones and limestones in the Roath Park Lake Member suggests the shoreline association of Goldring and Bridges (1973), deposited around normal wave-base. This is confirmed by the presence at the base of the Raglan Mudstone Formation of a fluvial distributary sandstone capped by calcrete overlying the member with an erosional contact.

Raglan Mudstone Formation

The Pøídolí Series comprises the lowest part of the Lower Old Red Sandstone sequence, which here entirely consists of the Raglan Mudstone Formation, formerly known in neighbouring districts as the Raglan Marl Group (Welch and Trotter, 1961). The sequence comprises about 300 m of red mudstones with calcretes and subordinate sandstones. It crops out in the northern part of Cardiff though it is poorly exposed. To the south, several water bores (Strahan and Cantrill, 1912; North, 1915b) have proved it beneath Triassic strata.

Although Storrie (1879) and Sollas (1879) make reference to the Downtonian of the district, the first detailed description was in the first edition of this memoir (Strahan and Cantrill, 1902), later amplified by Heard and Davies (1924). Rumney Borehole proved the nature of the junction of the Old Red Sandstone facies with the underlying Ludlow Series (Waters and White, 1980).

The junction with the underlying Roath Park Lake Member is gradational, but has been taken at the base of a thin bone bed that is correlated with the Ludlow Bone Bed. A 2 m-thick sandstone occurs about 0.3 m above the bone bed. It is coarse and pebbly at the base, fining upwards into fine-grained sandstone and siltstone at the top, in which occurs an immature calcrete profile. It appears to be very different, both lithologically and seclimentologically, from the basal sandstone—the Downton Castle Sandstone—in the Welsh Borderland, and so the use of this term around Cardiff is hardly appropriate.

Above this basal sandstone the formation comprises mainly structureless red mudstones with scattered thick sandstones up to 4 m thick arranged in fining-upwards cycles. Nodular calcrete profiles are common. The only recorded fossils are scattered fish fragments in the sandstones and lingulids from a unit of laminated mudstone and siltstone in Rumney Borehole.

The top of the formation is taken at the top of a well developed calcrete termed the Psammosteus Limestone; this is not exposed in the district but its presence has been inferred from old pits. The mapped boundary is thus largely conjectural.

The only comprehensive section is that provided by Rumney Borehole. Here, the basal bone bed is 6 cm thick, and comprises greyish purple, undulatory laminated siltstone with thin laminae up to 3 mm thick of thelodont scales and fish spines, with a few indeterminate shell fragments. Scattered small burrows have partly disrupted the laminae. The bed is sharply defined and is interbedded in laminated mudstones. Above it there is 0.3 m of strata consisting of dark purple mudstone with siltstone laminae and scattered fish scales, sharply overlain by a fish-bearing, sandy, mud-flake breccia and grading up into fine-grained laminated sandstones and some siltstones with scour-and-fill structures and lenticular bedding; a thin purplish red silty mudstone separates this from the basal sandstone.

The basal sandstone has an erosive base. In the lower part it comprises pale lilac to dark purplish brown, poorly cemented and feldspathic, medium- to coarse-grained, cross-bedded sandstone. It contains scattered mudflakes, indeterminate fish fragments and scattered quartz pebbles near its base. It grades up into purple, thinly-bedded, laminated and cross-laminated, fine-grained sandstones with some siltstone beds. Burrows are common while some poorly developed purple calcrete tubules are evident near its top. There is an upward gradation into red structureless mudstones.

The fining-upward cycles forming the bulk of the formation each begin with a major sandstone unit having an erosional base. The sandstones are commonly cross-bedded, and 0.9 to 4 m in thickness. They are micaceous and purplish brown to reddish purple in colour with green mottles. Grain size varies from coarse, in places with scattered quartz pebbles, to fine. Mudflakes are common in the basal part of the beds and above scoured surfaces within them. Internally the sandstones are cross-bedded or laminated, and fine upwards. Indeterminate fish remains were noted in the basal parts of sandstones in the lowest part of the sequence.

There are distinctive laminated packets, up to 4 m thick, of reddish purple, silty, micaceous mudstones with numerous thin beds and laminae of siltstone and fine-grained sandstone above some of the major sandstone units in Rumney Borehole. The mudstones contain scattered sand-filled, 5 mm diameter, burrows. The siltstones and very fine- and fine-grained sandstones are red-brown and micaceous; they form laminae and thin beds up to 0.28 m thick, though most are much thinner. The beds have sharp bases and are internally structureless, laminated or cross-laminated; sandstone dykes and other wet-sediment deformation structures are common. Two beds low in the formation contain plant fragments and lingulids, including Lingula cornea, in places preserved in life position.

Red-brown dominantly structureless mudstones occupy the upper parts of the fining-upward cycles, and make up most of the formation. They are micaceous and silty, in part grading to siltstone. Streaky lamination is locally preserved, whilst small listric surfaces are abundant. Desiccation cracks and small (5 mm) burrows have been noted.

Nodular calcrete profiles are present in the upper parts of the cycles, mainly in the structureless mudstones, and have been described in detail by Allen (1974b); multiple profiles may be present within a single unit. The carbonate forms irregular to tuberose nodules (glaebules), up to 10 mm in diameter, within the host sediment. Where elongate, the glaebules are orientated with their long axes at right angles to the bedding. They vary in colour from red and purplish blue to green mottled. In a well developed calcrete profile the percentage of nodules gradually increases upwards from 5–10 per cent to some 50–60 per cent over a thickness of between 0.2 and 0.8 m. In the upper parts of some of the profiles, local carbonate veins (crystallaria) connect glaebules. In the adjacent Newport district (Squirrel' and Downing, 1969) the Psammosteus Limestone at the top of the formation is reported to be a nodular calcrete in a mudstone host, passing up into a massive calcrete, 1.5 to 9 m thick.

The basal bone bed is part of an overall coarsening-upwards (shallowing) sequence, extending from the subtidal Roath Park Lake Member to the subaerial calcrete at the top of the basal sandstone of the formation. The bone-rich laminae appear to be winnowed lags in a relatively low-energy environment. Similar lags occur below the bone bed, whilst some of the sheet-sandstones in the upper part of the Roath Park Lake Member contain abundant scales. There is no direct evidence as to whether the bone bed is high subtidal or intertidal in origin.

The basal fish-bearing sandstone of the formation, fining-up into red structureless mudstone with calcrete, appears to be an alluvial channel-fill, probably of a distributory channel on a coastal plain. The lingulid-bearing laminated mudstones, siltstones and sandstones in the lower part of the formation are also suggestive of coastal mudflats.

The remainder of the formation presents little evidence of marine influence. The major sandstones are arranged in fining-upward cycles that are typical deposits of wandering alluvial channels (Allen, 1964), whilst the structureless mudstones with abundant calcretes suggest alluvial flood-plain deposits. However, the presence of the quasi-marine levels within the lower part of the formation suggests that the alluvial plains were near enough to the coast to suffer periodic marine influences.

Details

Pen-y-Lan Mudstone

An exposure in the front garden of a house in Lake Road East [ST 1864 7912] within the upper part of the exposed formation yielded a fauna dominated by small brachiopods including Cyrtia exporrecta, Dalejina hybrida, Dicoelosia biloba, Eoplectodonta sp., Skenidioides lewisii and Strophochonetes?. Further material in the National Museum of Wales from temporary sections in Lake Road East includes Dictyonelia capewellii, and from the junction of Pen-y-Lan Hill and Cyncoed Road Amphistrophia funiculata, Coolinia pecten, Eoplectodonta duvalii, Isorthis clivosa and Resserella canalis. A temporary section [ST 1978 7867], north-east of the junction of Waterloo Road and Dorchester Avenue, exposed some of the lowest beds of the formation. A varied fauna was recorded with corals including Favosites sp., brachiopods including Anastrophia deflexa, Meristina obtusa and Pentlandina lewisii plakodis, trilobites including Deiphon barrandei, and the graptolite Monograptus.

Scattered small sections in the lower part of the formation were seen in a gently northward-dipping sequence in the Eastern Avenue cutting at Pen-y-Lan [ST 1906 7875]–[ST 2019 7880]. Notes provided by Dr M. G. Bassett, made when the cutting was first opened, record blocky variably Shelly mudstones and siltstones with scattered thin sandstones and limestones throughout the length of the cutting. He noted three bentonites on the south side of the cutting, east of the Cyncoed road bridge [ST 1972 7878], these being 0.2, 0.2–0.06 m and 0.85 m thick respectively, the one of variable thickness having an irregular base. Most of the cutting is now obscured but a typical section exposed on the north side of the cut [ST 1906 7875] to [ST 1923 7876] is:

During the resurvey a collection made from mudstones in a temporary exposure on the north side of the cutting [ST 1976 7877] yielded Atrypa reticularis, Eospirifer radiatus, M. obtusa, orthaceans, Pentlandina cf. lewisii plakodis and Plectatrypa imbricata, together with decalcified bryozoans and phacopid trilobite fragments. Material in the National Museum of Wales from a nearby locality in the cutting, includes Cyrtia exporrecta, R. cf. canalis and Striispirifer plicatellus, together with trilobites including Acastocephala macrops, Dalmanites sp., Dicranopeltis salteri, Eophacops musheni and Leonaspis coronata.

Pen-y-Lan Quarry [ST 1981 7873] exposes some of the lowest beds in the formation and has provided much of the fossil material from the formation curated in the National Museum of Wales; it was, however, largely infilled during the construction of Eastern Avenue. A small part, now largely overgrown and filled with water, remains open on the south side of the road, where scattered small sections, each no more than 1.7 m high, cover a stratigraphic interval of about 10 m. Shelly calcareous silty mudstones occur in blocky bedded units 0.1 to 0.2 no thick, with scattered impure silty bioclastic limestones and sparse fine-grained sandstones. The following list is based mainly on BGS and National Museum of Wales collections: Small solitary Rugosa, brachiopods including Amphistrophia funiculata, Anastrophia deflexa, Atrypa reticularis, Coolinia pecten, Cyrtia exporrecta, Dalejina hybrida, Dicoelosia biloba, Eoplectodonta Eospirifer radiatus, Cypidula galeata, Howellella sp., Isorthis clivosa, Lepidoleptaena poulseni, Leptaena depressa, L. depressa kathekta Lingula sp., Megastrophia (Protornegastrophia) semiglohosa, Merisirna obtusa, Nucleospira pisum, Pentlandina lewisii plakodis, Plectatrypa imbricata, R. curtails, ?Salopina conservatrix, Sphaerirhynchia wilsonr, Spirigerina Streptis grayii, Striispirifer plicatellus and Strophonella euglypha, the gastropod Loxonema sp., the bivalves, Cypricardinia subplanulata, Modiolopsis sp., Nuculites sp.and Ptychopteria sp., the cephalopod Dawsonoceras annulatum, trilobilites including Acanthopyge cf. hirsuta, Acastocephala macrops, Bumastus? xestos, Dicranopeltis salteri, Encrinurus tuberculatus, ?Harpidella (H.) aitholix, Hemiarges scutatis?, Leonaspis cosonata, Platylichas grayii and the graptolite Monograptus flemingii.

Rumney Borehole proved 5.39 m of interbedded fine-grained sandstones and siltstones below the Rhymney Grit. Burrows were abundant, and scattered crinoid columnals, Atrypa reticularis, cf. Microsphaeridiorhynchus nucula and the alga Pachytheca sp.also occurred.

Borehole D28 [ST 2270 7801] drilled by University College Cardiff Geology Department (Anderson and Blundell, 1965) has been reclassified. Beneath 6.71 m of estuarine clay, it proved 3.96 m of fine- to medium-grained sandstones with scattered plant fragments representing the lower part of the Rhymney Grit. Below this, 22.64 m of Pen-y-Lan Mudstone was proved. Plant-like fragments were abundant in the top 10 m, whilst numerous thin graded siltstones and sandstones occurred in the uppermost part.

Cae Castell Formation

Dr Bassett has reported (personal communication) 7.28 m + of thickly bedded to massive, commonly coarse, buff, grey and maroon grit referable to the Rhymney Grit in a temporary section [ST 1843 7927] in Lake Road West. The top of the Rhymney Grit was exposed in the road opposite the path leading across Roath Park Lake dam, where it was overlain by 7.3 m of thinly bedded sandstones, siltstones and mudstones with sparse rhynchonelloids, crinoid columnals, Orbiculoidea sp.and locally abundant carbonaceous fragments.

On the eastern side of Roath Park, Storrie (1908, p.23) detailed a section exposed during the building of Lake Road East [ST 1862 7926] from the upper part of the Pen-y-Lan Mudstone through the Rhymney Grit, apparently some 19.2 m thick, into 5.43 m of the overlying part of the formation. The latter comprised thinly bedded sandstones and mudstones, with 0.18 m of yellow sandstone containing 'Ctenodonta subaequalis' lying 3.19 m above the top of the Rhymney Grit.

On the east bank [ST 2101 7898]–[ST 2097 7892] of the River Rhymney at Rumney scattered sections in strata above the Rhymney Grit are exposed around the high-tide mark. In the north [ST 2098 7895]–[ST 2101 7898] about 23 111 of the upper part of the formation are exposed, though there are some gaps in the section. The sharp junction with the overlying Cardiff Group is well exposed at the northern end of the section.

Rumney Quarry [ST 2148 7880] provides the following section:

Sollas (1879) reported about a further 7 m of Rhyrnney Grit below this section and described it as massive sandstone, locally becoming flaggy and rippled, and in places passing into a fine-grained conglomerate.

A borehole [ST 1977 7452] at the Crown Preserved Fuelworks at Roath Dock is fully detailed in Strahan and Cantrill (1912, p.99). The strata encountered between 195.8 and 261.2 m may belong to the Cae Castel' Formation.

Cardiff Group

The best sections in the Cardiff Group occur on the east bank of the River Rhymney at Rumney. Examination of samples from a trench [ST 2107 7919] excavated by Dr Bassett across the Cardiff Group/Raglan Mudstone Formation contact, shows that levels around the Eastern Avenue Member/Chapel Wood Member junction are in contact with the basal sandstone of the Raglan Mudstone Formation and that, therefore, the junction is faulted.

Farther south, a section [ST 2106 7918] high in the bank above the river exposes levels around the boundary between the Eastern Avenue Member and the Chapel Wood Member. It shows 0.45 m of green-grey blocky weathering siltstones and mudstones with a few brownish green weathering, thin, fine-grained sandstones. The strata contain a lowest Ludfordian fauna, the highest exposed in the district. This includes Favosites sp., bryozoans, Aegiria grayi, Atrypa reticularis, Coolinia pecten, Craniops implicatus, Dayia navicula, Howellella elegans, Isorthis clivosa, I. orbicularis, Leptostrophia filosa, Microsphaeridiorhynchus nucula, Orbiculoidea sp.Protochonetes ludloviensis, Salopina lunata, Shagamella minor, Sphaerirhynchia wilsoni, Kionoceras angulatum, Orthoceras' ibex and Dalmanites sp.Atrypa reticularis is represented by only a single specimen, whilst exposures immediately to the south (in older beds) yield this species in abundance.

High in the riverbank, a small pit [ST 2105 7917] exposes 2.5 m of the Eastern Avenue Member which are pale greyish green, very silty, tough, calcareous, micaceous mudstones, locally reddened along joints. They form blocky beds up to 0.13 m thick, with 'nodules' and thin beds of silty green-grey shelly crinoidal limestone containing scattered corals. The fauna is late Gorstian and includes Amphistrophia funiculata, Atrypa reticularis, Gypidula sp., Howellella cf. subinsignis, Isorthis orbicularis, Leptaena depressa, Leptostrophia filosa, Lingula?, Mesopholidostrophia lepisma, Microsphaeridiorhynchus nucula, Protochonetes ludloviensis, Shaleria aff. ornatella, Sphaerirhynchia wilsoni, Ptychopteria sp., Bembexia lloydii, Dalmanites? and a proetid. A similar section [ST 2105 7915], lower in the Eastern Avenue Member, yielded many elements of the above fauna as well as bryozoans, C. implicatus, H. elegans, I. clivosa, Salopina lunata and Calymene sp.

The Hill Gardens Formation is well exposed in the east bank of the river [ST 2102 7909] to [ST 2101 7898] (Plate 2):

In a bank [ST 2153 7895] at Ty Mawr Avenue, Rumney, 0.33 m of the Ty Mawr Ironstone, decalcified to a brown rottenstone, is overlain by 1.15 m of sandstones, comprising four beds 0.1 to 0.2 m thick, separated by purple shaly mudstone. The fauna from the ironstone collected during the present survey, together with that in National Museum of Wales, includes Syringopora?, Faoositella anolotichoides, F. interpuncta, Amphistrophia funiculata, Atrypa reticularis, G. galeata, Leptaena cf. depressa, Meristina obtusa, Spirigerina marginalis, Strophonella euglypha, Bellerophon wenlockensis?, Platyceras cornutum and Dalmanites sp.

A borehole [ST 2242 7832] at Pwll-Mawr proved the following sequence low in the Hill Gardens Formation, and probably including the equivalent of the Ty Mawr Ironstone:

Two deep boreholes south of the main inlier have proved the group beneath the thick Triassic succession of the central Cardiff area. Helen Street Borehole [ST 1992 7704] proved 46.5 m of Ludlovian strata overlain by Old Red Sandstone (North, 1915b). Material from the borehole curated in the National Museum of Wales has been re-examined. The lowest Old Red Sandstone encountered is the basal Raglan Mudstone Formation sandstone. The shelly fauna of the underlying Ludlow Series between 154.2 m (506') and the base (the few specimens between the base of the Raglan Mudstone sandstone and 154.2 m are missing) is late Gorstian in age and probably forms the, upper part of the Eastern Avenue Member. The Ludfordian is thus at most 9.75 m thick, and may be absent, possibly as a result of either faulting or, less probably, unconformity. Significantly North's log contains no mention of the conspicuous bone bed at the base of the Raglan Mudstone Formation or of the prominent gastropod-bearing limestones of the Roath Park Lake Member. Pellet Street Borehole [ST 1881 7611], drilled in 1920, proved 13.5 m of the Cardiff Group. Material from 131 m (430 feet) curated in the National Museum of Wales has yielded the acritarchs Diexallophasis denticulata (late Ludlow type processes), Eupoikilofusa cf. cantabrica (very faint striations), Veryhachium leintwardinensis, Leiosphaeridia spp., visbysphaera cf. microspinosa (late Ludlow style), Veryhachium sp.1. (3 processes, robust, ? reworked) and Polygonium gracilis (reworked); the chitinozoa Conochitina sp.and Sphaerochitina '.sphaerocephala'; and the miospores Ambitisporites sp.(small) and Archaeozonotriletes chulus. Dr Doming comments that the assemblage is Ludfordian in age, and in all probability equivalent to the Upper and Lower Leintwardine formations (Holland and others, 1980) of the Ludlow type-area. The reworked acritarchs are derived from the Lower Ordovician.

Raglan Mudstone Formation

A road cutting [ST 1994 8190]–[ST 1999 8112] in Cyncoed in the upper part of the formation exposed:

A temporary section in Cyncoed Road [ST 1925 7980], judged to be not far above the top of the basal sandstone, exposed:

Strahan and Cantrill (1912) recorded a bed of tine-grained pale dolomite in mottled marl, in a disused limestone quarry [ST 1948 8120] at Cyncoed; it probably belongs to the Psammosteus Limestone.

A temporary section [ST 1866 7952] on the west side of Roath Park, exposing contorted red and grey shales associated with two beds of coarse ferruginous quartz grit, was thought by Storrie (1908) and Strahan and Cantrill (1912) to lie within the Ludlow Series and to be overlain by further marine strata. However a recent temporary exposure nearby [ST 1864 7952] was in purple, rusty weathering, coarse-grained feldspathic sandstone, identical to the basal sandstone of the formation. It probably occupies the small faulted core of a syncline.

A locality [ST 2106 7919] on the east bank of the River Rhymney, where Storrie (1879) reported a temporary exposure of a fish-bearing 'red sandy bed' at the Ludlow Series/Old Red Sandstone junction, has been trenched by Dr Bassett. An examination of the material from the trench shows that the basal sandstone of the Raglan Mudstone Formation and the basal bone bed are faulted against the upper part of the Cardiff Group.

Helen Street Borehole (North, 1915b) proved 40 m of the formation directly beneath the Trias. An examination of the material curated in the National Museum of Wales appears to show that the basal sandstone, comprising 1.9 m of very coarse pebbly calcareous sandstone, is faulted against the underlying Cardiff Group.

Chapter 3 Devonian

Within the district, Lower and Upper Devonian strata comprise some 900 m of dominantly fluviatile red beds of Old Red Sandstone magnafacies that crops out in the core of the Cardiff–Cowbridge Anticline, on the flanks of the Rogerstone Anticline, and in small inliers at Cwrt-yr-ala and on the coast. Middle Devonian rocks are absent due to a major regional unconformity, the result of late Caledonian uplift and pre-Farlovian erosion. In South Wales and the Welsh Borders, the Silurian/Devonian boundary is difficult to place, but it is thought to lie somewhere near the top of the Raglan Mudstone Formation (Allen and Williams, 1981). The Devonian/Carboniferous boundary occurs in the uppermost part of the Old Red Sandstone (Gayer and others, 1973), with the latter passing up conformably into the Carboniferous Limestone through a few metres of marginal marine strata.

Past research in the district was undertaken by Sollas (1879), Howard (1894), Strahan and Cantrill (1902, 1912), Evans and Cox (1956), and Gayer and others (1973). Relevant descriptions of the sequence in nearby areas include those of Welch and Trotter (1961), in the Monmouth and Chepstow districts, and Squirrel' and Downing (1969) in the Newport district. Broader regional reviews include those of Allen (1965, 1974a, 1975, 1979) and House and others (1977). Previous and present classifications of the Devonian strata in the district are illustrated in (Figure 3); comparisons with classifications used in other areas are shown in (Figure 4).

St Maughans Formation

The St Maughans Formation comprises about 450 m of interbedded red mudstones with subordinate sandstones and scattered calcrete profiles. Its base is defined by the top of the Psammosteus Limestone, and its junction with the overlying Llanishen Conglomerate is taken at the lowest bed of conglomerate containing exotic pebbles. The formation crops out in St Mcllons and Llanishen, but exposure is poor.

The sandstones and mudstones are arranged in fining-upwards cycles, each many metres thick, in which the sandstone overlies an erosion surface and fines upwards into mudstone. Calcrete profiles are common in the mudstones in the upper part of cycles. Poor exposure makes it difficult to assess the total number and distribution of cycles within the sequence.

The sandstone units occupying the lower parts of the cycles are up to 3.5 m thick. Thin intraformational conglomerates of mudstone and calcrete pebbles are common at their bases. The sandstones are fine- to coarse-grained, red, purple and green mottled, and either cross-bedded or planar-laminated. The coarse-grained sandstones tend to be quartzitic, and commonly contain scattered intraformational clasts of up to pebble-size.

The mudstones forming the upper parts of the cycles are very silty and mainly structureless; most are red to purple in colour, but green mottling is common. They locally contain thin ribs of sandstone up to a few centimetres thick. Calcrete profiles are common in the mudstones, and consist of scattered, irregular, rounded or tuberose carbonate nodules up to a few centimetres in diameter. Mauve and violet mottling occurs in the mudstone around the nodules.

No fossils have been found during the present survey. However, typical Dittonian faunas are recorded from the adjacent Newport district (Squirrel] and Downing, 1969) where the Downtonian/Dittonian boundary was taken at the base of the formation, though with reservations (see Squirrel', 1973).

The formation and its equivalent in others areas forms a distal fluvial facies (Allen, 1963, 1964, 1970, 1974a, 1979). The sandstones, with their basal intraformational conglomerates, represent high sinuosity channel deposits; the mudstones represent the floodplain deposits; the caicretes indicate periods of prolonged subaerial exposure. The deposits appear to be derived from northern quadrants, and to have been deposited on a vast alluvial plain extending across South Wales and the Welsh Borders (Allen, 1974a).

Llanishen Conglomerate

The Llanishen Conglomerate comprises about 150 m of red sandstones, siltstones and mudstone, with beds of conglomerate containing mainly exotic pebbles. Calcrete profiles are common throughout.

It is exposed in the core of the Cardiff–Cowhridge Anticline, on the north-western limb of the Rogerstone Anticline and, in the south, in two fault-bounded inliers at Michaelston-super-Ely and Drope. Exposure is poor but much new information became available during the construction of the M4 motorway and from drilling along the Link Road from Capel Llanilterne to Culver House Cross. In the past the formation has been locally confused with the breccias of the Triassic marginal facies in the Capel Llanilterne area (Strahan and Cantrill, 1902) and, more recently, in the Michaelston-super-Ely inlier (Tucker, 1977).

No fossils are known from the formation, but it is considered from its stratigraphical position to be Dittonian to Brcconian in age (Squirrel' and Downing, 1969).

The base of the formation is not exposed, but it is taken at the base of the lowest bed of conglomerate with exotic pebbles. The top of the formation is taken at the entrance of the drab brown sandstones of the Brownstones; it was exposed in the excavations for the Pantmawr cutting on the M4 [ST 1487 8219], where the highest bed of conglomerate lay 5.6 m below the top of the formation, and the junction with the Brownstones was sharply defined.

The various lithologies are commonly organised in fining-upwards cycles, beginning with conglomerate, overlying an erosion surface which in most cases is cut in siltstone. The conglomerates fine up into sandstone, and this into siltstone and mudstone; some, however, are sharply overlain by siltstone. Some coarse members of the fining-upward cycles are multistorey, with individual units being separated by erosion surfaces. Calcrete profiles occur in the uppermost parts of the cycles.

The conglomerates are lilac to purple, and are composed of subrounded to subangular exotic pebbles with scattered cobbles. They are clast-supported with a medium- to coarse-grained sand matrix. Clasts are mainly 10–40 mm in diameter, though cobbles to 70 mm have been noted. The conglomerates are calcareously cemented, but appear as an unconsolidated gravel on weathering. Many of the clasts have a surficial dark purple irony stain. Individual beds of conglomerate range from a few centimetres to nearly 8 m in thickness. At Michaelston-super-Ely they make up about 40 per cent of the formation (Figure 5). The bases of the beds are distinctly erosional. Internally the conglomerates are flat-bedded or massive, with some pebble imbrication. There are thin laminae and beds of sandstone in the upper parts of some beds. The clasts have not been studied in any detail, but are comparable to those of the Newport district where about half were volcanic or pyroclastic rocks, the remainder being lithic sandstones and pink quartzites (Allen, 1975). Where intraformational conglomerates occur, they are predominantly composed of calcrete clasts.

The sandstones are purple with green mottles, strongly micaceous, and fine- to coarse-grained. Some are pebbly or contain thin pebble lenses. They are laminated or cross-bedded. Beds are up to 3 m thick, and either occur above conglomerate units or as separate units with erosive bases.

The siltstones and mudstones are characteristically bright red in colour, very micaceous, generally structureless, and typically weather to a greasy clay.

Calcrete profiles are very common, especially in the finer-grained lithologies. Most are composed of green and purple mottled carbonate nodules that increase in abundance up the profile. Where they occur in conglomerates, the pre-existing pebbles have suffered 'jig-saw' brecciation during the calcretisation process. Two particularly massive calcretes, up to 5 m thick, occur in coarse-grained sediments. They are mainly nodular, but contain some laminated zones. The lower one is only known from boreholes; (Figure 5) the upper one, lying 40 m below the top of the formation (Figure 5), is exposed in a railway cutting [ST 1094 7657]–[ST 1109 7656] and an old pit [ST 1098 7660]. One may equate with the Ruperra Limestone of the Newport district (Squirrell and Downing, 1969).

The sediments of the formation are of proximal alluvial facies (Allen, 1979). The conglomerates and sandstones are best regarded as the deposits of channelised, low sinuosity, stream systems that interfingered both laterally and downslope with flood-flats on which the siltstones were deposited. Calcrete presumably formed mainly on the flood- flats, but also in abandoned channels.

There is good evidence that the conglomerates were derived from the south (Squirrell and Downing, 1969; Allen, 1975): the source-area may possibly have been the 'Bristol Channel Landmass' (Tunbridge, 1983).

Brownstones

The dominantly arenaceous formation at the top of the Lower Old Red Sandstone in south-east Wales has been variously referred to as the Brownstones or the Brownstone Group (see Welch and Trotter, 1961; and Allen, 1974a, for discussion). Here, the former usage is followed. The sequence comprises drab brown to purple, fine- to coarse-grained sandstones, with numerous thin intraformational conglomerates and scattered beds of red siltstone and mudstone. It sharply overlies the Llanishen Conglomerate and is overlain unconformably by the (Upper Old Red Sandstone) Cwrt-yr-ala Formation.

It crops out on the northern limb of the Cardiff–Cowbridge Anticline between Craig-y-parc and Pantmawr, where it forms an escarpment that diminishes in height westwards, as the formation thins at outcrop due to overstep of the succeeding beds, from a calculated thickness of 190 m at Pantmawr to 55 m at Craig-y-parc. Exposure is poor, and the base of the formation has been taken at the base of the scarp; the junction with the overlying Cwrt-yr-ala Formation is not exposed and the mapped boundary is largely conjectural. Recent drilling in the core of the Cardiff–Cowbridge Anticline at Michaelston-super-Ely has provided considerable information about the lower part of the formation. South of the anticline, Cwrt-yr-ala Borehole [ST 1403 7339], provisionally thought to have ended in Upper Old Red Sandstone (Waters, 1978a), is now known to have proved 37 m of Brownstones (see (Figure 6)).

The lithologies of the formation are arranged in fining upwards units each comprising a multistorey sandstone up to 8 m thick and capped by thin siltstones and mudstoncs up to 2.9 m thick. In Cwrt-yr-ala Borehole (Figure 6) the sandstone to siltstone and mudstone ratio is 7:1. Each multistorey sandstone begins with an erosional surface cut in siltstone or mudstone, commonly overlain by intraformational conglomerate which may be anything from a few centimetres up to about 0.5 m thick; an exceptional one is 1.6 m thick in Cwrt-yr-ala Borehole. The conglomerate grades upwards into sandstone that is generally either cross-bedded or planar-laminated, though a few beds of fine-grained sandstone are cross-laminated. Similar sandstone beds, usually with basal intraformational conglomerate lags, make up the rest of the multistorey sandstone which eventually fines upward into siltstones and mudstones. Calcrete profiles are present locally in some of the latter.

The intraformational conglomerates are green and red-brown in colour, and composed of rounded to angular clasts of mudstone. Calcrete clasts are also present in some places and very rare exotic pebbles, mainly of quartz, occur in some beds. The matrix is coarse- to medium-grained sandstone.

The sandstones range from fine to coarse grained. The fine- and medium-grained beds are characteristically drab, brownish purple to purplish grey with scattered green mottles. They are poorly cemented and commonly very micaceous. The coarse-grained sandstones are generally pale purple to greyish purple, with whitish or pale green mottles in places. They are either poorly or well cemented, and some beds are quartzitic with rare scattered quartz granules and pebbles. Beds rich in intraformational mudstone clasts are common in most sandstone beds. Fish fragments and oxidised plant fragments have also been noted.

The siltstones are dark purple or green; they are laminated, and contain variable amounts of very fine-grained sandstone beds up to a centimetre thick which exhibit parallel- and cross-lamination. Some contain scattered plant fragments.

The mudstones are generally red or purple, silty, micaceous, and structureless. A few are green, and contain abundant carbonised plant debris.

Calcrete profiles are very uncommon; two, up to 0.3 m thick, were noted in Cwrt-yr-ala Borehole (Figure 6). They comprise scattered nodules of calcrete in mudstone or siltstone.

Only indeterminate fish fragments were collected from Cwrt-yr-ala Borehole. However, a grey-green plant-bearing mudstone in a borehole [ST 1112 7612] at Michaelston-super-Ely yielded the following miospore assemblage: Calamospora sp., Punctatisporites sp., Retusotriletes sp., Apiculiretusispora sp., Apiculatisporis sp.The general state of preservation was such that identification beyond generic level was not possible. No stratigraphically diagnostic taxa were recorded but simple azonate spores dominated. Elsewhere in Wales and the Borders the Brownstones are 'Dittonian' to 'Breconian' in age on evidence from fish and plants respectively (House and others, 1977). They have also yielded Siegenian spores and are thought to range up into the Emsian (House and others, 1977).

The multistorey sandstones are comparable to those described by Tunbridge (1981), from the formation in the Brecon Beacons; they fall between his proximal and medial facies. Each sandstone in the sequence, with its virtually ubiquitous intraformational lag, is a channel-fill. The facies can be interpreted as the result of repeated channel cutting, sandy infilling and avulsion, in a low sinuosity fluvial channel-system. The low percentage of siltstones and mudstones suggests that fine-grained overbank flood-plain deposits had low preservation potential, because the channels regularly combed across their flood-plains. The lack of calcrete profiles, compared to the amount of calcrete clasts in the intraformational conglomerates, suggests that calcrete profiles formed more widely than the rock record suggests. The thin fine-grained sandstones within the siltstones and mudstones may be coarse overbank deposits.

The Brownstones sequence in South Wales and the Borders exhibits a coarsening-up motif (Allen, 1974a; Allen and Dineley, 1976; Tunbridge, 1981), becoming conglomeratic with abundant exotic pebbles in the upper part in the Forest of Dean where they are thickest (Welch and Trotter, 1961; Allen, 1974a). This trend has been interpreted by Allen (1974a) as due to the southerly migration of a Lower Devonian 'fall line' in South Wales, resulting in the deposition of increasingly proximal alluvial facies. Although the Brownstones of the district compare to the upper part of the Brecon Beacon sequence, they are comparable to only the middle part of the Forest of Dean sequence as they lack the abundant exotic pebbles of that area. The Cardiff sequence is odd in that, even though it is 'downstream' of the Brecon Beacon sequence, it does not begin with a distal facies and pass up into a medial facies, as happens in the lower part of the Brownstones elsewhere. This may be because the Llanishen Conglomerate is partly equivalent to the lower part of the Brownstones elsewhere in South Wales and the Borders, and the Brownstones alluvial fans did not extend into the Cardiff district until the source area of the Llanishen Conglomerate ceased to provide any more sediment. The fact that the Brownstones begins with a proximal facies suggests that its local base considerably post-dates the base elsewhere. The upper pebbly part of the Brownstones may never have been deposited in the district, though it may be concealed by 'Upper Old Red Sandstone' overstep.

Cwrt-yr-ala Formation

The Cwrt-yr-ala Formation is a new stratigraphical term introduced here for the lowest part of the 'Upper Old Red

Sandstone' sequence, which is a dominantly fluviatile sequence in all up to 114 m thick. It oversteps the Lower Devonian deposits between Creigiau and the Taff valley where it cuts out 135 m of Brownstones in a south-westerly direction; the increase in overstep towards the axis of the Cardiff–Cowbridge Anticline may be due to Mid-Devonian movements along the Vale of Glamorgan Axis. It crops out on both limbs of the Cardiff–Cowbridge Anticline, and in small inliers to the south at Cyntwell and Cwrt-yr-ala.

Where it cannot be separated from the overlying Quartz Conglomerate Group the two units are depicted on the map as Upper Old Red Sandstone, undivided.

The formation comprises a sequence of thinly interbedded quartzitic sandstones, siltstones and mudstones, with subordinate, thick, commonly pebbly, sandstone units. Calcrete profiles are well developed. The designated type section is Cwrt-yr-ala Borehole (Waters, 1978a) [ST 1403 7339] (Figure 7).

Its base is taken at the unconformity with the Brownstones whilst the top is taken at the appearance of the multistorey sandstones with beds of quartz-conglomerate that characterise the Quartz Conglomerate Group. The formation varies in thickness from 15 m in the north to 73 m in Cwrt-yr-ala Borehole in the south.

The various lithologies are arranged in a series of fining-upwards cycles. Each cycle begins with an erosion surface overlain by a major sandstone up to 3 m thick, commonly with an intraformational conglomerate lag at the base. Each major sandstone unit itself fines upwards into thinly-interbedded sandstone and siltstone in varying proportions, and many cycles are capped by mudstone. Nodular calcrete profiles generally lie in the upper part of each cycle, but none is more than 1.8 m thick.

The intraformational conglomeratic lags are up to 0.3 m thick, and consist of angular to rounded sandstone, siltstone, mudstone and calcrete clasts, of up to pebble-size. The matrix is generally coarse sandstone, whilst scattered exotic pebbles, mainly quartz, occur in sonic beds.

The sandstone units forming the lower parts of the cycles vary from very coarse- to fine-grained, and most exhibit an overall upward-fining. Planar lamination or cross-bedding occurs in the main part of the unit, and cross-lamination may be present in the fine-grained sandstone at the top of the unit. In the lower parts of some units there are scattered exotic pebbles, mainly of quartz, up to 10 mm in diameter. Mudstone and siltstone intraclasts are also scattered in the lower parts of some units. Two types of sandstone have been noted. The most common is pale purplish grey, pale lilac or greyish pink with green mottles, hard, quartzose, well cemented and poorly micaceous (especially when compared to the very micaceous multistorey sandstones in the overlying Quartz Conglomerate Group). The second type is purplish brown with green mottles, sparsely micaceous, and not so well cemented. Burrows occur in the uppermost part of some of the sandstone units, whilst fish fragments commonly occur in the lower parts and in the intraformational conglomerates.

The thinly interbedded sandstones and siltstones occurring above the major sandstone units are either sandstone- or siltstone-dominant. Within these sequences, the sandstones occur in thin beds 5 mm–0.3 m thick. They are either red-brown or greyish purple, very fine grained and silty, or reddish to greyish pink with pale green mottles, quartzitic, and fine or, less commonly, medium grained. They form sharp-based beds that are planar laminated, cross-laminated or structureless, and have either sharp or gradational tops. The thicker sandstones contain scattered intraclasts in the lower part and a few exhibit cross-beds in sets to 0.15 m thick. The interbedded siltstones are up to 0.15 m thick, and mainly brick-red or, less commonly, greyish purple. They are sparsely micaceous and locally sandy. Cross-lamination and planar-lamination can be discerned in places, but much of the siltstone is structureless. Thin beds and laminae of disturbed micaceous mudstone also occur, while listric surfaces are abundant. A few sandy intraformational conglomerates, unrelated to major sandstone units, form beds up to 0.41 m thick; sonic contain scattered quartz pebbles and granules.

Within the thinly interbedded sandstones and siltstones load-casted ripples, load-balls with umbilical cords, and flame structures are common. Sparse plant impressions occur in the siltstones, whilst fish fragments occur in some of the sandstones. Small Beaconites burrows are locally present, and 2 mm wide cylindrical branching burrows are common.

The mudstones that cap some cycles are purplish or brownish red, silty, and contain scattered siltstone beds and laminae. A few thin, sharp-based, non-quartzitic sandstones, in beds up to 0.22 m thick, similar to those in the thinly interbedded sandstone-and-siltstone lithofacies occur. Burrows and plant impressions have also been noted.

The distribution of these different lithologies in Cwrt-yrala Borehole is illustrated in (Figure 7). The basal unconformity has been seen only in this hole. The lithological change was marked, drab brown sandstones of the Brownstones being sharply overlain by purplish red laminated fine-grained sandstones with thin interbeds of red-brown siltstone.

Miss S.V. Young (British Museum, Natural History) has identified a fragment of pectoral appendage of cf. Bothriolepis at 136.55 m, a Sauripteris scale at 169.39 m, and a plate fragment of Bothriolepis at 178.44 m from the lower half of the formation in Cwrt-yr-ala Borehole. The latter genera indicates an Upper Devonian age.

The lack of marine fauna and the presence of fining-upwards cycles associated with calcrete profiles, point to a fluviatile origin for the sediments. The major sandstone units, with their basal erosion surfaces commonly overlain by intraformational conglomerate, probably represent channel and lateral accretion deposits. The thinly interbedded sandstones and siltstones fall into Allen's (1970) 'alternating beds' facies of fluviatile fining-upwards cyclothems, interpreted as floodplain vertical accretion deposits. The sandstones represent crevasse splay and levee deposits. The mudstones similarly represent floodplain deposits but of a more distal nature, the scattered thin sandstones representing local crevasse-splay deposits. The dominance of vertical accretion deposits suggests that the formation represents the deposits of a high sinuosity fluviatile channel system.

Quartz Conglomerate Group

The Quartz Conglomerate Group (Heard and Davies, 1924) comprises locally pebbly multistorey sandstone units with very subordinate siltstones and quartz conglomerates. Although occurring throughout. as sparse thin beds, quartz conglomerates are commonest in the lower part in the north of the district. The group sharply overlies the Cwrt-yr-ala Formation and passes up gradationally into the Lower Limestone Shale Group. It crops out in all the Upper Old Red Sandstone inliers. On the northern limb of the Cardiff–Cowbridge Anticline it is 55–66 m thick but it thins southwards to 41 m in Cwrt-yr-ala Borehole (Figure 7).

Most of the group was deposited under a fluviatile environment, apart from locally, where the topmost part contains a rather restricted marine fauna. Its upper boundary is taken at a change from the dominantly sandstone succession of the Quartz Conglomerate Group to the dominantly grey shale and interbedded limestones of the Tongwynlais Formation.

The multistorey sandstones comprise bundles of planar-laminated and cross-bedded, mainly purple, brown, pale green and white sandstones separated by erosion surfaces. Each multistorey unit is separated by thin siltstones or mudstones, but these are sparse and some of the units are tens of metres thick. Above each erosion surface there is generally a purple and green mottled intraformational conglomerate composed of pebble-sized clasts of mudstone, siltstone and, less commonly, calcrete. The conglomerates have a sandy matrix and some contain scattered quartz pebbles; they are up to 0.5 m thick.

The sandstones between each erosion surface within a multistorey unit are mainly medium grained. However, they commonly fine upwards from the basal intraformational lag through coarse- to very coarse-grained sandstones with scattered quartz pebbles into medium-grained sandstones ending in fine-grained sandstones. Some sandstones are locally rich in mudstone intraclasts and scattered quartz pebbles. Cross-bedding with subordinate planar-lamination is the main structure in the coarse and medium sandstones; cross-lamination is dominant in the fine-grained sandstones. Fish fragments occur in some beds. The sandstones differ from those in the underlying Cwrt-yr-ala Formation in containing abundant strongly micaceous partings and laminae. They are also mainly poorly cemented. Most are quartzose, those that are well cemented being quartzites.

Units of purple, greenish grey and red-brown micaceous siltstone and silty micaceous mudstone are uncommon. They range up to 1.5 m in thickness and contain thin beds and laminae of fine-grained laminated sandstones. Fragmented plant material is abundant.

Individual units of quartz conglomerates are up to 7 m thick. The thickest occur near the base of the formation in the north. They mostly consist of quartz pebbles, up to 4 cm in diameter, set in a coarse-grained pale lilac to pale purplish grey quartzite matrix, in places with scattered mudstone intraclasts. They are cross-bedded and contain scattered lenses of quartzite.

Pink nodular calcrete profiles up to 1 m thick are sparsely present. They occur in the fine-grained sandstones at the top of the multistorey sandstone units.

At the top of the formation marginal marine shelly sandstones with subordinate siltstones, mudstones and sandy limestones occur locally between the multistorey sandstones and the overlying mudstones and limestones of the Tongwynlais Formation. In Cwrt-yr-ala Borehole the marine beds are 3.6 m thick. The sandstones in them are pale lilac, fine to medium grained, very calcareous, weakly laminated, and contain mudstone partings commonly associated with ?rootlets. Crinoid columnals and rhynchonellid brachiopods are scattered throughout. The associated siltstones are red-brown, very calcareous, and contain bioclastic laminae and beds, up to 2 cm thick, that contain rhynchonellids, crinoid columnals and bryozoa. A 0.1 m bed of reddish grey, slightly dolomitic, sandy packstone is present. At Tongwynlais [ST 1302 8245], the Tongwynlais Formation sharply overlies unfossiliferous fluviatile sandstones. The base of the Tongwynlais Formation in this section is taken 2.6 m lower than the position chosen by Gayer and others (1973), and lies at the base of their Bed 7.

The Upper Devonian fish Holoptychius nobilissimus has been reported from the middle of the Quartz Conglomerate Group at Tongwynlais (Evans and Cox, 1956). Gayer and others (1973) obtained spores which they regarded as very close to the Tnla-Tnlb boundary, about 18 m (their Bed 1) below the top of the group at Tongwynlais [ST 1302 8245]. This flora is earliest Famennian in age, following the redefinition of the Tournaisian/Famennian boundary (Conil and others, 1977). Since an earliest Courceyan conodont fauna has been obtained during the present survey from between 1.5 m and 3 m above the base of the Tongwynlais Formation (see p.28), the Devonian/Carboniferous boundary must lie somewhere between the uppermost part of the Quartz Conglomerate Group and the basal part of the Tongwynlais Formation. Plants from the top of the Quartz Conglomerate Group are described by Gayer and others (1973), but are not stratigraphically diagnostic.

The Quartz Conglomerate Group in the district can be interpreted as the deposits of low sinuosity streams that laid down little overbank sediment and regularly combed their floodplains to destroy most of such sediment that had accumulated (Allen, 1965). At the top of the sequence, the calcareous sandstones and siltstones with marine fossils and ?rootlets overlying cross-laminated fine-grained sandstones at the top of a multistorey sandstone unit, as seen in Cwrtyr-ala Borehole, are suggestive of marine sands being driven periodically by storms across abandoned fluvial lateral accretion deposits on a low lying coastal/delta plain. Farther north, as at Tongwynlais, the deposits of the basal marine transgression of the Tongwynlais Formation sit directly on fluvial sediments.

The quartz conglomerates in the lower part of the Quartz Conglomerate Group in the north of the district are best exemplified in the Newport district (Squirrell and Downing, 1969); they thin and eventually disappear southwards: possibly they pass laterally southwards into the Cwrt-yr-ala Formation. An analogous situation is the southward disappearance of the Quartz Conglomerate when traced from the Monmouth and Chepstow district into the Mendips where the Tintern Sandstone and Quartz Conglomerate are not separable and the Upper Old Red Sandstone is referred to as the Portishead Beds (Kellaway and Welch, 1955). From the descriptions of the Portishead Beds (Pick, 1964) it seems that lithologies similar to those of the Quartz Conglomerate Group in the south of the Cardiff district, overlie lithologies similar in parts to the Cwrt-yr-ala Formation.

Details

A stream section [ST 1944 8155]–[ST 1956 8162], north-east of Cyncoed showed about 50 m of purple and grey mottled mudstone with interbedded purple and dark green Fine- to medium-grained micaceous sandstone in beds up to 1.5 m thick. The section falls within the St Maughans Formation.

A disused quarry [ST 0997 8079] east of Craig-y-parc exposed the following sequence near the base of the Quartz Conglomerate Group:

A composite section of the Quartz Conglomerate Group at Tongwynlais [ST 1345 8227]–[ST 1373 8213], from a reinterpretation of Evans and Cox (1956) and the presently exposed lowermost part of the group and top of the Cwrt-yr-ala Formation, is given below:

Chapter 4 Lower Carboniferous (Dinantian)

The Lower Carboniferous (Dinantian) rocks of South Wales record the deposition of a southwards-thickening prism of predominantly carbonate shelf sediments, that accumulated and onlapped on to the southern edge of St. George's Land, a contemporary upland of older rocks (George, 1958, 1974), and are traditionally known as the Carboniferous Limestone. In the district, the Lower Carboniferous is conformable with the underlying Upper Old Red Sandstone. All the Dinantian regional stages (George and others, 1976) up to and including most of the Holkerian are represented by at least 950 m of strata. The upper part of the Holkerian sequence, the Asbian and the Brigantian are cut out by the sub-Namurian unconformity.

The Carboniferous Limestone crops out on both the northern and southern limbs of the Cardiff–Cowbridge Anticline. North of the axial trace of the anticline it forms the south-facing scarp that delimits the south crop of the South Wales Coalfield. To the south, it is largely covered by Triassic and Jurassic rocks, but crops out in several isolated inliers. The larger inliers form relatively high ground representing an exhumed Triassic topography. From the distribution of the inliers south of the Cardiff–Cowbridge Anticline and west of the Lavernock Fault, it is inferred that Carboniferous Limestone underlies this area, except along the projected subcrop of the Barry Docks Anticline where Upper Old Red Sandstone probably comes to subcrop.

Previous research

Apart from descriptions of the coastal area by Howard (1895), the first account of the Carboniferous Limestone of the district was by Strahan and Cantrill (1902), but it provided virtually no subdivision of the sequence. The succession along the south crop of the Coalfield between Ruthin and Risca was described by Dixey and Sibly (1918) and correlated with the zones of Vaughan (1905, 1906). More recent work (Bhatt, 1976; Hird and others, 1984) on this sequence east of Miskin, where it is largely dolomitised, has embraced the sedimentology, geochemistry and origin of the dolomites. The junction with the Upper Old Red Sandstone has been described by Gayer and others (1973) at Tongwynlais, immediately north of the district. South of the Cardiff–Cowbridge Anticline, early reconnaissance work was carried out by Sibly (1920). The sedimentology of the Gully Oolite in the St. Lythans Borehole (Waters, 1978b) has been described by Waters (1984). Accounts of the Dinantian rocks in adjacent areas include Squirrel] and Downing (1969); Whittaker and Green (1983) and Wilson and others (in press).

It has been suggested (Ramsbottom, 1973, 1977, 1979) that the South Wales sequence comprises a number of sedimentary cycles separated by non-sequences. The cycles (mesothems) were thought to be eustatic in origin, each recording an individual transgression followed by a regression. However, local tectonism has been regarded (George, 1978) as the dominant control of the cycles in South Wales.

Classification

The classification used by Dixey and Sibly (1918) in the north of the district and that used here is shown in (Figure 8). The lithostratigraphical nomenclature in current useage in the South-West Province has been summarised by George and others (1976). The district is situated between Gower and the Bristol/Mendip region and in both areas local names are used to describe the sequences. Commonly, units can be shown to have been continuous across the whole of the province and in these cases, the formation name with priority has been used. Where no suitable names are available, we have defined new ones.

The classification used is compared with that in adjacent regions in (Figure 8). The Lower Limestone Shale Group (Kellaway and Welch, 1955) has been divided into three new formations, namely the Tongwynlais Formation, the Castell Coch Limestone and the Cwmyniscoy Mudstone. The overlying bioclastic limestones are termed the Black Rock Limestone (Group), (Kellaway and Welch, 1955; but sensu Green and Welch, 1965 and George and others, 1976), rather than the Penmaen Burrows Limestone (J. V. Stephens, 1973) and its subdivisions proposed by George and others (1976), which are largely biostratigraphical. The group consists of three new formations, the Barry Harbour Limestone, the Brofiscin Oolite and the Friars Point Limestone. For the Caninia Oolite (Vaughan, 1905), the term Gully Oolite (Lloyd Morgan, 1889; Kellaway and Welch, 1955) is used. The distinctive thin 'Modiola phase' (Dixey and Sibly, 1918) at the top of the Gully Oolite is the same as the Caswell Bay Mudstone (J. V. Stephens, 1973) of Gower. For the bioclastic limestones above the Caswell Bay Mudstone, the Gower term High Tor Limestone U. V. Stephens, 1973) is employed, rather than the term Birnbeck Limestone (Whittaker and Green, 1983) of the Westonsuper-Mare district. Above the High Tor Limestone, an oolite broadly equivalent to the Goblin Combe Oolite of the Bristol/Mendip region is given the new name Cefnyrhendy Oolite because its exact correlation is uncertain. The oolitic sequences above the Cefnyrhendy Oolite are more litho-logically akin to the Hunts Bay Oolite U. V. Stephens, 1973) of Gower than to the equivalent Clifton Down Limestone (Kellaway and Welch, 1955) of the Bristol/Mendip region.

North of the Cardiff–Cowbridge Anticline the succession becomes increasingly dolomitised as it is traced eastwards. Dixey and Sibly (1918) could not trace most of their zonal divisions cast of Creigiau, and in the adjacent Newport district Squirrel] and Downing (1969) distinguished only limestone (of Holkerian age), and mapped the remainder of the sequence above the Lower Limestone Shale Group as dolomite (Rudry Formation of George and others, 1976). In the present survey, all the formations have been distinguished within the dolomite belt, except for the Cernyrhendy Oolite in the vicinity of the Taff valley and this is only due to poor exposure. The term Rudry Formation is therefore superfluous in the district.

Correlation with the Dinantian regional stages of George and others, (1976) has been achieved by use of corals, brachiopods, conodonts and foraminifera ((Figure 8)). The proposal to replace the Courceyan Stage by the Ivorian and Hastarian stages of Belgium (Ramsbottom and Mitchell, 1980) has not received unanimous acceptance in Britain (Clayton and Sevastopulo, 1981; see also Varker and Sevastopulo, 1985, for recent summary of conodont evidence). The Courceyan is therefore used in this account. The major cycles of Ramsbottom (1973) correspond broadly to the regional stages of George and others (1976), though some of the cycle boundaries suggested by Ramsbottom need refinement.

Lower Limestone Shale Group

Lithostratigraphy

The Lower Limestone Shale Group comprises 100–120 m of interbedded mudstones and bioclastic limestones with subordinate oolites, lying between the Quartz Conglomerate Group below and the limestone-dominated part of the Carboniferous Limestone above. At the base, the Tongwynlais Formation consists of 40–50 m of thin, dominantly skeletal, limestones with subordinate mudstones; in the middle, the Castell Coch Limestone consists of 9–25 m of thickly bedded oolitic and skeletal limestones, and at the top the Cwmyniscoy Mudstone comprises 40–60 m of mudstones with subordinate thin skeletal limestones.

The group crops out on the northern limb of the Cardiff–Cowbridge Anticline between Creigiau and the Taff valley, where the Castel] Coch Limestone forms a prominent scarp, and the softer Cwmyniscoy Mudstone and Tongwynlais Formation form slacks respectively below the

Black Rock Limestone and above the Quartz Conglomerate Group. On the southern limb of the Anticline in the St Lythans and St Andrews Major outcrops, there is a similar topographical expression, but the lower boundary of the Tongwynlais Formation is poorly defined because the Upper Old Red Sandstone does not form a scarp in this area. The group also crops out in small inliers in the Wenvoe valley, Cadoxton, and on the coast. Where the group cannot be subdivided because of poor exposure, it is represented on the map as 'undivided'.

Tongwynlais Formation

The Tongwynlais Formation consists predominantly of interbedded thin skeletal limestones and grey mudstones, with a basal unit, up to 12 m thick, comprising oolitic, peloidal, skeletal and sandy limestones, and interbedded mudstones, with subordinate calcite mudstones, calcretes and ironstones. The formation varies in thickness from 40 m in the north of the district to about 50 m in the south. The type locality is a road section [ST 1302 8245] at Tongwynlais, 300 m north of the district, described by Gayer and others (1973) and Burchette (1977). The junction with the Quartz Conglomerate Group is sharp but conformable. It is taken at the lithological change from the sandstone-dominated sequence below to the sandy limestone and grey mudstone sequence above. The formation is poorly exposed and detailed knowledge of it is based largely on the type section and Cwrtyr-ala Borehole [ST 1408 7339] (Waters, 1978a), about 9 km to the south (Figure 9).

The basal unit includes two distinctive lithologies, namely lagoonal beds and ironstones (ferruginous limestones). The lagoonal beds ('Modiola Phase' of Dixey and Sibly, 1918) comprise dark grey to grey-green shaly calcareous mud-stones with thin beds and laminae of calcareous siltstone, argillaceous micrite and micritic limestone. Wavy and lenticular bedding and burrowing are common. There is a restricted fauna, including modioliform bivalves and ostracods. A thin biostromal/biohermal unit of attached, erect, vermiform gastropods capped by stromatolites has been described from the type section (Burchette and Riding, 1977).

The ironstones (alpha limestones of Dixon and Vaughan, 1912) are hematitic, skeletal packstone/grainstones and grainstones. They dominantly comprise crinoid ossicles, but bryozoa, ostracods, gastropods, bivalves, brachiopods and fish fragments are also present. The iron minerals, dominantly hematite but also goethite and . possibly chamosite, occur within the skeletal voids and as thin coatings, and in places as a replacement of the skeletal material. Reworked phosphate granules are common. There is a complete gradation from ferruginous limestones to limestones with a few scattered ferruginised bioclasts.

The basal unit is different at the two main localities. At Tongwynlais (Figure 9) it comprises three parts with lagoonal sediments in the middle; the base has been taken 2.6 m lower than Gayer and others (1973), at the base of their bed 7, and 5 m lower than Burchette (1981). Here, 2.61 m of grey, marine, fine-grained, sandy, sparsely shelly packstones to calcareous sandstones, with silty partings, exhibit various wave-generated structures, and rest sharply on the fluviatile sandstones of the Quartz Conglomerate Group. Above, there is 1.71 m of coarse, sandy, skeletal packstone/grainstone interbedded with green mudstones; several nodular calcrete profiles are evident. The base of each coarse unit is erosional, and derived calcrete clasts are abundant. The middle part of the basal unit comprises 3.5 m of lagoonal beds that rest sharply on the underlying calcretised marine sediments. The Rhiwbina Ironstone (Rogers and others, 1861; Squirrell and Downing, 1969; Gayer and others, 1973) is 1.7 m of hematitic skeletal grainstone with mudstone partings, that sits erosively on the lagoonal sediments with a lag of derived, bored micrite pebbles at its base. A 1.86 m unit of ooid skeletal grainstone, passing up into thin bedded skeletal packstone/grainstone, completes the basal unit.

At Cwrt-yr-ala (Figure 9) a similar threefold division of the basal unit is evident. The lower part comprises a 3.5 m, coarsening-upwards sequence that rests sharply on the Quartz Conglomerate Group. It begins with marine, fossiliferous, grey, silty mudstones with thin beds and laminae of calcareous siltstone and silty packstone. Lenticular and wavy bedding are present, and burrows are common. Above, sheet-like beds of skeletal packstone pass up rapidly into locally sandy, ooid, peloid packstone/grainstone with mudstone interbeds. Sharply overlying this sequence are 2.5 m of lagoonal beds which contain less micritic limestone than the same facies at Tongwynlais. A 6 cm bed of bentonite (DX 1773)^‡1 

The interbedded limestones and mudstones above the basal unit exhibit little lateral variation. The mudstones are dark grey, shaly, silty, variably micaceous and sparsely fossiliferous. They contain thin beds and laminae of calcareous micaceous siltstone, commonly with abundant ostracods, crinoid debris, brachiopods, bivalves and small gastropods. Wavy and lenticular bedding is present in places.

The limestones are dominantly skeletal packstones with some packstone/grainstones. They contain a similar fauna to the calcareous siltstones, as well as bryozoa, orthocones, fish, and rare horny brachiopods. Phosphate granules, mainly derived from the phosphatised internal moulds of gastropods, are present in some beds. The packstones occur in sheet-like beds 1 cm to 1 m thick. They vary from coarse-to fine-grained, and many of the thicker ones (more than 4 cm) are graded, coarse, shelly, crinoidal lags occurring at the base of beds. Some thicker beds exhibit repeated grading, and others are composite beds separated by shaly partings. The bases of the beds are commonly erosional. Internally, the thicker beds exhibit hummocky cross-stratification and plane lamination, and are commonly capped by wave ripples. The tops of beds are commonly burrowed.

In the lowermost part of the sequence above the basal unit, thin limestones with scattered ferruginised bioclasts are common. A hematitic bioclastic limestone, up to 1 m thick, occurs in the uppermost part of the formation at Cwrt-yr-ala and on the coast north-east of Bendrick Rock.

A thin packet of sandstones is present in the lower part of the formation above the basal unit. At Cwrt-yr-ala it is 1 m thick, and comprises thin, laminated and cross-laminated beds interbedded with mudstones.

Castell Coch Limestone

The Castell Coch Limestone, formerly known as the 'Bastard Limestone' (Strahan and Cantrill, 1902), comprises thick to well bedded, commonly cross-bedded, skeletal and oolitic grainstones. In the north of the district a thin unit of lagoonal deposits overlain by skeletal packstone and packstonc/grainstone occurs at the top of the formation at Tongwynlais. The formation is up to 25 m thick in the north, thinning to 9 m on the coast. The type locality is Castell Coch Quarry [ST 1300 8264] (Figure 10), just north of the district. Exposure is generally good, as the crop is pitted with numerous disused quarries. The junction with the Tongwynlais Formation is taken at the top of the highest mudstone bed and is sharp but conformable.

The grainstone sequence shows all gradations from oolitic to skeletal. A few grainstone/packstone beds are present. Skeletal components are mostly brachiopod fragments and crinoid ossicles with subordinate bryozoa. In the north, oolite is the predominant lithology, but southwards it becomes subordinate to bioclastic limestone. Reddening of the original grey colouration is very common. It is locally pervasive, but elsewhere selective, picking out one particular component such as ooid cortices. The reddening is commonly accompanied by late-stage epigenetic dolomitisation which is either similarly selective or locally pervasive. Within the limestones, sedimentary structures are not commonly seen, but where picked out by weathering they include cross-bedding and herringbone cross-bedding, the latter in sets to 0.15 m thick. Mudstone partings within the limestones have been noted only in boreholes.

The lagoonal deposits, seen only at Castell Coch Quarry (Figure 10), rest sharply on the grainstone sequence and comprise 0.7 m of grey mudstones with thin beds of peloidal micrite and fine-grained, locally shelly and crinoidal packstone. They are erosively overlain by nearly 3 m of variably dolomitised crinoidal packstone and packstone/grainstone that rapidly passes up into the Cwmyniscoy Mudstone.

Cwmyniscoy Mudstone

The Cwmyniscoy Mudstone (Barclay, in press) is generally poorly exposed, and comprises dark grey mudstones with subordinate thin bioclastic limestones and calcareous siltstones. However, limestones are dominant in the top 5 m or so. The thickness varies from 40 to 46 m in the north and central parts of the district to about 60 m in the south. The junction with the underlying Castell Coch Limestone is taken at the lowest bed of fully marine mudstone above the oolitic and bioclastic limestones of the latter.

The mudstones are dark grey, silty, shaly and micaceous. They contain variable but subordinate laminae, and very thin beds (generally less than 20 mm) of laminated and cross-laminated calcareous siltstone, in part grading to silty, very fine-grained packstone. The mudstones are generally sparsely fossiliferous, though bedding planes strewn with brachiopods have been noted. Extensively burrowed beds are common.

Within the mudstones, the limestones form discrete, parallel-sided beds of variably silty, skeletal packstone, up to 0.24 m thick, with sharp bases. Many are graded, with coarse crinoidal basal lags containing subordinate brachiopods and scattered bryozoa, passing up into fine-grained laminated and cross-laminated skeletal packstone, in places with crinoid-rich laminae. Burrowing is common at the tops of the beds. Some of the coarser packstones contain fine-grained packstone intraclasts and rolled fragments of the colonial coral Vaughania. A few packstone beds consist dominantly of thin-shelled brachiopods. Many of the finer-grained packstones are silty and dark coloured, with disseminated pyrite. Units of amalgamated limestone beds are present in places, and can be up to 2.5 m thick, as in the lowest part of the formation at Barry. The limestones are locally affected by late-stage epigenetic dolomitisation.

Biostratigraphy

The Devonian–Carboniferous boundary lies within the uppermost part of the Quartz Conglomerate Group at Tongwynlais (Gayer and others, 1973) (see page 22). Conodonts obtained from the Lower Limestone Shale Group at Castel] Coch Quarry [ST 1300 8264] and from the coast section at Barry [ST 1353 6707]–[ST 1360 6706] are listed in (Figure 11) and those from Barry Harbour in (Figure 14). They were also obtained from the lower part of the Tongwynlais road section [ST 1302 8245]. The faunas from the Tongwynlais Formation, Castell Coch Limestone, and at least the lowest 7 m of the Cwmyniscoy Mudstone include Polygnathus spicatus, P. inornatus and rare siphonodellids, and are referable to the Siphonodella Zone of Groessens (1976). These faunas are typical of the shallow water facies of the zone, that for the most part lack the critical siphonodellids on which the refined early Dinantian siphonodellid zonation of Sandberg and others (1978) is based. However, Siphonodella cf. isosticha, indicative of the Siphonodella crenulata Zone which is equivalent to the youngest part of Groessens' Siphonodella Zone, is present at the base of the Cwmyniscoy Mudstone at Castell Coch Quarry.

In the early Courceyan in some parts of Britain and Ireland, Polygnathus spicatus, Patrognathus and bispathodids occur in high energy, shallow water deposits which are followed sharply by a transgressive packstone and shale facies with Polygnathus inornatus and siphonodellids (Varker and Sevastopulo, 1985). Such a relationship occurs at two levels in the conodont biostratigraphy of the district. In the Tongwynlais road section, Bispathodus aculeatus aculeatus, B. aculeatus–Clydagnathus transition, Patrognathus variabilis, Polygnathus communis communis, P. spicatus and Spathognathodus cf. stabilis were obtained from a composite channel sample between 1.5 m and 3 m above the base of the Tongwynlais Formation. The Rhiwbina Ironstone yielded Bispathodus a. aculeatus, Patrognathus variabilis and Polygnathus spicatus. Although no higher levels were sampled during the resurvey, Gayer and others (1973) described Siphonodella–P. inornatus Assemblage Zone (Rhodes and others, 1969) faunas including siphonodellids from the upper 23 m of the formation (they did not sample the remaining 12 m above the Rhiwbina Ironstone). Polygnathus inornatus was also recorded from the uppermost part of the formation at Barry. However it is absent in the barrier facies of the Castell Coch Limestone, but again present in the lagoonal and overlying beds at the top of the Castell Coch Limestone at Castell Coch Quarry and 1 m above the base of the Cwmyniscoy Mudstone at Barry where it persists for at least 7 m into the Cwmyniscoy Mudstone.

The two intervals with Polygnathus inornatus and siphonodellids are associated with two major transgressions (see page 4/3) that established offshore mudstone and packstone facies; one begins at the base of the Rhiwbina Ironstone, and the other at the base of the Cwmyniscoy Mudstone. This strongly suggests that the distribution of the conodonts is facies controlled, and throws doubt on the correlation by Gayer and others (1973) of the part of the Tongwynlais Formation containing Siphonodella–P. inornatus Assemblage Zone faunas with the Lower Limestone Shale that postdates the Bryozoa Bed in the Avon Gorge.

The middle part of the Cwmyniscoy Mudstone is not exposed and its age is uncertain. However, a sample 10 m below the top of the formation at Barry Harbour yielded a shallow water conodont fauna similar to those from the interzone between the Siphonoddla Zone and the Pseudopolygnathus multistriatus Zone of Varker and Sevastopulo (1985).

The macrofauna of the group is of limited value in correlation, but the eponymous taxon of the Vaughania vetus Assemblage Zone has been collected together with the brachiopods Cleiothyridina royssii and Pugilis vaughani from the upper part of the Cwmyniscoy Mudstone at Barry Harbour [ST 1043 6650]. The macrofauna of the type section of the Tongwynlais Formation has been described by Gayer and others (1973).

Details

Tongwynlais Formation

At Pant-y-caeau [ST 1057 8101], south of Pentyrch, Strahan and Cantrill (1912) described a trial for hematite in hematitic crinoidal limestone. This is considered to be the Rhiwbina Ironstone.

The topmost 1 m of the formation, including the junction with the Castell Coch Limestone, are exposed in a road section [ST 1101 7479] north of St Lythans (Figure 10) and a lane section [ST 1406 7333]–[ST 1411 7339] at Cwrt-yr-ala. The base of the latter section is about 1 m stratigraphically above the top of Cwrt-yr-ala Borehole [ST 1408 7339].

On the coast [ST 1368 6704] at Barry, north-east of the Bendrick Rock, the top of the formation and its junction with the Castell Coch Limestone is exposed (Figure 10).

Castell Coch Limestone

Between Creigiau and south of Pentyrch the formation is exposed in numerous small disused quarries and pits. The best complete section is immediately north of the district at Castell Coch Quarry [ST 1300 8264] (Figure 10), where 0.7 m of lagoonal beds is present in the upper part of the formation. The lagoonal beds are overlain by 1.84 m of reddened fine-grained dolomite with an erosional base, exhibiting large globular load casts. This in turn is overlain by 1.08 m of partly dolomitised skeletal packstone. A road section [ST 1087 7480] - [ST 1101 7479], north of St Lythans, exposed the upper and lower parts of the formation, and at Barry, north of the Bendrick Rock, the formation is exposed on the foreshore and cliffs [ST 1334 6703]–[ST 1368 6704] (Figure 10).

Cwmyniscoy Mudstone

The junction with the Barry Harbour Limestone is seen in a disused quarry [ST 0902 8117] at Pant-y-gorcd, south-west of Pentyrch, in a road section [ST 1087 7480], north of St Lythans (Figure 10), and at Palmerston [ST 1358 6925]. At Barry Harbour [ST 1043 6650] the following sequence is exposed on the foreshore between the base of the Barry Harbour Limestone and the Cold Knap Fault:

Black Rock Limestone (Group)

Lithostratigraphy

The Black Rock Limestone (Group) comprises skeletal, dominantly crinoidal limestones, thickening from 124 m in the north of the district to about 500 m in the south. The basal formation, the Barry Harbour Limestone, is thin bedded and exhibits hummocky cross-stratification and planar- and cross-lamination. The middle formation, the Brofiscin Oolite, thins rapidly southwards, and has not been mapped south of Michaelston-le-Pit. The Friars Point Limestone, the uppermost formation, is mainly thickly bedded, and generally lacks tractional lamination. Where the Brofiscin Oolite is absent, the distinctive Yorke Rock Bed, a thin unit of cross-bedded crinoidal limestone, defines the top of the Barry Harbour Limestone. However, this bed has not been identified in the poorly exposed St Andrews Major area, and here the group remains 'undivided'.

The group is completely dolomitised in the north of the district, but only the top 70–100 m of the Friars Point Limestone are prevasively affected south of the axis of the Cardiff–Cowbridge Anticline. Below this level, patchy epigenetic dolomitisation is locally present. Throughout the district, exposure is generally good.

Barry Harbour Limestone

The Barry Harbour Limestone comprises thin-bedded, dominantly fine but in part coarse-grained, commonly richly crinoidal, skeletal packstone with scattered thin beds and partings of shaly calcareous mudstone. Thin beds of oolite occur in the upper part of the formation in the north. Replacive chert lenses and silicification of the macrofauna are common. The limestones exhibit planar- and cross-lamination as well as hummocky cross-stratification: low-angle cross-bedding is present in places. On the northern limb of the Cardiff–Cowbridge Anticline the formation is 32 m thick, and pervasively dolomitised apart from the basal 7 m at the south-western end of the crop in the Pant-y-gored area [ST 092 812]. It thickens southwards to about 50 m on the southern limb of the anticline, and reaches about 80 In on the coast at Barry.

The type locality is the low cliffs and reefs [ST 1004 6646]–[ST 1043 6630] on the western side of the old Barry Harbour. The base of the formation is fairly sharp, being taken at the point where mudstone interbeds form only a minor part of the succession. The formation forms a prominent scarp above the Cwmyniscoy Mudstone slack. The top of the formation is taken at the sharp base of the Brofiscin Oolite where it is present, or at the top of the Yorke Rock Bed.

Two major lithofacies, A and B, are developed within the formation, the latter being restricted to the north of the district. Lithofacies A comprises predominantly graded beds of dark grey, skeletal packstone up to 0.45 m thick, but generally thinner, either amalgamated or separated by thin grey mudstone partings and beds. Each bed has an erosive base. Many begin with brachiopod- and crinoidal-rich lags up to 0.3 m thick, and grade up through fine-grained skeletal packstone, becoming argillaceous upwards, to mudstone. The fine-grained skeletal packstones exhibit undulatory lamination, some of which is hummocky cross-stratification (Harms and others, 1975), planar-lamination and cross-lamination, commonly developed in this order upwards. The upper parts of the beds are commonly extensively burrowed. Silicification is common, either as replacement of individual bioclasts in the lags, or as elongate replacive chert lenses up to several decimetres in length.

Lithofacies B consists of very fine- to fine-grained, pale to dark grey, partly laminated and cross-laminated secondary dolomite, in beds 8 cm to 0.9 m thick. Macro-skeletal debris is extremely sparse, and mainly confined to a few thin, locally oolitic, Shelly, crinoidal bryozoan lenses and beds up to 0.15 m thick, which are commonly silicified. The silicification predates the dolomitisation, the original limestone textures being preserved. Scattered burrows occur throughout.

The distribution of the two major lithofacies within the formation is shown in (Figure 12). At Barry, mudstone beds are more common in the lower third of the formation, one reaching 1 m in thickness. In the uppermost part of the formation below the Yorke Rock Bed, thin units of fine-grained, skeletal packstone up to 0.1 m thick, form single sets of planar cross-bedding.

The Yorke Rock Bed is a 3–5 m thick, coarse, well sorted crinoidal grainstone, equivalent to the upper part of the Brofiscin Oolite; its type locality is the reefs adjacent to Yorke Rock in Barry Harbour [ST 1071 6641]. The bed exhibits low angle, locally herringbone cross-bedding, with sets about 0.1 m thick, and both its boundaries are gradational. It consists mainly of crinoid ossicles, with some brachiopod and bryozoa fragments, micritic intraclasts and peloids. The micritic intraclasts are irregular in shape, but have rounded edges. They consist of either one or more skeletal fragments with micrite in the skeletal chambers and encasing them, and appear to have been derived from a micritic wackestone or packstone in which early cementation has been concentrated around the skeletal fragments. The crinoid ossicles commonly have micritic rims due to endolithic algal boring, and exhibit pronounced syntaxial overgrowths that form much of the spar content.

Between Creigiau and the Taff valley, lithofacies A forms the lower part of the formation and is largely dolomitised, except for the lowest 6.7 m around Pant-y-gored. A bed of dolomitised oolite up to 4 m thick, with relic ooids, occurs in the lower part of the overlying lithofacies B.

Brofiscin Oolite

The Brofiscin Oolite comprises predominantly ooid and in part skeletal grainstone. It has sharp junctions with both the Barry Harbour Limestone and the Friars Point Limestone. The type locality is Brofiscin Quarry [ST 0685 8122] (Figure 13), just west of the district. On the northern limb of the Cardiff–Cowbridge Anticline it is completely dolornitised, and is 12.6 to 16.4 m thick. Southwards, it thins to 8.6 m in the St. Lythans area, on the southern limb of the anticline. South of Michaelston-le-Pit it is thin or absent and has not been mapped.

The ooid grainstones are pale to purplish grey, and contain scattered crinoid ossicles and brachiopod fragments. They commonly exhibit local secondary reddening. Where only the ooid cortices are affected, this facilitates recognition of the formation in the dolomite belt between Creigiau and the Taff valley, for the dolomitisation post-dates the reddening. The dolomitised oolite appears as fine- to medium-grained dolomite. Relic ooids occur in two ways: as red-stained spheres, commonly with a white centre, scattered in a paler dolomite matrix, or as dark grey dolomite spheres in a paler dolomite matrix. However, some dolomite beds retain no oolitic texture. In thin section, the dolomites consist of a fairly equant mosaic of dolomite rhombs, commonly. zoned, 40–120 m in diameter. Ooids can be identified by impurities in their cortices that ghost their original shape. The formation is massive to thickly bedded in units 1–1.3 m thick. Planar cross-sets are locally present even in the dolomitised area.

Friars Point Limestone

The Friars Point Limestone lies with sharp junctions between the Brofiscin Oolite or Barry Harbour Limestone below, and the Gully Oolite above. The type locality is at Friars Point [ST 1070 6641]–[ST 1115 6586], Barry Island, where all but the top part of the formation is exposed. The formation thickens markedly across the Vale of Glamorgan Axis, being 76 to 83 m thick to the north between Creigiau and the Taff valley, and about 265 m thick to the south in the St Lythans/St Andrews Major inliers; farther south, it thickens to at least 408 m on the coast between Barry and Sully.

The formation consists predominantly of thickly bedded, grey to black, foetid, fine-grained, skeletal packstone with subordinate thin interbeds of shaly, argillaceous, skeletal packstone. Crinoid debris is abundant. North of the Cardiff–Cowbridge Anticline, the formation is completely dolomitised, apart from up to 20 m of the lower part in the west. South of the anticline only the uppermost 70 to 110 m are dolomitised. Dolomitisation of the uppermost part of the Black Rock Limestone is widespread in the South-West Province, where it has been variously termed laminosa Dolomite (Vaughan, 1905; Dixey and Sibly, 1918; George, 1933), Black Rock Dolomite (Kellaway and Welch, 1955; Whittaker and Green, 1983) and Langland Dolomite (George and others, 1976). As the dolomites are secondary, they have not been given separate lithostratigraphical status in this account, but have been depicted on the map as dolomitised Friars Point Limestone.

Lithofacies similar to those in the Barry Harbour Limestone occur in the lowermost part of the formation (Figure 12). In the north, between Creigiau and the Taff valley, lithofacies B, which is about 5 m thick and without cherts, sharply overlies the Brofiscin Oolite. It is overlain by about 10 m of dolomitised and poorly exposed lithofacies A. In the St. Lythans area, at least 9 m of lithofacies A occur above the Brofiscin Oolite. They are seen in Greenwood Quarry, but are dolomitised. At Barry, 21 m of lithofacies A lie above the Yorke Rock Bed.

The remainder of the formation gradationally overlies these basal developments. It comprises a fairly uniform lithofacies, termed lithofacies C, which consists of thick units of grey to black foetid packstone and packstone/wackestone up to 1.8 m thick, separated by thin, shaly packstone/wackestone beds, up to 0.3 m thick though generally thinner. The thick units are composite, comprising coarse bioclastic, dominantly crinoidal lags, interbedded with very fine- to fine-grained, partly streaky, laminated, skeletal packstone and subordinate packstone/wackestone. Individual beds within the composite units are up to 0.15 m thick. At the tops and bottoms of the packstone units, wispy, non-sutured stylolites (Wanless, 1979) and stylonodules (Logan and Semeniuk, 1976) are common.'Bioturbation, commonly as centimetre-sized burrows, varies from scattered examples to virtual homogenisation where only traces of bedding are preserved as clots and wisps of coarse bioclastic material floating in fine-grained packstone. Thalassinoides-type burrows are common in the shaly packstone/wackestone interbeds.

Petrographically, the packstones of lithofacies C consist of biotic fragments of all sizes, in a matrix of finer-grained biotic material, irregular micrite patches, variably abundant micrite pellets and microspar. The microspar is neomorphic and has absorbed the finest-grained skeletal material and micrite. Some sections show variable amounts of dolomite of similar grain-size to the microspar. Scattered silt-sized grains of angular quartz are abundant. Replacive chert nodules are only present in the lithofacies in a 10 m packet of beds in the lower part of the sequence at Barry (Figure 12).

The fauna of the lithofacies is diverse, and includes crinoids, fenestellid bryozoa, brachiopods and gastropods. Zaphrentoid corals occur thoughout, and large caninioid corals make their first appearance just above the base of the lithofacies. Large Syringopora colonies, up to 0.45 m in diameter, also occur. Most of the corals are not in life position.

Lithofacies C has been subdivided at Barry into two subfacies which occur in stratigraphical sequence (Figure 12). The lower one (Ci) is thinner-bedded, with more shaly partings; the bedding is fairly well defined within the packstone units, as the burrowing is not intense. In the upper lithofacies (Cii), the packstone units are thicker (up to 1.8 m), and virtually homogenised internally due to intense burrowing (Plate 3) and (Plate 4). This vertical sequence of lithofacies has not been noted to the north.

The dolomites comprising the main part of the formation north of the Cardiff–Cowbridge Anticline and the top 70–110 m to the south (Black Rock Dolomite of authors) can be recognised as dolomitised lithofacies C, for fabrics and allochem ghosts are commonly preserved. South of the anticline, the lower boundary with the undolomitised limestone is very gradational, there being a transition zone, commonly about 15 m thick, consisting of interbedded dolomites, dolomitic limestones and limestones. The area depicted on the map as dolomite excludes the transition zone, and the base of the dolomites is irregular, not following a stratigraphical horizon.

The dolomites are grey to dark grey, and fine to very fine grained. The macrofauna is variably preserved as hollow moulds, calcite or dolomite. Calcite geodes up to 3 cm diameter are locally common. The shaly partings in the undolomitised beds have generally been obliterated, but bedding-parallel sutured stylolites give the dolomites a well bedded appearance.

In the uppermost part of the formation north of the Cardiff–Cowbridge Anticline, dolomitised lithofacies C is overlain by about 5 m of fine-grained dolomites, generally lacking crinoids, and exhibiting faint undulatory and low-angle laminations; this comprises lithofacies D. It is in turn overlain by a red dolomitic clay up to 0.15 m thick, that rests on a slightly undulating surface. These two features are suggestive of a palaeosol and a palaeokarstic surface respectively, but most of the other diagnostic features are absent perhaps as a result of the dolomitisation. The clay is sharply overlain by dolomitised Gully Oolite. South of the Cardiff–Cowbridge Anticline dolomitised lithofacies C passes up, apparently gradationally, into dolomitised, fine-grained, skeletal packstones with thin oolitic beds of the Gully Oolite. However, a thin, red, calcareous, shaly clay, up to 6 cm thick, is present within this transition zone, and the top of the clay is taken as the top of the Friars Point Limestone.

Biostratigraphy

Correlation within the Black Rock Limestone of the South-West Province can be achieved by coral faunas, but conodonts and foraminifera assemblages are more valuable for correlations with the rest of Britain, Ireland and Belgium. The group ranges in age from mid-Courceyan to early Chadian.

The collecting of macrofossils was concentrated on the thickest sequence in the section exposed between Barry Harbour [ST 1004 6646]–[ST 1043 6630] and Friars Point [ST 1070 6641]–[ST 1115 6586], and the distribution of corals is given in (Figure 14). Preliminary results of the coral biostratigraphy of this section were outlined by Mitchell (1981). Macrofossil collecting was also undertaken at Sully Island to cover the top part of the group which is not exposed at Friars Point, and inland at Whitehall Quarry in the St Lythans Ether. No coral faunas were recorded from the dolomitised sequence north of the Cardiff–Cowbridge Anticline. The three distinctive Black Rock Limestone coral faunas, described by Mitchell and Green (in Green and Welch, 1965, table 1) from Buffington Combe in the Mendips and subsequently given formal assemblage -biozonal names by Ranisbottom and Mitchell (1980), have been recognised in the district. However, the correlation between coral and conodont biostratigraphy suggested here differs in part from that suggested by previous authors.

At Barry, the Barry Harbour Limestone falls entirely within the Zaphrentites delanouei Zone, the top of which occurs some 45 m above the base of the Friars Point Limestone. In addition to the zonal coral, Fasciculophyllum omaliusi, Michelinia favosa and Syringopora vaughani are also present. The small zaphrentoid Sychnoelasma clevedonensis is restricted to the highest beds referred to this zone, as it is in many of the Black Rock Limestone sections that have been studied elsewhere (Mitchell, 1981, p.580, fig. 1).

The other two coral zones are within the Friars Point Limestone. The Caninophyllum patulum Zone is apparently about 230 m thick, but this figure may have been overestimated because of a thrust in the section. The zone fauna appears abruptly in a 30 cm bed (Collecting No.7), and typical assemblages with Caninia cornucopiae, Cyathaxonia cornu, Cyathoclisia tabernaculum and Sychnoelasma konincki have been collected throughout. The range of the subspecies of Caninophyllum patulum follows the pattern described from all sections in the Bristol and South Wales region, with C. patulum greeni restricted to the lower part of the zone and being replaced (with no overlap) in the upper part of the zone by C. patulum patulum.

Some of the species common in the C. patulum Zone range up into the Siphonophyllia cylindrica Zone the base of which is taken at the first appearance of S. cylindrica at 276.24 m above the base of the formation (Collecting No.25) at Friars Point, where only 24 m of the zone are exposed.

At Sully Island the lowest 2 m of the section yield a C. patulum Zone fauna, and a S. cylindrica Zone fauna was obtained from beds 40 m above the base. The S. cylindrica Zone is, therefore, at least 80 m thick here, of which 40 m are pervasively dolomitised. In the St Lythans Inlier at Whitehall Quarry, S. cylindrica Zone faunas first occur 10 m above the base of the section and the remaining 102 m seen belong to this zone. About 30 m of Friars Point Limestone are present above the exposed section so the biozone is at least 132 m thick. It is likely that it is of similar thickness between Barry and Sully.

At Barry Harbour the other macrofaunal elements of the Barry Harbour Limestone include Buxtonia sp., Cleiothyridina royssii, Macropotamorhynchus mitcheldeanensis, orthotetoids, Pugilis vaughani, Rugosochonetes vaughani, smooth spiriferoid, Syringothyris cf. cyrtorhyncha, Unispirifer tornacensis and bryozoa. Those of the Friars Point Limestone include Buxtonia sp., Cleiothyridina glabristria, Leptagonia analoga, Macropotamorhynchus mitcheldeanensis, Megachonetes cf. magna, orthotetoids, Pustula sp., Rhipidomella michelini, Rugosochonetes vaughani, Schizophoria sp., smooth spiriferoid, Spiriferellina sp., Syringothyris sp., Tylothyris sp., U. tornacensis, Bellerophon sp., Euomphalus sp., Straparollus sp., turreted gastropods, bryozoa and trilobites.

Various attempts (Ramsbottom and Mitchell, 1980; Whittaker and Green, 1983) have been made to use the Tournaisian conodont zonation established in the distal ramp setting of the Dinant syncline in Belgium (Groessens, 1976) in the more proximal setting of the British South-West Province. However, as the conodont faunal distribution across the ramp appears to be partly facies or depth controlled, some of the correlations have been misleading (Clayton and Sevastopulo, 1981). The zonal and subzonal guides present in the immediate neighbourhood of the Waulsortian reefs (distal ramp) in Belgium tend to disappear as the shoreline is approached (Lees, 1982, fig. 5). Polygnathus communis carina extends farther than Scaliognathus anchoralis, but it appears increasingly later after the disappearance of the siphonodellids as it is traced shorewards, thus creating an expanding interzone. The zonal sequence used here is that suggested by Varker and Sevastopulo (1985) for the region north of the South-West England basinal sequences, and the position of the Belgian zonal and subzonal guides in the Cardiff succession are simply noted.

Conodonts have been obtained from Barry (Barry Harbour and Friars Point) (Figure 14), from the thin, largely dolomitised succession at Brofiscin Quarry [ST 0685 8122] (Figure 13), and from the Llantrisant road section [ST 0576 8094]–[ST 0566 8115] (Figure 15), the last two localities being situated immediately west of the district.

The Barry Harbour Limestone everywhere contains shallow water faunas that are of limited value for dating, but are typical of those that occur between the top of the Siphonodella Zone and the base of the overlying Pseudopolygnathus multistriatus Zone.

The base of the P. multistriatus Zone occurs 5–6 m below the top of the Brofiscin Oolite at Brofiscin Quarry, and at the base of the Yorke Rock Bed at Barry. The base of the Polygnathus mehli Zone occurs 6.5 m above the base of the Friars Point Limestone in Brofiscin Quarry at the incoming of lithofacies A, and at a similar level in the Llantrisant road section. However, at Barry the base of the P. mehli Zone occurs between 50 no and 73 m above the base of the Friars Point Limestone in lithofacies C.

The Belgian index Polygnathus c. carina first occurs in the uppermost part of the Brofiscin Oolite at Brofiscin Quarry, but at Barry its first appearance is 42 m above the base of the Friars Point Limestone about 10 m above the base of the deeper water lithofacies C and near the base of the C. patulum Zone. The facies control on P. c. carina can be demonstrated by the fact that farther south, in the more distal setting of the Knap Farm Borehole, Cannington (Mitchell and others, 1982), it first appears low in the Z. delanouei Zone. The eponymous fossil of the Eotaphrus cf. bultyncki Subzone has been noted 45–50 m above the base of the Friars Point Limestone at Barry, and that of the E. bultyncki Subzone 16–21 m above the base in the Llantrisant road section. The Dollymae bouckaerti Subzone is indicated by the presence of the eponymous species at the base of the cherts at Barry, 110 m above the base of the formation. As this subzone is never very thick, the base of the overlying S. anchoralis Zone is probably between samples 10 and 11 at Barry, but S. anchoralis has been recovered only in sample 20. In the Llantrisant road section, the base of the S. anchoralis Zone is probably about 37 m above the base of the formation for S. praeanchoralis, recorded here between 37 and 41 m, usually occurs just below and in the early part of the zone.

The placing of the Tournaisian- Viséan boundary is problematical. Mestognathus beckmanni, whose incoming defines the base of the M. beckmanni Zone, is the only Viséan taxon present. It has been recorded in sample 36 at Barry and in the north at Heol Goch Quarry [ST 123 821] 8 m below the top of the formation, but was not found in the Llantrisant road section. The highest Tournaisian conodont recorded at Barry and in the Llantrisant road section is P. c. communis recorded in sample 20 at the former (and tentatively as high as sample 25 on the evidence of a poorly preserved specimen) and 51 m above the base of the formation at the latter. M. cf. beckmanni, a taxon intermediate between the Tournaisian M. groessensi and the Visean M. beckmanni, first occurs in sample 27 at Barry and 15 to 20 m below the top of the formation in the Llantrisant road section; it is known to span the Tournaisian-Visean boundary, but its exact range is at present uncertain. In view of this, it is safer to suggest on conodont evidence that the base of the Visean lies somewhere within the range of M. cf. beckmanni as recorded in the district, and on balance at the lower end of that range. Correlation of the Tournaisian- Visean boundary within the Courceyan and Chadian stages remains uncertain.

Evidence from the Tears Point section in south-west Gower (Mitchell and others, 1986) indicates that the base of the Visean should be taken at the base of the Siphonophyllia cylindrica Zone which is drawn below sample 25 m the Barry section.

Details

Creigiau to the Taff valley

The Barry Harbour Limestone is exposed in crags [ST 0907 8118]–[ST 0928 8124] north of Pant-y-gored where lithofacies A is only partly dolomitised, the fine-grained packstones having been preferentially affected, and the base of the formation is exposed in a small pit [ST 0902 8117]. Completely dolomitised lithofacies A is seen in a small quarry [ST 0977 8117] at Pentwyn Farm. In Cwm-y-fuwch Quarry [ST 1038 8122] lithofacies B overlies dolomitised lithofacies A:

There are complete sections through the Brofiscin Oolite at the southern end of the disused railway tunnel [ST 1238 8194] at Morganstown, where it is 16.4 m thick, and in the disused Ty-Nant Quarry [ST 1270 8210] where the following sequence has been recorded:

A section low in the Friars Point Limestone is exposed in a disused quarry [ST 0886 8127] south-east of Creigiau:

The base of the formation is exposed in a disused quarry [ST 1031 8155] south of Pentyrch, where 1.88 m of lithofacies B, comprising medium grey, fine- to medium-grained dolomite with extremely sparse crinoid debris sharply overlies dolomitised Brofiscin Oolite. The base and a considerable thickness of the formation (completely dolomitised and structurally complex) is also seen in the unlined railway tunnel [ST 1240 8200]–[ST 1246 8222] west of Morganstown. Heol Goch (or Taff's Well) Quarry [ST 123 821] exposes the junction with the Gully Oolite and much of the Friars Point Limestone in a structurally complex sequence. It is estimated that the formation is 76 m thick between the railway tunnel and the quarry. The quarry section shows the following sequence below the Gully Oolite:

The Barry Harbour Limestone is in lithofacies A in this area. A small disused quarry [ST 1080 7453] at Downs exposes 2.13 m of the lithofacies. At Greenwood Quarry [ST 1140 7420] some 16 m of dolomitised lithofacies A is exposed below the Brofiscin Oolite.

The only complete section (8.6 m) through the Brofiscin Oolite is in Greenwood quarries [ST 114 742] NNE of St Lythans, where, due to the proximity of a major fault, there is variable epigenetic dolomitisation. Where completely dolomitised the Oolite weathers to a soft dolomite sand along joint planes.

In the northernmost [ST 1140 7420] of the three disused Greenwood quarries, the Friars Point Limestone is well exposed with lithofacies C overlying lithofacies A:

The middle quarry [ST 115 741] exposes about 54 m of lithofacies C with some local epigenetic dolomitisation. The southernmost quarry [ST 1160 7375] exposes about 16 m of the same lithofacies with particularly pronounced shaly partings. The disused Whitehall Quarry [ST 1170 7340] exposes lithofacies C, of S. cylindrica Zone age, passing up gradationally into dolomites that have replaced the topmost 110 m of the formation in this area. The section in the south face is as follows:

A similar transition into the dolomitised upper part of the formation is exposed at St Andrews Quarry [ST 143 713]. The St Lythans Borehole [ST 1054 7290] penetrated 20 m of dolomites at the top of the formation.

Barry to Sully

A complete section through the Barry Harbour Limestone (80 m thick) is seen at the type locality [ST 1004 6646]–[ST 1043 6630] in the low cliffs and reefs on the western side of Barry Harbour (Figure 12).

Many of the fine-grained packstones are selectively dolomitised. On the eastern side of Barry Harbour, at Yorke Rock [ST 1071 6641], the upper 25 m of the formation are exposed. Much of the formation is also exposed at Bendrick Rock [ST 1310 6680], where it is completely dolomitised except for the Yorke Rock Bed.

The succession in the type section [ST 1070 6641]–[ST 1115 6586] of the Friars Point Limestone at Friars Point, Barry Island, is shown in (Figure 12). The dolomitised upper part of the formation is not seen at Friars Point, but is calculated to be a maximum of 100 m in the Sully area. On Sully Island [ST 1671 6685]–[ST 1698 6694] about 80 m of partly dolomitiscd, dark grey, crinoidal packstones pass up into about 40 m of dark grey, fine-grained dolomites with scattered quartz veins containing hematite and galena. The packstones on Sully Island contain Caninophyllum patulum patulum, Cyathoclisia tabernaculum, Fasciculophyllum densum, Rugosochonetes sp.and Schizophoria sp.in the basal 8 m referable to the Caninophyllum patulum Zone, and Siphonophyllia sp.(cyclindrica group), C. patulum patulum, F. densum and Megachonetes sp.40–44 m above the base, referable to the S. cylindrica Zone.

Gully Oolite

Lithostratigraphy

The Gully Oolite comprises thick-bedded and commonly cross-bedded oolite, with subordinate beds of fine-grained, skeletal, peloidal packstone, and is 19–25 m thick in the north, increasing southwards to 83 m. South of the Cardiff–Cowbridge Anticline, it conformably overlies the Friars Point Limestone, but to the north it rests on an emergent surface at the top of the latter.

In the north, between Creigiau and the Taff valley, it is cut by abundant wide veins of late diagenetic dolomite, so the Ethological sequence is difficult to establish. It appears to consist mainly of pale grey ooid grainstone with very subordinate thin beds of fine-grained, skeletal, peloidal packstone. Where dolomitised, the ooid grainstones become a distinctive pale grey, commonly reddened, very coarse dolomite, in which the oolitic texture is mainly obliterated but cross-bedding is commonly preserved. These dolomites contain zebroid contraction fractures, and are locally preferentially mineralised by calcite, hematite or barytes.

South of the Cardiff–Cowbridge Anticline, the formation is 70 to 83 m thick, and the full sequence was proved in St Lythans Borehole (Waters, 1978b, 1984) (Figure 16). It consists of four coarsening-upward cycles, each beginning with thinly interbedded ooid packstone and skeletal, peloidal packstone, and passing up via thicker-bedded ooid packstone and packstone/grainstone with some skeletal/peloidal packstone into cross-bedded ooid grainstone. In the lower two cycles oolitic lumps (Flugel, 1982; termed composite ooids by Waters, 1984), comprising aggregates of ooids cemented by micrite, are very common and in places are the dominant allochem. In the upper parts of the top two cycles, calcrete profiles are manifested mainly by rhizoliths (Klappa, 1980; organosedimentary structures produced by roots). Dolomitisation within the formation is variable, but commonly affects the fine-grained, skeletal, peloidal packstones and fine-grained packstone matrix of ooid packstones in the lower part of the sequence; the lower two cycles in St Lythans Borehole are affected in this way. Beds immediately above the Friars Point Limestone are usually dolomitised.

North of Wenvoe, the top of the formation marks an important and widely-recognised non-sequnce (Riding and Wright, 1981; Wright, 1982), commonly referred to as the mid-Avonian unconformity (Dixon and Vaughan, 1912). The top of the oolite consists of an irregular and pitted surface overlain by an oolite breccia/conglomerate (a rubble of oolite clasts in a green clay matrix), passing up into mottled green clay. Textures and fabrics diagnostic of calcrete occur both in the uppermost part of the oolite and in the breccia/conglomerate, and include rhizoliths and laminated micrite crusts (Wright, 1982). The irregular surface and overlying breccia/conglomerate and clay are interpreted as a palaeokarst/palaeosol, the breccia/conglomerate clasts being regolith material. The profile becomes more mature and thicker as it is traced northwards, the palaeosol being 1.7 m thick in the Taff valley and 0.1 m in St Lythans Borehole. As the palaeosol represents a subaerial modification of the Gully Oolite, it is here included within the formation, in contrast to Riding and Wright (1981) who, at Miskin in the Bridgend district, placed it in the Caswell Bay Mudstone. The top of the Gully Oolite is drawn below the transgressive base of the Caswell Bay Mudstone.

In contrast, south of Wenvoe, the formation is sharply overlain by the High Tor Limestone without the intervening Caswell Bay Mudstone or any signs of emergence. At the contact with the High Tor Limestone, individual ooids in the Gully Oolite are planed off.

Biostratigraphy

The formation has a limited brachiopod fauna, including orthotetoids. Michelinia megastoma, Syringopora sp.and bellerophontoid gastropods were noted in St Lythans Borehole (Figure 16), either in the skeletal packstones of the lower part of the cycles or in skeletal-rich lags in the oolites.

Foraminifera and dasycladacean algae obtained from the top of the formation, in a small quarry [ST 1125 7168] on the east side of Goldsland Gorge, include ?Biseriella bristolensis, Palaeospiroplectammina sp.and primitive, single-walled Koninckopora, indicative of a Chadian age.

Details

Creigiau to the Taff Valley

A complete section of the Gully Oolite in Heol Goch Quarry [ST 1220 8220] shows:

Although the top palaeosol profile is completely dolomitised, all the macroscopic features are preserved. Its base is a well defined rolling surface, with a relief of 0.3 m, like a mamillated palaeokarstic surface (Walkden, 1974). Pipes up to 4 cm wide, filled with the clay, penetrate the oolite for a depth of 1.4 m, and lead off to thin horizontal clay-filled fissures. The breccia-conglomerate consists of a greenish/grey, shaly clay, grading to clayey dolomicrite with clasts of dolomitised oolite (now pale grey dolomite with locally preserved oolitic texture) from granule- to boulder-size. The larger blocks are angular, and commonly penetrated by clay-filled fissures. The larger oolite clasts have reddened, partly laminar, accretionary and replacive rinds. Smaller clasts are completely replaced by red dolomicrite, and have suffered jig-saw' brecciation. Samples from near the top of the breccia/conglomerate have pink calcrete glaebules within the clay. The micro-breccia at the top of the profile consists of grey dolomicrite nodules up to granule-size, set in a dolomicrite matrix and surrounded by irregular occluded fenestrae. Also present are scattered black chips and pebbles of altered oolite, locally with laminar rinds. Circumgranular cracks and loss of detail in the ooid clasts have been noted.

St Lythans Inlier

A disused quarry [ST 0290 7394] south of St. Nicholas showed:

The micrite crusts at the top of the oolite are best developed in complexly filled cracks up to 10 mm wide that occur to a depth of 4 cm at right-angles and subparallel to the top surface. These represent the occluded voids between oolite bedrock and platy clasts which have lifted off slightly. The clasts are riddled with small rhizoliths. The bases of the cracks are commonly lined with pale grey fenestral and peloidal micrite crust, as well as small oolite fragments which are similarly coated. Locally, the cracks, especially where they widen out, contain numerous calcrete ooids and pisoids as geopetal fills. The lower crust complex comprises a horizontal zone, 0.16 m thick, of some six or so similar fenestral and peloidal laminated micrite beds with scattered calcrete ooids. This represents either a lower calcrete profile or a calcretised fissure system below the main palaeokarst. Rhizoliths occur in the oolite between the micrite beds.

The St Lythans Borehole [ST 1054 7290], (Waters, 1978b) proved a complete section (Figure 16), comprising four coarsening upward cycles (A–D). Each cycle begins with interbedded fine-grained, skeletal, peloidal packstone and ooid packstone, and passes upwards via ooid packstones to ooid grainstones. Oolitic lumps are common in the lower two cycles.

Cycle A comprises three units, 1–3. At the base of Unit 1, 0.19 m of fine-grained, variably dolomitised, skeletal packstone, with thin beds containing crinoid ossicles and orthotetoids, overlies the clay at the top of Friars Point Limestone. The contact is steeper than the overlying bedding. In the rest of the unit, which is variably dolomitised, fine-grained, skeletal, peloidal packstone beds are up to 0.17 m thick, diffusely laminated, and contain scattered sharp-based beds up to 4 cm thick, rich in brachiopods and crinoid ossicles. They are interbedded with ooid, ooid lump packstone beds that are sharp-based, commonly graded, and up to 0.2 m thick. The second unit is thicker bedded, and partly cross-bedded with beds of ooid, ooid lump packstone and packstone/grainstone up to 1. 8 m thick which contain streaks and laminae of fine-grained, skeletal, peloidal packstone. Interbeds of fine-grained, skeletal, peloidal packstone are laminated or cross-laminated, and up to 0.3 m thick. The third unit comprises ooid, ooid lump grainstone to packstone/grainstone, with disarticulated orthotetoid brachiopods in laminae or in nests, commonly convex up, with fine-grained skeletal packstone in the underlying shelter cavities. Thin laminae of fine-grained, skeletal, peloidal packstone are sparse. The top of the cycle is abrupt.

Cycle B is similar to the previous cycle, consisting of three units. In the lowest unit, beds of fine-grained oolitic skeletal packstone, 2.5 cm thick, are interbedded with laminated, cross-laminated or structureless fine-grained, skeletal, peloidal packstone. Unit 2 contains scattered laminae and beds of fine-grained, skeletal, peloidal packstone up to 0.3 m thick throughout the ooid, ooid lump packstones and packstone/grainstones which are largely destratified due to burrowing. Centimetre-wide vertical burrows infilled from above are common. In the third unit, cross-bedding is preserved.

Cycle C also consists of three units. The lowest is again thinly-bedded, comprising fine-grained, skeletal, peloidal packstone beds up to 0.37 m thick, though generally thinner, interbedded with oolite. The former are mainly burrowed or feebly laminated in the lower half, and laminated or in places cross-laminated in the upper half of the unit. The oolite beds are up to 0.55 m thick, but averaging 0.1–0.2 m. The beds have sharp bases, and are commonly graded with basal lags of crinoid, shell debris and skeletal packstone intraclasts. In the lower part of the unit they are structureless to crudely laminated, or exhibit low-angle cross-bedding, and in the upper part they are laminated, cross-laminated or cross-bedded, with sets up to 0.15 m thick. Vertical burrows infilled from above are common. Some vertical escape burrows with 'V' structure have also been noted. Towards the top of the unit is a graded conglomerate, 0.6 m thick, with many subangular to subrounded intraclasts of oolite and skeletal, peloidal packstone up to cobble-size in an oolite matrix. The oolite pebbles exhibit truncated ooids at their margins. The overlying unit is mainly fine to coarse oolite but contains a few beds of fine-grained, oolitic, skeletal, peloidal packstone/grainstone. The main oolite beds are cross-laminated and cross-bedded, sets averaging 0.1 m though some are up to 0.2 m. Most sets are heterogeneous. In the lower half of the unit some oolite beds are conglomeratic with scattered oolite intraclasts. The third unit is mainly fine grained though some coarser beds of oolite occur. The lower half of the unit shows grain-size banding and low-angle cross-bedding which, near the top of the lower half, becomes diffuse and locally disappears as rhizoliths become common. The top of the lower half of the unit is a sutured stylolite, which may mask an erosion surface. The remainder of the unit is mostly structureless, with rhizoliths in the upper two thirds. The top of the unit is an erosion surface with a subvertical fissure, at least 0.5 m deep, penetrating the underlying beds, and infilled by paler and coarser oolite from the overlying cycle.

Cycle D is comparable to the upper part of Cycle C and comprises two units. In Unit 1 which is mainly cross-bedded ooid grainstone with grain-size banding, there is a basal skeletal-rich lag containing black micrite fragments (? from rhizoliths) and green mudstone clasts. About 1 m below its top, there are about four pale grey, fine-grained carbonate submarine crusts, in all 2 cm thick. The base of these crusts overlies a surface on which the ooids of the underlying bed are planed off. The individual crusts are 1–2 mm thick, and have irregular to crinkly micro-relief. Oolite between the crusts is darker, with spar-filled laminar fenestrae. In the upper crusts, depressions up to 6 mm deep have been infilled by paler oolite, and a new crust formed on the top. Above the crust complex the oolite is coarse, very dark coloured, and still cross-bedded. In the top of the unit the ooids are coated with an early isopachous fibrous fringing cement. The top of the unit is a channelled erosion surface along which both ooids and early cement have been planed off. In Unit 2 the low-angle cross-bedding and grain-size handing becomes more diffuse upwards. In the upper half it is structureless with rhizoliths throughout. However, an immature palaeosol is developed below a green clay seam towards the top of the unit. It comprises fine-grained oolite riddled with small rhizoliths with individual ooids floating in a green-grey micrite matrix. The cycle is capped by a 0.1 m clay palaeosol, which sits abruptly on a flat palaeokarstic surface of oolite that is solution-fretted and blackened, but there is no micrite crust. The oolite is reddened slightly in the topmost 8 cm below the palaeokarst.

In the borehole, most of the rhizoliths are 1–3 mm in diameter, circular in cross-section, lined with grey or black laminar micrite and occluded by sparry calcite. Some branch downwards. Larger rhizoliths are more complex, comprising tubes 1–2 cm in diameter, lined with laminar, commonly fenestral, micrite. The axial part of the tube is filled with homogeneous micrite with scattered globular, spar-filled fenestrae. A characteristic of these larger rhizoliths is the presence of dark, micritic haloes, up to 3 cm wide, which are due to the growth of micrite within the host sediment, mainly between the ooids.

A section in a track and disused quarry [ST 1124 7163] north-west of Goldsland shows:

Numerous small sections in the upper part of the formation occur to the north along the track, and exhibit cross-bedded ooid grainstone and very subordinate peloidal, skeletal packstone.

Caswell Bay Mudstone

Lithostratigraphy

The Caswell Bay Mudstone comprises up to 7 m of thin-bedded 'calcite mudstones', some rather argillaceous, and shales with a very restricted fauna. The 'calcite mudstones' vary from peloidal and bioclastic micrites, wackestones and packstones, to fenestral cryptalgal laminites. The formation is seen only to the north of Wenvoe, where the top of the Gully Oolite has suffered emergence; to the south, the High Tor Limestone rests directly on the Gully Oolite. Between Creigiau and the Taff valley it is poorly exposed except in Heol Goch Quarry where it varies from 3.2 to 6.9 m thick, and is completely dolomitised. South of the Cardiff-Cowbridge Anticline it is also poorly exposed, but was proved to be 1.64 m thick in St Lythans Borehole. The base of the formation is taken at the sharp transgressive contact with the palaeosol at the top of the Gully Oolite. The top of the formation is well defined by the erosive base of the thick-bedded bioclastic grainstones and packstones of the High Tor Limestone.

The only sequence available that is not completely dolotnitised is that of St Lythans Borehole (Figure 17). This shows an irregular base to the basal bed which comprises fine-grained dolomite with scattered angular pebble-sized clasts derived from the underlying palacosol; it is laminated in the upper part. Apart from cryptalgal laminites, the rest of the formation comprises thin beds, varying from structureless micrite to peloidal and skeletal wackestones and packstones. Colours vary from grey to green. Bedding is thin, and is commonly disrupted by burrowing. Some beds are very argillaceous, and there are complete gradations to fissile shales that form thin partings and beds. The allochems in the wackestones and packstones are mainly peloids and ostracods, hut bivalve fragments and coquinas of modioliform bivalves are present in some beds, as well as scattered calcispheres. Small, turreted gastropods and crinoid material occur in the topmost bed, together with fairly abundant ooids. Lamination is brought out by grain-size banding, and most wackestone and packstone beds and laminae are sharp based. Cut-and-fill structures are present at the base of some beds. Some beds are dolotnitised to dolomicrite.

Thin units of cryptalgal laminites (Aitken, 1967), with small laminar and globular fenestrae, occur interbedded with thin beds of micrite intraclast packstone and thicker beds of structureless micrite with tubular and globular fenestrae. Thin beds of peloidal wackestone are also present. Laminae and beds are disrupted by desiccation features.

Biostratigraphy

Although the restricted fauna contains no diagnostic fossils the formation has been regarded as either Chadian (Ramsbottom, 1973; George and others, 1976) or Arundian (Riding and Wright, 1981) on regional seclimentological grounds (sec p.49).

Details

The Caswell Bay Mudstone is very poorly exposed between Creigiau and the Taff valley, and does not even form a feature. However, it is completely exposed in Heol Goch Quarry [ST 1210 8220], where it comprises 3.2 to 6.9 m of dark to pale grey dolomicrite, partly laminated in beds 0.1 to 0.38 m thick, with dark grey dolomitic shale partings.

In the St Lythans Inlier, the formation crops out in small exposures, and is distinctive in field brash. In St Lythans Borehole, it is 1.64 m thick (Figure 17).

High Tor Limestone

Lithostratigraphy

The High Tor Limestone consists of thin- to thick-bedded, bioclastic limestone with a few thin beds and partings of shaly dolomite mudstone and siltstone. In most of the district it transgressively overlies the Caswell Bay Mudstone, but south of Wenvoe it rests sharply on the Gully Oolite. The formation passes up gradationally into the Cefnyrhendy Oolite, the top being taken where oolitic lithologies become dominant. It crops out between Creigiau and the Taff valley, where it is 65 m thick.

Little is known in detail of the sequence between Creigiau and the Taff valley, for it is poorly exposed and becomes increasingly dolomitised as it is traced eastwards. East of Pentyrch it has not been separated from the lower dolomitised part of the Hunts Bay Oolite.

In the St Lythans Inlier to the south, where it is 86 m thick, exposure is generally poor. The lower half of the formation was penetrated in the St. Lythans Borehole (Figure 18), and the upper part is exposed in numerous pits and crags to the west. The formation comprises five distinctive lithofacies in the St Lythans Inlier, arranged broadly in a stratigraphical sequence.

The basal lithofacies (A) is 7 m thick and comprises massive to well bedded, dominantly coarse, peloidal, skeletal packstone/grainstone and grainstone with subordinate thin beds of fine-grained skeletal packstone, 0.5 to 5 cm thick. The former are well sorted, and appear as 'pseudo-oolites' in the field; in thin section, however, they consist mainly of bioclasts, with thick, pale cream to white, micrite envelopes, and peloids mostly after micritised bioclasts. Coarse crinoid debris is common throughout, as are skeletal/peloidal, fine-grained packstone intraclasts up to pebble size. Burrowing is demonstrated by clots of fine-grained skeletal packstone in the coarse packstone/grainstone. Low-angle cross-bedding and lamination occur in places.

Above lithofacies A, the formation is dominated by packstones which are either medium grey and thick-bedded (lithofacies B) or dark and thin-bedded (lithofacies C). The latter is restricted to the lower part of the formation, where it occurs in a unit 9 m thick, above two prominent units of dolomite mudstones (Figure 18) that comprise a fourth, but minor, lithofacies (D). Lithofacies B comprises packstone units up to 4 m thick, separated by thin beds or partings of dolomite or dolomitic mudstone. The units mostly comprise pale grey to grey, fine-grained, skeletal packstone, with sparse scattered macrofossil material, notably crinoids, brachiopods and gastropods. Apart from a few scattered coarse beds, the units are mainly structureless, due to burrowing commonly manifest as clots of coarser material and mottling. However, in the upper half of the formation, a few isolated, low-angle cross-beds occur. Coarse beds fall into two types: (a) macrofossil-rich beds with a fine-grained packstone matrix; (b) better sorted skeletal packstone/ grainstone, commonly peloidal. They are 10 mm to 0.25 m thick, sharp-based, and commonly grade up into fine packstone. Some exhibit crude grain-size lamination. Scattered rounded, fine-grained skeletal packstone clasts occur in some beds. At their tops most packstone units fine up into a thin calcisiltite or dolomite/dolomitic mudstone bed or parting.

Lithofacies C consists of thin, dark grey bituminous, fine-to very fine-grained skeletal packstones, in beds 1 cm to 0.36 m thick, separated by thin argillaceous wackestone and shale beds and partings. The packstones contain scattered crinoids, brachiopods and some gastropods. The beds are sharp-based, some are graded, and all become argillaceous upwards. The argillaceous wackestone partings are 5 mm to 6 cm thick, and contain swarms of wispy black clay seams, comparable to non-sutured microstylolites (Wanless, 1979). The scams commonly define stylonodules. Some beds, however, grade both upwards and downwards into the argillaceous wackestone parting, and in a number of cases a thin shale, up to 3 cm thick, is also present. It is evident therefore that pressure solution has blurred many of the original bed junctions. Burrowing, seen as clots of coarser packstone in the shale or argillaceous partings, is common.

Lithofacies D occurs in units up to 1.6 m thick. The dolomite mudstones are dark grey, weathering to lilac and pale green; they are variably argillaceous and shaly, and grade to dolomitic mudstone. A few very thin dolomitised crinoidal packstone ribs are present in some units. Chondrites is widely present and other, larger burrows also occur. A few beds of dolomitic calcisiltite and argillaceous calcite mudstone are present but uncommon.

In the uppermost part of the formation, interbedded, coarse, well sorted, skeletal, peloidal packstone/grainstone, and fine-grained skeletal packstone appear as the highest lithofacies (E), and grade up into the Cefnyrhendy Oolite. They are bedded on a scale of 0.15 to 1 m, and exhibit planar and undulatory lamination as well as low-angle cross-bedding.

Biostratigraphy

The High Tor Limestone contains typical Arundian assemblages (Figure 19) with rich suites of algae and foraminifera, including Koninckopora inflata, Archaediscus sp . at the involutus stage, Eoparastaffella simplex, Glomodiscus miloni and G. oblongus . The lower part of the formation has a coral fauna with Cravenia cf. lamellata, Michelinia megastoma and Siphonophyllia garwoodi (diagnostic of Arundian), and the upper part contains Delepinea carinata, a brachiopod diagnostic of the middle part of the Arundian.

Details

Creigiau to the Taff valley

Between Creigiau and exposures [ST 0960 8180] just west of Pentyrch, numerous crags and small pits e.g. [ST 0880 8150]; [ST 0894 8156]; [ST 0890 8153] expose thick-bedded skeletal packstones with local development of 'slacks' probably due to dolomite mudstone units; patchy dolomitisation is present. The latter locality yielded .Siphonophyllia garwoodi , Linoprotonia cf. hemisphaerica, Syringothyris sp., Bellerophon sp., Straparollus sp., Archaediscus sp.(at the 'involutus' stage), Earlandia sp., Forschia parvula, Koninckopora inflata and Plectogyranopsis convexa. To the east of this locality, the High Tor Limestone is very poorly exposed, and east of Pentyrch is completely dolomitised and cannot be separated from the overlying dolomitised lower part of the Hunts Bay Oolite. Up to 3 m of dark grey, partly red-stained, fine- to medium-grained crinoidal dolomite were exposed above the Caswell Bay Mudstone in Heol Goch Quarry [ST 1220 8220].

St. Lythans Inlier

The top of the basal lithofacies (A) is seen in a disused quarry [ST 1023 7404] at Vianshill:

Two small disused quarries [ST 1050 7291] east of Nant-bran lie in the middle of the formation and typify lithofacies B. Here, 3 m of fine-grained skeletal packstones with scattered coarser beds show some low-angle cross-bedding, and one 0.6 m bed is rich in gastropods. The adjacent St Lythans Borehole proved 47.29 m of the formation above the Caswell Bay Mudstone (Figure 18). Similar thick-bedded skeletal packstones in crags [ST 1033 7298] east of Nant-bran Farm yielded Composita ambigua, Delepinea carinata, Linoprotonia sp., Megachonetes cf. hemisphaerica, M. ef. papilionaceus, Syringothyris elongata, Bellerophon sp., Earlandia sp., Eoparastaffella simplex, Forschia parvula, Glomodiscus oblongus, G. miloni, Koninckopora inflata, Mediocris breviscula and M. mediocris.

A cutting [ST 1019 7295] at Nant-bran in the uppermost part of the formation, exposes 6 m of lithofacies E comprising interbedded partly reddened, dolomitic, laminated, fine-grained, skeletal pack-stone, and coarse, reddened, crinoidal peloidal grain stone/packstone, which are laminated and exhibit low-angle cross-bedding.

A small disused quarry [ST 1124 7163], north-west of Goldsland Farm, exposes the basal lithofacies (A) which yielded Axophyllum sp., Cravenia ef. lamellata, Michelinia megastoma, Siphonophyllia garwoodi, Linoprotonia cf. hemisphaerica, Megachonetes sp.and Syringothyris sp.The section is as follows:

Hunts Bay Oolite (Group)

Lithostratigraphy

The Hunts Bay Oolite (Group) is a dominantly oolitic sequence, with subordinate skeletal, peloidal and oncolitic limestones and calcite mudstones, that crops out mainly between Creigiau and the Taff valley where it is 180 m thick. Southwards, it is exposed only at the western extremity of the St Lythans Inlier and in nearby small inliers where the highest beds are only just above the basal member, the Cefnyrhendy Oolite. This member is present both north and south of the Cardiff–Cowbridge Anticline, and is 13–25 m thick. In the north, much of the lower part of the group is affected by pervasive dolomitisation and the upper part is affected by abundant late diagenctic, commonly epigenetic, vein-dolomitisation. The topmost beds are missing in the north due to Namurian overstep.

The well established twofold division of the group (Dixon and Vaughan, 1912; George, 1933) into predominantly oolites below, and a heterogeneous assemblage above, comprising oolitic, skeletal and peloidal limestones, and characterised by the presence of calcite mudstones and oncoids, has been formally defined in the Bridgend district as the Cornelly Oolite and Stormy Limestone respectively (Wilson and others, in press) (Figure 8). However, these two subdivisions cannot be recognised in the district because of a facies change in the eastern part of the Bridgend district.

The Cefnyrhendy Oolite is a massive, partly cross-bedded ooid grainstone with scattered bioclastic material. Peloids are abundant at some levels. The junction with the underlying High Tor Limestone is not seen in the district, but at the type locality in Hendy Quarry [ST 0480 8130] it is gradational. The top of the member is also not exposed, but at Hendy Quarry is a palaeokarst overlain by a thin clay palaeosol. In the north, the member is 13 m thick, but it could not be identified for more than half a kilometre east of Creigiau due to poor exposure and obliteration by dolomitisation. In the St Lythans area it is about 25 m thick.

Above the Cefnyrhendy Oolite east of Creigiau, dolomitised beds, probably equivalent to the Argoed Limestone Member (Wilson and others, in press) are exposed in Creigiau Quarry [ST 0868 8162]. At Argoed Isha Quarry [ST 0677 8129], the type locality, the Argoed Limestone, consists of 7 m of thin- to medium-bedded, fine-to coarse-grained, crinoidal, skeletal packstones, locally rich in Lithostrotion martini, with a distinctive calcareous mudstone at the base, resting sharply on the Cefnyrhendy Oolite.

East of Creigiau about 37 m of dolomites are exposed above the level of the Argoed Limestone. They are grey, commonly reddened, well bedded, and vary from fine to coarse grained. The fine- and medium-grained types contain brachiopods and crinoids, suggesting that they were originally fine- and medium-grained skeletal packstones. However, the coarser dolomites exhibit no relic textures, but probably represent dolomitised ooid grainstone which commonly gives rise to a coarse grain-size on dolomitisation. East of Pentyrch they have not been separated on the map from the dolomites forming the upper part of the High Tor Limestone and Cefnyrhendy Oolite.

About 125 m of heterogeneous thick-bedded limestones, dominated by grainstones but including packstones and micrites, occur above. The sequence is cut by numerous thick veins of late diagenetic, commonly epigenetic dolomite. The grainstones are very variable. Allochems include ooids, micrite peloids from silt- to sand-size, bioclastic debris and intraclasts, especially of micrite, biomicrite and fenestral micrite. Oncoids are common, with cores formed of intraclasts or brachiopods. Grainstone composition varies from bed to bed and within beds, various allochem combinations being present. The commonest lithologies are ooid peloidal grainstone and oolitic intraclast peloidal grainstone. Oncoids are common in the coarse intraclast beds. Many of the peloidal and micrite intraclast grainstones have previously been termed 'chinastone'. The packstones tend to be dominantly bioclastic, but the range of allochems seen in the grainstones is present in some beds. Micrites occur as thin beds from about a centimetre to 0.4 m thick, and may be structureless, fenestral or laminated. Cryptalgal lamination is present in some beds, but digitate stromatolites (Aitken, 1967), occurring as black micrite 'fingers' within paler micrite, are more common. Thrombolitic stromatolites have also been noted. Micrites are more abundant in the upper part of the sequence.

Biostratigraphy

Foraminifera and algae from the type section of the Cefnyrhendy Oolite include Archaediscus aff. varsanofievae, Melarchaediscus sp., Eoparastaffella aff. restricts, Koninckopora tenuiramosa and Rectodiscus sp. , an assemblage that is characteristic of the late Arundian (Figure 19). Lithostrotion martini, Palaeosmilia murchisoni (common) and Linoprotonia hemisphaerica are present in the Cefnyrhendy Oolite and also indicate an Arundian age. The overlying Argued Limestone contains Holkerian foraminifera including Archaediscus at the concavus stage, Koskinotextularia sp., Nodosarchaediscus sp. and Nibelia nibelis.

The remainder of the Hunts Bay Oolite has a rich macro-fauna, of which Axophyllum vaughani, Lithostrotion aranea, Davidsonina carbonaria and Linoprotonia corrugatohemispherica are diagnostic of the Holkerian (Figure 19).

Details

The two westernmost Creigiau quarries provide a composite succession through the Hunts Bay Oolite (Group) with some overlap. The two sections are:

From small quarries and crags [ST 090 820] immediately north-east of the two large quarries noted above, a composite faunal list from the Hunts Bay Oolite above the Argoed Limestone includes Chaetetes depresses, C. septosus, Axophyllum vaughani, Haplolasma cf. subibicina, Lithostrotion cf. sociale, Syringopora cf. reticulata, Composita ficoidea, Davidsonina carbonaria, Gigantoproductus sp., Linoprotonia corugatohemispherica, Megachonetes sp.(papilonaceus group), Plicochonetes sp., Productus sp. and Pustula sp.

Crags [ST 0894 8163] 310 m south-west of Pen-llwyn expose 1.2 m of slightly skeletal ooid grainstone of the Cefnyrhendy Oolite with Palaeosmilia murchisoni, Linoprotonia cf. hemisphaerica, Syringothyris sp.and Straparollus sp.

The Cefnyrhendy Oolite is exposed in crags [ST 0965 7425] at Dyffryn House, which show about 22 m of pale grey, variably skeletal, ooid grainstones. Ooid lumps and ooid/peloid packstone and micrite intraclasts occur locally. Much of the skeletal debris has micrite envelopes. The fauna includes P. murchisoni, Syringopora cf. reticulate, Linoprotonia sp.and Straparollus sp.

Dinantian–conditions of deposition

The Tongwynlais Formation marks the continuation of a marine transgression (Major Cycle 1 of Ramsbottom, 1973), initiated during the deposition of the uppermost part of the Quartz Conglomerate Group, and migrating northwards onto St. George's Land. Correlation of the basal unit of the Tongwynlais Formation between Tongwynlais and Cwrt-yrala is best achieved by an examination of the event stratigraphy, which suggests that the transgression was made up of at least three pulses.

At Cwrt-yr-ala, the marine mudstones at the base of the coarsening upward sequence below the lagoonal beds point to a rapid drowning of the coastal/delta plain that was already receiving marine deposits during periodic storm events. However, this transgression appears to have been short-lived, the appearance of oolitic and peloidal sheet-like beds at the top of the coarsening upwards sequence suggesting shoaling culminating in the initiation of the overlying lagoon. The next pulse of the transgression is demonstrated by the overlying marine succession. The basal erosion surface beneath the packstone sheets with their basal lag of reworked micrite pebbles is probably a ravinement surface (Swift, 1968), resulting from the transgressive passage of a barrier shoreface across the lagoon. This transgressive pulse also soon waned, for the sequence rapidly coarsens upwards from offshore mudstones into a grainstone shoal. Although the latter is capped by an iron-stained crust, there is no evidence for emergence at this level. The overlying mudstone with the basal ironstone containing grainstone pebbles marks the final transgression.

There are also three pulses in the Tongwynlais sequence. The first is manifested by the dominantly fine-grained packstone sequence, probably representing an interdistributory bay/distal mouth bar complex resting sharply on fluviatilc channel sandstones of the Quartz Conglomerate Group. Evidence for rapid shoaling is provided by the appearance of the coarse, mixed carbonate/siliciclastic packstone/grainstones that quickly became emergent, as demonstrated by the calcretc. Repeated additions of similar sediment to a background of green alluvial mudstone, and repeated calcrete profiles, suggest periodic ?storm events driving marine sediment across the coastal delta plain. The base of the overlying lagoonal deposits can be interpreted as transgressive over the underlying calcretised sediments. The third transgressive pulse is demonstrated by the erosion surface at the base of the Rhiwbina Ironstone with its basal lag of lagoonal micrite pebbles. This surface, and the lag, have been interpreted as a ravinement horizon (Burchette, 1977, 1981).

It is likely that the first transgression at Cwrt-yr-ala and Tongwynlais is the same event, but the Cwrt-yr-ala lagoon probably correlates with the calcrete levels at Tongwynlais.

The Tongwynlais lagoon was probably developed behind a barrier formed by the grainstone shoal at the top of the second transgressive sequence at Cwrt-yr-ala. The overlying ironstone would then correlate with the Rhiwbina Ironstone, and together these would mark the beginning of the third transgressive pulse.

The ferruginised bioclasts comprising the ironstone are reworked, for they occur together with non-ferruginised bioclasts in high-energy grainstone lithologies and storm-generated packstone sheet sands. The fauna suggests that they were derived from the nearshore shelf in a specialised environment where iron minerals could be precipitated. It is likely that the primary iron mineral was chamosite, later oxidised to hematite during reworking (Burchette, 1977), and not pyrite as suggested by Squirrell and Downing (1969). As the reworked ferruginised bioclasts occur in the lower part of the transgressive sequences, it is probable that iron precipitation occurred during regression or stillstand. This is further supported by the position of the iron-stained crust capping the 'regressive' ooid grainstone shoal at Cwrt-yr-ala, for this appears to have been one of the sites of iron precipitation.

The main part of the formation above the Rhiwbina Ironstone and its correlative was deposited in a muddy shelf environment that supported a rich marine fauna. The parallel-sided, graded packstone beds show all the features to suggest that they are carbonate analogs of sublittoral sheet sandstones (Goldring and Bridges, 1973) deposited by storm events (Don and Bourgeoise, 1982).

The cross-bedding and grainstone lithologies of the Castel] Coch Limestone are typical of a high-energy shoal environment. It is likely that this regional shoaling event at the top of the Tongwynlais Formation is due to the southward pro-gradation of shoal complexes (Burchette, 1981) rather than a major marine regression. A lagoon locally developed behind the barrier complex in the north of the district.

The Cwmyniscoy Mudstone marks renewed transgression. (In the north, the skeletal packstones overlying the lagoonal beds within the top of the Castel] Coch Limestone represent the basal beds of the Cwmyniscoy Mudstone transgression but are included within the former formation for convenience). The graded, sheet-like packstone beds of the Cwmyniscoy Mudstone are similar to those in the Tongwynlais Formation, and of similar origin. The fact that they are generally thinner and much less abundant than in the latter suggests that the Cwmyniscoy Mudstone was deposited farther out on the shelf. Evidence of shallowing is provided by the increased number of storm-generated units in the uppermost part of the formation, and the rapid transition into the Barry Harbour Limestone.

The packstones of lithofacies A, that everywhere form the lower part of the Barry Harbour Limestone again show all the features of storm-generated event deposits. However, the fact that the mudstone interbeds are extremely thin and decrease in abundance upwards, suggests that the lithofacies was deposited not far below normal wave base. Lithofacies B, lacking the abundance of coarse bioclastic lags and macro-fauna, notably crinoids, suggests a semi-restricted environment perhaps locally developed to the north of the Vale of Glamorgan Axis. Final shoaling to within wave base is manifested by the appearance of the cross-bedded, high-energy Brofiscin Oolite, whose southward-thinning suggests that it was a southward-prograding shoal complex. Farther south on the shelf, the upward shoaling cycle is capped by the Yorke Rock Bed, which marks a brief period of deposition within wave base, probably at the acme of Brofiscin Oolite progradation.

The Friars Point Limestone marks a major transgression in which deeper shelf conditions were rapidly established. In the south, both the Yorke Rock Bed and the Brofiscin Oolite are overlain by the relatively deeper lithofacies A. However lithofacies A did not arrive in the north until considerably later, for the upper part of the Brofiscin Oolite and thin overlying lithofacies B are equivalent in age to the lower part of the Friars Point Limestone at Barry. This suggests that the upper part of the Brofiscin Oolite is transgressive in the north. The general lack of traction current structures, apart from winnowed lags in the overlying lithofacies C, suggests that it was the result of slow in situ accumulation of bioclastic debris and carbonate mud below wave base, but periodically stirred by storm activity which resulted in a winnowing of the fines and a concentration of the coarse material into lags (Specht and Brenner, 1979). The environment appears to have been deeper and more distal than that of lithofacies A. The reason for the increase in bioturbation that gave rise to the initiation of lithofacies Cii at Barry is not clear: it may possibly broadly represent the early S. anchoralis Zone regressive event recognized in the outer shelf facies by Lees (1982).

Mitchell (1981) suggested that the base of the C. patulum and S. cylindrica zones are paraconformities (breaks of major time significance). As evidence, he pointed out that the base of the former at Barry is represented by an irregular rolling bedding-plane above which large caninioid corals appear for the first time. However, irregular bedding-planes are a common feature of many of the packstone/shaly wackestone bed-junctions which are also commonly modified by pressure solution; nor is there any evidence for a hardground at this level. The sudden appearance of caninioid corals is more probably associated with the establishment of the quieter, more muddy conditions of lithofacies C. Similarly, the evidence for a major break at the base of the S. cylindrica Zone is not convincing, there being no lithological change or hardground at this level either.

North of the Cardiff–Cowbridge Anticline, the appearance of lithofacies D at the top of the Friars Point Limestone represents shallowing to within wave base followed by emergence suggested by the probable palaeokarst and palaeosol. South of the Cardiff–Cowbridge Anticline, the lack of discernible evidence for shallowing at the top of the dolomitised Friars Point Limestone and the rapid gradation into the overlying Gully Oolite suggest a conformable and progradational junction, although the origin of the red clay remains enigmatic. Ramsbottom (1973) took the top of his Major Cycle 1 at the top of the dolomite at the top of the Black Rock Limestone and noted a faunal non-sequence (Mitchell, 1972) between Major Cycle 1 and Major Cycle 2 as the succession was traced northwards across the Mendip–South Wales shelf. There is no evidence for such a non-sequence in the Cardiff conodont fauna. The later suggestion by Ramsbottom (1977) that there is a major cycle (mesothem) boundary coincident with the Courceyan/Chadian junction is difficult to verify as the position of the latter junction is uncertain (see p.38).

The dolomitisation at the top of the Friars Point Limestone south of the Cardiff–Cowbridge Anticline has been discussed by Hird and others (1984). More than one phase has been recognised, later phases enhancing earlier ones. The earliest phase is due to meteoric influence probably in a mixing zone. (Hanshaw and others, 1971; Folk and Land, 1975), rather than to seepage reflux (Adams and Rhodes, 1960) from the north as suggested by Ramsbottom (1973). Emergence at both the top of the Friars Point Limestone and later at the top of the Gully Oolite could have allowed an input of meteoric water into the formation. The last stage of dolomitisation probably occurred during post-Dinantian burial.

The southward thickening of the Black Rock Limestone comprises a twofold increase in the Barry Harbour Limestone (including the Brofiscin Oolite in the north) and a fivefold increase in the Friars Point Limestone. The thickening appears to be due to hinged movement along the Vale of Glamorgan Axis, probably in response to a major fault at depth (Waters, 1984), rather than to northward attenuation due to beds missing at the top of the group and internal paraconformities (Mitchell, 1981).

The early- to mid-Chadian regression responsible for emergence north of the Cardiff–Cowbridge Anticline probably initiated the Gully Oolite to the south by lowering wave-base. This allowed ooid shoals to become established and progade southwards, thus accounting for the gradational junction with the Friars Point Limestone, notwithstanding the enigmatic red clay. The lower parts of cycles A and B appear to represent offshore skeletal, peloidal sands, deposited within storm wave-base, and downcurrent from an ooid shoal periodically receiving oolitic material perhaps during storm events. The middle parts were probably deposited on a shoal margin, and the upper part was deposited close to the active zone. The sharp tops to the cycles reflect shoal abandonment.

The lower part of Cycle C again comprises offshore, storm-generated event beds, conglomeratic material being derived from early cemented oolite sediment (hardground) (Dravis, 1979) by erosion during such events. The middle and upper parts of Cycle C appear to represent the transition from shoreface to beach deposits, the low-angle cross-bedding in the upper part being comparable with the swash cross-stratification of modern foreshores, and the overlying levels with rhizoliths, represent the subaerial colonisation of beach deposits by plants in a semi-arid climate. Cycle D is of similar origin to the middle and upper parts of the previous cycle. The junction between cycles C and D appears to be a modified palaeokarst, and the base of Cycle D is undoubtedly transgressive.

The cycles represent at least four separate progradations south of the Vale of Glamorgan Axis. In contrast, to the north of the axis, the thin grainstone-dominated sequence marks the passage of a transgressive shoal complex over the emergent Friars Point Limestone. This transgression may correspond with the base of Ramsbottom's (1977) mesothem D2b. However, rapid progradation led to the area soon becoming emergent again; by late Chadian, a palaeokarstic surface must have stretched from the Taff valley to just south of St Lythans; it represents the mid-Avonian unconformity. It has been suggested that pulsed subsidence south of the axis led to the progradation (Waters, 1984). It is likely that the transgression that eventually drowned the emergent Friars Point Limestone north of the axis equates with either Cycle C or Cycle D, or both.

Following the transgression at the base of the Caswell Bay Mudstone (Riding and Wright, 1981), that flooded the karst platform at the top of the Gully Oolite, a lagoonal/peritidal environment with calcite mudstone lithologies was established behind a temporary barrier. The lack of emergent features in the peloidal/skeletal packstones and wackestones suggests a subtidal environment, probably in a shallow lagoon, the sharp-based packstone and wackestone beds and laminae probably representing storm-generated units. The beds associated with the fenestral, cryptalgal laminites are intertidal by comparison with modern, flat, laminated, algal-bound sediments (Ginsberg, 1975). The basal High Tor Limestone lithofacics (A) represents a relatively high energy nearshore environment. It has been suggested that it records the northward passage of the barrier (Riding and Wright, 1981), whereby barrier shoreface retreat during the transgression reworked the upper part of the Caswell Bay Mudstone and all the barrier sediments. Whether this ravinement is responsible for the absence of the Caswell Bay Mudstone south of Wenvoe is not clear. The lack of evidence of emergence at the top of the Gully Oolite here may mean that this area remained subtidal throughout, or that the evidence has been removed during ravinement.

The rapid transition into the thin-bedded, dark packstones of lithofacies C via units of dolomite mudstone (lithofacies D) suggests a deepening environment as the transgression progressed northwards up the shelf. The commonly graded beds in lithofacies C are probably of storm origin. Although bioturbation has destroyed most of the sedimentary structures in the overlying thick-bedded packstones of lithofacies B, enough is present to suggest that each unit of the lithofacies represents an in-situ accumulation of skeletal material and carbonate mud, laid down below wave-base but periodically stirred by storm events. The decrease in the abundance of steal interbeds, and the generally cleaner packstones, suggest a shallower environment than lithofacies C. The appearance of the low-angle, cross-bedded and undulatory laminated skeletal, peloidal packstone/grainstone (lithofacies E) at the top of the formation appear to represent the progradation back across the shelf of a higher-energy shoreface environment. This shoaling event continued with the Cefnyrhendy Oolite, a high-energy shoal complex that eventually became emergent in the north of the district.

The Holkerian transgression (Major Cycle 4 of Ramsbottom, 1973) drowned the emergent Cefnyrhendy Oolite in the north to deposit the Argoed Limestone. The latter probably represents a partly barred environment with restricted circulation, rather than open shelf conditions. West of the district, an ooid shoal complex represented by the Cornelly Oolite was established, and gradually prograded southwards with the back barrier lagoonal and peritidal deposits of the Stormy Limestone accumulating behind (Wilson and others, in press). In the district dolomitisation makes an assessment of the group difficult, but the presence of peritidal micrites and oncolitic beds throughout suggests a lagoonal/peritidal environment with a variety of sub-environments, including inter- and supra-tidal flat, tidal channel and delta, washover fan and deep lagoon.

Chapter 5 Upper Carboniferous (Silesian)

Within the district, the Upper Carboniferous rocks are restricted to those of Namurian age, which are traditionally referred to as the 'Millstone Grit'. Namurian strata crop out in two small poorly exposed areas near Creigiau. As a result of uplift along the Usk Axis (George, 1956), first shown by the appearance of east–west facies changes in the Holkerian, little or no strata were deposited in the district from the late Holkerian until the late Marsdenian (R_2c transgression (Mesothem N10 of Ramsbottom, 1977). Thus late Namurian strata in the district rest unconformably on Hunts Bay Oolite, though without noticeable angular discordance. The Namurian of adjacent districts is described by Woodland and Evans (1964) and Squirrell and Downing (1969), and more regional accounts include those of Jones (1974) and Kelling (1974).

Only the lowermost 20 m crops in the district; these comprise grey mudstones with a thin bed of quartzitic sandstone up to 0.3 m thick at the base. The Bilinguites superbilinguis Marine Band (R_2c) occurs at the base and Yeadonian (G₁) faunas are known immediately to the north in the Newport district (Squirrell and Downing, 1969).

Details

The basal unconformity is exposed in Creigiau Quarry [ST 0880 8205], hut is currently inaccessible. However, a section [ST 0870 8203] was recorded by Squirrell and Downing (1969) and is given below:

Chapter 6 Triassic

Following the Hercynian deformation in South Wales there was a prolonged period of erosion in the district that continued through Permian into late Triassic times. During this period tensional faulting developed in relation to the opening of the Atlantic (Hallam, 1971), and the Central Somerset basin was established (Whittaker, 1973; Whittaker and Green, 1983) together with its probable western extension along the Bristol Channel Syncline (Whittaker, 1975). The Triassic succession in the Cardiff district was deposited just beyond the northern margin of this basin, defined by South Wales and the Mendips, and is considerably thinner than the sequence within the basin.

Triassic sedimentation in the district was governed by the Triassic topography. The latter was dominated by the effects of erosion on the Cardiff–Cowbridge Anticline; Carboniferous Limestone and Upper Old Red Sandstone on the two limbs of this fold formed major ridges, and soft Lower Old Red Sandstone was easily eroded along its axis to form a major valley that connected south-eastwards with the main Triassic basin (Figure 20) and was limited north-eastwards by Silurian strata. In the south-west of the district isolated hills of Carboniferous Limestone were present at Barry, Cadoxton and Sully.

Only late Triassic deposits are preserved in the district. They dominantly comprise up to 200 m of lacustrine and continental sediments, of the Mercia Mudstone Group, that progressively onlap over the irregular Triassic topography. Towards the end of the Triassic period there was a change from lacustrine to marine conditions, and with continuing onlap the whole district was finally covered by late Triassic strata. This dominantly marine sequence, up to 12 m thick, is known as the Penarth Group and comprises mudstones with thin subordinate limestones. The Triassic–Jurassic boundary occurs in the lowermost part of the overlying marine Blue Lias, but for convenience this highest part of the Triassic is described in Chapter 7.

Early publications on the Triassic sequence in the district include those of De la Beche (1846), Wright (1860), Bristow (1864), Etheridge (1872) and Howard (1895, 1897). The first complete account was in the first edition of this memoir (Strahan and Cantrill, 1902). Further detailed descriptions were provided by Richardson (1905, 1911). A review of the Triassic sequence in South Wales was provided by Ivimey-Cook (1974). A study of the sedimentology of the marginal facies of the Mercia Mudstone Group has been made by Tucker (1977, 1978), and one of the Penarth Group by Mayall (1979, 1981, 1983). (Figure 21) summarises past classifications of the strata and sets out the one adopted in this account.

Mercia Mudstone Group

The greatest thickness of the Mercia Mudstone Group in the district is at least 162 m calculated at Cardiff. The sequence includes an argillaceous facies and a contrasting heterogeneous marginal facies (Figure 22). The former comprises calcareous or dolomitic mudstones with evaporites and sparse limestones. It has been divided into an informal stratigraphical unit of 'red mudstones', up to at least 146 m thick, and an overlying Blue Anchor Formation, up to 16 m thick, of mainly green mudstone. The marginal facies consists of breccia, conglomerate, sandstone, calcarenite, siltstone, mudstone, evaporite, calcrete, limestone and dolomite that are arrayed in a number of distinctive lithofacies; previously it has also been termed the 'Dolomitic Conglomerate' or 'Litoral deposits' (Strahan and Cantrill, 1902, 1912). This marginal facies rests with marked angular discordance on the Palaeozoic rocks (Plate 5) and varies from a few cm up to 35 m thick. It passes laterally into both units of the argillaceous facies and usually underlies the 'red mudstones' or, more rarely, the Blue Anchor Formation, due to major onlap of the latter. The Blue Anchor Formation is itself locally overlapped around some of the Palaeozoic highs by the overlying Penarth Group which here may rest with a slight disconformity on marginal facies equivalent in age to the Blue Anchor Formation.

The 'red mudstones'

These red, rather massive, monotonous, dolomitic mudstones, with common gypsum nodules and veins, were formerly known as the 'Keuper Marl'. Locally, where the 'red mudstones' pass laterally into marginal facies, there are thin beds of calcisiltite, fine calcarenite and siltstone. The junction with the overlying Blue Anchor Formation is sharply marked by a change to predominantly green mudstones and by the appearance of better defined bedding. Thicknesses range from a feather edge (due to overlap by the Blue Anchor Formation) to over 146 m calculated in the Cardiff area.

The outcrop everywhere produces low, generally flat land, commonly covered by superficial deposits; exposures are generally poor and are confined to the topmost 40 m or so. Good sections, however, are available on the coast between Penarth and Lavernock, and at Barry, as well as in inland brickworks quarries. The mudstones are dominantly red-brown, silty, calcareous or dolomitic, and unfossiliferous. They are generally compact and hard, breaking with a conchoidal or subconchoidal fracture. Typically, at outcrop, they appear structureless, and occur in units between 0.5 m and 3 m thick which are blocky in the lower part and become more fissile towards the top. In the St Fagans Borehole [ST 1169 7813], however, the 'red mudstones' locally exhibited a streaky diffuse lamination and contained thin laminae and beds of silt. Some of the blocky units may have resulted from the in-situ destruction of laminites, as described from the lower part of the 'Keuper Marl' of Cheshire (Arthurton, 1980, p.49). Meigh (1976) noted a cyclic pattern in the engineering strength characteristics of the mudstones in the Cardiff area which he related to depositional variations. At Lavernock Point, Alkattan (1976) showed that both calcite and dolomite are more abundant than in other British 'Keuper Marls' (cf. Davis, 1967), probably reflecting the ready availability of these minerals from local Carboniferous and earlier Triassic deposits.

The 'red mudstones' contain scattered green mottles and spots, and sporadic beds of grey-green mudstone up to 0.2 m thick occur throughout. Some individual green beds persist over large areas, as for example several which can be traced for almost 3.5 km in the cliff-section between Lavernock Point and Penarth Head. It has not, however, proved possible to correlate the green beds of separate sections. Apart from their colours the green mudstones seem identical to the red mudstones, and the two pass laterally and vertically into one another.

Where the 'red mudstones' are adjacent to the marginal facies there is a transition zone of variable thickness in which the mudstones become 'gritty', containing scattered quartz grains and Carboniferous Limestone clasts up to coarse sand and, in places, to granule size. Beds of grey, red- and green-mottled dolomitic and calcareous siltstone, calcisiltite and fine-grained calcarenite also appear within the mudstones. They mostly vary from 1 to 50 cm thick, though one prominent bed on Barry Island, about 1 m below the base of the Blue Anchor Formation, is up to 1.6 m thick. The beds are generally parallel-sided, but thicken locally where they fill small channels up to 0.9 m deep. The thinnest beds are commonly structureless or diffusely laminated, while the thicker ones are variably laminated and cross-laminated or rarely, cross-bedded. Mudcracks and raindrop imprints have been described from mudstones in the transition zone at Barry and Sully (Klein, 1962).

White to pale pink gypsum occurs within the 'red mudstones' as discrete nodules and as beds of nodules. In the cliff sections they can be seen to lie at relatively well defined and constant levels in beds up to 0.5 m thick; a nodule up to 0.63 m diameter was proved in the St Fagans Borehole. In surface exposures, the nodular gypsum has commonly suffered extensive dissolution, resulting in the formation of voids in the cliffs (Plate 6) or of collapse breccias. In places the voids are partly occluded by scalenohedral crystals of calcite. In the St Fagans Borehole, gypsum nodules commonly exhibited calcite at their margins. Meigh (1976) has described the lack of gypsum, and only 80–90 per cent core recovery in the upper 38 m of high quality boreholes in the upper part of the sequence in central Cardiff, as due to the leaching of gypsum. It has been suggested that gypsum dissolution in the dolomitic mudstones has led to localised dedolomitisation, now manifested as thin beds, 3–16 cm thick, of degraded mudstone as seen in Llandough Quarry [ST 166 736] (Hawkins, 1979). Fibrous gypsum (satin spar) is present in the 'red mudstones' as scattered veins and also as ramifying networks around gypsum nodules, probably resulting from secondary emplacement by hydraulic injection (Shearman and others, 1972). Little is known about the distribution of gypsum nodules in the thicker sequence beneath Cardiff.

Celestite was recorded from the 'red mudstones' at Cogan (Cox and Trueman, 1936, p.50), and celestite and baryte have been recorded from Lavernock Point (Alkattan, 1976). The celestite occurs at about the level of the Severnside Evaporite Bed (Nickless and others, 1976).

Fossils are extremely rare in the 'red mudstones'. Sporadic palynomorphs recovered from 134.01 m in the St Fagans Borehole (i.e. about 6 m below the Blue Anchor Formation) include Ovalipollis pseudoalatus, a form indicative of mid to late Triassic (Ladinian to Rhaetian) age; other samples from this borehole and from outcrops at Barry Docks and north of Lavernock Point proved barren. In north Somerset, palynomorphs of late Norian (?) and Rhaetian age are known from the highest 74 m of the Mercia Mudstone Group (Warrington in Whittaker and Green, 1983).

Details

Cardiff City centre and docks

The sequence here is known from boreholes mainly drilled around the turn of the century (Strahan and Cantrill, 1912; North, 1915b), ((Figure 23) and (Figure 24)). The greatest thickness proved is in the east, where a borehole [ST 2052 7518], starting about 20 m below the highest 'red mudstones' exposed to the north at Pen-y-Lan, proved 126.5 m of mudstones without reaching their base. Near Penarth, 103 m of red mudstones were drilled above the marginal facies in the Cogan Brickworks Borehole [ST 174 274]; the calculated total thickness of the 'red mudstones' at Cogan is about 130 m.

In general the lithological descriptions in the old records are very vague. In the east of the area they refer to beds of 'conglomerate', 'grey rock' or 'white rock', up to 7.4 m thick (Strahan and Cantrill, 1912). These 'rock' beds are probably calcareous sandstones or dolomitic calcarenites, and in Helen Street Borehole [ST 1992 7704] (North, 1915b) and another borehole nearby [ST 2034 7556], lie between 60 m and 70 m above the base of the 'red mudstones' (Figure 24); calculations based on figures given by Strahan and Cantrill (1912, p.97) suggest that the 'upper water-bed', a water bearing bed of sandstone, about a metre thick (Howard, 1894), is at about this height above the base of the sequence (e.g. 62 m above the base in Brain's Brewery Borehole, St Mary Street [ST 1832 7611]), and that the 'rock' beds are probably a continuation of it. The 'white rock' in another borehole [ST 2081 7799] is apparently lower in the sequence. Material from Helen Street Borehole, curated in the National Museum of Wales, includes red-brown silty mudstones with green reduction spots, listric surfaces and gypsum nodules in the upper 40 m, and similar mudstones, but siltier and with scattered quartz grains to coarse-sand size, in the lower 35 m. Gypsum is recorded from several levels in this and other old records.

Barry and Barry Island

The 'red mudstones' sequence here onlaps southwards onto the Carboniferous Limestone ridges of Friars Point and Nell's Point, and passes into marginal facies. It is up to 35 m thick (proved in boreholes) in the dock area, and thins southwards where it is overlapped by the Blue Anchor Formation. All the sections on Barry Island are transitional between the 'red mudstones' and the marginal facies, and the former contain siltstones and sandstones. The following section at Jackson's Bay [ST 1202 6658]–[ST 1213 6669] is typical:

Lavernock Point to Penarth Head

About 15 m of red mudstones below the Blue Anchor Formation are well exposed in the cliffs and foreshore north of Lavernock Point

[ST 187 684] and again near Penarth Head. Thin green beds can be traced from Lavernock to Penarth. Four layers of gypsum nodules up to 0.5 m thick are present.

Ely to Leckwith

There are major inland sections in beds immediately below the Blue Anchor Formation at the disused quarries of the Ely Brickworks [ST 141 754] north of Caerau Wood, at Cogan [ST 176 720] and at Llandough Brickworks [ST 166 736].

Blue Anchor Formation

The Blue Anchor Formation, formerly known as the 'Tea Green Marls', overlies the 'red mudstones' and is disconformably overlain by the Penarth Group. It is well exposed in the coastal sections and a complete sequence was proved by St Fagans Borehole. Where it is overlain by the Penarth Group and the St Mary's Well Bay Formation, it occupies the lower slopes of the feature formed by the latter. In the Dyffryn and Creigiau areas, it overlaps the 'red mudstones', and overlies and passes laterally into the marginal facies.

The sequence comprises up to 16 m of grey to green, less commonly red-brown, variably dolomitic bedded mudstones, with subordinate thin dolomites and limestones. Silt-laminated mudstones, intraformational breccias and beds of evaporite nodules are also present.

The base of the formation is fairly sharp and is generally determined by the relatively sharp colour change and the onset of good bedding. At Lavernock Point it has been taken at the base of 0.2 m of green mudstone containing gypsum nodules which rests on the highest red bed of the underlying 'red mudstones'. The top of the formation is taken at the sharp incoming of the dark grey shales of the Westbury Formation. The transition into marginal facies is not exposed but appears to be fairly rapid. Where the formation overlies the marginal facies it contains scattered Carboniferous Limestone chips in its basal few metres. In St Fagans Borehole scattered quartz grains and grains of Carboniferous Limestone ranging up to coarse sand size were commonly scattered throughout, reflecting the close proximity of the marginal environment.

The distribution of lithologies at Lavernock Point is shown in (Figure 25), and other provings throughout the district are fairly similar. Although the sequence is dominantly argillaceous, with sparse burrows at several levels, dolomites and limestones increase in importance upwards. There is also a gradual upward colour change, from greenish yellow and dark green at the base through grey-green to dark grey at the top; a unit of alternating pinkish red and green beds, 1.3 m thick, about one third of the way up the section, can be traced throughout the district (this is the 'pink band' of Strahan and Cantrill (1912)).

The mudstones occur in bedded units averaging 0.5 m thick, and range from fissile to 'blocky' and massive with a conchoidal fracture, reflecting increasing carbonate content.

The latter variety grades to dolomites and dolomitic limestones that form beds up to 0.2 m thick which weather out prominently in natural sections. The mudstones and dolomites are mainly either structureless or exhibit a diffuse lamination. Some units of silt-laminated mudstones locally exhibit desiccation cracks. The clay mineralogy of the mudstones has been described from Lavernock Point by Mayall (1981).

Thin intraformational breccias up to 0.3 m thick are common, and comprise mudstone and dolomite clasts in a mudstone matrix. Some of those high in the formation yield sporadic vertebrate remains.

Displacive gypsum occurs in beds of commonly coalescing nodules up to 0.2 m thick in the lowest 10 m of the formation. Gypsum is also present as scattered veins of fibrous satin spar. In natural outcrops it has commonly been completely dissolved, leaving voids which are locally partly filled by calcite, but the areal distribution of the dissolution is very variable. In St Fagans Borehole some gypsum nodules had been partially replaced by calcite at their margins.

Throughout the formation, many of the laminated units exhibit micro-faulting, together with fluidisation and dewatering structures that involve the destratification, homogenisation or brecciation of the beds, and in some cases the upward injection of homogenised material via narrow dykes. In some cases these processes have caused brecciation of the injected beds. Many of the dolomite beds are overlain by breccias or have brecciated tops caused by dewatering of the underlying sediments by upward injection of pore fluids; some are penetrated by mudstone-filled cracks leading down from the breccia above. In many cases it is difficult to distinguish true intraformational breccias from such autobreccias.

Fish, possible amphibians and reptile remains and plants are present in the upper part of the Blue Anchor Formation (Storrie, 1894; Ivimey-Cook, 1974). Palynological samples from the topmost part of the formation include recognisable plant debris, including vascular elements and cuticles. Organic residues from the rest of the formation are sparse and, though probably carbonised plant tissues are represented, there is seldom structured debris that is clearly of plant origin.

Of 12 samples from exposures in cliffs north of Lavernock Point only one, from 1.7 m below the top of the formation, yielded a sparse miospore assemblage (Figure 26). Twelve samples were examined from the formation in St Fagans Borehole and four, from the top 3.3 m, yielded sparse miospore associations including the taxa recorded from Lavernock and specimens of Leptolepidites argenteaeformis , A lisporites thomasii , Vesicaspora fuscus , Quadraeculina anellaeformis and Granuloperculatipollis rudis. Comparable, but more varied, palynomorph assemblages, some of which include organic-walled microplankton (acritarchs and dinoflagellate cysts) were recorded by Orbell (1973) from the Lavernock section. These associations comprise taxa which are major components of assemblages from the succeeding Penarth Group (Figure 26) and are indicative of a Rhaetian age.

Details

At Creigiau the formation is inferred from evidence in the adjacent Bridgend district to have overlapped the 'red mudstones' and to rest directly on marginal facies of 'Blue Anchor Formation age'.

Between St Fagans and Ely, St Fagans Borehole proved the formation to be 13.94 m thick. The upper part of the Ely Brickworks quarries [ST 141 754] also provides a complete section through the formation. Other complete sequences are exposed in the disused Llandough quarry [ST 166 736], in the cliffs between Cogan and Penarth Head, and in much of the cliff section southwards to Lavernock Point. The best accessible section is immediately north of Lavernock Point [ST 1875 6815] (Figure 25).

In the St Nicholas–Wenvoe area the formation overlaps the 'red mudstones' and rests directly on the marginal facies 1.5 km west of Goldsland; the contact is seen in a road cutting [ST 0969 7213] where dolomitic mudstones with a few thin dolomites rest on fine-grained calcarenites of the marginal facies. The lithology of the beds in the basal part of the formation is unusual in a track-section [ST 0853 7361] west of Brooklands. The grey-green mudstones contain beds and numerous irregular buff-weathering nodules of finely crystalline dolomite which form up to half of the sequence; the nodules are up to 0.3 m thick and occur both isolated and in semi-continuous beds. There is a gap of about 1 m between this exposure and another in the underlying marginal facies which here consists of fine-grained dolomite with scattered spar-replaced granules of Carboniferous Limestone (see p.68). The formation thins rapidly to the north of Brooklands where it has probably been overlapped by the Westbury Formation [e.g. [ST 0817 7383]. Strahan and Cantrill (1912, p.45) reported Westbury Formation ('Rhaetic') shales overlying 'hard yellow rock' (presumably dolomitic fine-grained calcarenite of the marginal facies) in a pit at Redland [ST 0778 7373] in the neighbouring Bridgend district.

At Barry Island the lower part of the formation is exposed in the cliffs on the west side of Whitmore Bay [ST 111 664] overlying the 'red mudstones' . It is also seen in faulted cliff sections between Jackson's Bay [ST 1210 6698] and Barry Docks [ST 1211 6698]. On the southern part of Nell 's Point it overlaps the 'red mudstones' and rests directly on the marginal facies. On the southern part of Nell's Point overlap reduces the thickness of the formation to 3 m.

Marginal facies

The term 'litoral deposits' was used by Strahan and Cantrill (1902, 1912) to describe all the lithologies of the marginal facies, although they did suggest that some breccias might represent subaerial talus. More recently, however, the marginal facies has been divided into continental and lacustrine shore-zone subfacies (Tucker, 1977, 1978) (Figure 22). The lithologies present in both subfacies are dominated by breccias and conglomerates, commonly variably dolomitised. The only lithologies exclusive to one subfacies are cryptalgal and fenestral carbonates, which occur only in the lacustrine shore-zone subfacies. The major subfacies can, however, be distinguished by their distinctive component lithofacies, by the presence of exhumed Triassic palaeogeomorphological features, and by their relationship to the 'red mudstones' and the Blue Anchor Formation. Although the continental subfacies is dominantly in the lower parts of the local succession, and occurs mainly in the northern part of the district, the two subfacies are locally interbedded and pass laterally into each other. The Blue Anchor Formation marginal facies is entirely in the lacustrinc shore-zone subfacies.

1. Continental subfacies

Beds of this subfacies predominantly consist of red conglomerates, breccias and sandstones, with some subordinate siltstones and mudstones. Nodular calcrete is locally present. Four distinctive lithofacies can be recognised: well sorted conglomerates and sandstones; ill sorted angular breccias; matrix supported conglomerates; and thinly bedded sandstones/calcarenites, siltstones and mudstones.

1a. Well sorted conglomerates and sandstones: This lithofacies comprises mainly red and grey conglomerates. The matrix varies from sand to sandy calcarenite or locally calcite spar. Clasts are mostly of Carboniferous Limestone with some Old Red Sandstone locally. They are mainly pebble to cobble grade, though some attain 0.2 m diameter.

They are well sorted and subangular to rounded. Bed thicknesses vary from 0.15 m upwards. The thicker units comprise massive multi-storied sequences in which beds are definable only by the presence of siltstone partings and scoured horizons; an 18.7 m thick massive unit of conglomerate was proved in St Fagans Borehole. Some of these units show tabular cross-bedding and erosional bases. Trough cross-bedded or plane-laminated, pebbly and coarse sandstones and calcarenites also occur as lenses or thin beds. Fine-grained sandstones and siltstones form lenses within the coarse sandstones. Desiccation cracks are present locally.

(1b) Ill sorted angular breccias: These consist of red or grey, generally chaotic, massive breccias with clasts mainly of Carboniferous Limestone up to 2 m across. The matrix is variable, ranging from calcarenite to red siltstone, locally with nodular calcrete. The breccias generally rest on, or are banked against, vertical cliffs of Carboniferous Limestone. Locally the Carboniferous Limestone below the plane of the unconformity is split into thin sheets to a depth of 30 cm, with spaces between filled by Triassic sediment.

(1c) Matrix-supported conglomerates: In these conglomerates, angular clasts, up to 0.5 m across, 'float' in a fine-grained matrix. The beds are up to 3 m thick and are generally structureless. The lithofacies is uncommon.

(1d) Thinly bedded sandstones/calcarenites, siltstones and mudstones: These comprise laterally persistent beds of graded sandstone and calcarenite, 3 to 15 cm thick, interbedded singly or as multiples with laminated or massive fine-grained sandstone, siltstone and mudstone. Small pebbles and granules are common at the bases of the graded beds, which exhibit local scours. Ripples and desiccation cracks are common on the tops of the beds, and dinosaur footprints have been noted at Barry (Tucker and Burchette, 1977). The interbedded strata form units up to several metres thick. Symmetrical ripple marks and desiccation cracks are common at the top of these. Locally, horizons of red nodular dolomite in mudstone show a well developed polygonal pattern on bedding surfaces. The nodules contain evidence for precursor evaporite (Tucker, 1977). Nodular calcretes are also locally present.

(2) Lacustrine shore-zone subfacies

This subfacies comprises predominantly well sorted breccias, conglomerates, calcarenites, calcisiltites and mudstones, with subordinate nodular dolomites and both fenestral and cryptalgal carbonates. The breccias and conglomerates are well sorted and composed of Carboniferous Limestone clasts. The calcarenites and calcisiltites are grey, weathering buff, commonly dolomitic and include some scattered small pebbles and granules.

The breccias, conglomerates and calcarenites characteristically occur in thin tabular beds. A rapid lateral change in grain-size from conglomerates and breccias through finer-grained parts of the subfacies into the 'red mudstones' is characteristic of the shore-zone subfacies. This passage can occur in as little as 10 m. For example, at Barry Island, ill sorted breccias of the continental subfacies pass laterally into well bedded breccias and fine-grained calcarenites with wave ripples and desiccation cracks, and in turn these pass into the 'red mudstones'. On Sully Island (Figure 27) a wedge of coarse sediments resting unconformably on a planar surface of Carboniferous Limestone thickens downslope with a corresponding decrease in grain size from breccias at the thin end of the wedge to red pebbly calcarenite and calcisiltites at the thick end of the wedge over a horizontal distance of 10.20 m.

Triassic platforms, (termed terraces by Tucker, 1978), backed by low cliffs, and cut into both Carboniferous Limestone and the shore-zone subfacies, are typically present where the subfacies is developed. They are up to 15 m wide and slope at up to 5°. The cut platforms are best displayed in coastal sections such as at Barry Island (Figure 28). They are overlain by tabular bedded Triassic shore-zone conglomerates, breccias, calcarenites, etc. Several platforms are commonly present, occurring in flights, the back features forming vertical to subvertical cliffs up to 5 m high cut in the Carboniferous Limestone. Locally, ill sorted continental subfacies breccias banked up against cliffs of Carboniferous Limestone pass laterally into well sorted, tabular bedded, shore-zone breccias overlying a platform. Some of the latter cut across both Carboniferous Limestone and earlier Triassic continental subfacies breccias, truncating large boulders within the latter.

Evaporites were originally present in these shore-zone sequences but they have now been largely replaced. Such a replaced unit occurs in the Sully area where it is up to 12 m thick. It occurs above shore-zone breccias, calcarenites and siltstones, and comprises red, structureless, sandy, silty, dolomitic mudstone passing up into red nodular and irregularly layered dolomite and dolomitic mudstonc. The latter contain scattered quartz nodules with relic anhydrite inclusions (Tucker, 1978). Both the 'chicken wire' and enterolithic textures of modern anhydrite deposits are present.

Fenestral and cryptalgal carbonates (Plate 8) and (Plate 9) mainly occur in the shore-zone subfacies adjacent to the 'red mudstones' or Blue Anchor Formation. They are well developed on the coast at Sully where their petrology has been detailed by Tucker (1975). Nodular calcretes occur within them locally. Fenestral calcarenites are the most common, being very fine- to medium-grained, and red to grey but buff-weathering. They comprise sand- and silt-sized fragments of Carboniferous Limestone, peloids and carbonate mud. Some contain scattered granules and small pebbles of Carboniferous Limestone. They are commonly dolomitic. Laminoid and equant fenestrae are sporadically present. The fenestrae are filled by calcite spar, commonly weathered out. The calcarenites can be laminated, wave rippled, structureless or show wet-sediment deformation structures.

Cryptalgal carbonates (Aitken, 1967) are represented by flat-laminated and columnar varieties. Large pisoids or oncoids have been noted at one locality. Pseudo-anticlines or tepees (Assereto and Kendall, 1977), formed by the lateral expansion of the sediment by the growth of carbonate during subaerial exposure, are locally present.

Details

St. Fagans–Radyr–Llandaff–Ely

This area occupies the core of the C ardiff–Cowbridge Anticline, and is bounded to the north and south by escarpments of Carboniferous Limestone and Upper Old Red Sandstone, and to the east by the River Taff.

Within it, the marginal facies rests on the Old Red Sandstone, and comprises reddish brown well sorted continental subfacies conglomerates with very subordinate thin mudstone beds and partings. The conglomerates contain both Old Red Sandstone and Carboniferous Limestone clasts.

The sequence is best known in the St Fagans area, where St Fagans Borehole proved 20 m of massive pebble-cobble-conglomerates overlying the Old Red Sandstone, and containing mudstone partings only in their' uppermost part. This sequence is capped by 0.6 m of pebbly red-brown dolomitic siltstone with calcrete glaebules in the upper part. The top of the conglomerate has also been calcretised (probably as part of the same profile), individual pebbles exhibiting jig-saw' brecciation fabrics. The nodular calcrete is sharply overlain by a thick succession of the 'red mudstones'. The sequence is interpreted as the product of stream-flood deposition on an alluvial fan draining the core of the Cardiff–Cowbridge Anticline and receiving coarse detritus from the surrounding escarpments. The calcrete that caps the sequence suggests that fan deposition ended suddenly.

A similar thickness of conglomerate is known from Radyr and Llandaff, south of the Taff gorge, which, it has been suggested, was a Triassic canyon from which an ephemeral stream supplied sedi ment to the fan (Tucker, 1977). The conglomerates were formerly extensively worked as a building stone–the 'Radyr Stone'–at Radyr Quarry [ST 138 796]. Here Strahan and Cantrill (1912) recorded a thickness of 18.3 m of Triassic strata above the Old Red Sandstone; thin red silty mudstone and siltstone interbeds occur in the sequence which is dominated by massive uneven beds of reddish purple pebble-cobble-conglomerates. The clasts are subrounded to angular and consist mainly of pebbles of Carboniferous Limestone together with pebbly grits, red quartzitic sandstones and micaceous sandstones probably derived from the Old Red Sandstone; the matrix is composed of sand-sized quartz grains. Beds near the base of the sequence, exposed in the River Taff [ST 1400 7956], contain subangular boulders of Carboniferous Limestone. The conglomerates are well exposed in quarries south of the River Ely in Plymouth Great Wood. The largest of these [ST 1282 7688] shows 10 m of massive red-brown, slightly bouldery, pebble-cobbleconglomerate with well developed large-scale tabular cross-bedding; the clasts are of Carboniferous Limestone and Old Red Sandstone rocks.

Central Cardiff

The St Fagans fan conglomerate passes eastwards under, and laterally into, the 'red mudstones' that crop out in central Cardiff. Contours on the base of the Mercia Mudstone Group under Cardiff are shown in (Figure 23). They suggest that the area forms part of the north-western rim of a basin that is separated from the Mercia Mudstone Group underlying the Wentlooge Level by a ridge of Silurian strata. At the northern margin of the outcrop of the Mercia Mudstone Group, between Whitchurch and Pen-y-Lan, the 'red mudstones' rest directly on Palaeozoic rocks with no intervening marginal facies. Anderson and Blundell (1965) suggested that this boundary may be a fault, but borehole evidence suggests that this is unlikely unless an intra-Triassic fault is now covered by later Triassic sediments.

Water boreholes put down around the turn of the century (Strahan and Cantrill, 1912; North, 1915b) established that continental subfacies conglomerates underlie the thick 'red mudstone' sequence (Figure 24). They are from 22 to 35 m thick (the latter in the south-west around Ely).

The records from the old boreholes do not usually indicate detailed lithologies but some specimens from them are curated in the Natural Museum of Wales and have been examined. In general the marginal facies becomes finer grained to the east, the massive conglomerates around St Fagans passing laterally into a sequence of interbedded conglomerates, sandstones and siltstones beneath the city centre. Curated samples from two boreholes [ST 1992 7704]; [ST 1881 7611] on the east side of the city show that there the marginal facies consists of interbedded pebbly sandstones and siltstones; the sandstones contain granules and small pebbles of quartz and, less commonly, Carboniferous Limestone. They are probably the deposits of sheet-and stream-floods deposited at the distal end of the St Fagans fan.

The marginal facies thins northwards to a feather-edge. Its northernmost occurrence in this area is in a borehole [ST 2081 7799] where 'rock' underlies 71 m of red mudstones; in a nearby borehole [ST 2078 7837] 41 m of 'red mudstones' rest directly on the Pen-y-Lan Mudstone.

The best documented borehole sequence is Helen Street Borehole [ST 1992 7704] (North, 1915b). The following log from between the base of the 'red mudstones' at 80.5 m to the unconformable base of the Triassic at 104.2 m has been redescribed from curated material.

Shallow boreholes beneath Wentlooge Level prove that a thin marginal facies flanks the north-eastern edge of the Silurian ridge. However, this outcrop cannot be traced east of the Bronze Foundry [ST 226 791] at Rumney where the 'red mudstoncs' appear to directly overlie Old Red Sandstone. A little farther south-east, however, shallow boreholes near Newton have proved up to 3.6 m of pale green and brown, medium- to coarse-grained sandstones with scattered quartz granules and Carboniferous Limestone pebbles beneath the 'red mudstones'; presumably these sandstones are overstepped by the 'red mudstoncs' in this short distance.

Wenvoe valley

Between the Carboniferous Limestone inliers of St Lythans and St Andrews Major, the Wenvoe valley is floored mainly by the 'red mudstones' ; these overlie and pass laterally into a thin outcrop of marginal facies that flanks the valley sides at various levels. Marginal facies also occurs at depth, flooring the bottom of the valley beneath the 'red mudstones', and this sequence is exposed at the northern end of the valley.

The marginal facies in the valley floor succession is of continental subfacies, predominantly the well sorted conglomerate lithofacics. It comprises granule- and pebble-grade breccias and conglomerates which are in places interbedded with red mudstones and siltstones. They are exposed only in the Wenvoe railway cutting where they extend for 450 m south from the tunnel portal [ST 1249 7396]; just inside the tunnel, granule breccia unconformably overlies massive Carboniferous Limestone. The beds dip gently southwards, the angle of dip increasing southwards to about 12°. The following sequence is exposed:

The section is overlain by the 'red mudstones'. It has been suggested (Tucker, 1977) that the 1.2 m of breccia and conglomerate above the lowest conglomerate may be a mudflow. The other conglomerates and breccias are stream- and sheet-flood deposits interbedded with playa lake mudstones.

The marginal facies deposits flanking the valley sides belong to both the continental and lacustrine shore-zone subfacics. The former comprises massive or crudely stratified angular to subangular partly dolomitised ill-sorted limestone breccias. The breccias are generally of pebble-cobble grade, but contain some boulders. They are interpreted as scrces. The lacustrine shore-zone subfacies comprises tabular bedded, fairly well sorted, fine- to very coarse-grained calcarenites and limestone breccias interbedded with subordinate dolomitic siltstones and red silty mudstones. The breccias are generally of pebble grade but locally cobbly; they commonly show grain-size banding. Laminoid fenestrae are common in the finer grained calcarenites. Shore-zone breccias and calcarenites rest on platforms cut across the Carboniferous Limestone in places e.g. [ST 1240 7390].

The relationship between the screes and the shore-zone subfacies is apparent in a small quarry [ST 1105 7474] near Tumbledown, where 2.4 m of grey, yellow weathering, pebble-cobble scree-breccia pass upwards and laterally into tabular bedded, pebbly, grey, locally fenestral calcarenite. It appears that, after a period of stream-flood and sheet-flood deposition, the 'red mudstones' lake entered the Wenvoe valley from the south and north, and a shore-zone subfacies was laid down along the valley sides at successively higher levels with progressive onlap. Subaerial screes were locally developed on the valley sides during this time.

Cuvrt-yr-ala–Divas Pawls

The marginal facies flanks the eastern and southern sides of the St Andrews Major Carboniferous Limestone Inlier. Several sections show ill-sorted scree breccias together with lacustrine shore-zone breccias and carbonates and platform features, suggesting that the exposed marginal facies correlates with the upper part of the Wenvoe valley sequence.

At the eastern end of Cwm George [ST 1496 7196], 8 m of locally bouldery pebble-cobble Carboniferous Limestone scree-breccia rest against Carboniferous Limestone on both sides of the valley.

Around Dinas Powis the lacustrine shore-zone subfacies is commonly exposed. For example, 1.9 m of fenestral fine-grained calcarenites and dark red mudstones with thin sandstones crops out on the western side of Dinas Powis Castle [ST 1524 7166], as does 1.2 m of pale reddish grey limestone with small columnar stromatolites on the western side of Mill Road [ST 1535 7133] (Plate 9).

St Nicholas–Dyffryn–Goldsland

Exposure is poor between Wenvoe Castle and Dyffryn, but four steep-sided valleys [ST 115 716] filled with marginal facies at Wenvoc Castle Golf Course may represent partly exhumed Triassic wadis; the upper part of the infill may be of lacustrine shore-zone subfacies as this is present at the same stratigraphical level farther east in the Wenvoe valley.

North of Dyffryn, the 'red mudstones' are overlapped by the Blue Anchor Formation, which rests directly on marginal facies that mantles the gentle southern and western slopes of the St Lythans Carboniferous Inlier. Traced farther north the Blue Anchor Formation thins due to progressive overlap, and north of Brooklands is overlapped by the Penarth Group. Thus the marginal facies in this area is, to a large extent, equivalent in age to the Blue Anchor Formation.

Exposure in the area is poor and limited to small sections, but the lacustrine shore-zone subfacies has been identified over a wide area and is likely to be predominant. Breccias and calcarenites are the main lithologies present, but there are red-brown sandstones, siltstones and mudstones in the upper part of the sequence in the south, and these may represent marginal facies equivalent to the 'red mudstones'. Calcretes and pisolites are also present in a few isolated exposures.

The breccias occur in tabular beds up to 0.2 m thick; they are generally well sorted and contain subangular to subrounded pebbles set in a variably dolomitic, fine- to coarse-grained, yellow-green or red, calcarenite matrix. Nearly all the pebbles are composed of Carboniferous Limestone, and they are generally less than 40 mm in diameter; some are partly or completely replaced by coarsely crystalline calcite.

The calcarenites are usually at least partially dolomitised and range from coarse- to very fine-grained; many are pebbly, while scattered Carboniferous Limestone granules (some replaced by calcite spar) are almost ubiquitous. They vary in colour from green-grey to pale grey, usually weathering yellow. Tabular beds from 10 mm to 0.2 m thick are common, and channels up to 0.4 m deep occur [ST 0965 7310] in grey dolomitic calcarenites south-east of Tinkinswood. Globular and laminar spar-filled desiccation fenestrae are locally present in fine-grained calcarenites; examples occur west of St Lythans [ST 1012 7280]; [ST 0954 7338]; [ST 1011 7223]; and [ST 1018 7390].

Pisolite is known only from crags [ST 0953 7335] at Tinkinswood where 0.2 m of pisolite underlies grey tabular bedded breccia; it has a spar cement and contains pisoids up to 30 mm in diameter with nuclei composed of Carboniferous Limestone, green dolomite mudstone intraclasts or calcarenite, and cortices composed of cream dolomitic carbonate. The fine-grained dolomite intraclasts also form patches of flat-pebble conglomerate.

Calcrete occurs at two localities, both within a sequence of interbedded breccias and calcarenites: 70 mm of green nodular calcrete occurs [ST 0948 7328] at Tinkinswood, and 120 mm of red and green mottled calcrete is present [ST 0953 7335] on the other side of the valley.

In the Tinkinswood area Strahan and Cantrill (1912) reported a buff cavernous dolomitic limestone that they termed the 'Cromlech Bed' since it was used in the building of the local cromlechs. They suspected that it was of tuffaceous origin because of its cavernous character, but it is a dolomitised, very fine-grained calcarenite that weathers distinctively. It is pale greenish grey when fresh, but weathers, commonly patchily, to a soft yellow rock with, in places, microcavernous texture. The dolomite locally contains scattered fragments of Carboniferous Limestone or calcite (replacing the former) which vary in size from sand-grade through granules to pebbles; the finer-grained fragments are the most abundant, and pebbles are very uncommon. It is the dissolution of the sand-sized grains that has produced the microcavernous texture of weathered specimens. Larger cavities and surface depressions up to 0.15 m in diameter are caused by modern weathering. The 'Cromlech Bed' occurs immediately beneath the Blue Anchor Formation in the Tinkinswood area on the upper part of the interfluve between the River Waycock and its easterly tributary; the ridge is capped by a small outlier of the Blue Anchor Formation. A stream section [ST 0853 7361] exposes a 2 m section in the 'Cromlech Bed', which here lies less than 1 m beneath the base of the Blue Anchor Formation.

Barry Island–Sully Island

On Barry Island the sides of the Carboniferous Limestone promontories of Cold Knap, Friars Point and Nell's Point locally expose Triassic platforms cut into both the Carboniferous Limestone and the Triassic marginal facies, many of which are covered by marginal facies. Where the tops of the promontories are in Carboniferous Limestone they are flattish and may represent an exhumed Triassic topography. The platforms occur in flights, of which the greatest number in any one place is five. They range in height from 1 to 3 m, and are up to 15 m wide. They dip up to 5° away from the promontories. The cliffs forming the backs of each platform are vertical to subvertical and locally overhang. The surface of the platforms is commonly sheeted.

Patches and larger areas of ill sorted angular breccias, interpreted as screes, are banked against vertical cliffs on the promontories. Some are chaotic whilst others exhibit a crude foreset bedding. Calcrete is locally present. The scree-breccias rapidly pass laterally into tabular bedded breccias and calcarenites of lacustrine shore-zone subfacies which rest on the platforms. The sequence of platforms, shore-zone breccias and screes is best displayed on the western side of Friars Point [ST 109 663] (Figure 28). At the back of one of the platforms, there is a wave-cut notch (Plate 7) with some Triassic sediment adhering to it and demonstrating that it is not a modern feature.

On the eastern side of Friars Point [ST 111 663] similar sequences are exposed. At one locality, scree-breccias pass, over a short distance, into tabular breccias and calcarenites with symmetrical ripple-marks orientated subparallel to the Triassic cliff against which the scree is banked. Traced northwards the shore-zone subfacies, comprising locally pebbly, fine- to coarse-grained, variably dolomitised calcarenites, contains a 0.7 m bed with abundant irregular and laminoid fenestrae. Farther north the calcarenites become less pebbly and less fenestral, and they contain dewatering structures and thin beds of dolomitic mudstone. By reference to the cliffs at the head of Whitmore Bay immediately to the north, this sequence can be seen to be the equivalent of the 'red mudstones' immediately below the Blue Anchor Formation.

On the western side of Nell's Point [ST 118 663], the same associations of platforms and sediments are again exposed on the foreshore below the cliff path, the section providing the best example of shore-zone sediment passing rapidly into the 'red mudstones'. Immediately above the cliff path 2.4 m of buff weathered, fine-grained, shore-zone calcarenites and calcisiltites, with abundant scattered granules and rare pebbles of Carboniferous Limestone, pass laterally within a metre or so into a sequence of red mudstones with thin dolomitic siltstone beds that belongs to the 'red mudstones'. Within these shore-zone calcarenites large Carboniferous Limestone boulders at the base of the section presumably rest on an unexposed platform.

The best example of thin graded sandstones (sheet-flood deposits) and of the fine-grained sandstone, siltstone and mudstone (floodplain and playa) assemblage (Figure 27) is exposed between Bendrick Rock [ST 1291 6690] and Hayes Point [ST 1415 6728]. Two beds of coarse well sorted conglomerate (stream-flood deposits) are also present. In the lower part of the section, replaced evaporites occur as dolomite beds. Dinosaur footprints attributed to Anchisauripus are present in the sheet-flood deposits (Tucker and Burchette, 1977).

A dip slope of grey dolomitic and fenestral shore-zone subfacies carbonates is exposed along the foreshore from Hayes Point to Sully Island [ST 1415 6728] to [ST 16944 6690]. The sequence is best seen on the southern side of Sully Island (Figure 27) where a wedge of shore-zone breccias and calcarenites above the Carboniferous Limestone (Plate 5) fines rapidly and thickens southwards into pebbly calcarenites, calcarenites and calcisiltites. Above comes a sequence of nodular dolomites (replaced evaporites) and red mudstones which is overlain by fenestral calcarenites (Plate 8), cryptalgal limestones and calcretes.

Mercia Mudstone Group–conditions of deposition

It has been suggested that the 'Keuper Marl' of Britain was deposited either in a generally subaqueous environment, such as a hypersaline lake or an epeiric sea (Strahan and Cantrill, 1902; Warrington, 1970), or in a generally subaerial environment, such as a megaplaya or desert plain or supra-tidal flat (Bosworth, 1913; Wills, 1976). An examination of both the major facies of the group in the Cardiff district leads to the conclusion that the group was deposited in a major lake that suffered considerable periods of lowered lake level when much of the district suffered prolonged periods of subaerial exposure.

The environments of the continental subfacies of the marginal facies have been discussed by Tucker (1977). The well sorted conglomerates and sandstones can be interpreted as alluvial fans formed by stream-floods in a semi-arid climate. The ill sorted angular breccias compare well with modern subaerial screes. The sheeting of the Carboniferous Limestone immediately below the plane of the unconformity has been interpreted as due to insolation under a desert climate (Tucker, 1974). The matrix supported conglomerates are best explained as mudflows and debris-flows. The graded sandstones/calcarenites in the thinly bedded sandstone/calcarenite, siltstone and mudstone lithofacies appear to have been deposited by sheet-floods during waning flow conditions. Sheet-floods are features of semi-arid regions and they develop on alluvial fans and pediments during rainstorms. The intervening fine-grained sediments are probably the deposits of playa lakes, in which periods of shallow water were interspersed with periods of subaerial exposure as witnessed by the desiccation cracks, dinosaur footprints and calcretes. The nodular dolomites, interpreted as replaced sulphate, were probably evaporite crusts as suggested by their polygonal pattern.

The nature of the lacustrine shore-zone subfacies has been demonstrated by Tucker (1978). The rapid lateral change from scree deposits via shore-zone elastics to red mudstones at places like Barry and Sully Islands is best explained as a passage from a littoral to a sublittoral environment. The fenestral and cryptalgal carbonates represent subaqueous lacustrine sediments that suffered periodic subaerial exposure as witnessed by the laminoid desiccation fenestrae. Evidence that there were major changes in lake level is proved by the platforms which appear to represent shore platforms cut by wave action, and possibly by solution as well, during periods of stillstand. Regressions led to the backs of the platforms being covered by screes whilst subsequent rises in lake-level led to the truncation of earlier screes. The presence of replaced evaporites in the shore-zone sediments and gypsum in the 'red mudstones' provide further evidence of repeated drops in lake-level allowing sulphates to be precipitated as in modern coastal sabkhas. The thicker calcarenites and calcisiltites in the 'red mudstones' within the transition zone into the shore-zone sediments probably represent sheets of shore-zone sediments deposited during regressive phases.

Local palaeogeography was variable when the 'red mudstones' were being deposited. Along the Cardiff–Cowbridge Anticline and the Wenvoe valley, alluvial facies, mainly comprising stream-flood deposits, were laid down. Later these areas were invaded by the 'red mudstones' lake, and shore-zone sediments associated with screes formed where the lake impinged against the remaining Palaeozoic 'highs'. Between Barry and Sully, screes and local sheet-flood and stream-flood deposits are intimately associated with shore-zone facies in latest 'red mudstones' times.

The marginal facies of the Blue Anchor Formation is all in lacustrine shore-zone subfacies, suggesting that much of the district was drowned by the end of 'red mudstones' times, and that the remaining highs were not capable of generating fluvial sediment. Periods of low lake-level led to subaerial exposure, and the formation of desiccation cracks and sabkha type sulphates in the argillaceous facies. The silt laminae also presumably reflect periodic marginal influence when the lake-level was low. Mayall (1981) has interpreted the clay mineralogy as representing detrital minerals derived from a surrounding landmass under increasingly humid climatic conditions. Furthermore the more consistent occurrence and greater diversity of miospores in samples from the upper part of the Blue Anchor Formation, together with the comparable distribution of recognisable plant tissues, may reflect a change from dominantly arid climatic conditions to ones more favourable to the development and diversification of a land flora.

Although the Blue Anchor Formation is basically lacustrine, a marine influence is indicated by the presence of sparse organic-walled microplankton in the upper part of the formation at Lavernock (Orbell, 1973). This marine influence reflects the advent of a marine transgression better manifested in the overlying Westbury Formation. There is, however, no evidence in the district of the presence of the Williton Member (Mayall, 1981), a thin unit of marine shales and sandstones with a disconformity at the base, present at the top of the formation in north Somerset (but see p.71).

Penarth Group

The Penarth Group comprises up to 12 m of dominantly marine dark grey and grey shales with subordinate sandstones, siltstones and limestones (Warrington and others, 1980). The name replaces the former term 'Rhaetic Beds'. The group does not appear to pass into a marginal facies as it does in the adjacent Bridgend district (Strahan and Cantrill, 1904), though where it rests on the marginal facies of the Blue Anchor Formation it thins to as little as 4 m.

The sequence comprises several distinctive stratigraphical units of regional significance (Warrington and others, 1980) (Figure 29). The lowest, the Westbury Formation, comprises dark grey shales with a few thin limestones and sandstones resting, with a marked erosional contact, on the underlying Blue Anchor Formation. The overlying Lilstock Formation includes the Cotharn Member (grey and green calcareous mudstones with thin siltstones, sandstones and limestones) and the Langport Member (a basal unit of porcellanous limestones overlain by grey mudstones with a few thin limestones). The latter member is sharply, but conformably, overlain by a thick 'paper shale' that marks the base of the Blue Lias.

Inland exposures are generally poor but the group is best displayed in cliff sections both north and south of Penarth Head, at St Mary's Well Bay, and just north of Lavernock Point; it has been proposed that a 'composite section' of these exposures should form the type for all the component units (Warrington and others, 1980). Accounts of one or other of these sections have been compiled by numerous authors for example Bristow (1864); Strahan and Cantrill (1902, 1912); Richardson (1905); Ivimey-Cook (1962); Whittaker (1978); Mayall (1979, 1983). The St Fagans Borehole provided a complete 11.41 m section through the group.

The macrofaunas are considerably more profuse and varied than those in the Mercia Mudstone Group reflecting the establishment of marine conditions in the area. A progressive upward increase in diversity occurs in the Westbury Formation and into the base of the Cotham Member. However, although faunas found in the upper part of the Cotham Member are more restricted, they diversify again in the Langport Member. Microfaunas include both foraminifera and marine ostracods, some of which are of freshwater affinity (Anderson, 1964). Palynomorph assemblages from Lavernock were studied by Orbell (1973) and a further analysis of this section is given by Warrington (Figure 26) herein. These assemblages are more varied and profuse than those in the Blue Anchor Formation; they include miospores from land plants as well as marine organic-walled microplankton. They diversify upwards through the Westbury Formation and Cotham Member but become more restricted in higher beds. The changes observed are probably related to the late Triassic marine transgression. Climatic amelioration resulting from the proximity of a large expanse of water may have allowed the colonisation of formerly largely arid land areas by an increasing diversity and profusion of plants giving diverse miospore associations. The subsequent decrease in miospore diversity may record the progressive inundation of low-lying land and the consequent elimination of elements of the land flora which inhabited only such areas.

Westbury Formation

The Westbury Formation (also known as the ‘contorta Shales’ auctt) comprises dominantly dark grey fissile, commonly pyritous shales, with a few thin beds of limestone and calcareous sandstone. It ranges in thickness from 6.6 m at Lavernock to 5.1 m in St Fagans Borehole (Figure 29), though it may be thinner in the Creigiau area. It contains a marine fauna dominated by bivalves, but also present are several ossiferous sandstones or 'bone beds' generally rich in fish fragments.

The base of the formation is sharply defined by an erosion surface cut in the underlying Blue Anchor Formation. At Creigiau and south of Bonvilston the formation rests directly, and presumably disconformably, on Blue Anchor Formation marginal facies. South of a line from Cadoxton through Dinas Powis to the coast just north of Lavernock Point, the basal beds comprise a thin unit of limestones and shales, with a restricted bivalve fauna dominated by Liostrea bristovi, that thickens southwards from a feather edge to 1.13 m at St Mary's Well Bay. These beds, termed the 'bristovi limestones' by Ivimey-Cook (1962, 1974), are included in the formation as they sharply overlie the Blue Anchor Formation along an erosional contact. They may, however, be the time correlative of the marine Williton Member, which is taken as the topmost unit of the Blue Anchor Formation in north Somerset (Mayall, 1981).

Overlying the bristovi limestones' is a very thin, ossiferous, locally conglomeratic, coarse-grained sandstone, commonly referred to as a bone bed. It rests on an eroded surface and in the north-west of the district it oversteps the bristovi limestones' to rest on the Blue Anchor Formation. At Lavernock, where it is known as the 'fish bed' (Storrie, 1883), it is markedly conglomeratic but patchily developed, mainly occupying a series of channels at the top of the bristovi limestones'. In St Fagans Borehole, vertical burrows 8 mm wide penetrate up to 6 cm into the underlying Blue Anchor Formation and are filled by the overlying basal ossiferous sandstone which is 4 cm thick. Here the bed also contains laminae of black mudstone and 7 mm wide vertical burrows. It passes up, via burrowed mudstones with medium- to coarse-grained sandstone laminae and fish scales, into shales locally with grey siltstone lenses. The basal bed is much thinner at Seven Sisters Bay [ST 187 699], almost lacking the quartz and coarser components as seen at Laver-flock, and this is generally true of most of the other outcrops in the district (Sykes, 1977).

In the major part of the formation, above the basal ossiferous sandstone, the shales are dark grey to black, and are variably silty, varying also from fissile to structureless. Pyrite is present as small scattered nodules commonly in aggregates or as pyritic laminations. The fauna of bivalves and some fish remains is variably distributed within the sequence. Uncemented bivalve rich layers up to 8 mm thick occur, and burrowing may be partly picked out by pyrite. One unit of calcareous sandstone, 0.3 m thick is present. It is fine to coarse grained, locally with scattered quartz granules, and contains scattered vertebrate fragments. Parallel lamination and lenticular bedding have been observed, but burrowing has commonly obscured much of the structure. Other sandstones are variably calcareous and much thinner, being less than a centimetre. The limestones vary from grey argillaceous wackestone to coquinoidal packstone/grainstone, and are 3 cm to 0.3 m thick. Layers of 'beef' (fibrous calcite) up to 10 mm thick are common above or below the thinner limestones.

Six depositional cycles (Ivimey-Cook, 1962, 1974) have been recognised within the formation (Figure 30) and can be traced laterally within south-east Glamorgan. Each begins with a fine sandstone or silty limestone, and fines upwards into shales with bivalve rich layers. The bivalve dominated fauna is most varied in the lower half of each cycle, but in the upper part the taxa commonly decrease both in variety and number until the fauna is sparse; within the formation as a whole however the marine fauna also becomes more diverse upwards. Some of the cycles have eroded tops, but where they are more complete a thin transition, back through silty shale, is present. Similar cycles are known in the Penarth Group in the Bristol region (Hamilton, 1962, 1977) and from Chilcompton, Somerset (Duffin, 1980).

The erosion surface at the base of the Westbury Formation points to a substantial shoreline regression at the top of the Blue Anchor Formation. The bristovi limestones' are the deposits of a quasi-marine pulse of the major late Triassic transgression that only invaded the south-eastern part of the district at this juncture. Further shoreline regression at the top of the bristovi limestones' was followed by the main marine transgression now manifested by the erosion surface at the top of the bristovi limestones'. This transgression appears to have inundated the whole district, the St Nicholas ridge, one of the last upstanding Palaeozoic ridges, finally being covered. Because of post-Jurassic faulting it is not known whether the transgression crossed into the Coalfield, but the lack of a marginal facies in the Creigiau area suggests that it is likely to have clone so.

Storrie's 'fish bed', like other basal Westbury Formation 'bone beds' elsewhere in south-west Britain, can be interpreted as a strand line deposit (Sellwood and others, 1970; Hamilton, 1977). The presence of foraminifera, marine ostracods, the brachiopod Orbiculoidea, echinoid fragments and ophiuroids, cirripedes and various marine invertebrates indicates a diverse community in a shallow marine environment. However, the rare occurrence of the ostracod Darwinula in Cycle 3 may indicate input of fresh water, as modern species of this genus are associated with brackish/freshwater conditions. The relatively coarse lithologies at the base of each cycle, and the uncemented shell accumulations in shales have been regarded as representing more turbulent phases in an otherwise low energy environment (Ivimey-Cook, 1962, 1974). Such phases may be best interpreted as due to periodic storms when the movement of coarse material took place below normal wave base. It has been suggested that Cycle 4 contains evidence that periods of no turbulence led to slightly anoxic bottom conditions inimical to colonisation by benthos (Bazley, 1968).

Lilstock Formation

The Lilstock Formation comprises grey and grey-green mudstones with very subordinate siltstones, sandstones and limestones. It is divided into the Cotham Member and the Langport Member.

Cotham Member

This includes grey and grey-green calcareous mudstones and shales, and laminated and lenticular-bedded calcareous siltstones. Beds of coarse-grained calcareous, commonly oolitic, sandstone occur in the upper part. The member is 1.16 to 1.7 m thick, and is divisible into two along a persistent horizon of large desiccation cracks filled by calcareous sandstone.

The junction between the Westbury Formation and overlying Cotham Member is sharp and often shows channelling. The calcareous, silty pale grey mudstones at the base of the member contrast sharply with the dark grey shales below. At some localities the channels contain abundant dark mudstone clasts derived from the Westbury Formation. This kind of channel fill is well seen in Seven Sisters Bay [ST 187 699] where the channels are up to 40 cm deep. Ivimey-Cook (1962, 1974) took the top of the Westbury Formation 0.7 m above this channelled level at the horizon of the desiccation cracks, on the basis that the mudstones between contain the abundant bivalve fauna, including Rhaetavicula contorta, of the Westbury Formation (Figure 30).

The succeeding beds of the lower part of the member are calcareous mudstones with thin beds and laminae of calcareous siltstone that locally make up 50 per cent of the thickness. Wavy and lenticular bedding is well developed.

The mudstones are medium to pale grey, becoming grey-green upwards. Locally they are so calcareous that they grade into argillaceous micrite. The lenticular bedded siltstones exhibit straight to locally bifurcating symmetrical wave-ripple marks. The desiccation horizon lies on the top of a unit of pale grey-green mudstone, up to 0.15 m thick at Lavernock Point, that contains synsedimentary deformation structures including microfaults and slump folds. This bed is clearly widespread, for both the deformation structures and desiccation cracks have been recorded from a similar bed in north Somerset (Mayall, 1983).

The base of the upper part of the member is a bed of coarse-grained, commonly oolitic sandstone, normally up to a few centimetres thick but locally reaching up to 26 cm, that fills large polygonal desiccation cracks which reach down in places as far as the top of the Westbury Formation. The cracks are up to 8 cm wide and the polygons up to 90 cm across at Lavernock. The base of the sandstone bed is sharp and commonly erosional. It is cross-laminated and plane-laminated, commonly with straight crested, locally bifurcating ripple marks on its upper surface. The sandstone is blue-grey when fresh but weathers to a distinctive buff colour. It comprises ooids, with nucleii of quartz grains or shell fragments, and peloids.

Above the oolitic sandstone come lenticular and wavy bedded fine-grained sandstones and siltstones with grey-green calcareous mudstones, locally grading to argillaceous micrite. Flaser-bedding is locally present but not common. The micritic Cotham Marble, known to the east of the Bristol Channel, has not been recognised in the district. The top of the member is locally marked by an erosion surface, as in St Fagans Borehole, but elsewhere is sharp or gradational with the overlying basal porcellanous limestones of the Langport Member.

The mudstones in the channel lag deposit at the base of the member contain both fossils derived from the underlying formation and indigenous taxa. Above, they yield a diverse fauna with Cardinia and abundant Gervillia praecursor in addition to most of the bivalve taxa found in the Westbury Formation. The foraminiferan Eoguttulina liassica and comminuted slender echinoid spines are also common. A rapid decline in variety and numbers occurs as the sequence is ascended, until in the grey-green mudstones in the upper part the marine fauna is greatly reduced. The thickness of these near barren strata varies according to the amount cut out by erosion at the horizon of the desiccation cracks. Above this horizon there are few macrofossils, but the presence of both foraminifera (Dentalina, E. liassica, Lingulina tenera and Nodosaria) and ostracods (Darwinula liassica) indicate a marine environment with fresh water influxes.

The erosion surface at the base of the Cotham Member represents a sudden shallowing due to uplift (Mayall, 1979) but there is no evidence of emergence at this level. The lowermost part of the member represents a very shallow and partly restricted marine environment that was rapidly transformed into a lagoonal setting. The lenticular bedded siltstones and sandstones were formed by wave action in this shallow water environment subject to periodic desiccation. The deformed beds and the overlying desiccation horizon have been interpreted as due to contemporaneous earthquake activity followed by uplift in the Bristol Channel area resultant on the tectonic activity (Mayan, 1983). The widely developed oolitic sandstone and overlying beds signify a renewed inundation of the lagoon.

Langport Member

This member, which is 3 to 6.5 m thick, comprises two distinctive facies. The lower, up to 0.65 m thick, is of porcellanous limestones ('Langport Beds' or 'White Lias' of Richardson, 1905, 1911). The remainder of the member comprises grey calcareous mudstones with thin ribs and lenses of fine-grained sandstone, siltstone and limestone. These are the Watchet Beds of Richardson (1911) and Watchet Member of Ivimey-Cook (1974). However Whittaker (1978), comparing them with the beds in the Watchet area, concluded that the two sequences did not correlate.

The lower, 'White Lias', facies contains several thin beds of very pale greyish white weathering, green to green-grey, porcellanous micrite. These are individually up to 11 cm thick, and are separated by thin shaly mudstone partings up to 5 cm thick. The limestones contain thin bioclastic packstone beds with marine micro- and macrofaunas as well as peloids and intraclasts. Identifiable bivalves are commonest on the upper surfaces of the limestones, and include Modiolus, Dimyopsis (Atreta) and Liostrea hisingeri. The mudstones contain foraminifera, ostracods (mainly of marine genera), and fragments of echinoids, fish, and plant tissue. Recrystallised fragments of simple corals are not uncommon in the limestones and Montlivaltia rhaetica occurs at Penarth Head. The facies varies in thickness from 0.13 m in St Fagans Borehole to 0.65 m at Porthkerry. It represents the final marine inundation of the Cotham Member lagoon. The fauna appears to be very shallow subtidal, the thin coquinoidal beds representing storm events.

The 'Watchet Beds' facies is composed dominantly of pale to medium grey, blocky, bedded, calcareous silty mudstones that are rather structureless apart from burrow mottling and a sporadic diffuse lamination noted in St Fagans Borehole. A few thin beds, up to 1.5 cm thick, vary from calcareous siltstone to fine-grained sandstone, and are plane-laminated, cross-laminated or lenticular bedded. Also present are scattered thin shelly argillaceous limestones up to 9 cm thick, some of which are graded. A few limestone nodules centred on oysters are locally present. The fauna is dominated by D. intusstriata, L. hisingeri and Modiolus hillanoides with sporadic Plicatula and Tutcheria. In the basal few centimetres, echinoid fragments and foraminifera, similar to those in the 'White Lias' fades, are commonly abundant. However, the micro- and macrofauna becomes sparser upwards, but quantitive data are difficult to obtain because the fauna is preserved as indistinct casts in these porous silts. The facies rests sharply on the underlying 'White Lias' facies, and is sharply overlain by paper shales at the base of the Blue Lias. It varies in thickness from 1.93 m at St Mary's Well Bay to 4.95 m in St Fagans Borehole. The facies can be interpreted as shallow marine offshore mudstones, with the scattered thin sandstones and limestones comparable to the storm-generated sublittoral sheet sandstones of Goldring and Bridges (1973).

Details

Creigiau

South of Creigiau the outcrop of the Penarth Group is drift covered, but relationships immediately to the west in the Bridgend district suggest that it overlaps northwards onto marginal facies of the Blue Anchor Formation.

Peterston-super-Ely

In Alt Isaf the stream cuts down into the Penarth Group and exposes ripple-marked sandy limestones [ST 0796 7806] of the Lilstock Formation; the Westbury Formation is exposed to the west.

St Nicholas

The Penarth Group overlaps the Blue Anchor Formation northwest of Brooklands [ST 0817 7383], and rests on the latter's marginal facies. North of the point of overlap the Penarth Group sequence attenuates rapidly, and is probably only about 4 m thick though it is not exposed. Immediately to the south, both the constituent formations are present.

Leckwith, Penarth, Lavernock, St Mary's Well Bay

Reasonably complete sections occur in disused quarries at Ely [ST 141 754] and Llandough [ST 166 736]. The best sections in the district occur between the north side of Penarth Head, Lavernock Point and at St Mary's Well Bay. The sections at the two last named localities are shown in (Figure 29), and the fauna from near Lavernock in (Figure 30); the main details of the faunal and lithological successions are from Ivimey-Cook (1962) and Bazley (1968).

The 'bristovi limestones' at the base of the Westbury Formation, were formerly best seen in St Mary's Well Bay [ST 176 677] where the 1.13 m-thick sequence is the maximum known. This section is currently obscured by landslip. Channels cut into the Blue Anchor Formation to a depth of 0.1 m are overlain by 0.23 m of grey-green shale containing 8 to 23 cm of mudstone conglomerate with an argillaceous matrix. Above this are 0.9 m of grey coquinoidal limestone rich in Liostrea bristovi; the bed has a lenticular shale parting and the upper segment of limestone is massive, more argillaceous and contains broken and commonly bored Liostrea. Also present are Rhaetavicula contorta, sporadic specimens of a large Modiolus, Sphaerodus teeth and wood fragments. The top of the limestone is argillaceous, silty and shows minor erosion beneath the overlying fine-grained ossiferous sandstone. A palynomorph assemblage from the ‘bristovi limestones’ is dominated by miospores, principally Ovalipollis pseudoalatus and Classopollis spp.; a few specimens of the dinoflagellate cyst Rhaetogonyaulax rhaetica and other organic-walled microplankton (acritarchs) are present (Figure 26).

In the coast section [ST 188 681] just north of Lavernock Point, the 'bristovi limestones' sequence (Cycle 1 of Ivimey-Cook, 1962) is thinner and the coquinoidal limestones are missing. On the foreshore, the top dolomite bed of the Blue Anchor Formation has an eroded top which Maya]] (1981) reported contains large 'U' burrows up to 10 cm deep; it also has other cracks and fissures. These are all filled with mudstone, some quartz grains and rare fish fragments. Some 15–20 cm of dark shale above contains rare Rhaetavicula and Protocardia. It is capped by 5–10 cm of fine-grained dolomitic limestone which forms a continuous bed on the foreshore to the south-east but closer to the cliffs it is deeply eroded and in the cliff it becomes a row of nodules. Traced north both it and the underlying shale are rapidly cut out by a minor non-sequence.

Inland, thin sequences of the 'bristovi limestones' are exposed in a lane section south of Murch [ST 1652 7090]–[ST 1655 7085] and others were recorded by Richardson (1905) and Ivimey-Cook (1962), eg, near Cross Farm [ST 157 706]. Liostrea bristovi occurs just above the base of the Formation at Dinas Powis [ST 155 710] and traces of the unit may be present north of Cadoxton as a few centimetres of arenaceous limestone.

Higher in the Lavernock cliff section [ST 188 681], above the 'bristovi limestones' sequence, there is only a trace of the ossiferous sandstone known as Storrie's 'Fish Bed'. On the foreshore there was a 60 x 20 m area, now substantially destroyed by recent marine erosion, where the top limestone of the 'bristovi limestones' had been eroded to leave a maze of steep-sided and relatively flat-bottomed channels filled with coarse-grained ossiferous sandstone and conglomerate. Well rounded quartz grains, granules and pebbles of quartzite, chert, limestone and mudstone up to 5 cm diameter, together with bones and fish remains, up to 25 cm long, occur. The sandstone contains up to 50 per cent fragments of teeth, scales and bones of fish and reptiles, with some coprolites. Taxa present include the teeth of Acrodus, Birgeria, rare Ceratodus, Dalatias, Hybodus, Lissodus, ‘Sargodon tomicus’, Saurichthys, 'Sphaerodus' and the long fin-spines of Hybodus and Nemacanthus monilifer. Gyroleptis and numerous unidentified scales and bones occur including reptilian vertebrae and other bones. Many of the larger remains have been broken, eroded and the softer parts packed with sediment; pyrite crystals are common.

This ossiferous sandstone forms the base of Cycle 2 (Figure 30). It fines rapidly upwards into siltstone and dark shale with thin lenses of grey siltstone together with pyrite as scattered nodules and laminae and some comminuted fish fragments. The shales contain a sparse bivalve fauna but already Eotrapezium, Protocardia rhaetica and R. contorta, the most characteristic taxa of the formation, are present. The squashed remains of a small ?neritoid gastropod (Natica oppelii) are here preserved in fibrous calcite, but more usually in pyrite. Three plates of the small pedunculate cirrepede Eolepas were also found. Like those present in more abundance at higher levels they are small compared with E. rhaetica, but identifiable, fragmentary, cardinal, tergal and scutal plates occur. The higher levels of the mudstones in this cycle are sparsely fossiliferous but do contain brown coprolite-like pellets, as 0.5 to 1 mm-long rectangular cylinders with some markings, compared (Bazley, 1968) with Bactryllium striolatum'. These also occur in Cycles 2 and 4 inland.

The third Cycle is both lithologically and faunally distinctive. Its base is a slight erosion surface overlain by discontinuous fine-grained limestone lenses with Eotrapezium and fish fragments. The overlying 0.6 m consists of alternations of shales and pyritic sandy limestones. It includes the 'bone bed proper' (Storrie, 1883), a thin (1 cm) layer of arenaceous pyritic limestone packed with comminuted fish fragments. The other limestones vary up to 10 cm in thickness, generally show load casts and other surface features and some include flat to slightly curved and broken rafts of somewhat indurated laminated mudstone. The fauna is rich in the marine ostracods Metacytheropteron and Rhombocythere (Anderson, 1964) as well as frequent Eolepas, especially in the somewhat carbonaceous mudstones. Bivalves include numerous Eotrapezium concentricum, E. cf. germari, Pleurophorus elongatus and rare Rhaetavicula. Above, grey blocky mudstones yield R. contorta, Lyriomyophoria postera, 'Modiolus' sodburiensis and some 'N' oppelii; however, the bivalves become rare upwards whilst ‘Natica’ becomes abundant. The top of the cycle is marked by a rapid change to calcareous mudstone with corn-minuted shell fragments.

The fourth Cycle begins with calcareous siltstones passing rapidly into massive, locally nodular, fine-grained limestone. Similar siltstones occur between the nodules and for a few centimetres above. They contain numerous bivalves, especially Chlamys valoniensis, Lyriomyophoria (mainly above and below limestone), 'M.' sodburiensis, Placunopsis alpina, abundant comminuted R. contorta and diademopsid plate and spine fragments. The fauna is abundant through the overlying mudstones and into a bed of calcareous mudstone about 1 m above the base. Tutcheria cloacina is particularly common and these beds also yield ophiuroids and more echinoid fragments (Cycle 4a of Bazley, 1968). Above however, there are 0.4 m of dark hard mudstone with a sparse bivalve fauna. Inland these upper beds yield E. concentricum, 'M. ' sodburiensis, 'N.' oppelii and Bactryllium aff. striolatum.

Cycle 5 is thinner, extending up from a thin basal calcareous silty mudstone, into the lower 'Pecten Bed' limestone–with limestone clasts in its base and calcite 'beef on top. Above, grey, slightly silty, mudstones yield abundant Chlamys valoniensis, Placunopsis alpina and some ophiuroids in the more calcareous and coarsest beds. Rare Liostrea and large Modiolus occur. Higher mudstones have layers rich in R. contorta, Eotrapezium and L. postera but these become rapidly rarer upwards. Simultaneously, 'N.' oppelii increases in abundance upwards to a level about 0.08 m below a 1 cm-thick shell fragmental and silty ferruginous sand layer.

Cycle 6 starts with a thin sand parting which is rich in fragments of R. contorta. Above, calcareous mudstones with numerous P. rhaetica and R. contorta, pass up into a thin argillaceous limestone and shale with layers of fibrous calcite up to 10 mm thick forming the 'Upper Pecten Bed'; it is richly fossiliferous, especially with C. valoniensis and P. alpina. The suceeding 15 cm of mudstone has abundant Lyriomyophoria, Protocardia, Tutcheria and some Rhaetavicula. Higher, all bivalves become scarce, whilst clusters of 'N'. oppelii are abundant. The uppermost 0.6 m of mudstones are amongst the most barren in the sequence but in the top 20 cm a bivalve fauna, first with Protocardia, then Eotrapezium and Rhaetavicula reappears on some bedding planes with 'N.' oppelii. In the Westbury Formation the upward diversification of the palynomorph associations first seen in the 'bristovi limestones' continues. Those from the lower part of the sequence are dominated by miospores but include a few organic-walled microplankton, mainly dinoflagellate cysts and acanthomorph acritarchs (Figure 26). In the higher beds the assemblages are dominated by the dinoflagellate cyst Rhaetogonyaulax rhaetica. Herkomorph acritarchs, (Cymatiosphaera) are present in most assemblages; polygonomorph acritarchs (Veryhachium) and Tasmanaceae occur sporadically. Insoluble test linings of foraminifera have been recorded from the highest beds.

The Lilstock Formation is also well exposed in the cliffs at Laver-flock and St Mary's Well Bay although less accessible near Penarth. Faunally the formation is very variable and although principally marine, there is a dominance amongst the bivalves of the shallow water attached forms Dimyopsis, Liostrea, Modiolus and, more sporadically, Plicatula and Tutcheria with other taxa being generally very sporadic in occurrence (Figure 30). The palynomorph assemblages continue to diversify into the Gotham Member. The two lower assemblages (below the desiccation level) are marine and dominated by Rhaetogonyaulax rhaetica; they are similar to those from the top of the Westbury Formation (Figure 26). The assemblage from the upper part is dominated by miospores, principally Gliscopollis meyeriana, and contains very few organic-walled microplankton.

The assemblage from the 'White Lias' facies at the base of the Langport Member contains only a few specimens of the miospores Gliscopollis meyeriana, Classopollis spp.,Deltoidospora and some indeterminate bisaccates. The overlying 'Watchet Beds' facies contains assemblages dominated by a few miospore taxa, Classopollis spp., G. meyeriana, Quadraeculina anellaeformis and Kraeuselisporites reissingeri. In composition and character these are closer to those from the Lias than from the rest of the Penarth Group. The organic-walled microplankton associations contain very few dinoflagellate cysts and mainly comprise acanthomorph acritarchs (Micrhystridium). The marked change in character of these assemblages is possibly controlled by environmental changes; it does not serve to identify the Triassic/Jurassic System boundary.

Porthkerry–Barry

A faulted broad anticline in the foreshore reefs at Bull Cliff exposes only the Lilstock Formation beneath the Blue Lias. The sand-filled desiccation-crack horizon in the Cotham Member is well displayed. In reefs in the old Barry Harbour [ST 1055 6665] the same part of the sequence is also well displayed. Here the oolitic sandstone is at its thickest in the district (0.26 m).

Chapter 7 Jurassic: Lower Lias

The late Triassic transgression continued into the early Jurassic and widespread marine sediments were deposited in a shallow continental (epeiric) sea that covered much of southern Britain. Within the district only the lower part of the Lower Lias, which rests conformably on the Penarth Group, is preserved. It comprises about 110 m of Hettangian and early Sinemurian sediments. These consist predominantly of an 'offshore facies' of calcareous mudstones (marls) and argillaceous calcilutites known as the Blue Lias. In the adjacent Bridgend district, remnant Carboniferous massifs, inundated during this period, were fringed by a marginal (shoreline) facies of conglomerates and calcarenites formerly known as 'litoral' facies (Strahan and Cantrill, 1902, 1904). Continued uplift along the Vale of Glamorgan Axis sustained this marginal facies throughout much of the Hettangian and early Sinemurian (Wilson and others, in press). Within the Cardiff district the marginal facies only occurs in the north where it was the result of a major progradation into the offshore facies during the lower Sinemurian, probably reflecting changes in the relative rates of uplift and drowning over the Vale of Glamorgan Axis. It comprises skeletal, oolitic and peloidal limestones that are excluded from but enveloped by the Blue Lias.

The Blue Lias crops out round Barry, where it forms the eastern end of the main Vale of Glamorgan outcrop, and in outliers between Lavernock Point and Leckwith. In the north-west of the district the Lower Lias in both marginal and Blue Lias facies occupies scattered outliers, now faulted into the core of the Cardiff–Cowbridge Anticline.

Apart from a brief mention of the Lower Lias at Penarth Head by Wright (1860), the first description of the Lower Lias was in the vertical sections drawn up by Bristow and Etheridge (1873) for the Penarth and Lavernock area. A further description was given by Woodward (1893, p.121). The first detailed description of the Lower Lias of the district was in the Geological Survey memoir for the Cardiff area (Strahan and Cantrill, 1902). Further, largely biostratigraphical, descriptions of the succession were provided by Trueman (1920, 1922). More recently, sedimentological and faunal studies of the Lower Lias in the Vale of Glamorgan are contained in Hallam (1960) and Wobber (1965, 1967, 1968a, 1968b), though these are mainly restricted to the coastal sections. Ager (1974) has subsequently presented a review of work on the Lower Jurassic of South Wales.

The present classification of the Lower Lias is set out and compared with that used by Trueman (1920) in (Figure 31) (opposite). Correlation with other areas is best achieved by means of ammonites. On the basis of the local ammonite fauna, the figure shows the relationship of the lithostratigraphical units and the standard ammonite biozones of the Lias of North-West Europe (Dean and others, 1961; Ivimey-Cook and Donovan in Whittaker and Green, 1983). In common with current practice, the base of the Jurassic System is taken at the base of the planorbis Subzone (Cope and others, 1980), which is about 5 m above the base of the Blue Lias in the district. This basal 5 m of strata, which are Triassic in age, are described in this chapter for reasons of practicability and because they are lithologically part of the Blue Lias.

Blue Lias

The term Blue Lias has been variously used but has recently been formally defined with a type area between Bath and Bristol (Cope and others, 1980). Typically it consists of alternating thin-bedded argillaceous calcilutites and calcareous mudstones (marls) in varying proportions. Some 88 m are exposed in the south of the district, whilst in the core of the Cardiff–Cowbridge Anticline it is about 78 m thick (excluding 32 m of marginal facies in the uppermost part of the succession). Nowhere in the Vale of Glamorgan is there a sequence of Lower Lias shales overlying the Blue Lias. The thickest known sequence is in a borehole at Bridgend which proved nearly 150 m of Lower Lias (Francis, 1959); this includes some marginal facies in the upper part.

Within the district the Blue Lias is conformable upon the Penarth Group, the base being drawn at the base of a distinctive bed of fissile shale above which the alternating limestones and mudstones enter abruptly, contrasting markedly with the mudstone-dominated Lilstock Formation.

The sequence has been subdivided into three formations on the basis of limestone–mudstone ratios (Figure 31). The lowest, the St Mary's Well Bay Formation, comprises mudstones with significant, but just subordinate, limestones; the Lavernock Shales above are mostly mudstones; in the overlying Porthkerry Formation, limestones dominate.

The Blue Lias is well exposed in the cliffs between Porthkerry and Barry, and between St Mary's Well Bay and Penarth Head. Inland exposure is poor apart from a few disused quarries but the component formations can generally be mapped inland since the limestone-rich ones produce strong topographic features; and the Lavernock Shales occupy the lower part of a major escarpment formed by the Porthkerry Formation. Where the features are absent the Blue Lias cannot be subdivided. The St Fagans Borehole, sited in one of the fault-bounded inliers in the core of the Cardiff–Cowbridge Anticline, provides the thickest Lower Lias succession in the district, comprising Blue Lias with marginal facies forming the uppermost 30 m (Lawrence and Waters, 1978).

St Mary's Well Bay Formation

The St Mary's Well Bay Formation is a stratigraphic term, here introduced, to cover some 16 to 18 m of mudstones with slightly subordinate limestones that forms the lower part of the Blue Lias. The designated type-section is the cliffs between St Mary's Well Bay and Lavernock Point [ST 1765 6776]–[ST 1874 6813] (Plate 1) where the formation is 16.2 m thick, and is completely exposed and accessible (Figure 32). Its base is taken at the base of a distinctive 20 cm unit of laminated very fissile shale, the 'Paper Shales' (bed 92 of Richardson, 1905). The top of the formation is gradational, but has been defined at the top of a prominent limestone bed (bed 86 of (Figure 32)). The formation ranges in age from the topmost Triassic to the lower part of the liasicus Zone. It is known to be thicker (18.2 m) in St Fagans Borehole.

Two types of limestone and mudstone are present. In the most prevalent type the limestones are bluish grey argillaceous calcilutites, containing sparse comminuted bioclastic debris as well as scattered bivalves, ammonites, brachiopods and echinoderm debris. Apart from burrows, they are mostly structureless. The mudstones are bluish grey, variably calcareous, and shaly; they contain a similar fauna and are burrowed. Mudstone–limestone boundaries are gradational due to diagenetic modifications. The thicknesses of individual limestone beds vary from about 5 cm to 0.3 m; the mudstones vary in thickness from partings to 0.7 m. The limestone beds are both parallel-sided or undulatory, the two types being gradational, while isolated limestone nodules or beds of nodules also occur.

The second type of limestone and mudstone is characterised by lamination and a general lack of burrowing. The mudstones are finely laminated, variably bituminous and become fissile on weathering. The limestones are argillaceous calcilutites with dark carbonaceous laminae; they locally pass laterally into laminated shale. Laminated bituminous shales form the lowest member of the typical Blue Lias cycle (Hallam, 1960) that comprises, in upward order, laminated shale, burrowed calcareous mudstone (marl) and limestone. However, laminated lithologies only occur at three levels in the formation.

The Bull Cliff Member is a new stratigraphic term here given to a 3.1 to 3.6 m unit of generally parallel-sided limestones and subordinate mudstones, rich in oysters, lying at the base of the formation. The type-section is at Bull Cliff, Porthkerry [ST 0920 6670]. The distinctive 'Paper Shales' lies at the base and consists of 0.17 to 0.24 m of delicately interlaminated pale grey calcareous siltstone and grey calcareous slightly micaceous silty mudstone. The siltstone laminae are up to 3 mm thick and contain abundant Liostrea hisingeri, Modiolus sp.and echinoderm debris. Above the 'Paper Shales' the individual limestones are 3 cm to 0.2 m thick and unlaminated, the mudstones range from thin partings to beds up to 0.22 m thick. The shallow water attached suspension feeders Liostrea hisingeri and Modiolus dominate the bivalve macrofauna while diademopsid spines are abundant at some levels. The bivalves occur both in the mudstones and limestones, either scattered or as winnowed coquinas up to 3 cm thick. Laminae of fine-grained shell-hash less than 1 mm thick are common in the mudstones, and represent winnowed lags. The top of the member is taken at the base of the lowest nodular bedded limestone, informally termed the 'Dual bed', which has one or, in places, two mudstone partings within it. Oysters are less common above this level.

Above the Bull Cliff Member the percentage of limestone decreases (Figure 32) though beds of limestone nodules are common in the upper part. Three distinctive units can be used as marker beds (see (Figure 32)): they are the Planorbis Mudstones, the Lower Laminated Beds and the Upper Laminated Beds, the latter two having been recognised as far afield as Dorset (Hallam, 1964).

The Planorbis Mudstones (bed 30 of Trueman, 1920) lie 6 to 7 m above the base of the formation. At Lavernock the unit consists of 0.65 m of generally very fissile, dark grey, mudstones with abundant, commonly very poorly preserved, impressions of the ammonite Psiloceras planorbis on the bedding surfaces. The unit is the same thickness at Barry Harbour [ST 1044 6649]–[ST 1070 6683] and Bull Cliff [ST 0920 6670], but at the latter locality 0.08 m of limestone occurs within it. In St Fagan's Borehole, brown traces on the surfaces of a laminated mudstone bed at 95.47 m, 7.35 m above the Penarth Group, include very poorly preserved specimens of P. planorbis.

The Lower Laminated Beds occur approximately 8 to 9 m above the base of the formation. They consist of one or more thin parallel-sided laminated limestones that pass locally into very fissile laminated shale. At Barry Harbour and Bull Cliff the unit is 0.08 m thick; it attains its maximum of 0.7 m at Lavernock Point (part of bed 33 of Trueman, 1920) where laminated shale forms the upper part. In St Fagans Borehole only laminated shales are present at this level. The Upper Laminated Beds occur about 10 to 11 m above the base of the formation and are up to 0.9 In thick. They consist of two or three thin parallel-sided laminated limestones with intervening fissile to laminated shales. West of Barry they contain three limestones; the middle one locally passes into shale, and the topmost one is laminated only on Barry Island [ST 1137 6702] to [ST 1189 6703]. A distinctive 0.08 m limestone ((Figure 32), bed 59) (bed 39 of Trueman, 1920) is laminated throughout the cliff-section between St Mary's Well Bay and Lavernock, and is overlain by 0.13 m of laminated shales with a thin locally laminated limestone bed above.

The formation is remarkably constant in the coastal outcrops, and it is possible to correlate groupings of beds in the separate sections, although there is some lateral variation within individual beds, particularly the nodular limestones. Even so, the sections at Bull Cliff [ST 0920 6670] and Laver-flock, which are some 10 km apart, can be matched almost bed for bed (Figure 32).

Inland exposure is extremely poor apart from a few disused quarries. Temporary sections [ST 1007 6845]; [ST 0984 6817] near Barry have yielded a johnstoni Subzone fauna, whilst another [ST 1039 6817] contained Waehneroceras, indicating the liasicus Zone. Downswood quarry [ST 1720 7020], south of Penarth, exposed 8.5 m of beds in the lower part of the formation. These sections show a sequence little different from the coast sections. The thickest succession is in St Fagans Borehole (Figure 32), but it cannot be correlated in such detail, though the Bull Cliff Member and the three marker beds are all present.

Lavernock Shales

The term Lavernock Shales was introduced by Strahan and Cantrill (1902, p.68) for some 12 m of 'shales or marls, almost devoid of limestones', within the Blue Lias of the district. Trueman (1920, pp.99–100) gave a composite log of the formation for the Penarth and Lavernock areas, stating that it was about 12 m thick (Trueman, 1920, p.95) and composed of his Waehneroceras beds and part of his angulata beds. In the present account this term has been redefined and the lower boundary has been taken at a higher level. The formation is entirely within the liasicus Zone.

The type-locality is the coastal section between St Mary's Well Bay and Lavernock Point [ST 1798 6782] to [ST 1857 6808] (Plate 1) where the sequence consists of approximately 12 m of blue-grey, very calcareous, shaly mudstones with very subordinate pale blue-grey calcilutites, commonly as nodular beds and limestone nodules, occurring between the St Mary's Well Bay Formation and the Porthkerry Formation. The base is gradational, but has been taken at the top of a particular continuous limestone bed at the type section ((Figure 32), bed 86). Although this bed can be recognised at Bull Cliff, Porthkerry [ST 0920 6674], it could not be recognised in St Fagans Borehole. In this section the base of the formation has been taken at a point where the percentage of mudstone increases markedly. The top has been taken where the dominantly mudstone sequence gives way (over about 2 m) to mudstones and nodular limestones in roughly equal proportions (Porthkerry Formation). At the above mentioned coastal localities and at Penarth Head the formation is for the most part inaccessible and the sections partly degraded, though there are exposures in a faulted block in the near-vertical cliff at Penarth Head. The succession in St Fagans Borehole, where it was 11.31 m thick, is the only one which has been examined in detail. Inland although the formation can be mapped easily by the topographical features it produces, it is very poorly exposed; though a degraded section in the old railway cutting [ST 1789 0858] to [ST 1810 6881] exposes about 7 m of dark grey, brown-weathering shaly mudstones with thin beds of nodular limestone.

The mudstones are medium to dark grey, calcareous and variably fissile; they occur in beds up to 2.25 m thick. They are burrowed, and are variably fossilifcrous, containing crinoid and echinoid fragments, ostracods, bivalves and, less commonly, ammonites. Sparse very thin laminated mud-stones were noted in St Fagans Borehole. The limestones are similar to those in the underlying formation but occur as nodules and nodular beds up to 0.15 m thick, and are commonly less than 0.1 m thick. They are generally poorly fossiliferous, but contain sonic lenses rich in bivalves and are bioturbated in places.

Porthkerry Formation

The Porthkerry Formation is a stratigraphic term here introduced for that part of the Blue Lias succession in the Vale of Glamorgan lying above the Lavernock Shales and is thus equivalent to the earlier informal 'upper limestone series' (Trueman, 1922). The formation consists of alternating limestones and subordinate mudstones broadly similar to those predominating in the St Mary's Well Bay Formation; no laminated lithologies have been noted. The limestone–mudstone ratio gradually increases from 50 per cent in the lower part to 80 per cent in the upper. The base of the formation is transitional over about 2 m from the dominantly mudstone sequence below to limestones and mudstones in approximately equal proportions above; its top is not preserved in Glamorgan. Between Dunraven [SS 885 730], in the adjacent Bridgend district, and Porthkerry it has a minimum thickness of 105 m (Wilson and others, in press). Within the district it ranges in age from late liasicus Zone to late semicostatum Zone. The designated type-locality is the cliffs west of Porthkerry where about the lower half of the formation, (some 52 m but excluding about 8 m at the base), is visible and accessible between The Bulwarks [ST 0810 6612] and Dams Bay, 700 m to the west in the adjacent Bridgend district (Figure 33). In the north-west, the lower 80 m of the formation are exposed in the core of the Cardiff–Cowbridge Anticline, but this includes a wedge of marginal facies, up to 32 m thick in the upper part. The St Fagans Borehole proved 31.1 m of marginal facies overlying 42 m of the Porthkerry Formation (Figure 34). Immediately to the north of the borehole the marginal facies is overlain sharply by about 5 m of Blue Lias that is included within the formation.

On the coast between Whitemore Stairs [SS 898 712], in the adjacent Bridgend district, and Porthkerry, the formation has been divided into four informal lithostratigraphic units (A to D) (Wilson and others, in press) which differ markedly in their gross bedding characteristics and limestone–mudstone ratios. Only the lower two of these are exposed on the coast within the district.

Unit A is about 36 m thick between Dams Bay and Porthkerry (Figure 33), including an estimated 8 m for the gap between the base of the cliffs at The Bulwarks and the top of the Lavernock Shales exposed on the east side of Porthkerry Bay. It consists of impersistent nodular limestones, on average 0.08 to 0.2 m thick, with wavy bedding planes, interbedded with shaly mudstone beds that are usually up to 0.2 m thick. Limestone–mudstone ratios are on average 55:45 m the lower part, rising to 65:35 m the upper part. There are few distinctive marker beds, but individual mudstone beds that are thicker ttian normal (up to 0.48 m) can be used, as can the thicker limestones, the thickest of which is 0.5 m, though such limestones are composite, having shaly partings. In the upper part of Unit A, beds crowded with Gryphaea are common, whilst Thalassinoides burrows are common on the under-surfaces of limestone beds. Schlotheimia occurs in beds 7, 23 and 25 indicating the angulata Zone.

Only the lowest 21 m of Unit B are exposed between The Bulwarks and Dams Bay (Figure 33) (Plate 10). The base of the Unit is a very distinctive 3 m-thick 'package' of 16 or 17 rather planar-bedded limestones, in beds 0.07 to 0.21 m thick with interbedded mudstones up to 0.1 m thick. This 'package' (bed 33) commonly projects as a ledge in the cliffs. The Unit contains thicker individual limestone beds than those of Unit A. There are also composite limestone beds, up to 1.68 m thick, which are made up of up to seven rather nodular limestones separated by commonly anastomising mudstone partings; these composite beds commonly have planar tops and bottoms. The mudstones between the composite units are up to 0.46 m thick, and are on the whole thicker than those of Unit A. Most of the beds in Unit B can be traced along the cliffs, and one has been given the informal name of the 'Sandwich bed' (bed 53). The thickest limestone in Unit B (bed no 59) has been informally termed the 'Main Limestone'. Some of the limestone beds contain more bioclastic material than is normal in the Blue Lias, and Thallasinoides burrows are common on the under-surfaces. The occurrence of Vermiceras in beds 38 to 41 indicates the conybeariSubzone, and Coroniceras cf. rotiforme in bed 49 indicates the rotiformeSubzone of the bucklandi Zone.

Units A and B have not been recognised inland, but St Fagans Borehole shows the same upward increase in limestone content above the Lavernock Shales. It is estimated that Unit A is about 34 m thick in the borehole, and Unit B about 9 m thick. The rest of Unit B together with Unit C, is represented by marginal facies. On the coast in the Bridgend district, Unit C comprises 7 m of massive composite limestone units with thin wispy mudstone partings and beds; the limestone–mudstone ratio is 85:15. Caloceras, Waehneroceras and Psilophyllites occur in the lowest part of Unit A in St Fagans Borehole indicating a liasicus Zone age, and Schlotheimia in the upper part an angulata Zone age.

In the north-west of the district a wedge of marginal facies, that probably all belongs to the bucklandi Zone, is intercalated high in the Porthkerry Formation. The upper contact of this wedge can be seen in Slanney quarry [ST 1155 7845] (Figure 34), immediately north of St Fagans Borehole, where Blue Lias rests sharply on a bored surface at the top of a marginal facies oolite. The Blue Lias comprises 4.5 m of alternating limestones and mudstones with ammonites and abundant Gryphaea, with 0.3 m of mudstone overlying the junction. The borings are filled by brown ooids and very sparse quartz granules and the lowest limestone contains scattered ooids. The limestones are individually up to 0.23 m thick; the thickest mudstone is 1.22 m thick. These strata are the youngest Porthkerry Formation in the district and contain the bucklandi–semicostatum boundary. They correlate with Unit D on the coast in the Bridgend district. The latter comprises alternating limestones and mudstones with a limestone-mudstone ratio of 60:40; the bucklandi–semicostatum boundary occurs in the lowermost part (Wilson and others, in press).

Marginal facies

The marginal facies of the Lower Lias in the district consists of massive to thinly bedded oolitic, peloidal and skeletal limestones with scattered mudstone partings, local chert nodules, and selective silicification. It is developed only within the upper part of the Porthkerry Formation in the north of the district where it outcrops in fault-bounded outliers in the core of the Cardiff–Cowbridge Anticline. It is probably all of bucklandi Zone age. In the St Fagans outlier it is up to 32 m thick and it appears to be about the same thickness elsewhere. It is exposed in numerous small disused quarries and pits, but the only complete section is St Fagans Borehole (Figure 35).

Its lower junction with the Porthkerry Formation is gradational whilst its upper boundary is sharp. The lower boundary is taken at the incoming of limestones in which abundant allochems, especially ooids and peloids, are visible; mudstones within the facies are mostly mere partings. In St Fagans Borehole the lithologies are arrayed in a shoaling-upwards sequence from Blue Lias calcilutites, through oolitic, skeletal and peloidal packstones, to oolitic skeletal grainstones.

The packstones in the facies are buff-weathering, medium grey, fine grained, and very variable in composition. The proportion of ooids varies considerably from being barely significant to being dominant. The commonest lithology is one in which ooids are subordinate to bioclastic and peloidal components. The bioclastic material is mainly echinoderm debris, but bivalve and gastropod material is also present. Detrital quartz is present as very fine sand grade grains. Bioclastic material, especially echinoderm fragments, is locally silicified. Replacive chert nodules are also common, as are carbonised wood fragments. The packstones occur both in thick units up to 1.5 m thick and in beds 0.15 to 0.6 m thick that are separated by undulatory bedding planes, some of which have associated mudstone or argillaceous packstone partings. Undulatory bedding planes are characteristic and are well seen in surface exposures. Current structures include lamination, low angle cross-bedding and cross-lamination. Scoured surfaces overlain by coarse bioclastic material are apparent in the thicker units, and burrowing is common.

The ooid skeletal grainstones are cream to pale orange weathering, but grey when fresh. The ooids are mainly 0.3 mm in diameter and micritic. Peloids, shell fragments, echinoid plates and spines, and crinoid columnals are also present in variable quantities, and scattered carbonised wood fragments are common. Many of the bioclasts have micrite envelopes. Locally, bioclasts have been silicified, whilst chert nodules have been noted in surface exposures. A few beds of skeletal packstone, up to 4 cm thick, occur in places within the grainstones. In St Fagans Borehole the grainstones are massive or laminated, but trough cross-bedding is seen in some exposures. Mudstone partings are uncommon; the few that are present are up to 3 mm thick. Some burrows and escape traces have been noted. In the St Fagans area the top of the grainstone sequence is sharp, as in Slanney quarry [ST 1153 7844], immediately to the north of St Fagans Borehole, where it has a rolling surface with numerous 2–4 mm-wide straight to slightly curving cylindrical Trypanites borings which truncate ooids. The ooid grainstone is iron-stained in the top few centimetres.

In the Creigiau area the marginal facies consists of variably oolitic, skeletal, and peloidal packstone. It forms a small poorly exposed outlier bounded to the east by the Crcigiau Fault. The Blue Lias below is also poorly exposed; the outcrop is largely drift covered and the position of the Lavernock Shales is uncertain; the exact position of the marginal facies in the sequence is thus not known.

Details

Creigiau

In the railway cutting at Creigiau station [ST 0833 8138] to [ST 0837 8125] scattered sections, up to 1.2 m high, expose well-bedded, pale grey, fine-grained, variably oolitic, peloidal and skeletal packstones.

Peterston-super-Ely to Stockland

In shallow pits [ST 0827 7742] to [ST 0834 7738], south-west of Palla Farm, up to 5.1 m of very fine-grained skeletal packstone, in beds 0.15 m to 0.6 m thick with undulatory bedding surfaces, are interbedded with more argillaceous units less than 7 cm thick. Lamination, in places steeply dipping and/or overturned, is visible on weathered surfaces.

Farther east about 12 m of marginal facies, gradationally overlying the Porthkerry Formation, are exposed in a disused railway cutting [ST 0990 7875] to [ST 1014 7813] near Tregurnog. The succession consists of pale grey, very fine-grained packstones containing scattered spicular and nodular chert, and with partings and beds of mudstone up to 0.1 m thick. The limestones contain undulatory bedding planes, and in the upper part of the succession are coarser grained with recognisable ooids; individual beds increase in thickness from 0.08 to 0.38 m at the base to 0.38 to 0.61 m at the top. The junction with the underlying Porthkerry Formation has been placed at the base of the cutting at its northern end but, since there appears to be a downwards passage from the marginal facies into the Porthkerry Formation, marked by increasing amounts of interbedded mudstone and argillaceous calcilutite and a decrease in packstone beds, the boundary is imprecise. A disused quarry [ST 0990 7865], immediately to the south-west in beds stratigraphically higher than the railway cutting, exposes 4.5 m of pale grey oolite with chert nodules.

St Fagans

Coedbychan Quarry [ST 127 774], immediately east of St Fagans, exposes the following sequence near the base of marginal facies.

The sequence is poorly fossiliferous, but yielded crinoid columnals, Camptonectes, Gryphaea cf. arcuata (juv .), ostreid fragments, Pholadomya?, Pronoella, Pseudolimea pectinoides and Amberleya (Eucyclus) sp.The horizon is likely to be of late angulata to bucklandi Zone age.

Biostratigraphy

The Triassic rocks that form the lowest few metres of the Blue Lias are without ammonites and were termed the pre-planorbis Beds by Richardson (1904). They are 4.7 m thick at Lavernock Point, 5 m at the Bull Cliff, Porthkerry, and up to 7.3 m in St Fagans Borehole. They include the Bull Cliff Member, in which the macrofauna at the type section is dominated by two shallow-water epifaunal bivalves, the attached suspension feeders Liostrea hisingeri and Modiolus. There are scattered specimens of Meleagrinella, Plagiostoma and Pteromya, and some levels contain abundant diademopsid spines, indicating a carbonate mud community. Above the 'Dual bed', burrowing bivalves, Pholadomya, Pinna and the large semi-buried Plagiostoma occur together with various pectinaceans, including Camptonectes, Chlamys pollux, Oxytoma and Terquemia arietis, indicating both an attached and motile epifauna. A similar faunal change occurs in St Fagans Borehole; the basal unit of laminated silty mudstones with Meleagrinella decussata and Pteromya, passes upwards through limestones and shales with Cuneigervillia, Liostrea and Modiolus minimus into a more diverse assemblage. Palynomorph assemblages from the Pre-planorbis Beds are rather sparse and are less varied than those in the 'Watchet Beds' facies, at the top of the Langport Member. One from 3.5 m above the base of the unit (1.2 m below the base of the Jurassic) is dominated by Classopollis spp.and Kraeuselisporites reissingeri. Associated organic-walled microplankton largely comprise Micrhystridium, though Veryhachium? and Tasmanaceae are also present (Figure 26).

The remainder of the Lower Lias has been zoned by its ammonite fauna, and the relationship of the constituent zones and subzones to the stratigraphical subdivisions is set out in (Figure 34). The oldest ammonite-bearing beds within the St Mary's Well Bay Formation fall within the planorbisSubzone, which has been collected from at Lavernock Point and Bull Cliff, Porthkerry. The earliest ammonites are indeterminate psiloceratids found both by Whittaker (Cope and others, 1980, p.21) and in the present survey, in bed 30 at Lavernock and bed 22 at Bull Cliff. They are taken to mark the base of the planorbisSubzone (and the Jurassic). In the fissile calcareous mudstones (bed 32 at Lavernock, and beds 31–33 at Bull Cliff), there are abundant flattened brown impressions of Psiloceras planorbis. Higher in the sequence there is a slight difference between the two sequences, as the plicate form (Psiloceras plicatulum) is common in beds 32–33 at Bull Cliff, whilst at Lavernock (beds 39–41) the specimens are dominantly of the smooth P. planorbis. Other taxa include crinoid and echinoid fragments, Calcirhynchia calcaria (occurring very early at Bull Cliff in bed 24), Astarte, Cardinia?, Liostrea hisingeri, Modiolus laevis, M. minimus, Pholadomya, Plagiostoma giganteum and the large gastropod Pleurotomaria. The subzone is 3.4 m thick at Bull Cliff, 4.6 m at Lavernock and about 3 m in St Fagans Borehole. A palynomorph assemblage from this subzone is dominated by Classopollis and Gliscopollis with only sparse representatives of other taxa, including organic-walled microplankton (Figure 26).

The johnstoniSubzone ranges through about 4 to 5 m of strata extending upwards from the first appearance of the genus Caloceras, which is found in bed 41 at Bull Cliff and bed 52 at Lavernock, to the appearance of the schlotheimiid ammonite Waehneroceras in bed 58 and 77 respectively: it thus falls within the St Mary's Well Bay Formation. The two sections are very similar faunally, containing both surface living bivalves such as Camptonectes, Chlamys, Cuneigervillia, Liostrea, Modiolus and Pseudolimea, and shallow burrowing genera including Mactromya, Pholadomya and Pinna (St Fagans Borehole), and Plagiostoma. High in the subzone some specimens of Caloceras have more closely ribbed inner whorls, and can be referred to C. cf. intermedium. In St Fagans Borehole the oldest Caloceras was found at 92.4 m and specimens probably from this subzone occur up to 88.26 m, though no schlotheimiids were found other than in much younger beds. Echinoid and crinoid fragments occur in' all localities, and the small attached inarticulate brachiopod Discinisca has been recorded from St Fagans Borehole. A sparse palynomorph assemblage comparable to that from the Pre-planorbis Beds (Figure 26) was recovered from this sub-zone.

The Alsatites liasicus Zone, although very well displayed in the cliffs, is now largely inaccessable, but may well be about 30 m thick. Of this about the lower 4 m or so lies in the St Mary's Well Bay Formation, and the rest in the Lavernock Shales and the lower part of the Porthkerry Formation. The base of the zone can be proved by the presence of Waehneroceras, and in St Mary's Well Bay this genus continues up into the Lavernock Shales. Trueman (1920) collated sections from several localities, but his published sequence contains records which suggest that it contains errors either of identification or correlation. There have also been subsequent revisions of nomenclature (Dean and others, 1961). Trueman (1920) recorded Alsatites cf. laqueolus in his bed 70 which appears to relate to one of two specimens of A. liasicus in his collection (now in the National Museum of Wales), though the exact provenance of these is not clear. The Museum also has a specimen referred to Schlotheimia gallica from bed 65, which Professor Donovan has identified as Waehneroceras portlocki. In St Fagans Borehole few ammonites were found in the zone, though Caloceras occurs sporadically up to 73.32 m, where the late C. cf. bloomfieldense is found. Pyritic, sutured nucleii of Psilophyllites occur be tween 79.14 and 77.43 m, and Waehneroceras? occurs at 77.09 m with some fragments of indeterminate schlotheimiids. The earliest Schlotheimia sp.was at 58.17 m, which gives the latest position for the base of the overlying angulata Zone. There is inadequate evidence to determine the limits of the two subzones within the liasicus Zone. The remaining fauna of the liasicus Zone in the borehole includes both echinoid and crinoid debris, Discinisca, and a wide variety of bivalves including Cardinia oxalis (especially between 75–85 m), Chlamys, Gryphaea (above 60.5 m), and the two morphotypes of Liostrea, one with a large attachment area (L. irregularis) and one with a small area (L. hisingeri). Both surface-living and shallow burrowing forms are present. No liasicus Zone ammonites have yet been found near The Bulwarks, but it seems likely that part of the exposure below bed 7 is of this age although the fauna found (including a nautiloid) does not establish this zone. Palynomorph assemblages from the lower part of the zone (Figure 26) are comparable in composition with those from the upper part of the Pre-planorbis Beds.

Fossils of the Schlotheimia angulata Zone are found about 20 m up in the lower part of the Porthkerry Formation but the zone is poorly accessible along the coast; its thickness is estimated to be about 30 m. It is exposed in the higher parts of the cliffs at St Mary's Well Bay, in Bull Cliff, and west of The Bulwarks. Very few ammonites are known in situ, but Schlotheimia has been recovered from beds 7, 23 and 25 between The Bulwarks and Dams Bay and also between 58.17 m and 38.48 m in St Fagans Borehole, where the zone may extend up to the base of the marginal facies. The remaining fauna of the angulata Zone includes Calcirhynchia calcaria (in bed 7), Cardinia cf. hybrida and common Gryphaea arcuata. Large specimens of Plagiostoma giganteum on the seabed became attachment points for ostreids. A rather sparse fauna was recovered from St Fagans Borehole. Chondrites type burrows occur commonly, with scattered crinoid columnals, Camptonectes, Gryphaea, Liostrea, Mactromya, Modiolus, Pinna (many large and some vertical in the sediment), Plagiostoma, Pleuromya, Pseudolimea, Pseudopecten, Protocardia, Unicardium, sporadic gastropods, fish scales and wood. Above 38.48 m in the borehole the fauna is very sparse and commonly silicified; it contains crinoid columnals, Gryphaea, Lingula and crustacean fragments. The assemblage is a shallow water one, with both burrowing and surface-living taxa.

The base of the Arietites bucklandi Zone, which is the base of the Lower Sinemurian, is taken at the earliest arietitid ammonite found at 1.8 m above the base of bed 38 on the coast between Dams Bay and The Bulwarks. About 15 m of the zone occur east of Dams Bay. The presence of the conybeariSubzone is established by Vermiceras conybeari in bed 39 and Vermiceras sp.between beds 41 and 38. The top of the Sub-zone is taken at the base of bed 49 which has yielded Coroniceras sp.(of rotiforme group), probably indicating the younger rotiforme Subzone. The rest of the fauna at the locality is similar to that of the angulata Zone, though burrowing bivalves (Pholadomya and Pleuromya) are commoner and broken fragments of the large turbinate coral Stylophyllopsis occur in several beds. In St Fagans Borehole and the adjacent Slanney Quarry the marginal facies yields a sparse fauna with crinoid fragments, Camptonectes, Liostrea, Oxytoma and plant fragments, together with rare Lingula, Pinna and simple coral fragments. Low in the sequence some of the silicified limestones contain concentrations of spicules. In Slanney Quarry the Blue Lias overlying the marginal facies yielded Coroniceras cf. rotiforme in muddy limestone with rare brown ooids. This indicates the highest beds of the rotiforme Subzone as the exposure also yielded Arnioceras close to this horizon.

Following the redefinition of the bucklandi/semicostatum Zone boundary (Ivimey-Cook and Donovan in Whittaker and Green, 1983), the earliest occurrence of Arnioceras is now taken to mark the base of the Arnioceras semicostatum Zone. Arnioceras has been identified by Professor Donovan from the National Museum of Wales specimen 21–400–912, from Slanney Quarry; although the exact bed provenance is in doubt, it certainly came from the limestone–mudstone sequence in the upper part of the quarry which also yielded Coroniceras hyatti, crinoid fragments, Gryphaea arcuata, Liostrea, Meleagrinella?, Modiolus and Plagiostoma. Thus these highest beds of the Porthkerry Formation range into the semicostatum Zone.

Conditions of deposition

The late Rhaetian transgression continued into the early Jurassic, and led to the establishment of a shallow epeiric sea across much of Southern Britain.

Whether the limestone–shale rhythms of the Blue Lias are primary or secondary has been much debated (Hallam, 1964). Evidence for a primary origin rests on the widespread lateral extent of individual beds, the presence of bituminous shales, interpreted as anaerobic deposits punctuating a succession deposited largely under aerobic conditions, and the piping by burrowing of pale limestone down into dark mudstone. Individual cycles in Dorset and Somerset have been interpreted as comprising shallowing events following an initial deepening, responsible for the anaerobic shale (Hallam, 1960, 1964; Whittaker and Green, 1983). It has further been suggested that they are similar to less diagenetically altered Lower Lias cycles in the Belemnite Marls which show evidence of winnowing in the limestone at the top of shallowing-upwards sequences (Sellwood, 1970). An alternative mechanism to sea-level changes may simply be climatic oscillations controlling terrigenous input.

In Glamorgan, evidence for the primary origin of the cycles is commonly lacking, and the burrowing relationships are not developed. Bituminous shales are virtually absent and where present, contain beds and lenses of limestone. Evidence for diagenetic redistribution of carbonate in both a vertical and lateral sense is abundant. It includes irregular bedding planes bounding limestone beds, horizons of ellipsoidal nodules, and anastomosing mudstone beds and par tings which are especially well developed in the composite beds of Unit B in the Porthkerry Formation. The latter are probably caused by pressure solution. Hallam (1964) has commented that the greater abundance of limestones in the conybeari Subzone (upper Porthkerry Formation) in Glamorgan compared to North Somerset points to secondary segregation; if every limestone represented a shallowing event this should manifest itself in the marginal facies, but does not appear to do so. In conclusion, apart from those with laminated shales, primary cycles are difficult to recognise, if present at all in Glamorgan, because of the heavy diagenetic overprint. In Glamorgan, the Blue Lias was deposited within storm wave base with predominantly aerobic bottom conditions. Evidence of winnowing due to storm activity is present in the Bull Cliff Member and in Unit C of the Porthkerry Formation. Elsewhere, if present, it has been destroyed by burrowing. Any interpretation of Lavernock Shales rests on the nature of the mechanism controlling carbonate production and distribution in the Blue Lias and is therefore enigmatic.

By the beginning of St Mary's Well Bay Formation deposition all the Carboniferous massifs in the district appear to have been inundated, for the formation is everywhere conformable on the Penarth Group and no marginal facies are known at this level; for example the St Mary's Well Bay Formation north-west of Brooklands lies practically at the summit of the St Nicholas massif. This situation contrasts with that in the Bridgend district where conglomeratic marginal facies form fringes of shoreface deposits around the massifs that were sustained from late Triassic to late bucklandi Zone times (Trueman, 1922), possibly because of localised fault activity related to uplift along the Vale of Glamorgan axis, (Wilson and others, in press). It is also likely that the Coalfield to the north was submerged during the Lower Lias, for there is a lack of Millstone Grit and Coal Measures detritus in the marginal facies (Wobber, 1966; Owen, 1967).

The bucklandi Zone marginal facies in the district developed in the Porthkerry Formation exhibits a distinctive shoaling upwards motif. The upwards gradation from Blue Lias into the packstones of the lower part of the marginal facies marks the earliest introduction of allochems such as ooids onto the shelf, presumably transported by storms from active ooid shoal areas. Many of the marginal facies packstones beds that have scoured bases can be interpreted as storm-generated beds in which the bulk of the material has been introduced from adjacent shoreface areas, and the packstone sequence was probably dominantly deposited within storm wave base. However, the upward coarsening into ooid grainstones, locally with cross-bedding, and the virtual disappearance of mudstone partings points to a shift to an environment within normal wave-base associated with active ooid shoals. The bored iron-stained surface capping the marginal facies in the St Fagans area is a hardground, suggesting a significant pause in deposition though there is no evidence of subaerial exposure at this level. Ooid grainstones pass laterally into conglomeratic marginal facies, and commonly form large lenticular bodies which probably represent former ooid shoal areas, within the offshore facies of the Porthkerry Formation at comparable levels in the Bridgend district (Wilson and others, in press).

This major progradation of marginal facies during the bucklandi Zone may be associated with the upward increase in limestone content in the Porthkerry Formation on the coast, the increase in both limestone content and bed thickness probably reflecting a general shallowing. Unit C in the Bridgend district caps this shoaling-upwards sequence in the offshore facies and contains abundant coarse winnowed bioclastic lags (Wilson and others, in press). It is considered that this regional shoaling event was the result of uplift along the Vale of Glamorgan Axis.

This bucklandi Zone regression was ended by an abrupt return to off-shore facies everywhere in Glamorgan indicating a transgression initiated late in bucklandi Zone times and continuing into the semicostatum Zone.

Chapter 8 Quaternary

Two glaciations have been recognised in South Wales (Bowen, 1973, 1974). The first is undated, but preceded the Ipswichian interglacial. Although deposits of this glaciation are known from both sides of the Bristol Channel, including Gower and the Newport district in South Wales (Bowen, 1973, 1974; Squirrell and Downing, 1969) and Fremington in North Devon (N. Stephens, 1973), there are none in the district. However, the possibly pre-Ipswichian deposits in the Ewenny valley near Bridgend (Strahan and Cantrill, 1904; Bowen, 1970, 1974; Wilson and others, in press), together with more circumstantial evidence (Griffiths, 1939; Mitchell, 1960; Crampton, 1961, 1966) have been used by many authors, including Bowen (1970, 1973, 1974) to suggest that the Vale of Glamorgan was glaciated at this time. Rare, westerly-derived erratics, supposedly deposited by Irish Sea ice, occurring as far east as Cardiff (Griffiths, 1940) have not been confirmed during the resurvey. Indeed, in the district south of the Devensian limit there is a complete absence of erratics apart from those in the fluvioglacial deposits. Contamination by exotic heavy mineral suites in the Vale of Glamorgan soils south of the Devensian limit has been used by Crampton (1960, 1961) as evidence of pre-Ipswichian, Irish Sea ice crossing the Vale of Glamorgan. However, it is more likely that the exotic minerals are of aeolian origin. Crampton (1966) further suggested that Lower Lias cobbles within soil profiles overlying Lower Lias bedrock are of glacial origin. Such cobbles, however, are simply derived from the in-situ deep weathering of Lower Lias nodular limestones. In conclusion there appears to be no evidence that pre-Ipswichian ice entered the district south of the Devensian limit.

Similarly there is no evidence for Ipswichian deposits in the district. Crampton (1966) has pointed out that shallow, freely-drained soils developed on the Lower Lias in the Vale of Glamorgan are akin to terra fusca soils of more temperate regions, and therefore may be inherited from an interglacial period like the terra rossa soils he has described in the neighbouring Bridgend district. However, Wilson and others (in press) suggest that the latter are modern soil profiles, whilst the terra fusca soils appear to be related to limestone-rich parts of the Lower Lias succession. Trenhaile (1971) has described ledges along the Vale of Glamorgan coast, that range from low tide to 12 m above OD, and that he considered may represent the remains of interglacial shore platforms. The two he describes at Sully Hospital and Sully Island are, however, probably not marine benches, the Sully Island example being a result of differential subaerial weathering of soft Triassic mudstones and hard limestones and dolomites. Trenhaile (1972) further argued that the modern shore-platforms between Lavernock and St Mary's Well Bay owe their width to the destruction of higher planation levels, but no evidence has been found to support this view.

In the Cardiff district, the glacial deposits are all referable to the late-Devensian, largely because of the freshness of the glacial landforms. The deposits are limited to the north of a line that runs from St Nicholas eastwards to the Ely valley north of Leckwith, and thence south-eastwards to the coast, following the western margin of the combined floodplain of the Ely and Taff (Figure 36).

They form part of a linear belt of gravelly glacial drift, commonly with a distinctive topography, that skirts the southern rim of the coalfield from Swansea Bay to near Newport, and defines the southern limit of the Devensian glaciation (Charlesworth, 1929). The main deposit is gravelly till, but there are lenses of sand and gravel and laminated clay within it. Fluvioglacial outwash gravels occur on the fringes of the deposit both within and outside the Devensian limit, particularly in the river valleys, many of which, including those of the Cadoxton, Ely, Taff and Rhymney, contain buried valleys cut when sea-level was much lower than it is today.

Post-glacial (late-Devensian and Flandrian) deposits mainly comprise estuarine clays with subordinate sands, gravels and peats. These were deposited in the major river estuaries and along the coastal plain (the Wentlooge Level) when the sea-level rose as the ice-sheets melted (Figure 36). Outside the limit of Devensian glaciation, Head deposits (mainly hillwash) are probably predominantly Flandrian in age, though some may be earlier. Deposits of calcareous tufa of similar age commonly occur on outcrops of Lower Lias. Contemporary deposits include the present-day beaches, including storm beach gravels and coastal dunes, and the tracts of river alluvium.

Although the first systematic description of the Quaternary deposits of the district was in the first edition of this memoir (Strahan and Cantrill, 1902), earlier references include David (1883), Howard and Small (1901) and Dutton (1904). Strahan and Cantrill recognised that the southern edge of the glacial drift in the district corresponded with the southern limit of glaciation. They classified the glacial drift mainly as sand and gravel, whereas it has now been shown to be gravelly till ('morainic drift' of Squirrell and Downing, 1969). Later, Charlesworth (1929) suggested that these deposits were part of a South Wales end-moraine defining the limit of the 'Newer Drift' (late-Devensian), that was deposited by a piedmont glacier in the Vale of Glamorgan. The buried channels of the major rivers in the district and their sediment fills have been described by Anderson and Blundell (1965) and Anderson (1968, 1974). The Flandrian sequences along the coast, were well exposed during construction of the docks at Cardiff and Barry at the turn of the century; accounts are given in Strahan (1896) and Strahan and Cantrill (1902, 1912). Pertinent regional reviews of the drift deposits of South Wales and the Bristol Channel area are contained in Bowen (1970, 1973, 1974, 1977, 1981), Kidson (1977), Kidson and Heyworth (1973) and Peake and others (1973).

Devensian

Till and glacial sand and gravel

Although the predominant type of glacial drift is gravelly till, lenses of contemporaneous sand and gravel occur within it; apart from large masses, their differentiation on the map was not practicable.

These tills are very variable in composition, ranging from stiff, stony, silty clay to clayey gravel. Gravel-grade material is usually in the pebble–cobble range, although boulders are common. The matrix consists of varying mixtures of sand, silt and clay. The fresh till is either brown or reddish brown depending on the amount of Triassic or Old Red Sandstone material it contains. Clasts are mostly of Pennant sandstones, but there are also many local lithologies. The presence of 'flints' in the drift of the Ely Valley and around Wenvoe has involved much discussion (David, 1883; Howard and Small, 1901; Strahan and Cantrill, 1912; Bowen, 1973): Strahan and Cantrill (1912) suggested that they are more likely to have been derived from the local Carboniferous and Jurassic than from Irish Sea Chalk. The till commonly gives rise to sandy and gravelly soils, for in the acid soil conditions many of the Pennant sandstone and Old Red Sandstone pebbles tend to break down to a sand.

The lenses of glacial sand and gravel comprise grey to buff sands and sandy pebble-cobble gravels. The sands are commonly laminated or cross-laminated with scattered beds of laminated silt and clay; some are slightly clayey and contain scattered pebbles and cobbles. The gravels contain scattered boulders. Internally they vary from being well-bedded to chaotic with little evidence of bedding, and collapse structures are common. Some beds grade to sandy, gravelly till. Gravel–till contacts and bedding within the gravels commonly dip at around 45°.

Determining the relative amounts of sand and gravel and till is difficult without good exposure. However, the excavations for the M4 motorway in 1976 enabled this to be done across much of the northern part of the district; gravelly till proved to be the dominant deposit. Elsewhere, reliance has had to be placed on boreholes, but Fookes and others (1978), in a study of glacial drift in the Taff Valley north of the district, showed that samples from boreholes always had coarser average gradings than those from trial pits, because fines were lost during recovery of samples by drop tool methods. Borehole data, therefore, tend to exaggerate the amount of sand and gravel that is present.

The glacial deposits form three distinctive landforms that partly grade into one another. The first comprises high-relief mounds and irregular ridges with intervening kettles, other irregular closed depressions, and sinuous drainage systems; the second forms low-relief mounds and undulating terrain with low, open depressions; it is gradational to the third, which comprises a gently sloping to flattish plain. The distribution of these landforms in the district is shown in (Figure 36). The sediments and landforms are characteristic of those developed at ice-sheet margins of modern Arctic glaciers (Boulton, 1972), where supraglacial processes are dominant and lodgement till sparse.

It is probable that ice flowing southwards out of the Coalfield halted at the scarp between St Nicholas and Leckwith. The ice just topped the escarpment around St Nicholas whilst, to the east, small tongues penetrated down the Wenvoe and Cwrt-yr-ala valleys for a kilometre or so. It stopped at the north side of the Jurassic plateau around Leckwith, but to the east it may have flowed into the Bristol Channel, for north-east of Peterstone Wentlooge gravelly drift is present beneath the estuarine clays.

Glacial lake clay

Brown laminated clay in lenses up to 7 m thick was proved in excavations for the M4.

Fluvioglacial sand and gravel (including fluvioglacial terraces)

Though most of the Devensian outwash lies within the Devensian limit, and was deposited during the northward retreat of the ice, meltwater escaped southwards beyond the maximum limit of the ice. From the St Nicholas ridge it flowed southwards down the drainage system provided by the River Waycock, for fluvio-glacial sand and gravel floors the upper reaches of its tributaries. Farther east, some meltwater probably escaped down the Wenvoe valley to join the Cadoxton River system via the Wrinstone Brook, a tributary that occupies a deep gorge cut through the Dinantian inlier at St Andrews Major. No outwash gravels have been proved however, in the Cadoxton River system, even under the Flandrian alluvium. The St Andrews Major ridge is cut by another gorge, Cwm-slade [ST 135 736], now a dry valley floored by sand and gravel of uncertain origin, and sloping both north-east and south-west. The gravel is probably Devensian outwash, but an earlier age cannot be excluded since it is topographically higher than the Cwrt-yr-ala gorge. The latter has been interpreted by Crampton (see Bowen, 1967) as an overflow channel, whilst Bowen (1967) regards it as possibly due to river capture.

Meltwater also cut through the supraglacial till complexes rimming the Coalfield, utilising the valleys of the Ely, Taff and Rhymney. A major outwash fan appears to have been deposited at the confluence of the Ely, Taff and Rhymney rivers at Cardiff. Downcutting by rivers in the late-Devensian has left these outwash gravels as scattered terraces within the valleys, while the fan at the confluence is preserved only in dissected remnants. The terraces have their back-features cut in Devensian till and are overlain by Flandrian estuarine alluvium.

Peat

As the ice retreated and the climate ameliorated, peats accumulated in some of the kettle-holes on the supraglacial till complex. A kettle-hole, exposed in 1976 during excavations for the M4 Coryton interchange [ST 1397 8170], showed 0.6 m of pinkish brown to pale grey silty clay with some pebbles, on 0.15 to 1.0 m of dark brown partly clayey reedy peat, on 0.07 m of pale brown to grey organic-rich clay, on pale blue gravelly till. The peat has been dated radiometrically as 11 837 ± 50 BP.

Buried valleys

The modern deposits of the main rivers conceal deep buried valleys filled by sands and gravels and, in their estuarine reaches, also by Flandrian clays. Along the coastal plain eastwards of the Taff there is a broad, locally-dissected, rock platform similarly draped by gravels or Flandrian clays, and commonly by both. The buried valleys of the Ely, Taff and Rhymney cut across this platform as tributaries to the buried channel of the Severn (North, 1929; Anderson and Blundell, 1965; Williams, 1968; Anderson, 1968; Evans, 1982). The buried channels were cut during a period or periods of low sea-level, and it has even been suggested (Kidson, 1977) that at the Devensian glacial maximum the whole of the Bristol Channel area was dry land.

The date of the downcutting and subsequent infilling by pre-Flandrian sands and gravels is problematical (Anderson and Blundell, 1965; Williams, 1968). There is no evidence in the district for pre-Devensian downcutting, but several phases probably occurred during the Devensian. Firstly, all the major rivers suffered downcutting leading up to the Devensian glacial maximum when most of the Bristol Channel was land. They were then all glaciated and, therefore, further modified. An example of such modification can be demonstrated in the Taff at Morganstown where a narrow trench-like feature, whose base falls to at least 7.4 m below OD, forms a closed depression below the base of the main buried valley (see p.93). As the trench is much lower than the main buried valley farther downstream, it is interpreted as being due to subglacial erosion by meltwater under hydrostatic pressure. The trench and the remainder of the buried valley are filled by fluvioglacial sand and gravel.

Withdrawal of the Devensian ice into the Coalfield probably occurred fairly soon after the Devensian glacial maximum, allowing a long period for aggradation in the valleys as the sea-level gradually rose. Eventually the buried channels were filled, and gravel spread out as a fan over the interfluves cutting a seaward-sloping platform in the bedrock between the Taff and the Rhymney. Isostatic readjustment may have led to a relative fall in sea-level and to renewed downcutting, and may explain the fluvioglacial terraces of central Cardiff and in the lower reaches of the valleys. The buried valley of the Cadoxton River contains no outwash gravels, but it probably formed at the same time as that in the Taff, Ely and Rhymney.

Flandrian

The Flandrian sediments are mostly estuarine and marine, and formed during the post-Devensian rise in sea-level. The dominant sediment is clayey estuarine alluvium. Others include modern beach deposits, river alluvium, head, blown sand and calcareous tufa.

Estuarine alluvium

This is mostly clay with very subordinate silts, sands and gravels. A diachronous basal peat is commonly present as it is in the Flandrian of the Somerset Levels (Kidson and Heyworth, 1973; Kidson, 1977). The estuarine clays form the modern coastal plain between Cardiff and St Bride's Wentlooge, and also occur in the estuaries of all the major rivers. They are soft to very soft, blue-grey when fresh, and brown when weathered. They contain scattered diffuse silt laminae, and locally contain variable amounts of organic debris and scattered thin peats. A surface weathered zone, some 2 m thick, is stiffer than the clays at depth.

Gravels occur in the estuarine reaches of the major rivers (apart from the Cadoxton) at or near the base of the clays, and are impossible to distinguish in boreholes from the older fluvioglacial gravels unless separated from them by the basal peat.

Head

Head is generally regarded as a periglacial deposit, but most of the deposits so designated within the district appear to be due to hillwash; this is especially true of those with associated tufa, which are probably Flandrian.

Variably stony loams and clays occur in most of the dry valleys south of the Devensian ice-limit. The sides of the larger valleys such as the Waycock are, in contrast, remarkably free of head. On the Blue Lias (mainly the Porthkerry Formation) the upper reaches of many small valleys, some of which are dry, are mantled by light brown clay with tufa fragments, for example, north of Merthyr Dyfan [ST 124 701]. Flat-bottomed dry valleys on the Carboniferous Limestone in the St Lythans area and around Creigiau and Pentyrch are filled with loams containing scattered angular Carboniferous Limestone clasts.

Storm gravel beach deposits

Major storm beaches are present at Porthkerry and northwest of Cold Knap Point, Barry. They consist predominantly of Blue Lias limestone cobbles. At Porthkerry the storm beach has blocked off the valley, so that the river now drains to the sea by percolation through the gravel bar. There are minor narrow storm beaches between Barry Island and Lavernock Point.

Marine beach deposits

These dominantly comprise sands and gravels. Most of the beaches are gravelly, the deposits commonly occurring as a very thin skin on the wave-cut platform. The only major sandy beach is Whitmore Bay, Barry Island.

Calcareous tufa

Calcareous tufa is commonly deposited on the Lavernock Shales and the upper part of the St Mary's Well Bay Formation by streams originating from the Porthkerry Formation. The tufa has mainly been deposited around organic fragments such as twigs. The deposits contain variable admixtures of clay and other alluvial material. Mapped examples occur on the southern slope of the Waycock valley [ST 081 687], north of Porthkerry, and on the north-eastern side of the Leckwith plateau [ST 748 161], the latter being in the form of an alluvial cone.

In many of the small valleys draining the Porthkerry Formation, similar calcareous tufa is present beneath their narrow alluvial tracts on the Lavernock Shales and upper St Mary's Well Bay Formation, either as earlier tufaceous alluvium or slope wash deposits. Such deposits occur in most of the tributaries of the Porthkerry valley, one example being in Cwm Barry [ST 0988 6736] where 2.1 m of white tufa lies beneath 0.6 m of clay alluvium.

Alluvium and alluvial fans

Alluvium is associated with all the major streams and rivers; it consists variably of clay, sand and gravel. Alluvial fans are locally present on the Blue Lias, where streams confined in narrow valleys flowing off the escarpment of the Porthkerry Formation have spread veneers of silt, clay and calcareous tufa over the gentle slopes of its base.

Blown sand

The only deposit of blown sand is on Barry Island, shoreward of Whitmore Bay, where it overlies low ground occupied by the Mercia Mudstone Group.

Landslips

Two small landslips [ST 750 159] occur on the steep northeastern edge of the Leckwith plateau. They are probably rotational, and involve the St Mary's Well Bay Formation, the Penarth Group and the Blue Anchor Formation, the back-feature being in the former.

Made ground

Extensive made ground occurs in the urban areas of Cardiff and Barry. The most extensive tracts are due to dock construction. At Barry the docks were constructed by the partial filling of the estuary of the Cadoxton River. In Cardiff, dock construction involved the coastline being extended southwards for about a kilometre. Since the docks were built the tidal flats of the Ely, Taff and Rhymney have been tipped on so as to reclaim land for industrial use. Other areas of made ground in the urban areas include landfill projects in disused docks, clay pits, canals and quarries.

Details

Till

In the north-east, between Creigiau and the Taff valley, the high ground formed by the Old Red Sandstone and Carboniferous Limestone is relatively free from glacial drift. Immediately south is an east–west belt of extremely thick glacial drift, extending from Capel Llanilterne to Radyr and thence south-eastwards to Radyr Court. The drift is up to 24 m thick adjacent to the Taff valley. The topography consists of high-relief mounds with kettles and irregular enclosed depressions, but becomes more subdued towards Radyr Court. The M4 cuttings made in 1976 between Capel Llanilterne and Morganstown showed that up to 9 m of pinkish red gravelly till rest on laminated clays near Capel Llanilterne [ST 1015 7997]–[ST 1935 8015]. Sections east of Tyla-Morris exposed dark brown stiff silty and sandy gravelly till up to 8 m thick, weathering in the upper part to buff soft clayey silty pebbly sand; large lenses of clayey sandy gravel are present within the till in places e.g. [ST 1190 8100]. The sand and gravel mapped immediately to the east [ST 1225 8125] overlies pinkish till.

East of the Taff, the belt of moundy drift continues in the Coryton area. It ends abruptly just north of Whitchurch Hospital, and is replaced to the south by an extensive flat to gently sloping, drift-covered plain. The moundy drift was extensively exposed during the building of the M4 Coryton interchange; it consists of red-brown, variably gravelly till, with sand and gravel beds up to 8 m thick that locally dip at up to 45°. The glacial drift is up to 25 m thick adjacent to the Taff valley, but thins to the north and east.

South of the moundy belt and west of the Taff, through Pentrebane, Llandaff and Ely, the glacial drift is till with some sand and gravel; it is up to 10 m thick, and has a relatively subdued relief with gentle slopes and a few local mounds. Similarly, east of the Taff, through Whitchurch, Birchgrove, Llanishen and Heath, the surface of the glacial drift slopes gently southwards as a slightly dissected plain, until it is terminated by the fluvioglacial terraces. The area is mainly urban, and sections are few. Boreholes suggest that there is considerable sand and gravel within the till in this area.

Another east–west belt of particularly thick glacial drift occupies the slopes and escarpment south of Peterston-super-Ely and Caerau. In the central part of this belt, boreholes have proved up to 24 m of drift. It reaches the top of the escarpment around St Nicholas, and tongues southwards down the Wenvoe valley. Substantial mounds and irregular ridges occur; these enclose kettles and irregular-shaped depressions commonly filled with peaty alluvium. Along the top of the escarpment, around St Nicholas, the moundy topography is replaced southwards by gentle slopes. The deposit consists dominantly of gravelly till with lenses of sand and gravel.

The north–south Pen-y-Lan–Cyncoed ridge between the valleys of Roath Brook and the Rhymney is largely drift-free, except for its north-eastern part which is covered by till including some sand and gravel lenses. In this area, in the Nant Glandulais valley, the glacial drift is up to 8.5 m thick, and reddish in colour.

The high ground east of the Rhymney valley between Rumney and St Mellons forms a dissected ridge that has a flat top in the south capped by red-brown pebbly till. In the north the ridge is completely drift-covered, thicknesses of up to 12.5 m being proved in boreholes. Eastwards the ridge passes into undulating terrain, with low mounds between St Mellons and Trowbridge. Borehole data have enabled an elongate area of sand and gravel with till lenses to be delimited within the predominant till cover. The thickness of drift decreases from 14 m in the north-west to 3 m in the south-east towards Peterstone Wentlooge. Beneath the estuarine clays of the Wentlooge Level gravelly glacial drift (? till) is present east of Sluice House Farm, but absent to the west.

Glacial sand and gravel

The glacial sand and gravel mapped at Morganstown was exposed in the M4 cuttings [ST 1224 8126]–[ST 1290 8146]. The deposit is very variable, comprising up to 15 m of chaotic sandy pebble-cobble gravel with lenses up to several metres across, of locally pebbly sand, laminated clay and gravelly till. Steeply inclined bedding is common, and collapse structures were noted. Boreholes prove the sand and gravel to be at least 21 m thick.

At St George's, on the south bank of the Ely [ST 0940 7658]–[ST 0985 7680], up to 6 m of brownish orange, very sandy pebble-cobble gravel is exposed. Pebble imbrication is common, and there is a cement of either iron or calcium carbonate. One lens of till was noted within the gravel.

In the St Mellons area, the glacial sand and gravel has been delineated entirely from borehole data.

Glacial lake clay

A small lens of laminated clay has been mapped at Capel Llanilterne, where it was exposed during excavations for the M4. It floors the motorway cutting [ST 1015 7997]–[ST 1035 8015], and rests on rockhead. Boreholes proved that it is up to 7 m thick and overlain by till. In the south-eastern end of the cutting the clay dipped at 30° against the local bedrock, with signs of collapse.

Fluvioglacial sand and gravel and fluvioglacial terraces

Between St Nicholas and Vianshill the upper reaches of the tributaries of the River Waycock are flat-bottomed dry valleys containing gravel rich in Pennant sandstone. These valleys end abruptly upstream against the till that marks the Devensian ice-limit. Downstream, the outwash gravels are mostly hidden by the modern alluvium, but are exposed at various localities along the Waycock valley e.g. [ST 0946 7275]; [ST 0949 7025] where they comprise clayey sandy pebble-cobble gravels, commonly partly cemented.

South of the Taff gorge between Morganstown and Radyr, the River Taff cuts through the glacial drift complex in a valley whose slopes begin some 30 m above the modern floodplain. Up to two levels of fluvioglacial terrace are present at about 2 m and about 6 m above the modern flood-plain.

One fluvioglacial terrace at the confluence of the Ely and the Taff centred on Canton, and another at the confluence of the Taff and Rhymney centred on Cardiff city centre, make up the exposed part of an outwash fan. They form a very gently southward-sloping area, locally mildly dissected, Temporary sections in the city centre show that the terraces comprise pebble-cobble gravel with an orange-buff, slightly clayey sand matrix. The gravels are up to 7.6 m thick, but average 4 to 6 m. Some thin gravel beds are matrix-free, whilst thin beds of sand are locally present. Most sections and boreholes show that the terraces are locally overlain by up to 1 m of sparsely pebbly orange-brown clayey silt and very fine clayey sand. One section [ST 1854 7638] showed silt filling a channel in the upper surface of the gravel. Frost wedges, picked out by zones of vertical pebbles, are common in the upper part of the gravel as in the bank of the Rhymney [ST 2155 7554] at Pengam, where 5 m of gravel are exposed.

The gravels rest on a gently southward-sloping rock-head that is locally dissected. Farther SSE the fluvioglacial gravels on which the city centre stands, are overlain by the Flandrian estuarine clays of the docks and Pengam Moors. The Rhymney and the Taff have both cut through the terraces and removed the gravels, for rock-head under the modern flood plains is usually much deeper than beneath the adjacent terraces.

Buried valleys

Prior to the building of Barry Docks, the Cadoxton River entered the sea via two channels between Barry Island and the mainland, one on the east side of the island, and one between Cold Knap and Little Island (Strahan, 1896; Strahan and Cantrill, 1912). Borehole data show that the buried valley is situated in the channel to the north of the island. The deepest point proved is 13.8 m below OD [ST 1236 6732]; west of this there are no borehole data, but Evans (1982) has reported that the channel is as low as 50 m below OD offshore.

In the Ely valley, a kilometre east of St Fagans, rockhead is at 3 m above OD in the valley bottom. Locally at Ely, the buried valley lies south of the present alluvial belt, and is filled by till. South-east of Ely the buried valley occurs on the south-east side of the estuary, with the Leckwith escarpment marking its steep south-western limit below the floodplain. The greatest known depth of the buried channel is 12.5 m below OD, near the outlet of the modern river to the sea.

In the Taff valley there are few boreholes upstream from Llandaff North. However, a cross-section across the Taff at Morganstown has been provided by the site investigation boreholes for the M4 crossing. Here the buried valley is a narrow trench-like feature, up to 100 m wide on the western side of the valley, and reaching down to 7.4 m below OD. The remainder of the buried valley floor is above 9 m above OD. As the base of the channel does not fall below OD for a further 6 km downstream (at Cathays Park), the trench at Morganstown is a closed depression. It may be due to scour by a subglacial stream rather than to ice-scour. The trench is filled by 3 m of possible till, overlain by 31 m of fluvioglacial sand and gravel and modern alluvium. Downstream from Llandaff North the axis of the buried valley lies on the south side of the modern floodplain as far as Cathays Park, where it swings east. At Cardiff railway station it is immediately east of the present river, and from here it trends south-east to enter the sea on the south-east side of the Queen Alexandra Dock. The lowest point proved in the buried valley is 14.7 m below OD on the north-western edge of the dock.

The buried valley of the Rhymney appears to be a broad-bottomed feature, whose deepest point is only 8.6 m below OD at the coast, south-west of Maerdy Farm. The eastern edge of the gravel-filled buried valley is well defined where it cuts across the

Wentlooge Level, for Flandrian clays, rest directly on rockhead to the east. Where borehole information is available, in the southwestern part of the Wentlooge Level, a buried valley has been noted, north-east of Maerdy Farm, and falls to 9.6 m below OD at its deepest point. This valley cannot be followed beyond the limit of Flandrian clay. Borehole data show local patches of red silt and gravel along the bottom of this valley in an area which is otherwise free of till or outwash gravels beneath the Flandrian clays.

Estuarine alluvium

Flandrian clays fill the buried valley of the Cadoxton estuary. Borehole data prove the presence of a basal peat over much of the area. It generally overlies rockhead, but locally rests on a very thin gravel. During the construction of Barry docks, temporary sections to rockhead were exposed (Strahan, 1896). The sections showed the basal peat, with trunks and stumps of trees with roots, extending down into an underlying soil. Above were estuarine clays with two further peat beds. Strahan (1896) recorded a basal peat in two isolated sections; he regarded them as two separate beds because they occur at different depths below OD, but they are probably the same diachronous basal peat, as described by Kidson (1977) in the Somerset Levels.

Above the fluvioglacial gravels, which fill the buried valleys of the Taff and Rhymney, are estuarine clays with a little interbedded sand and gravel. Similarly, the coastal plain between the two estuaries is formed of clay with a basal peat overlying the fluvioglacial gravels. The basal peat is well developed under the old East Moors Steelworks, but is rarely present beneath the estuaries. The excavations for Cardiff Docks provided many temporary sections, including the new South Dock, where the basal peat contained vertical tree-trunks in the growth position (Strahan and Cantrill, 1912). Up to two higher peats have been recorded.

The estuarine clays of the Wentlooge Level rest on rockhead in the south-west, and on glacial deposits in the north-east. The basal peat is present locally and higher peats occur within the sequence in places.

Chapter 9 Structure

The earliest recognised structural events within the district are late Caledonian intra-Devonian regional uplift accompanied by localised faulting, tilting and erosion. Intra-Carboniferous movements, exerting a tectonic control on early Dinantian sedimentation, and late Dinantian to Namurian uplift, preceded the main Variscan deformation. The latter gave rise to the west–east-trending fold axes and the predominant NW–SE and WNW–ESE faults that affect the Palaeozoic rocks. Mesozoic movements seem to have exerted structural controls on sedimentation during the Triassic and early Jurassic. Later movements were confined to faults; these are difficult to date precisely but are certainly post-Jurassic. The main structural elements of the district are shown in (Figure 37).

Within the district the structure is dominated by a major Variscan fold, the Cardiff–Cowbridge Anticline (Strahan and Cantrill, 1912), with an east–west axial trace. Some authors (George, 1956; Owen, 1974) have suggested that the anticline continues north-eastwards to join the Usk Anticline on the eastern margin of the Coalfield to form a major curvilinear fold, but Squirrel] and Downing (1969) claimed that the Usk Anticline was quite unrelated to the Cardiff–Cowbridge Anticline, a view that is supported by the present studies. These latter authors (Squirrell and Downing, 1969, plate II) suggested that the Usk Anticline (Walmsley, 1959) swings south-westwards for some distance down the south-eastern edge of the Coalfield as a south-west orientated anticline, but this Anticline—here termed the Rogerstone Anticline—does not appear to be connected to the Usk Anticline to the north or the Cardiff-Cowbridge Anticline to the south; it seems to be a discrete structure trending at an angle to the two major folds.

The latter anticlines seem to mark two major axes of repeated movement which are probably related to early basement fractures. Both have affected sedimentation, from time to time, up to as late as the Lower Jurassic. The axis of movement that is broadly coincident with the axial trace of the Cardiff- Cowbridge Anticline is here called the Vale of Glamorgan Axis.

Late Caledonian movements

The district forms part of the virtually undeformed platform of South Wales and the Borders, south-east of the paratectonic Caledonian fold-belt to the north. Late Caledonian orogenic uplift is manifested by the absence of Middle Devonian sediments from the district. Deformation, recognised by localised Upper Devonian overstep, was confined to arching along roughly north- south aligned axes in south-east Wales, such as the Usk Axis, and to movement on some east-west aligned faults in Pembrokshire (Allen, 1965, 1974a), while the progressive south-westward overstep of the Upper Devonian onto successively lower levels of the Brownstones in the north of the district (see p.20) is due to similar movements parallel to the Vale of Glamorgan Axis, probably reflecting vertical movements along a reactivated basement fracture. It is possible that movement on a similar fracture to the south in the Bristol Channel accounts for the enigmatic 'Bristol Channel Landmass' that was the provenance of the southerly-derived Llanishen Conglomerate. No late Caledonian minor structures have been identified in the district.

Intra-Carboniferous movements

The major southward thickening of the Dinantian, especially the Black Rock Limestone and Gully Oolite, across the Vale of Glamorgan Axis (p.49) was probably due to movement on a major synsedimentary fault (trending parallel to or coincident with the later Variscan Cardiff-Cowbridge Anticline) which was probably a reactivated basement structure.

The east-west facies changes in the Holkerian, and the progressive eastward overstep of the Namurian in south-east Wales, have been interpreted as due to uplift along the Usk Axis (George, 1956). This movement has generally been regarded as due to Variscan compression resolved in an east-west direction (Walmsley, 1959; Owen and Weaver, 1983).

Variscan structures

The main compressional Variscan deformation was late Silesian in age. It generated a major east-west fols, the Cardiff-Cowbridge Anticline, and several other meso-folds of similar orientation. Other associated structures include 'cross faults', orientated NW-SE to WNW-ESE, and local thrusts.

Folds

Most of the major and meso-scale folds trend approximately east-west to ESE-WNW, though the Rogerstone Anticline trends NE- SW. Minor folds, apart from those associated with faults, are virtually unknown, and there is no cleavage.

The Cardiff-Cowbridge Anticline is the largest fold in the district. It plunges gently westwards, is slightly asymmetric and open, and verges northwards. Average dips on the northern limb are 30° and on the southern limb 20°. Its core is largely obscured by Triassic and Jurassic rocks, and it is difficult to locate its axial trace exactly, especially beneath Cardiff.

South of the Cardiff- Cowbridge Anticline only mesoscale west-trending open folds occur; they plunge gently westwards perhaps as part of a major synclinal structure complementing the Cardiff-Cowbridge Anticline. One of these folds, the Barry Anticline, brings up Old Red Sandstone in its core.

North of the Cardiff-Cowbridge Anticline there are several meso-scale folds; a west-trending pericline at Pentyrch and three open folds with an ESE-NNW trend in the Ludlow and Wenlock outcrops at Rumney. A geniculation in the axial trace of the two most northern folds is interpreted as due to interference with the north-east-plunging Rogerstone Anticline.

The Rogerstone Anticline trends NE-SW and is almost symmetrical. It is well defined both in the Old Red Sandstone and the northern part of Ludlow and Wenlock Series outcrop. It seems likely that it intersects the Cardiff-Cowbridge Anticline at an angle of about 70°.

The fact that the Vale of Glamorgan Axis and the axial trace of the Cardiff- Cowbridge Anticline are largely coincident suggests that they are related. Wilson and others (in press) have suggested that the Axis was reactivated as a major thrust within the basement rocks and controlled the development of the Cardiff-Cowbridge Anticline during the Variscan orogeny. Further thrust displacements may have occurred in the cover along major west-east faults, particularly in the hinge zone of the Anticline in the Bridgend district. In the Cardiff district, these faults change orientation to WNW-ESE where they impinge on the Creigiau cross fault. They are largely concealed by later strata and have suffered post Jurassic reactivation; there is therefore little evidence of northward-directed thrusting.

Thrust-faults

Northward-dipping thrusts with associated folds occur along the southern margin of the Coalfield (Woodland and Evans, 1964; Squirrell and Downing, 1969). Two such thrusts are present in the Cardiff district within the Dinantian sequence. At Creigiau the Pen-y-garn Thrust (Squirrell and Downing, 1969) overthrusts Hunts Bay Oolite over Namurian shales. It has a southward transport direction and, although the thrust-plane is not exposed, it is thought to be subparallel to the bedding. The relationships are well displayed in the Creigiau quarries [ST 088 821]. In Heol Goch Quarry [ST 1245 8222], within dolomitised Friars Point Limestone, a thrust-plane dips 15°–45° NW/064°, becoming steeper upwards, The southward direction of overthrusting is demonstrated by a southward-verging, strongly asymmetric pair of folds, beneath the sole of the thrust. A thrust breccia, up to 0.5 m thick, is developed along the thrust plane. The folds associated with such thrusts were regarded by Woodland and Evans (1964) as being formed by the drag due to the northward movement of the footwalls of the thrusts, the latter being therefore termed underthrusts.

A northward-directed thrust occurs on Friars Point [ST 1095 6613], Barry Island, in the Friars Point Limestone dipping 45°S, just steeper than the southward-dipping bedding.

Other faults

The only faults that can be proved to be of Variscan age are those that do not cut Triassic and later strata, but it is very likely that some faults with post-Triassic movement are due to renewed movement along Variscan faults. This can be demonstrated by the amount of throw changing, or even the direction of throw changing, as a fault is traced from Palaeozoic to Triassic and younger strata. Some faults that cut only Palaeozoic rocks vary in trend from NNW–SSE to NW–SE; others vary from WNW–ESE to E–W. The former trend is similar to that of the 'cross-faults' of the Coalfield, regarded by Woodland and Evans (1964) to largely post-date the thrusting, and to be broadly synchronous with the development of folds within the Coalfield. The main cross-faults in the Cardiff district are the Tongwynlais Fault (Squirrell and Downing, 1969), the Wenvoe Fault and the Creigiau Fault, though the latter has suffered later reactivation. Both the Tongwynlais and the Creigiau Faults originate in the Coalfield.

The north-west-trending Tongwynlais Fault occupies the Taff valley and throws down to the west. It has been suggested (Gayer and others, 1973) that it suffered dextral strike slip movement during the folding and thrusting, followed by later slight sinistral strike-slip movement and that finally, during the later Hercynian tensional phase, the major western downthrow occurred. Their evidence for dextral movement during the compressional phase, comes mainly from the fact that in the Carboniferous Limestone on the east of the valley, the dominant thrust direction and the asymmetry of the folds is to the south, whilst to the west of the valley the asymmetry of the folds is markedly northwards. However, new sections in Heol Goch quarry show southward-verging folds and a southward-directed thrust (see above), which casts doubt on any suggestion of an early dextral movement on the fault.

The Wenvoe Fault is the largest of a number of NNW–SSE faults in the Wenvoe valley, most of which are concealed by Triassic strata. The Wenvoe Fault throws down to the east, repeating the lower half of the Dinantian succession to the east of the valley in the St Andrews Major Inlier. This fault-zone cannot be traced north of Wenvoe, but may be the southern extension of the Creigiau Fault; the intervening ground forms the core of the Cardiff–Cowbridge Anticline, now cut by numerous post-Lower Jurassic faults, and largely covered by Mesozoic sediments.

The Variscan Front

The evidence used by various authors to define the position of a front to the Variscan orogeny has been reviewed by

Matthews (1974) and Dunning (1977). Some authors (e.g. Freshney and Taylor, 1980, fig. 9:3) have placed the northern boundary of the Variscan front through the Vale of Glamorgan. However, overturned or generally tight folds and axial plane cleavage, used by various authors (Matthews, 1974; Hancock and others, 1983) to define the orogenic belt, are absent from the district, as is a well defined zone of northward-directed thrusting. There are two possibilities. Possibly the Front is represented by a major undetected decollement surface beneath the South Wales Coalfield (Shackleton and others, 1982, fig. 6), or it lies to the south beneath the Bristol Channel, where geophysical evidence points to a major structural lineament (Brooks and others, 1983).

Post-Variscan structures

Following the Variscan deformation, a major tensional regime was initiated related to the opening of the Atlantic (Hallam, 1971). In the Bristol Channel–Somerset area a major west-orientated fault-bounded Permo/Triassic basin was initiated (Whittaker, 1975), the northern margin of which was the Vale of Glamorgan. During the period of late Triassic continental and lacustrine sedimentation some of the Palaeozoic highs may have been sustained by synsedimentary faulting, for it seems more than fortuitious that parts of the margins of the highs are visibly faulted. For example, the eastern edge of the Carboniferous Limestone Inlier at St Andrews Major is fault-bounded to the north by the Penarth Fault, though it is apparently unfaulted to the south. However, to the south of the inlier, but on the line of a southward projection of its eastern margin, the Lavernock Fault has the same sense of throw to the east. The Penarth Fault (south of Cyntwell), and the Lavernock Fault, together with the apparently unfaulted margin of the inlier, around Michaelston-le-Pit, lie on a NNW–SSE line east of which there are no Palaeozoic inliers and the Triassic succession thickens rapidly.

Wilson and others (in press) have described similar active faults sustaining Palaeozoic massifs in the Lower Jurassic of the Bridgend district. The major regression towards the end of the bucklandi Zone in the district was probably due to a regeneration of the Vale of Glamorgan Axis, resulting in uplift along the latter and increased subsidence to the south in the Bristol Channel Basin.

Post-Lower Jurassic faulting is most intense in the western part of the core of the Cardiff–Cowbridge Anticline. Eastwards the faults tend to die out or curve southeastwards. The major trend in the core is WNW–ESE to NW–SE, with a less important complementary set at WSW–ENE to NE–SW. Two of the most important ones, the Stockland and St Fagans Faults, have downthrows of as much as 200 m.

The Creigiau Fault in the Coalfield to the north is a NNE–SSW cross-fault throwing down to the west about 190 m (Squirrell and Downing, 1969). In the Creigiau area the Carboniferous Limestone appears to be thrown down to the east, and the Mesozoic strata to the west. Unless the fault has some strike-slip component, the eastward throw is probably Variscan, and the westward throw due to post Jurassic reactivation. Traced southwards it bends into a WNW trend, continuing to throw down to the south-west. Possibly the rejuvenation was inoperative across the strong zone of WNW–ESE faults in the core of the Cardiff–Cowbridge Anticline, but ended at the northern most major WNW–ESE fault. However, there may have been local reactivation south of the core of the anticline along the Penarth Fault north of Michaelston-le-Pit and the Lavernock Fault to the south.

South of the axis of the Cardiff–Cowbridge Anticline, post-Jurassic faults are much less numerous but of similar trend. The largest is the WSW–ENE Cold Knap Fault that throws down to the north; the fault zone is well displayed in Barry Harbour [ST 104 665], whilst there are many complementary north-west-trending minor faults in the Blue Lias cliffs immediately to the north.

There is no clear evidence of the age of the post-Jurassic faulting. A Miocene age has been suggested for some of them by Anderson and Owen (1968) and Gayer and Criddle (1970).

Chapter 10 Economic geology

Aggregate

The district generates a considerable demand for aggregate. Most comes from crushed hard rock derived exclusively from the Carboniferous Limestone; the remainder comes from marine-dredged material.

Various hard rock aggregate surveys (Harrison and others, 1983; Adlam and others, 1984; Harrison, 1984) have described the physical and mechanical properties of the Carboniferous Limestone in the district and its environs. All the predominantly limestone formations provide aggregate strong enough for most construction purposes, such as base and sub-base roadstone and concrete aggregate, and the Black Rock Limestone is particularly strong. The fine- and medium-grained dolomites in the north of the district (Barry Harbour Limestone, Friars Point Limestone, and parts of the High Tor Limestone) are of uniformally good strength and also highly resistant to abrasion. Their low wet attrition values make them highly suitable for railway ballast. However, the coarser dolomites, mainly occurring in veins cutting the Gully Oolite and Hunts Bay Oolite, are weaker and much less durable.

Aggregate production is now concentrated in two large quarries, Wenvoe Quarry [ST 133 742] and Heol Goch Quarry [ST 122 822] (also known as Taffs Well or Walnut Tree Quarry). The former works the Black Rock Limestone and the latter dolomitised strata from the Friars Point Limestone to the High Tor Limestone. Greenwood [ST 114 742], Twynyr-odyn [ST 1160 7375], Whitehall [ST 177 734] and St Andrews quarries [ST 143 713], which all worked aggregate in the past, have closed in the last twenty years. Crushed rock sales in South Glamorgan in 1981 totalled 1.484 million tonnes and in West and Mid-Glamorgan 3.338 million tonnes (Anon, 1983).

Sandstones in the Old Red Sandstone have not been worked for aggregate but have been examined as possible sources (Adlam and others, 1984). The limited test results available suggest that most would produce relatively weak aggregate.

Sand and gravel has not been worked onshore in any quantity within the district, and an assessment of its northern part (James and Thornton, 1983) concluded that the potential deposits are extremely limited, mainly due to the variable nature of the glacial drift. The main source of sand is from material dredged from the Bristol Channel. In 1981, 0.546 million tonnes of dredged aggregate were landed at Barry and Cardiff, most of which was sand (Anon, 1983).

Building stone

In the past, stone was quarried extensively from the Carboniferous Limestone, the Triassic conglomerates (Radyr Stone) and the Blue Lias (mainly from the Bull Cliff Member); minor amounts also came from the Quartz Conglomerate Group. Both the local and imported building stones have been described in detail by Perkins (1984).

Brick and pottery clays

Marls from the Raglan Mudstone Formation and the Mercia Mudstone Group were formerly used for brickmaking. The former were worked at Maendy [ST 1745 7825], and the latter at several localities including Llandough, Cogan, Cadoxton and Ely.

Estuarine clays have been extensively worked for brickmaking and pottery on the coastal plain south-east of Cardiff city-centre and at Cadoxton.

Chemical-grade limestone and dolomite

Although very subordinate to its use as an aggregate, the Carboniferous Limestone is also worked for agricultural lime, ferrous metallurgical flux and dolomite refractories, whilst in the past it has been worked for cement. In the southern half of the district much of the sequence is composed of high-grade chemical limestone, the Gully Oolite and Cefnyrhendy Oolite being classified as very high purity limestones (Harrison and others, 1983). In the north the dolomites, although variable, are predominantly of high chemical purity, containing 17.5–22 per cent MgO, and comprising a major resource of material suitable for metallurgical and refractory use (Adlam and others, 1984). Heol Goch Quarry [ST 122 822] works the dolomitised Friars Point Limestone for flux used in sintering and for dolomite refractories. Creigiau Quarry [ST 086 808] selectively worked the Hunts Bay Oolite for both limestone and dolomite which was formerly used for flux in the now closed East Moors Steelworks.

The lower part of the St Mary's Well Bay Formation in the Blue Lias has been worked in the past for cement at Lower Penarth [ST 174 695]. Carboniferous Limestone was added to the quarry product derived from the more shaly parts of the sequence in order to increase the limestone content in the feed-stock. Production is now concentrated in the Bridgend district where the Porthkerry Formation is used, and most of this formation is a cement resource in the Vale of Glamorgan (Harrison, 1984).

Gypsum

The gypsum nodules in the Mercia Mudstone Group mudstones do not occur in commercial quantities. They were, however, formerly worked for ornamental purposes at Penarth as Penarth Alabaster' (Cox and Trueman, 1936; Sherlock and others, 1938; Notholt and Highley, 1975).

Metalliferous minerals

Lead

Galena is present in the Carboniferous Limestone along the axis of the St Lythans Anticline. A small open-cast working [ST 1056 7190] east of Dyffryn was sited in a west-trending vein where the Gully Oolite is faulted against Triassic pebbly calcarenite. Farther east there are old trials [ST 1101 7192] in Triassic breccias and old excavations [ST 1712 6797], said to have been for lead, occur in the Gully Oolite near Lavernock (Strahan and Cantrill, 1912). Small quantities of lead ore were raised at Barry in the eighteenth century (Lewis, 1967). There is no record of the exact location of the working, but galena has been noted in the Barry Harbour Limestone at Mark Rock [ST 1381 6680] and it is associated with hematite in small veins, cutting dolomitised Friars Point Limestone on Sully Island [ST 1697 6692]. The age of the mineralisation is unknown, but it is apparently post-Triassic.

Hematite

A small amount of hematite was worked north-east of Wenvoe [ST 1343 7423] between 1859 and 1864 (Sibly, 1927). The deposit lies along a WNW-trending fault in the Barry Harbour Limestone that branches off a larger fault that throws down Triassic marginal facies immediately to the east. Hematite has also been worked from a small deposit at Peny-garn [ST 0980 8188] that lies parallel to the bedding in the dolomitised Hunts Bay Oolite (Strahan and Cantrill, 1912). Both these occurrences appear to be similar to the larger deposits present farther west in the Bridgend district (Sibly, 1927), where the hematite occurs both as a replacement and a cavity-fill and is regarded as either hydrothermal or supergene in origin (Gayer and Griddle, 1970).

A trial level at Pant-y-caeau [ST 1057 8101], south of Pentyrch, is reported to have been in the bedded Rhiwbina Ironstone that occurs in the lower part of the Tongwynlais Formation (Strahan and Cantrill, 1912).

Limonite

Vertical ochreous limonite veins, up to 0.3 m thick, trending 020°, 040°, 090° and 120°, occur in Heol Goch Quarry [ST 1245 8222]. They lie about 500 m south-east of the workings of Garth hematite mine (Squirrell and Downing, 1969) and are associated with calcite and barytes veins of similar trends.

Water supply

Although the majority of the district's supply comes from surface water sources outside the area, there is one groundwater abstraction for public supply and several for industrial purposes. The average annual rainfall of the district varies from 950 mm along the Severn estuary to 1300 mm inland; annual potential evapotranspiration is around 530 to 535 mm.

Wenlock and Ludlow

The predominantly argillaceous nature of these strata, together with the fact that the interbedded sandstones and siltstones are well cemented, means they have low porosities and permeabilities. Only one borehole is known to have been drilled entirely in Silurian strata. Drilled to a depth of 46 m in the Pen-y-Lan Mudstone at Pen-y-Lan [ST 195 789], this 100 mm-diameter borehole only yielded 0.7l/s. A borehole at Crwys Road sidings, Cardiff [ST 1845 7850], originally tapped the 'red mudstones' of the Mercia Mudstone Group and was later deepened by 83 m to penetrate 13 m of Ludlow/Wenlock, but the supply did not increase and the borehole is now disused.

Old Red Sandstone

The mudstone-dominant Lower Old Red Sandstone formations are largely impermeable. However, several deep boreholes in the Cardiff area have entered them after passing through the basal Triassic aquifer. A 32 m-deep well into Raglan Mudstone Formation beneath till at Whitchurch [ST 1685 8086] supplied a continuous yield of 0.15 l/s and was tested at 0.3 l/s. Three 200 to 250 mm-diameter boreholes into the St Maughans Formation beneath till at the Unigate Dairies at Marshfield [ST 263 820], varying in depth from 36 m to 91 m, all initially yielded 6.3 l/s (in the case of the deepest for a negligible drawdown) and are currently licensed together for a total of 250 Ml/a, although only two are used. A chemical analysis of the water is shown in (Figure 38).

Little is known of the hydrogeological properties of the Llanishen Conglomerate. However, a 100 mm-diameter borehole drilled at Llanishen to a depth of 20 m into bands of conglomerate, rock and marl beneath 10 m of lined out gravelly glacial drift was pumped dry in five minutes, but recovered nearly immediately (Howard, 1894; Strahan and Cantrill, 1912).

The sandstone dominant Brownstones and Upper Old Red Sandstone are variably cemented, and groundwater flow occurs mainly through fissures generally developed within 50 m of the ground surface. There are no records of wells into these formations, but their narrow and mainly drift-covered outcrops receive little recharge and hence yields are likely to be low. Springs occur within the Quartz Conglomerate Group south of Pentyrch. The quality of the water is good but hard.

Carboniferous

Although the 'Main Limestone' of the Carboniferous Limestone is the most important aquifer in South Wales, it is little used in the district. Drilling for water is highly speculative as the limestones have minimal primary porosities and permeabilities, and groundwater movement is restricted to fissures enlarged by solution. The latter are neither regularly spaced nor extensively interconnected and boreholes failing to intersect them are commonly dry. The outcrop is largely drift-free and recharge equals effective precipitation. Beneath Triassic cover, the water is confined by mudstones of the Mercia Mudstone Group, and Triassic deposits frequently infill the fissures.

Springs issue from around the junction of the 'Main Limestone' and the Lower Limestone Shale Group [e.g. [ST 1034 8106]; [ST 081 752]. Other springs occur on the 'Main Limestone' outcrop at the lower ends of dry valleys. There is a record of a 15 m-deep well at Wenvoe [ST 1242 7350], and at St Andrews [ST 1355 7185] a 12 m-deep well, with a 2 m borehole in the base, yielded 0.28 l/s. Present day abstraction is limited to a 115 mm-diameter borehole, 46 m deep, at St Lythans [ST 1141 7314] licensed for 4.5 Ml/a.

The quality of water from the Carboniferous Limestone is hard but good under low flow conditions. However after heavy rain, supplies from both wells and springs may become turbid and polluted. Beneath Triassic cover, concentrations of iron and manganese may be significant. Waters are not highly mineralised at depth. A thermal spring at Taff's Well [ST 119 838] just north of the district, which probably originates from the limestone at a depth of 700 m, only contains 450 mg/I total dissolved solids.

Triassic

The marginal facies at the base of the Mercia Mudstone Group is relatively permeable due to the presence of solution fissures, mainly joint-controlled, in the limestone conglomerates, breccias and calcarenites, and to localised dolomitisation. Where the marginal facies overlies Carboniferous Limestone, they are expected to be in hydraulic continuity.

Both in the past and today the marginal facies is the main aquifer used in the district. The largest abstraction is from two large diameter wells at Biglis [ST 147 698], used by the Welsh Water Authority for public supply, licensed for a total of 1818 Ml/a. Both penetrated 2 to 3 m of alluvium above limestone conglomerate. One is 9 m deep and encountered a 'strong spring' at the base of the alluvium which used to overflow at the surface in winter. Its yield decreased to 15.8 l/s after the excavation of Barry Docks. The other is a 3.4 m-diameter shaft, 12 m deep which yielded 53 l/s when constructed.

At Cardiff there are records of at least twenty five boreholes into the marginal facies beneath the thick sequence of 'red mudstones'. Here the water is often reported to have come from the 'Lower Water Bed', a sandstone in the upper part of the marginal facies (Howard, 1894; Boulton, 1910; Strahan and Cantrill, 1912). Yields from boreholes into the marginal facies in the district varied from dry holes at Helen Street, Cardiff [ST 1992 7704] and Cog Moors [ST 1548 6975] to in excess of 10 l/s from a 200 mm-diameter, 30 m-deep borehole at Biglis [ST 1458 6995], a 200 mm-diameter, 131 m-deep borehole in Pellett Street, Cardiff [ST 1880 7610] and a 150 mm-diameter, 130 m-deep borehole at Grangetown Gasworks [ST 1759 7445]. Today only a few of the boreholes that exploited the marginal facies in the past, carry a current abstraction licence. These include a 125 mm-diameter, 68 m-deep borehole at the Ely Paper Mill [ST 1496 7677] licensed for 682 Ml/a, a 250 mm-diameter, 20 m-deep borehole in Andrews Road, Llandaff North [ST 1482 7897] licensed for 68 Ml/a, a borehole in Crawshay Street, Cardiff [ST 1811 7573] licensed with a shallow well into the drift for a total of 436 Ml/a and three boreholes in St Mary Street, Cardiff [ST 183 761] all about 100 m deep, licensed respectively for 11 Ml/a, 82 Ml/a and 114 Ml/a. One of these last three boreholes encountered water in the 'Lower Water Bed' at the base of the hole which then rose to within 10 m of the ground surface.

The quality of the water is good but very hard with a total hardness as high as 500 mg/1 recorded. Sources near the coast may have high chloride concentrations and the borehole at Grangetown [ST 1759 7445] yielded moderately saline water with total dissolved solids of 4000 mg/l. Analyses from four boreholes into the marginal facies are shown in (Figure 38).

The mudstones of the Mercia Mudstone Group are not very permeable and groundwater movement is restricted to thin, poorly cemented sandstones and beds of gypsum that have undergone variable dissolution. Around Cardiff, water has been obtained from a thin bed of 'loose sand', about 1 m thick, known as the 'Upper Water Bed', that occurs in the upper part of the 'red mudstones' (Howard, 1894; Strahan and Cantrill, 1912). It is, however, reported to have yielded only small quantities of water (Boulton, 1910). A 185 mm-diameter borehole in Frederick Street [ST 1846 7636] yielded 2.5 l/s from this horizon at a depth of 36 m. Although the latter was also encountered in other borings in the Cardiff area, in at least two cases it yielded insufficient water and the boreholes were continued into the 'Lower Water Bed' of the marginal facies. A 2.4 m-diameter shaft, 19 m-deep into the 'red mudstones' at Crwys Road Sidings, Cardiff [ST 1845 7850], reached water at 8 m and overflowed at 0.5 l/s before being pumped at 2.6 l/s. There are no records of any wells abstracting water from the largely argillaceous Penarth Group. However, a borehole at the South Wales Portland Cement and Lime Works at Lower Penarth [ST 182 695] encountered a 'strong sulphurous spring' at a depth of 10 m in a pyritic bed of stone, possibly part of the Westbury Formation. This borehole was continued to a depth of about 70 m through the mudstones of the Mercia Mudstone Group which yielded 2.5 l/s from a 3 m-thick band of hard red marl at a depth of 24 m and 'very much water from gypsum bands' near the base.

Scattered springs occur in the 'red mudstones', as at Goldsland, Wenvoe [ST 1119 7129], and in the Blue Anchor Formation south of Dyffryn [ST 0924 7178].

Water from the mudstones of the Mercia Mudstone Group is of poor quality with high sulphates. The water from the well at Crwys Road Sidings [ST 1845 7850] had a total hardness of about 26 000 mg/1 of which 14 000 mg/l were temporary. Near the coast the water becomes brackish and tidal effects are seen in many wells in the area.

Lower Jurassic

The interbedded limestones and mudstones of the Lower Lias are relatively impermeable, but could yield small quantities of water if fissures were intercepted. No boreholes abstracting from the Blue Lias are known in this district but several successful ones have been drilled in the Bridgend district, to the west. The marginal facies in the north of the district is unlikely to be of importance because of its limited outcrop.

Numerous small springs occur along the thin outcrop of the Lavernock Shales where they crop out on steep slopes beneath the Porthkerry Formation. The water tends to be calcareous and ferruginous.

Superficial deposits

In the past, superficial deposits were extensively utilised for water supplies, particularly the fluvioglacial terrace gravels underlying central Cardiff. However, because of the risk of pollution from surface sources, rivers or the sea, they are no longer used for public supply. Private abstractions have also declined and currently there is only one licensed abstraction which is definitely wholly from the superficial deposits. This is a 3 m-diameter and 9 m-deep well into estuarine alluvium underlain by gravels which yielded 12.5 l/s at Crawshay Street, Cardiff [ST 1814 7562]; the water is used for cooling. Two other sources are licensed to abstract from holes penetrating a few metres of Mercia Mudstone Group mudstones beneath the gravels. One is a 300 mm-diameter, 11 m-deep borehole at the Initial Laundry in Cardiff [ST 1673 7717] which is licensed for 78 Ml/a. The quality of the water is moderately hard (see (Figure 38)). The other is an 11 m-deep well at the Ely Paper Mill [ST 1504 7678] which is licensed for 6.8 Ml/a. A further licence granted to the Central Electricity Generating Board is for 1364 Ml/a from a total of eleven (although only seven are used) 8 m-deep wells interconnected by pipes and adits at Ely [ST 140 770]. No geological logs of the wells are available but it is thought that most of the water comes from the gravels although the most westerly well is sited on the predicted boundary between the 'red mudstones' of the Mercia Mudstone Group and the marginal facies, beneath superficial deposits; it could have a significant contribution from the Trias. The others would be into mudstones if they extend below the drift. The wells are pumped into one sump, so there are no records of individual yields or the quality of the water. A while ago the water became grossly polluted, but when two wells were bricked up the quality improved, thus implying that the remaining sources are not in direct hydraulic continuity with the Ely River. An analysis is shown in (Figure 38).

Other sources in superficial deposits used in the past include a 9 m-deep shaft in Cardiff [ST 1992 7703], which yielded 16.3 l/s from gravels for cooling and washing, and two 200 mm-diameter, 9 m-deep boreholes at Ely [ST 1442 7661] and [ST 1458 7705], which yielded 1.9 Us and 1.6 Us respectively from gravel overlain by marl.

The chemical quality of unpolluted sources is similar to, but harder than, that of any surface water with which they are in hydraulic continuity. Iron and manganese levels can be undesirably high, depending on the composition of the underlying deposits, but tend to decrease with time as the percentage of induced recharge increases with the length of time the well or borehole is pumped.

References

ADAMS, J. E. and RHODES, M. L. 1960. Dolomitisation by seepage refluxion. Bull. Am. Assoc. Pet. Geologists, Vol. 44, 1912–1920.

ADLAM, K. A. Mc.L., HARRISON, D. J., and WILD, J. B. L. 1984. The hard rock resources of the country around Caerphilly, South Wales: description of parts of 1:50 000 sheets 249 and 263. Miner. Assess. Rep. Inst. Geol. Sci., No. 141.

ACER, D. V. 1974. The Jurassic Period in Wales. 323–339 in The Upper Palaeozoic and post-Palaeozoic rocks of Wales. OWEN, T. R. (editor). (Cardiff: University of Wales Press.)

AITKEN, J. D. 1967. Classification and environmental significance of cryptalgal limestones and dolomites, with illustrations from the Cambrian and Ordovician of southwestern Alberta. J. Sediment. Petrol., Vol. 37, 1163–1178.

ALKATTAN, M. M. 1976. The geochemistry of lower Keuper Marl at Lavernock Point, Glamorgan, South Wales. Unpublished MSc thesis, University of Wales.

ALLEN, J. R. L. 1963. Depositional features of Dittonian rocks: Pembrokeshire compared with the Welsh Borderland. Geol. Mag., Vol. 100, 385–400.

ALLEN, J. R. L. 1964. Studies in fluviatile sedimentation: six cyclothems from the Lower Old Red Sandstone, Anglo-Welsh Basin. Sedimentology, Vol. 3, 163–198.

ALLEN, J. R. L. 1965. Upper Old Red Sandstone (Farlovian) palaeogeography in South Wales and the Welsh Borderland. J. Sediment. Petrol., Vol. 35, 167–195.

ALLEN, J. R. L. 1970. Studies in fluviatile sedimentation: a comparison of fining upwards cyclothems with special reference to coarse-member composition and interpretation. J. Sediment. Petrol., Vol. 40, 298–323.

ALLEN, J. R. L. 1974a. The Devonian rocks of Wales and the Welsh Borderland. 47–84 in The Upper Palaeozoic and post-Palaeozoic rocks of Wales. OWEN, T. R. (editor). (Cardiff: University of Wales Press.)

ALLEN, J. R. L. 1974b. Studies in fluviatile sedimentation: implications of pedogenic carbonate units, Lower Old Red Sandstone, Anglo-Welsh outcrop. Geol. J., Vol. 4, 181–208.

ALLEN, J. R. L. 1975. Source rocks of the Lower Old Red Sandstone: Llanishen Conglomerate of the Cardiff area, South Wales. Proc. Geol. Assoc., Vol. 86, 63–76.

ALLEN, J. R. L. 1979. Old Red Sandstone facies in external basins, with particular reference to Southern Britain. 65–80 in The Devonian System, International Symposium on the Devonian System, Bristol, 1978. SCRUTTON, C. T. and BASSETT, M. G. (editors). Special Papers in Palaeontology, No. 23. (London: Palaeontologists Association.)

ALLEN, J. R. L. and DINELEY, D. L. 1976. The succession of the Lower Old Red Sandstone (Siluro-Devonian) along the Ross-Tewkesbury Spur Motorway (M50), Hereford and Worcester. Geol. J., Vol. 11, 1–14.

ALLEN, J. R. L. and WILLIAMS, B. P. J. 1981. Sedimentology and stratigraphy of the Townsend Tuff Bed (Lower Old Red Sandstone) in South Wales and the Welsh Borders. J. Geol. Soc. London, Vol. 138, 15–29.

ANDERSON, F. W. 1964. Rhaetic ostracoda. Bull. Geol. Surv. G.B., Vol. 21, 133–174.

ANDERSON, F. W. 1968. The concealed rock-surface and overlying deposits of the Severn Valley and estuaries from Upton to Neath. Proc. South Wales Inst. Eng., Vol. 83, 27–47.

ANDERSON, F. W. 1974. The buried channels, rock-floors and rock-basins and overlying deposits of the South Wales valleys from Wye to Neath. Proc. South Wales Inst. Eng., Vol. 88, 1–25.

ANDERSON, F. W. and BLUNDELL, C. R. K. 1965. The sub-drill rock surface and buried valleys of the Cardiff district. Proc. Geol. Assoc. , Vol. 76, 367–378.

ANDERSON, F. W. and OWEN, T. R. 1968. The structure of the British Isles. (Oxford: Pergamon Press.)

ANON. 1983. 1981 aggregate minerals survey. South Wales Working Party on Aggregates. (Cardiff: Mid-Glamorgan County Council.)

ARTHURTON, R. S. 1980. Rhythmic sedimentary sequences in the Triassic Keuper Marl (Mercia Mudstone Group) of Cheshire, northwest England. Geol. J., Vol. 15, 43–58.

ASSERETO, R. L. A. M. and KENDALL, C. G. S. C. 1977. Nature, origin and classification of peritidal tepee structures and related breccias. Sedimentology, Vol. 24, 153–210.

BARBER, C. A. 1892. Nematophycus storriei, nov. sp. Ann. Bot., Vol. 6, 329.

BARCLAY, W. J. In press. The geology of the South Wales Coalfield. Part 2. The country around Abergavenny. Mem. Br. Geol. Surv., Sheet 232.

BASSETT, M. G. 1969. The age of the oldest Silurian beds of the Rumney (Cardiff) Inlier. Geol. Mag., Vol. 106, 90–92.

BASSETT, M. G. 1974. Review of the stratigraphy of the Weulock Series in the Welsh Borderland and South Wales. Palaeontology, Vol. 17, 745–777.

BASSETT, M. G. COCKS, L. R. M., HOLLAND, C. H., RICKARDS, R. B. and WARREN, P. T. 1975. The type Wenlock Series. Rep. Inst. Geol. Sci., No. 75/13, 1–19.

BAILEY, R. A. B. 1968. A study of palaeoecology in the Lower Rhaetic of the Vale of Glamorgan and an account of some members of the fossil assemblages. Unpublished PhD thesis, University of Wales, Cardiff.

BHATT, J. J. 1976. Geochemistry and petrology of the Main Limestone Series (Lower Carboniferous), S. Wales, UK. Sediment. Geol., Vol. 15, 55–86.

BOSWORTH, T. O. 1913. The Keuper Marls around Charnwood. (Leicester: Leicester Literary and Philosophical Society.)

BOULTON, G. S. 1972. Modern Arctic glaciers as depositional models for former ice sheets. J. Geol. Soc. London, Vol. 128, 361–393.

BOULTON, W. S. 1910. The underground water supply of Cardiff. Trans. Cardiff Naturalists Soc., Vol. 43, 32–35.

BOWEN, D. Q. 1967. On the supposed ice-dammed lakes of South Wales. Trans. Cardiff Naturalists Soc., Vol. 93, 4–17.

BOWEN, D. Q. 1970. South-east and central South Wales. 197–227 in The glaciation of Wales and adjoining regions. LEWIS, C. A. (editor). (London: Longman.)

BOWEN, D. Q. 1973. Pleistocene history of Wales and the Borderland. Geol. J., Vol. 8, 207–224.

BOWEN, D. Q. 1974. The Quaternary of Wales. 373–426 in The Upper Palaeozoic and post-Paleozoic rocks of Wales. OWEN, T. K. (editor). (Cardiff: University of Wales Press.)

BOWEN, D. Q. 1977. The coast of Wales. 223–256 in The Quaternary history of the Irish Sea. KIDSON. C. and TOOLEY, M. J. (editors). (Liverpool: Seel House Press.)

BOWEN, D. Q. 1981. The 'South Wales end-moraine': fifty years after. 60–67 in The Quaternary in Britain. NEALE, J. and FLENLEY, J. (editors). (Oxford: Pergamon Press.)

BRISTOW, H. W. 1864. On the Rhaetic or Penarth Beds of the neighbourhood of Bristol and the south-west of England. Geol. Mag., Vol. 1, 236–239.

BRISTOW, H. W. and ETHERIDGE, R. 1873. Vertical sections of the Lower Lias and Rhaetic or Penarth Beds of Glamorgan, Somerset and Gloucestershire. Vertical Sections Sheet No. 47. Geol. Surv. G. B.

BROOKS, M., MECHIE, J. and LLEWELLYN, D. J. 1983. Geophysical investigations in the Variscides of southwest Britain. 186–197 in The Variscan fold belt in the British Isles. HANCOCK, P. (editor). (Bristol: Adam Hilger.)

BURCHETTE, T. P. 1977. The sedimentology and palaeoecology of the Lower Limestone Shales in the Forest of Dean and a part of South Wales. Unpublished PhD thesis, University of Wales, Cardiff.

BURCHETTE, T. P. 1981. The Lower Limestone Shales. 13–27 in A field guide to the Carboniferous Limestone around Abergavenny. WRIGHT, V. P., RAVEN, M. and BURCHETTE, T. P. Department of Geology, University College of Wales, Cardiff.

BURCHETTE, T. P. and RIDING, R. 1977. Attached vermiform gastropods in Carboniferous marginal marine stromatolites and biostromes. Lethaia, Vol. 10, 17–28.

CHARLESWORTH, J. K. 1929. The South Wales end-moraine. Q. J. Geol. Soc. London, Vol. 85, 335–358.

CLAYTON, G. and SEVASTOPULO, G. D. 1981, Discussion on the recognition and division of the Tournaisian Series in Britain. J. Geol. Soc. London, Vol. 138, 104–105.

COCKS, L. R. M., HOLLAND, C. H., RICKARDS, R. B. and STRACHAN, I. 1971. A correlation of Silurian rocks in the British Isles. J. Geol. Soc. London, Vol. 127, 103–136.

CONIL, R., GROESSENS, E. and PIRLET, H. 1977. Nouvelle charte stratigraphique du Dinantien type de la Belgique. Ann. Soc. Geol. Nord. , Vol. 96, 363–371.

COPE, J. C. W., GETTY, T. A., HOWARTH, M. K., MORTON, N. and TORRENS, H. S. 1980. A correlation of Jurassic rocks in the British Isles. Part One: Introduction and Lower Jurassic. Spec. Rep. Geol. Soc. London, No. 14.

COX, A. H. and TRUEMAN, A. E. 1936. The Mesozoic rocks. 19–59 in Glamorgan County History, No. 1, Natural History. TATTERSALL, W. M. (editor). (Cardiff: William Lewis.)

CRAMPTON, C. B. 1960. Analysis of heavy minerals in the Carboniferous Limestone, Millstone Grit and soils derived from certain glacial gravels of Glamorgan and Monmouth. Trans. Cardiff Naturalists Soc. , Vol. 87, 13–22.

CRAMPTON, C. B. 1961. An interpretation of the micromineralogy of certain Glamorgan soils: the influence of ice and wind. J. Soil Sci., Vol. 12, 158–171.

CRAMPTON, C. B. 1966. Certain effects of glacial events in the Vale of Glamorgan, South Wales. J. Glaciology, Vol. 6, 261–266.

DAVID, J. W. E. 1883. On the evidence of glacial action in south Brecknockshire and east Glamorganshire. Q. ,J. Geol. Soc. London, Vol. 39, 39–54.

DAVIS, A. G. 1967. The mineralogy and phase equilibrium of Keuper Marl. Q. J. Eng. Geol., Vol. 1, 25–38.

DEAN, W. T., DONOVAN, D. T. and HOWARTH, M. K. 1961. The Liassic ammonite zones and subzones of the North-West European Province. Bull. Br. Mus. (Nat. Hist.) Ser. Geol. , Vol. 4, 435–505.

DE LA BECHE, H. T. 1846. On the formation of the rocks of South Wales and south-western England. Mem. Geol. Surv. G. B., Vol. 1, 1–296.

DIXEY, F. and SIBLY, T. F. 1918. The Carboniferous Limestone Series on the south-eastern margin of the S. Wales Coalfield. Q.,/ Geol. Soc. London, Vol. 73, 111–164.

DIXON, E. E. L. and VAUGHAN, A. 1912. The Carboniferous succession in Gower (Glamorganshire). Q. J. Geol. Soc. London, Vol. 67, 477–571.

DOTT, R. H. JR. and BOURGEOISE, J. 1982. Hummocky stratification: significance of its variable bedding sequences. Bull. Geol. Soc. Am., Vol. 93, 663–680.

DRAVIS, J. 1979. Rapid and widespread generation of recent oolitic hardgrounds on a high energy Bahamian platform, Eleuthera Bank, Bahamas. J. Sediment. Petrol., Vol. 49, 195 - 208.

DUFFIN, C..1. 1980. The Upper Triassic section at Chilcompton, Somerset, with notes on the Rhaetic of the Mendips in general. Mercian Geol., Vol. 7, 251–268.

DUNNING, F. W. 1977. Caledonian- Variscan relations in northwest Europe. 165–180 in La Chaine Varisque d'Europe Moyenne et Occidentale. Editions du C.N.R.S., No. 243. (Paris: C.N.R.S.)

DUTTON, G. H. 1904. Notes on glacial and alluvial deposits near Cardiff. Trans. Cardiff Naturalists Soc., Vol. 36, 109–115.

ETHERIDGE, R. 1872. On the physical structure and organic remains of the Penarth (Rhaetic) beds at Penarth and Lavernock, also with description of the Westbury-on-Severn section. Trans. Cardiff Naturalists Soc., Vol. 3, 39–64.

EVANS, C. D. R. 1982. The geology and superficial sediments of the inner Bristol Channel and Severn estuary. 35–42 in Severn Barrage. Proceedings of a symposium organised by the Institution of Civil Engineers, London, 1981. (London: Thomas Telford.)

EVANS, W. D. and Cox, A. H. 1956. An Old Red Sandstone-Carboniferous Limestone Series junction at Tongwynlais, north of Cardiff. Geol. Mag., Vol. 93, 431–434.

FLÜGEL, E. 1982. Microfacies analysis of limestones. (Berlin: Springer-Verlag.)

FOLK, R. L. and LAND, L. S. 1975. Mg/Ca ratio and salinity: Two controls over crystallization of dolomite. Bull. Am. Assoc. Pet. Geol., Vol. 59, 60–68.

FOOKES, P. G., HINCH, L. W., HUXLEY, M. A. and SIMONS, N. E. 1978. Some soil properties in glacial terrain-the Taff valley, South Wales. 93–116 in The engineering behaviour of glacial materials. (Norwich: Geo Abstracts.)

FRANCIS, E. H. 1959. The Rhaetic of the Bridgend district. Proc. Geol. Assoc. , Vol. 70, 158–170.

FRESHNEY, E. C. and TAYLOR, R. T. 1980. The Variscides of southwest Britain. 49–57 in United Kingdom: Introduction to general geology and guides to excursions 002, 055, 093 and 151. OWEN, T. R. (editor). 26th International Geological Congress, Paris, 1980. (London: Institute of Geological Sciences.)

GAYER, R. A. and CRIDDLE, A. J. 1970. Mineralogy and genesis of the Llanharry iron ore deposits, Glamorgan. 605–626 in Mining and Petroleum. Proc. 9th Commonwealth Mining and Metallurgical Congress 1969, Vol. 2. (London: Institute of Mining Metallurgy.)

GAYER, R. A. ALLEN, K. C., BASSETT, M. G. and EDWARDS, D. 1973. The structure of the Taff Gorge area, Glamorgan, and the stratigraphy of the Old Red Sandstone-Carboniferous Limestone transition. Geol. J., Vol. 8, 345–374.

GEORGE, T. N. 1933. The Carboniferous Limestone Series in the west of the Vale of Glamorgan. Q. J. Geol. Soc. London, Vol. 89, 221–271.

GEORGE, T. N. 1956. The Namurian Usk Anticline. Proc. Geol. Assoc., Vol. 66, 297–316.

GEORGE, T. N. 1958. Lower Carboniferous palaeogeography of the British Isles. Proc. Yorkshire Geol. Soc., Vol. 31, 227–318.

GEORGE, T. N. 1974. Lower Carboniferous rocks in Wales. 85–115 in The Upper Palaeozoic and post-Palaeozoic rocks in Wales. OWEN, T. R. (editor). (Cardiff: University of Wales Press.)

GEORGE, T. N. 1978. Eustasy and tectonics: Sedimentary rhythms and stratigraphical units in British Dinantian correlation. Proc. Yorkshire Geol. Soc., Vol. 42, 229–262.

GEORGE, T. N. JOHNSON, G. A. L., MITCHELL, M. PRENTICE, J. E., RAMSBOTTOM, W. H. C., SEVASTOPULO, G. D. and WILSON, R. B. 1976. A correlation of Dinantian rocks in the British Isles. Spec. Rep. Geol. Soc. London, No. 7.

GINSBERG, R. N. 1975. Tidal deposits. (New York: Springer-Verlag.)

GLASS, N. 1861. Silurian strata near Cardiff. The Geologist, Vol. 4, 168- 169.

GOLDRING, R. and BRIDGES, P. 1973. Sublittoral sheet sandstones. J. Sediment. Petrol., Vol. 43, 736–747.

GREEN, G. W. and WELCH, F. B. A. 1965. Geology of the country around Wells and Cheddar. Mem. Geol. Surv. G.B., Sheet 280.

GRIFFITHS, J. C. 1939. The mineralogy of the glacial deposits of the region between the rivers Neath and Towy, South Wales. Proc. Geol. Assoc. , Vol. 50, 433–462.

GRIFFITHS, J. C. 1940. The glacial deposits west of the Taff. Unpublished PhD thesis, University of London.

GROESSENS, E. 1976. Distribution de Conodontes dans le Dinantien de la Belgique. International Symposium on Belgian micro-palaeontological limits from Emsian to Vision, Namur, 1974. Publication No. 17. Ministry of Economic Affairs, Administration of Mines-Geological Survey Belgium.

HALLAM, A. 1960. A sedimentary and faunal study of the Blue Lias of Dorset and Glamorgan. Philos. Trans. R. Soc. London, Ser. B, Vol. 243, 1–14.

HALLAM, A. 1964. Origin of the limestone-shale rhythm in the Blue Lias of England: A composite theory. J. Geol. , Vol. 72, 157–169.

HALLAM, A. 1971. Mesozoic geology and the opening of the North Atlantic. J. Geol., Vol. 79, 129–157.

HAMILTON, D. 1962. Some notes on the Rhaetic sediments of the Filton Bypass Substitute, near Bristol. Proc. Bristol Naturalists Soc., Vol. 30, 110–118.

HAMILTON, D. 1977. Aust Cliff. 110–118 in Geological excursions in the Bristol District. SAVAGE, R. J. G. (editor). (Bristol: University of Bristol Press.)

HANCOCK, P. L., DUNNE, W. M. and TRINGHAM, M. E. 1983. Variscan deformation in southwest Wales. 47–73 in The Variscan fold belt in the British Isles. HANCOCK, P. L. (editor). (Bristol: Adam Hilger.)

HANSHAW, B. B., BACK, W. and DEIKE, R. G. 1971. A geochemical hypothesis for dolomitization by groundwater. Econ. Geol., Vol. 66, 710–724.

HARMS, J. C., SOUTHARD, J. B., SPEARING, D. R. and WALKER, R. G. 1975. Depositional environments as interpreted from primary sedimentary structures and stratification sequences. Soc. Econ. Palaeontol. Mineral., Short Course Notes No. 2.

HARRIS, W. H. 1884. Fossil plants [Pachytheca sphaerica, Hooker] in the Silurian Formation near Cardiff. Sci. Gossip, Vol. 20, 28–30.

HARRISON, D.J. 1984. The limestone resources of the Bridgend area- a reconnaissance survey. Description of 1:50 000 geological sheet 262. Open file Report, Institute of Geological Sciences.

HARRISON, D.J. MURRAY, D. W. and WILD, J. B. L. 1983. Reconnaissance survey of Carboniferous Limestone resources in Wales. Open file Report, Institute of Geological Sciences.

HAWKINS, A. B. 1979. Case histories of some effects of solution/dissolution in the Keuper rocks of the Severn estuary region. Q. J. Eng. Geol., Vol. 12, 31–40.

HEARD, A. and DAVIES, R. 1924. The Old Red Sandstone of the Cardiff district. Q. J. Geol. Soc. London, Vol. 80, 489–515.

HIRD, K., TUCKER, M. E. and WATERS, R. A. 1984. Petrology, geochemistry and origin of Dinantian dolomites from south east Wales. 22–25 in European Dinantian environments, 1st Mtg. 1984, Abstr. (Milton Keynes: Department of Earth Sciences, Open University.)

HOLLAND, C. H. 1980. Silurian series and stages: decisions concerning chronostratigraphy. Lethaia, Vol. 13, p.238.

HOLLAND, C. H. and RICHARDSON, J. B. 1977. The British Isles. 35–44 in The Silurian -Devonian Boundary. MARTINSSON, A. (editor). I. U. G. S. Ser. A. , (Stuttgart). No. 5.

HOLLAND, C. H. LAWSON, J. D., WALMSI.EY, V. G. and WHITE, D. E. 1980. Ludlow Stages. Lethaia, Vol. 13, p.268.

HOUSE, M. R., RICHARDSON, J. B., CHALONER, W. G., ALLEN, J. R. L., HOLLAND, C. H. and WESTOLL, T. S. 1977 A correlation of Devonian rocks of the British Isles. Spec. Rep. Geol. Soc. London, No. 7.

HOWARD, F. T. 1894. Underground water circulation of Cardiff and its neighbourhood. Trans. Cardiff Naturalists Soc., Vol. 26, 95–104.

HOWARD, F. T. 1895. The geology of Barry Island. Trans. Cardiff Naturalists Soc. , Vol. 27, 45–55.

HOWARD, F. T. 1897. Notes on the base of the Rhaetic Series at Lavernock Point. Trans. Cardiff Naturalists Soc., Vol. 29, 64–66.

HOWARD, F. T. and SMALL, E. W. 1901. Notes on ice action in South Wales. Trans. Cardiff Naturalists Soc., Vol. 32, 44–48.

HURST, J. M. 1975. The diachronism of the Wenlock Limestone. Lethaia, Vol. 8, 301–314.

HURST, J. M. HANCOCK, N. J. and MCKERROW, W. S. 1978. Wenlock stratigraphy and palaeogeography of Wales and the Welsh Borderland. Proc. Geol. Assoc., Vol. 89, 197–226.

IVIMEY-COOK, H. C. 1962. On the relationship between the fauna and the sediments of the Rhaetic in south-east Glamorgan. Unpublished PhD thesis, University of Wales, Cardiff.

IVIMEY-COOK, H. C. 1974. The Permian and Triassic deposits of Wales. 295–321 in The Upper Palaeozoic and post-Palaeozoic rocks of Wales. OWEN, T. R. (editor). (Cardiff: University of Wales Press.)

JAMES, J. W. C. and THORNTON, M. H. 1983. A preliminary study of the sand and gravel deposits of part of south-east Glamorgan and south-west Gwent (1:25 000 Sheets ST 08, 18, 27 and 28). Open file report, Institute of Geological Sciences, Keyworth.

JONES, D. G. 1974. The Namurian Series in South Wales. 117–132 in The Upper Palaeozoic and post-Palaeozoic rocks of Wales. OWEN, T. R. (editor). (Cardiff: University of Wales Press.)

KELLAWAY, G. A. and WELCH, F. B. A. 1955. The Upper Old Red Sandstone and Lower Carboniferous rocks of Bristol and the Mendips compared with those of Chepstow and the Forest of Dean. Bull. Geol. Surv. G. B. , No. 9, 1–21.

KELLING, G. 1974. Upper Carboniferous sedimentation in Wales. 185–224 in The Upper Palaeozoic and post-Palaeozoic rocks of Wales. OWEN, T. R. (editor). (Cardiff: University of Wales Press.)

KIDSON, C. 1977. The coast of South West England. 257–298 in The Quaternary history of the Irish Sea. KIDSON, C. and TOOLEY, M. J. (editors). (Liverpool: Seel House Press.)

KIDSON, C. and HEYWORTH, A. 1973. The Flandrian sea-level rise in the Bristol Channel. Proc. Ussher. Soc., Vol. 2, 565–584.

KLAPPA, C. F. 1980. Rhizoliths in terrestrial carbonates: classification, recognition, genesis and significance. Sedimentology, Vol. 27, 613–629.

KLEIN, G. DE V. 1962. Sedimentary structures in the Keuper Marl (Upper Triassic). Geol. Mag., Vol. 99, 137–144.

LAWRENCE, D. J. D. and WATERS, R. A. 1978. St Fagans Borehole. 10–11 in IGS Boreholes, 1977. Rep. Inst. Geol. Sci., No. 78/21.

LEES, A. 1982. The palaeoenvironmental setting and distribution of the Waulsortian facies of Belgium and Southern Britain. 1–16 in Symposium on the palaeoenvironmental setting and distribution of the Waulsortian facies. BOLTON, K., LANE, H. R. and LEMONE, D. V. (editors). (El Paso: El Paso Geological Society and University of Texas.)

LEWIS, W. J. 1967. Lead mining in Wales. (Cardiff: University of Wales Press.)

LLOYD, M. C. 1889. The geology of Tytherington and Grovesend. Proc. Bristol Nat. Soc. New Ser., Vol. 6, 1–17.

LOGAN, B. W. and SEMENIUK, V. 1976. Dynamic metamorphism; processes and products in Devonian carbonate rocks, Canning Basin, Western Australia. Spec. Publ. Geol. Soc. Australia, No. 16.

MATTHEWS, S. C. 1974. Exmoor thrust? Variscan front? Proc. Ussher. Soc., Vol. 3, 82–94.

MAYALL, M. J. 1979. Sedimentology of the Rhaetic (Upper Triassic) in south-west Britain. Unpublished PhD thesis, University of Reading.

MAYALL, M. J. 1981. The late Triassic Blue Anchor Formation and the initial Rhaetian marine transgression in south-west Britain. Geol. Mag., Vol. 118, 377–384.

MAYALL, M. J. 1983. An earthquake origin for synsedimentary deformation in a late Triassic (Rhaetian) lagoonal sequence, south-west Britain. Geol. Mag., Vol. 120, 613–622.

MEIGH, A. C. 1976. The Triassic rocks, with particular reference to predicted and observed performance of some major foundations. Geotechnique, Vol. 26, 391–452.

MITCHELL, G. F. 1960. The Pleistocene history of the Irish Sea. Adv. Sci., Vol. 17, 313–325.

MITCHELL, M. 1972. The base of the Visean in south-west and north-west England. Proc. Yorkshire Geol. Soc., Vol. 39, 151–160.

MITCHELL, M. 1981. The distribution of Tournaisian and early Visean (Carboniferous) coral faunas from the Bristol and South Wales areas of Britain. Acta. Palaeontologica Polonica, Vol. 25, 577–585.

MITCHELL, M. STRANK, A. R. E., THORNBURY, B. M. and SEVASTOPULO, G. D. 1986. The distribution of platform conodonts, corals and foraminifera from the Black Rock Limestone (late Tournaisian and early Visean), of Tears Point, Cower, South Wales. Proc. Yorkshire Geol. Soc..

MITCHELL, M. WHITTAKER, A., SCRIVENER, R. C., REYNOLDS, M. J., LALOUX, M., OWENS, B., LEES, A. and HENNEBERT, M. 1982. The geology of the IGS deep borehole (Devonian-Carboniferous) at Knap Farm, Canninington Park, Somerset. Rep. Inst. Geol. Sci., No. 82/5, 1–37.

NICKLESS, E. F. P., BOOTH, S. J. and MOSELEY, P. N. 1976. The celestite resources of the area north-east of Bristol with notes on occurrences north and south of the Mendip Hills and in the Vale of Glamorgan: Description of 1:25 000 resource sheet ST 68 and parts of ST 59, 69, 79, 58, 78, 67 and 77. Miner. Assess. Rep. Inst. Geol. Sci. , No. 25.

NORTH, F. J. 1915a. Note on the Silurian Inlier near Cardiff. Geol. Mag., New Ser., Vol. 2, 385–387.

NORTH, F. J. 1915b. On a boring for water at Roath, Cardiff, with a note on the underground structure of the pre-Triassic rocks of the vicinity. Trans. Cardiff Naturalists Soc., Vol. 58, 36–49.

NORTH, F. J. 1929. The evolution of the Bristol Channel. (2nd edition, 1955). (Cardiff: National Museum of Wales.)

NOTHOLT, A. J. G. and HIGHLEY, D. E. 1975. Gypsum and anhydrite. Miner. Dossier Miner. Resour. Consult. Comm., No. 13.

ORBELL, G. 1973. Palynology of the British Rhaeto-Liassic. Bull. Geol. Surv. G.B., Vol. 44, 1–44.

OWEN, T. R. 1967. 'From the South': a discussion of two recent papers in the proceedings. Proc. Geol. Assoc., Vol. 78, 595 - 599.

OWEN, T. R. 1974. The Variscan orogeny in Wales. 285–294 in The Upper Palaeozoic and post-Palaeozoic rocks of Wales. OWEN, T. R. (editor). (Cardiff: University of Wales Press.)

OWEN, T. R. and WEAVER, J. D. 1983. The structure of the main South Wales Coalfield and its margins. 75–87 in The Variscan fold belt in the British Isles. HANCOCK, P. L. (editor). (Bristol: Adam Hilger.)

PEAKS, D. S., BOWEN, D. Q., HAINS, B. A. and SEDDON, B. 1973. Wales. 59–67 in A correlation of Quaternary deposits in the British Isles. MITCHELL, G. F., PENNY, L. F., SHOTTON, F. W. and WEST, R. G. Spec. Rep. Geol. Soc. London, No. 4.

PERKINS, J. W. 1984. The building stones of Cardiff. geological trail guides. (Cardiff: University College Cardiff Press.)

PICK, M. C. 1964. The stratigraphy and sedimentary features of the Old Red Sandstone, Portishead coastal section, north-east Somerset. Proc. Geol. Assoc., Vol. 75, 199–221.

RAMSBOTTOM, W. H. C. 1973. Transgressions and regressions in the Dinantian: a new synthesis of British Dinantian stratigraphy. Proc. Yorkshire Geol. Soc., Vol. 39, 567–607.

RAMSBOTTOM, W. H. C. 1977. Major cycles of transgression and regression (mesothems) in the Namurian. Proc. Yorkshire Geol. Soc., Vol. 41, 261–291.

RAMSBOTTOM, W. H. C. 1979. Rates of transgression and regression in the Carboniferous of N.W. Europe. J. Geol. Soc. London, Vol. 136, 147–153.

RAMSBOTTOM, W. H. C. and MITCHELL, M. 1980. The recognition and division of the Tournaisian Series in Britain. J. Geol. Soc. London, Vol. 137, 61–63.

RHODES, F. H. T., AUSTIN, R. L. and DEUCE, E. C. 1969. British Avonian (Carboniferous) conodont faunas and their value in local and international correlation. Bull. Br. Mus. (Nat. Hist.), Geol. , Suppl. 5.

RICHARDSON, L. 1904. A handbook to the geology of Cheltenham and neighbourhood. (Cheltenham: Norman Sawyer. Reprinted Chicheley: Paul P. B. Minet, 1972.)

RICHARDSON, L. 1905. The Rhaetic and contiguous deposits of Glamorgan-shire. Q. J. Geol. Soc. London, Vol. 61, 385–422.

RICHARDSON, L. 1911. The Rhaetic and contiguous deposits of West, Mid and part of East Somerset. Q. J. Geol. Soc. London, Vol. 67, 1–72.

RIDING, R. and WRIGHT, V. P. 1981. Paleosols and tidalflat/lagoon sequences on a Carboniferous carbonate shelf: sedimentary associations of triple disconformities. J. Sediment. Petrol., Vol. 51, 1323–1339.

ROGERS, E. and OTHERS. 1861. The iron ores of Great Britain. Part 3, Iron ores of South Wales. Mem. Geol. Surv. G.B., 165–236.

SANDBERG, C. A., ZIEGLER, W., LEUTERITZ, K. and BRILL, S. M. 1978. Phylogeny, speciation and zonation of Siphonodella (Conodonta, Upper Devonian and Lower Carboniferous). Newsl. Stratigr., Vol. 7, 102–120.

SELLWOOD, B. W. 1970. The relation of trace fossils to small scale sedimentary cycles in the British Lias. 489–504 in Trace fossils. CRIMES, T. P. and HARPER, J. C. (editors). (Liverpool: Seel House Press.)

SELLWOOD, B. W. DURKIN, M. K. and KENNEDY, W. J. 1970. Field meeting on the Jurassic and Cretaceous rocks of Wessex. Proc. Geol. Assoc., Vol. 81, 715–732.

SHACKLETON, R. M., RIES, A. C. and COWARD, M. P. 1982. An interpretation of the Variscan structures in S.W. England. J. Geol. Soc. London, Vol. 139, 533–541.

SHEARMAN, D. J., MOSSOP, G., DUNSMORE, H. and MARTIN, M. 1972. Origin of gypsum veins by hydraulic fracture. Trans. Inst. MM. Metall., London, Vol. 81, 149–155.

SHERLOCK, R. L., SMITH, B. and HOLLINGWORTH, S. E. 1938. Gypsum, anhydrite, celestite and strontianite. Spec. Rep. Miner. Resour. G.B., Vol. 3, (3rd edition).

SIBLY, T. F. 1920. The Carboniferous Limestone of the Cardiff district. Proc. Geol. Assoc., Vol. 31, 76–92.

SIBLY, T. F. 1927. The haematites of the Forest of Dean and South Wales. Special reports on the mineral resources of Great Britain, Vol. 10. 2nd edition. Mem. Geol. Surv., G.B.

SOLLAS, W. J. 1879. On the Silurian district of Rhymney and Pen-y-Lan, Cardiff. Q. J. Geol. Soc. London, Vol. 35, 475–507.

SPECHT, R. W. and BRENNER, R. L. 1979. Storm-wave genesis of bioclastic carbonates in Upper Jurassic epicontinental mudstones, east-central Wyoming. J. Sediment. Petrol., Vol. 49, 1307- 1322.

SQUIRRELL, H. C. 1973. The Downtonian-Dittonian boundary. J. Geol. Soc. London, Vol. 129, 205–206.

SQUIRRELL, H. C. and DOWNING, R. A. 1969. Geology of the South Wales Coalfield, Part 1, the Country around Newport (Mon). Mem. Geol. Surv. G. B. , Sheet 249.

STEPHENS, J. V. 1973. Geological Survey 1-inch map 247 (Swansea) (revised). Inst. Geol. Sci.

STEPHENS, N. 1973. South West England. 36–45 in A correlation of Quaternary deposits in the British Isles. MITCHELL, G. F., PENNY, L. F., SHOTTON, F. W. and WEST, R. G. Spec. Rep. Geol. Soc. London, No. 4.

STORRIE, J. 1879. The Upper Silurian bone bed at the Rumney river section. Trans. Cardiff Naturalists Soc., Vol. 10, 92–93.

STORRIE, J. 1883. The fossils at Penarth. Trans. Cardiff Naturalists Soc., Vol. 14, 100–103.

STORRIE, J. 1892. On the occurence of Pachytheca and a species of Nematophycus in the Silurian Beds at Tymawr Quarry, Rumney. Rep. Br. Assoc. for 1891, 652–654.

STORRIE, J. 1894. Notes on the tooth of a species of Mastodnnsaurus, found with some other bones near Lavernock. Trans. Cardiff Naturalists Soc., Vol. 26, 105–106.

STORRIE, J. 1908. Geology of the Roath Park, Cardiff. Trans. Cardiff Naturalists Soc., Vol. 40, 18–25.

STRAHAN, A. 1896. On submerged land surfaces at Barry, Glamorgan, with notes on the fauna and flora by Clement Reid. Q. J. Geol. Soc. London, Vol. 52, 474–489.

STRAHAN, A. and CANTRILL, T. C. 1902. The geology of the South Wales Coalfield. Part III. The country around Cardiff. Mem. Geol. Surv. G.B., (1st edition), Sheet 263.

STRAHAN, A. and CANTRILL, T. C. 1904. The geology of the South Wales Coalfield. Part VI. The country around Bridgend. Mem. Geol. Surv. G.B., Sheet 262.

STRAHAN, A. and CANTRILL, T. C. 1912. The geology of the Smith Wales Coalfield. Part III. The country around Cardiff. Mem. Geol. Surv. G.B. , (2nd edition), Sheet 263.

SWIFT, D. J. P. 1968. Coastal erosion and transgressive stratigraphy. J. Geol., Vol. 76, 444–456.

SYKES, J. H. 1977. British Rhaetic Bone-Beds. Mercian Geol., Vol. 6, 197–239.

TRENHAILE, A. S. 1971. Lithological control of high water ledges in the Vale of Glamorgan, Wales. Geog. Annlr., Vol. 53A, 56–59.

TRENHAILE, A. S. 1972. The shore-platforms of the Vale of Glamorgan. Trans. Inst. Br. Geogr. , Vol. 56, 127–144.

TRUEMAN, A. E. 1920. The Liassic rocks of the Cardiff district. Proc. Geol. Assoc., Vol. 31, 93–107.

TRUEMAN, A. E. 1922. The Liassic rocks of Glamorgan. Proc. Geol. Assoc., Vol. 33, 245–284.

TUCKER, M. E. 1974. Exfoliated pebbles and sheeting in the Triassic. Nature, London, Vol. 252, 375–376.

TUCKER, M. E. 1975. Vadose diagenetic fabrics in Triassic lacustrine limestones from South Wales. Extract des publications du IX me Congres Internat. Sediment. (Nice), Vol. 7, 217–224.

TUCKER, M. E. 1977. The Marginal Triassic deposits of South Wales: continental facies and palaeogeography. Geol. J., Vol. 12, 169- 188.

TUCKER, M. E. 1978. Triassic lacustrine sediments from South Wales: shore-zone elastics, evaporites and carbonates. 205–224 in Modern and ancient lake sediments. MATTER, A. and TUCKER, M. E. (editors). Spec. Publ. No. 2 Int. Assoc. Sediment.

TUCKER, M. E. and BURCHETTE, T. P. 1977. Triassic dinosaur footprints from South Wales: their context and preservation. Palaeogeogr., Palaeoclimatol., Palaeoecol., Vol. 22, 195–208.

TUNBRIDGE, I. P. 1981. Old Red Sandstone sedimentation-an example from the Brownstones (highest Lower Old Red Sandstone) of south central Wales. Geol. J., Vol. 16, 111–124.

TUNBRIDGE, I. P. 1983. The Bristol Channel landmass. Proc. Ussher. Soc., Vol. 5, p.490.

VARKER, W. J. and SEVASTOPULO, G. D. 1985. The Carboniferous System: Part 1-Conodonts of the Dinantian Subsystem from Great Britain and Ireland. 167–209 in The stratigraphical index of conodonts. HIGGINS, A. C. and AUSTIN, R. L. (editors). (Cardiff: Ellis Horwood Ltd.)

VAUGHAN, A. 1905. The palaeontological sequence in the Carboniferous Limestone of the Bristol area. Q. J. Geol. Soc. London, Vol. 61, 181–305.

VAUGHAN, A. 1906. The Carboniferous Limestone Series (Avonian) of the Avon Gorge. Proc. Bristol Nat. Soc., Ser. 4, Vol. 1, 74–168.

WALKDEN, G. M. 1974. Palaeokarstic surfaces in upper Visean (Carboniferous) Limestones of the Derbyshire Block, England. J. Sediment. Petrol., Vol. 44, 1232–1247.

WALMSLEY, V. G. 1959. The geology of the Usk Inlier, Monmouthshire. Q. J. Geol. Soc. London, Vol. 114, 483–521.

WALMSLEY, V. G. 1962. Upper Silurian-Devonian contacts in the Welsh Borderland and South Wales. 288–295 in International Symposium on the Silurian -Devonian boundary and the stratigraphy of the Lower and Middle Devonian. Second symposium, Bonn and Brussels, 1960. (Stuttgart: Schweizerbart.)

WANLESS, H. R. 1979. Limestone response to stress, pressure solution and dolomitization. J. Sediment. Petrol., Vol. 49, 437–462.

WARRINGTON, G. 1970. The stratigraphy and palaeontology of the 'Keuper' Series of the central Midlands of England. J. Geol. Soc. London, Vol. 126, 183–223.

WARRINGTON, G. AUDLEY-CHARLES, M. G., ELLIOTT, R. E., EVANS, W. B., IVIMEY-COOK, H. C., KENT, P., ROBINSON, P. L., SHOTTON, F. W. and TAYLOR, F. M. 1980. A correlation of Triassic rocks in the British Isles. Spec. Rep. Geol. Soc. London, No. 13.

WATERS, R. A. 1978a. Cwrt-yr-ala Borehole. 9 in IGS Boreholcs 1977. Rep. Inst. Geol. Sci., No. 78/21.

WATERS, R. A. 1978b. Nant Bran, St Lythans Borehole. 9–10 in IGS Boreholes 1977. Rep. Inst. Geol. Sci., No. 78/21.

WATERS, R. A. 1984. Some aspects of the Black Rock Limestone and Gully Oolite (Dinantian) in the eastern Vale of Glamorgan. Proc. Geol. Assoc., Vol. 95, 391–392.

WATERS, R. A. and WHITE, D. E. 1980. Rumney Borehole. 9–10 in IGS Boreholes 1978. Rep. Inst. Geol. Sci., No. 79/12.

WATKINS, R. and AITHIE, C. J. 1980. Carbonate shelf environments and faunal communities in the Upper Bringewood Beds of the British Silurian. Palaeogeog. Palaeoclimatol. Palaeoecol., Vol. 29, 341–368.

WELCH, F. B. A. and TROTTER, F. M. 1961. Geology of the country around Monmouth and Chepstow. Mem. Geol. Surv. G. B.

WHITTAKER, A. 1973. The Central Somerset Basin. Proc. Ussher Soc., Vol. 2, 585–592.

WHITTAKER, A. 1975. A postulated post-Hercynian rift valley system in Southern Britain. Geol. Mag., Vol. 112, 137–149.

WHITTAKER, A. 1978. The lithostratigraphical correlation of the uppermost Rhaetic and lowermost Liassic strata of the Somerset and Glamorgan areas. Geol. Mag., Vol. 115, 63–67.

WHITTAKER, A. and GREEN, G. W. 1983. Geology of the country around Weston-super-Mare. Mem. Geol. Surv. G.B. , Sheet 279 with parts of 263 and 295.

WILLIAMS, G. J. 1968. The buried channel and superficial deposits of the Lower Usk, and their correlation with similar features in the Lower Severn. Proc. Geol. Assoc., Vol. 79, 325–348.

WILLS, L. J. 1976. The Trias of Worcestershire and Warwickshire. Rep. Inst. Geol. Sci., No. 76/2.

WILSON, D., DAVIES, J. R., FLETCHER, C. J. N. and SMITH, M. In press. Geology of the country around Bridgend. Explan. 1:50 000 sheet, Br. Geol. Surv. , Sheet 262, England and Wales.

WOBBER, F. J. 1965. Sedimentology of the Lias (Lower Jurassic) of South Wales. J. Sediment. Petrol., Vol. 35, 683–703.

WOBBER, F. J. 1966. A study of the depositional area of the Glamorgan Lias. Proc. Geol. Assoc. , Vol. 77, 127–137.

WOBBER, F. J. 1967. Post-depositional structures in the Lias, South Wales. J. Sediment. Petrol., Vol. 37, 166–174.

WOBBER, F. J. 1968a. Microscdimentary analysis of the Lias in South Wales. Sediment. Geol., Vol. 2, 13–49.

WOBBER, F. J. 1968b. A faunal analysis of the Lias (Lower Jurassic) of South Wales (Great Britain). Palaeogeogr. Palaeoclimatol. Palaeoecol., Vol. 10, 269–308.

WOODLAND, A. W. and EVANS, W. B. 1964. The geology of the South Wales Coalfield, part 4. The country around Pontypridd and Maesteg. Mem. Geol. Surv. G.B., Sheet 248.

WOODWARD, H. B. 1893. The Jurassic rocks of Britain, Vol. 3, The Lias of England and Wales (Yorkshire excepted). Mem. Geol. Surv. G. B.

WRIGHT, T. 1860. On the zone of Avicula contorta, and the Lower Lias of the south of England. Q J. Geol. Soc. London, Vol. 16, 374–411.

WRIGHT, V. P. 1982. The recognition of palaeokarsts: two examples from the Lower Carboniferous of South Wales. J. Sediment. Petrol., Vol. 52, 83–4.

Fossil index

No distinction is made here between a positively determined genus or species and examples doubtfully referred to them (i.e. with the qualifications aff., cf. or ?).

• Acanthopyge hirsuta (Fletcher)
• Acanthotriletes spp.
• Acastocephala macrops (Salter)
• acritarchs,
• Acrodus
• Aegiria grayi (Davidson)
• Alisporites thomasii (Couper) Nilsson 1958
• A. spp.
• Alsatites laqueolus (Schloenbach)
• liasicus (d'Orbigny)
• Amberleya (Eucyclus) sp.
• Ambitisporites sp.13
• Ammarchaediscus sp.
• ammonites
• Amphistrophia funiculata (McCoy)
• Anastrophia deflexa (J. de C. Sowerby)
• Anchisauripus
• Apatognathus sp.
• Apiculatisporis sp.
• Apiculiretusispora sp.
• Aplocoma (Ophiolepis) damesi (Wright)
• ‘Arca’ pullus Terquem
• Archaediscus varsanofievae Grozdilova & Lebedeva
• A. sp. (concaves stage)
• A. sp. (involutes stage)
• Archaeozonotriletes chutes (Cramer) Richardson & Lister, 1969
• Arnioceras
• Astarte
• Atrypa reticularis (Linnaeus)
• Axophyllum sp.
• vaughani (Salée)
• 'Bactryllium striolatum' Heer
• Baltisphaeridium spp.
• Beaconites
• Bellerophon wenlockensis J. de C. Sowerby
• B. sp.
• bellerophontids
• Bembexia lloydii (J. de C. Sowerby)
• Bilinguites superbilinguis (Bisat)
• Birgeria
• Biseriella bristolensis (Reichel)
• bispathodids
• Bispathodus aculeatus aculeatus (Branson)
• B. aculeatus-Clydagnathus transition
• B. aculeatus plumulus (Rhodes, Austin & Druce)
• B.-Pseudopolygnathus transition
• B. stabilis (Branson & Mehl)
• B. sp.
• bivalves
• Bothriolepis
• brachiopods
• bryozoa
• Bumastus? xestos Lane & Thomas
• burrows
• Buxtonia sp.
• Calamospora sp.
• Calcirhynchia calcaria S. S. Buckman
• Caloceras bloomfieldense (Donovan)
• intermedium (Portlock)
• Caloceras
• Calymene sp.
• Camptonectes
• Ganinia cornucopiae Michelin
• Caninophyllum patulum (Michelin)
• C. patulum greeni Mitchell
• C. patulum patulum (Michelin)
• C. Cardinia hybrida (J. Sowerby)
• C. ovalis (Stutchbury)
• C. regularis Terquem
• C. sp.
• Carnisporites spp.
• Catenipora
• cephalopods
• Ceratodus
• Chaetetes depressus (Fleming)
• C. septosus (Fleming)
• Chasmatosporites spp.
• chitinozoa
• Chlarnys Pollux (d'Orbigny)
• C. valoniensis (Defrance)
• Chlamys
• Chondrites
• Cingulizonates rhaeticus (Reinhardt) Schulz 1967
• Classopollis torosus (Reissinger) Balme 1957
• C. spp.
• Cleiothyridina glabristria (Phillips)
• C. royssii (Davidson)
• Cleistosphaeridium mojsisovicsii Morbey 1975
• Clisiophyllum multiseptatum Garwood
• Clydagnathus sp.
• Colonograptus colones (Barrande)
• C. varians (Wood)
• Composita ambigua (J. Sowerby)
• C. C. ficoidea (Vaughan)
• Composita
• Conochitina sp.
• conodonts
• Converrucosisporites luebbenensis Schulz 1967
• Convolutispora microrugulata Schulz 1967
• Coolinia pecten (Linnaeus)
• corals
• Cornutisporites regulates Schulz 1967
• Coroniceras hyatti Donovan
• C. rotiforme (J. de C. Sowerby)
• C. sp.
• Craniops implicates (J. de C. Sowerby)
• Cravenia lamellate Howell
• crinoids/columnals/debris/fragments/crustaceans
• 'Ctenodonta subaequalis' (McCoy)
• Cuneigervillia
• Cyathaxonia corne Michelin
• Cyathoclisia tabernaculum Dingwall
• Cymatiosphaera polypartita Morbey 1975
• Cymatiosphaera
• Cypricardinia subplanulata Reed
• Cyrtia exporrecta (Wahlenberg)
• Dacryomya titei (Moore)
• Dalatias
• Dalejina hybrida (J. de C. Sowerby)
• Dalmanites sp.
• Dapcodinium priscum Evitt 1961
• Dapsilidinium langii (Wall) Lentin & Williams 1981
• Darwinula liassica (Brodie)
• Darwinula
• dasycladacean algae
• Davidsonina carbonaria (McCoy)
• Dawsonoceras annulatum (J. Sowerby)
• Dayia navicula (J. de C. Sowerby)
• Deiphon barrandei Whittard
• Delepinea carinata (Garwood)
• D. sp.
• Deltoidospora spp.
• dendroid graptolite
• Dentalina
• diademopsid fragments
• Dicoelosia biloba (Linnaeus)
• Dicranopeltis salteri (Fletcher)
• Dictyonella capewellii (Davidson)
• Diexallophasis denticulata (Stockmans & Williere) Loeblich 1970
• Dimyopsis (Atreta) intusstriata (Emmrich)
• Dimyopsis (Atreta)
• Dimyopsis
• disacciatriletes
• Discinisca
• Dollymae bouckaerti Groessens
• Earlandia sp.
• echinoderm debris/fragments
• echinoid fragments/spines
• Encrinurus tuberculatus (Buckland)
• Eoguttulina liassica (Strickland)
• Eolepas rhaetica (Moore)
• E. sp.
• Eoparastaffella restricta Postoialko & Garini
• E. simplex (Vdovenko)
• Eophacops musheni. (Salter)
• Eoplectodonta duvalii (Davidson)
• E. sp.
• Eospirifer radiatus (J. de C. Sowerby)
• Eotaphrus bultyncki (Groessens)
• Eotrapezium concentricum (Moore)
• E. germari (Dunker)
• Eotrapezium
• Euomphalus sp.
• Eupoikilofusa cantabrica (Cramer) Cramer 1970
• Fasciculophyllum densum (Carruthers)
• E. omaliusi (Milne-Edwards & Haime)
• Favositella anolotichoides Oakley
• E. interpuncta (Quenstedt)
• Favosites sp.
• favositoids
• Fenestella sp.
• fish fragments/scales/spines
• flora
• foraminifera
• Foraminisporis jurassicus Schulz 1967
• Forschia parvula Rauser
• gastropods
• 'Gervillia' praecursor (Quenstedt)
• Gigantoproductus aff.(Vaughan)
• G. sp.
• Gliscopollis meyeriana (Klaus) Venkatachala 1966
• Gliscopollis
• Glomodiscus miloni (Pelhate)
• G. oblongus (Conil & Lys)
• G. sp.
• Gnathodus bulbosus Thompson
• G. cuneiformis Mehl & Thomas
• G. delicatus Branson & Mehl
• G. delicatus-G. bulbosus-transition
• G. pseudosemiglaber Thompson & Fellows
• G. simplicatus Rhodes, Austin & Druce
• goniatite
• Granuloperculatipollis rudis Venkatachala & Goczan emend Morbey 1975
• graptolites
• Gryphaea arcuata Lamarck
• Gryphaea
• Gypidula galeata (Dalman)
• G. sp.
• Gyrolepis
• Haplolasma subibicina (McCoy)
• Harpidella (H.) aitholix Thomas
• Hemiarges scutalis (Salter)
• Hindeodus cristulus (Youngquist & Miller)
• Holoptychius nobilissimus Agassiz
• Homoceratoides fortelirifer Ramsbottom
• homalonotid
• Howellella elegans (Muir-Wood)
• H. subinsignis (Reed)
• H. sp.
• Hybodus
• hyolithid
• Isorthis clivosa Walmsley
• H. orbicularis (J. de C. Sowerby)
• Kionoceras angulatum (Wahlenberg)
• Koninckophyllum sp.
• Koninckopora inflata (de Koninck)
• K. tenuiramosa Wood
• K. sp.
• Koremagraptus sp.
• Koskinotextularia sp.
• Kraeuselisporites reissingeri (Harris) Morbey 1975
• Leiosphaeridia spp.
• Leonaspis coronata (Salter)
• Lepidoleptaena poulseni (Kelly)
• Leptaena depressa U. de C. Sowerby)
• L. depressa kathekta Bassett
• Leptagonia analoga (Phillips)
• Leptolepidites argenteaeformis (Bolkhovitina) Morbey 1975
• Leptostrophia filosa (J. de C. Sowerby)
• Limbosporites lundbladii Nilsson 1958
• Lingula cornea J. de C. Sowerby
• L. sp.
• lingulids
• Lingulina tenera Bornemann
• Linoprotonia corrugatohemispherica (Vaughan)
• L. hemisphaerica (J. Sowerby)
• L. sp.
• Liostrea bristovi (Richardson)
• L. hisingeri (Nilsson)
• L. irregularis (Munster)
• L. sp.
• Lissodus
• Lithostrotion aranea (McCoy)
• L. martini Milne-Edwards & Haime46,
• L. sociale (Phillips)
• L. sp.
• Loxonema gracile (Goldfuss)
• L. hydropica (Sollas)
• L. sp.
• Lunatisporites rhaeticus (Schulz) Warrington 1974
• Lycopodiacidites spp.
• Lyriomyophoria postera (Quenstedt)
• Lyriomyophoria
• Macropotamorhynchus mitcheldeanensis (Vaughan)
• Mactromya
• Mediocris breviscula (Ganelina)
• M. mediocris (Vissarionova)
• Megachonetes hemisphaerica (Sernenow)
• M. magna Rotai
• M. papdionaceus (Phillips)
• M. sp.
• Megastrophia (Protomegastrophia) semiglobosa (Davidson)
• Melarchaediscus sp.
• Meleagrinella decussata (Munster)
• M. fallax (Pflucker)
• Meleagrinella
• Meristina obtusa (J. Sowerby)
• Mesopholidostrophia lepisma (J. de C. Sowerby)
• Mestognathus beckmanni Bischoff
• M. groessensi Belka
• Metacytheropteron
• Michelinia favosa (Goldfuss)
• M. konincki Vaughan
• M. megastoma Phillips
• M. megastoma Z₂ Vaughan
• M. sp.
• Micrhystridium licroidium Morbey 1975
• M. lymense Wall 1965
• M. sp.
• Microreticulatisporites fuscus (Nilsson) Morbey 1975
• Microsphaeridiorhynchus nucula U. de C. Sowerby)
• miospores
• Modiolopsis sp.
• Modiolus hillanoides (Chapuis & Dewalque)
• M. laevis (J. Sowerby)
• M. minimus J. Sowerby
• ‘M.’ sodburiensis Vaughan
• M. sp.
• molluscs
• monograptids
• Monograptus flemingii (Salter)
• Montlivaltia rhaetica Tomes
• Murchisonia elegans Sollas
• Watica' oppelii Moore
• Nemacanthus monitifer L. Agassiz
• Nevesisporites bigranulatus (Levet-Carette) Morbey 1975
• Nibelia nibelis Durkina
• Nodosarchaediscus sp.
• Nodosaria
• Nucleospira pisum (J. de C. Sowerby)
• Nuculites sp.
• Orbiculoidea sp.
• organic-walled microplankton
• 'Orthoceras' ibex J. de C. Sowerby
• orthocones
• orthotetoids
• ostracods
• ostreid fragments
• Ovalipollis pseudoalatus (Thiergart) Schuurman 1976
• Oxytoma
• oysters
• Pachytheca sp.
• Palaeosmilia murchisoni Milne-Edwards & Haime
• Palaeospiroplectammina sp.
• Parallelodon sp.
• Patrognathus variabilis Rhodes, Austin & Deuce
• Patrognathus
• Pentlandina lewisii plakodis Bassett
• Perinosporites thuringiacus Schulz 1962
• Phaulactis sp.
• Pholadomya
• Pinna
• Placunopsis alpina (Winkler)
• Plagiostoma giganteum J. Sowerby
• P. valoniensis (Defrance)
• Plagiostoma
• plants
• Platyceras cornutum (Hisinger)
• Platylichas grayii (Fletcher)
• Plectatrypa imbricate (J. de C. Sowerby)
• Plectogyranopsis convexa (Rauser)
• Pleuromya sp.
• Pleurophorus' elongatus Moore
• Pleurotomaria
• Plicatula sp.
• Plicochonetes sp.
• Polycingulatisporites bicollateralis (Rogalska) Morbey 1975
• Polygnathus bischoffi Rhodes, Austin & Deuce
• P. communis caring Hass
• P. communis communis Branson & Mehl
• P. communis dentatus Druce
• P. inornatus Branson
• P. mehli Thompson
• P. spicatus Branson
• Polygonium gracilis Vavrdova 1966
• Porcellispora longdonensis (Clarke) Scheuring emend Morbey 1975
• Prioniodina oweni Rhodes, Austin & Druce
• Productus sp.
• Pronoella
• Protocardia rhaetica (Merian)
• Protocardia
• Protochonetes ludloviensis Muir-Wood
• Protognathodus cordiformis Lane, Sandberg & Ziegler
• Protohaploxypinus microcorpus (Schaarschmidt) Clarke 1965
• Pseudolimea pectinoides (J. Sowerby)
• Pseudolimea
• Pseudopecten
• Pseudopolygnathus conili Bouckaert & Groessens
• P. dentilineatus Branson
• P. minutus Metcalfe
• P. multistriatus Mehl & Thomas
• P. sp.
• P. triangulus pinnatus Voges
• Psiloceras planorbis (J. de C. Sowerby)
• P. plicatulum (Quenstedt)
• Psiloceras
• Psilophyllites
• Pteromya sp.
• Ptychopteria sp.
• Pugilis vaughani (Muir-Wood)
• Punctatisporites sp.
• Pustula sp.
• Quadraeculina anellaeformis Maljavkina 1949
• Rectodiscus sp.
• reptile fragments/remains
• Resserella canalis U. de C. Sowerby)
• Retusotriletes sp.
• Rhaetavicula contorta (Portlock)
• Rhaetavicula
• Rhaetipolli s germanicus Schulz 1967
• Rhaetogonyaulax rhaetica (Sarjeant) Loeblich & Loeblich emend Harland, Morbey & Sarjeant 1975
• Rhipidomella michelini (Leveille)
• Rhombocytheie
• rhynchonellids
• rhynchonelloid
• Ricciisporites tuberculatus Lundblad 1954
• rootlets
• Rugosa
• Rugosochonetes vaughani Muir-Wood
• Rugosochonetes sp.
• Saetograptus chimaera salweyi (Lapworth)
• S. leintwardinensis (Lapworth)
• S. sp.
• Salopina conservatrix (McLearn)
• S. lunata (J. de C. Sowerby)
• 'Sargodon tomicus' Pleininger
• Saurichthys
• Sauripteris
• Scaliognathus anchoralis Branson & Mehl
• S. praeanchoralis Lane, Sandberg & Ziegler
• Schizophoria sp.
• Schlotheimia gallica S. S. Buckman
• Schlotheimia
• scolecodonts
• Shagamella minor (Salter)
• Shaleria ornatella (Davidson)
• Siphonodella isosticha (Cooper)
• siphonodellids
• Siphonophyllia cylindrica (Scouler)
• S. garwoodi Ramsbottom & Mitchell
• S. sp.
• Skenidioides lewisii (Davidson)
• Solisphaeridium sp.
• Spathognathodus scitulus (Hinde)
• S. stabilis (Branson & Mehl)
• S. sp.
• Sphaerirhynchia wilsoni (J. Sowerby)
• Sphaerochitina 'sphaerocephala' (Eisenack) Eisenack 1955
• 'Sphaerodus'
• Spirifer sp.
• Spiriferellina sp.
• spiriferoid
• Spirigerina marginates (Dalman)
• Stenoscisma isorhyncha (McCoy)
• Straparollus sp.
• Streptis grayii (Davidson)
• Striispirifer plicatellus (Linnaeus)
• stromatolites
• Strophochonetes
• Strophonella euglypha (Dalman)
• Stylophyllopsis
• Sychnoelasma clevedonensis Mitchell
• S. konincki (Milne-Edwards & Haime)
• Syringopora reticulata Goldfuss
• S. vaughani Hudson
• S. spp.
• Syringothyris cyrtorhyncha North
• S. elongata North
• S. sp.
• Tasmanites spp.
• Terquemia arietis (Quenstedt)
• Thalassinoides
• Todisporites minor Couper 1958
• trilobites
• Trypanites
• Tsugaepollenites? pseudomassulae (Madler) Morbey 1975
• Tutcheria cloacina (Quenstedt)
• Tutcheria
• Tylothyris sp.
• Unicardium
• Unispirifer tornacensis (de Koninck)
• Vaughania vetus Smyth
• Vaughania
• Venniceras conybeari (J. Sowerby)
• Vermiceras vertebrate remains/fragments
• Veryhachium leintwardinensis Doming81
• V. sp.
• Vesicaspora fuscus (Pautsch) Morbey 1975
• Visbysphaera microspinosa (Eisenack) Lister 1970
• Vitreisporites pallidus (Reissinger) Nilsson 1958
• Waehneroceras portlocki (Wright)
• Waehneroceras
• wood fragments
• Zaphrentites delanouei Milne-Edwards & Haime
• Zebrasporites interscriptus (Thiergart) Klaus 1960

Figures and plates

(Figure 1) Silurian nomenclature

(Figure 2) Rumney Borehole: Silurian stratigraphy

(Figure 3) Old Red Sandstone: lithostratigraphical nomenclature. Vertical ruling indicates unconformity

(Figure 4) Chronostratigraphy of Devonian strata of the Cardiff district and adjacent areas. Vertical ruling indicates unconformity. Numbers in brackets refer to BGS 1:50 000 geological sheet numbers

(Figure 5) Composite section of the Llanishen Conglomerate of the Michaelston-super-Ely Inlier

(Figure 6) Graphic log of Brownstones in Cwrt-yr-ala Borehole

(Figure 7) Graphic log of Upper Old Red Sandstone in Cwrt-yr-ala Borehole

(Figure 8) Nomenclature and classification of Dinantian sequence. *The Caswell Bay Mudstone was regarded as Chadian by George and others, (1976) and Whittaker and Green, (1983)

(Figure 9) Sections of the Tongwynlais Formation at Tongwynlais and Cwrt-yr-ala

(Figure 10) Sections in the Castell Coch Limestone between the Taff valley and the coast

(Figure 11) Conodont faunas from the Lower Limestone Shale Group. Section at Castell Coch Quarry begins at base of Castell Coch Limestone J = juvenile

(Figure 12) Distribution of lithofacies in the Black Rock Limestone (Group)

(Figure 13) Lithofacies and conodonts from Brofiscin Quarry. Key as in (Figure 12)

(Figure 14) Range of corals and conodonts from the Cwmyniscoy Mudstone, Barry Harbour Limestone and Friars Point limestone at Barry Harbour and Friars Point (trefers to distance in metres below the base of the Barry Harbour Limestone)

(Figure 15) Conodont faunas from the Black Rock Limestone (Group) of the Llantrisant Road section

(Figure 16) Gully Oolite in St Lythans Borehole

(Figure 17) Graphic section of Caswell Bay Mudstone in St Lythans Borehole and details of contact with the Gully Oolite

(Figure 18) High Tor Limestone (lower part only) in St Lythans Borehole

(Figure 19) Distribution of significant Visean fossils from the Dinantian of the Cardiff district

(Figure 20) Distribution of lithologies in Mercia Mudstone Group marginal facies

(Figure 21) Nomenclature and classification of the Triassic strata

(Figure 22) Schematic diagram of facies relationships in the Mercia Mudstone Group

(Figure 23) Contours on base of (A) Mercia Mudstone Group and (B) base of 'red mudstones' in the Cardiff area

(Figure 24) Deep boreholes in the Cardiff area. Boreholes numbered as in (Figure 23) 1 Farmers and Dairymen Borehole, Newport Road [ST 2078 7837] 2 Schweppes Borehole, Newport Road [ST 2081 7799] 3 Helen Street Borehole [ST 1992 7704] 4 Pellett Street Borehole [ST 1881 7611] 5 Crawshay Street Borehole [ST 1810 7573] 6 Grangetown Gasworks Borehole No.2 [ST 1761 7437]

(Figure 25) Blue Anchor Formation north of Lavernock Point

(Figure 26) Distibution of palynomorphs in the late Triassic and early Jurassic succession, Lavernock

(Figure 27) Typical sections in Triassic continental and lacustrine shore-zone subfacies Continental subfacies: 1a. Well sorted conglomerates and sandstones 1b. Ill sorted angular breccias 1(d) Thin-bedded sandstones/calcarenites, siltstones and mudstones with units of nodular dolomite. Lacustrine shore-zone subfacies 2a. Conglomerates and calcarenites 2b. Fenestral and cryptalgal carbonates 2c. Red mudstones with nodular dolomites and gypsum 

(Figure 28) Horizontal sketch sections through the marginal facies on the west side of Little Island, Barry Island, to show the relationship of the marginal facies to the Triassic platforms and Carboniferous Limestone

(Figure 29) Vertical sections of Penarth Group

(Figure 30 Principal microfauna of the Penarth Group near Lavernock Point

(Figure 31) Nomenclature of Lower Lias in the coastal belt

(Figure 32) Vertical sections of the St Mary's Well Bay Formation

(Figure 33) Porthkerry Formation in cliffs between Dams Bay and The Bulwarks, Porthkerry

(Figure 34) Biostratigraphy of the major sections in the Lower Lias

(Figure 35) St Fagans Borehole: Lower Lias marginal facies

(Figure 36) Drift deposits of the Cardiff district

(Figure 37) Structural elements of the Cardiff district

(Figure 38) Typical chemical analyses of groundwaters in the Cardiff district

Plates

(Plate 1) (Frontispiece) Coastal section in the Blue Lias between Lavernock Point and St Mary's Well Bay (A12619) The cliff section exposes a sequence, from the St Mary's Well Bay Formation to the lower part of the Porthkerry Formation, disposed in a broad open syncline

(Plate 2) Hill Gardens Formation, east bank of the River Rhymney, Rumney (A12912) Thin sheet-like siltstones and sandstones interbedded with mud-stones

(Plate 3) Friars Point Limestone, Friars Point, Barry Island (A12923) Lithofacies Cii, comprising thick units of skeletal packstone separated by shaly, skeletal wackestone/packstone partings

(Plate 4) Bioturbation in Friars Point Limestone (lithofacies Cii), Friars Point, Barry Island (A12740) Remnants of thin, coarse, crinoidal lags in intensely burrowed, skeletal packstone

(Plate 5) Sub-Triassic unconformity, Sully Island (A12623) Mercia Mudstone Group marginal facies, comprising shore-zone subfacies breccias and calcarenites overlain by red mudstones with nodular evap°rites, resting unconformably on steeply dipping Carboniferous Limestone

(Plate 6) Mercia Mudstone Group 'red mudstones' with voids due to the dissolution of gypsum nodules, Jackson's Bay, Barry Island (A12742) The structureless red mudstones contain green reduction streaks whilst the voids are lined by calcite.

(Plate 7) Mercia Mudstone Group lacustrine shore-zone subfacies, Friars Point, Barry Island (A12738) In the foreground a Triassic platform is cut across ill sorted scree breccias. The platform terminates in a Triassic cliff, cut in Carboniferous Limestone and exhibiting a wave cut notch. Tabular bedded, well sorted, shore-zone breccias overlie a further platform cut in Carboniferous Limestone

(Plate 8) Fenestral limestone in Mercia Mudstone Group marginal facies, Sully Island (A12906)

(Plate 9) Columnar stromatolites in Mercia Mudstone Group marginal facies, Dinas Powis (A12907)

(Plate 10) Coast section in Porthkerry Formation, west of Porthkerry (A12918). Unit B comprising limestones with subordinate mudstones.

Figures

Figure 38 Typical chemical analyses of groundwaters in the Cardiff district

* Data provided by the South Eastern Division of the Welsh Water Authority