Content and licensingview original scan buy a printed copy

Geology of the country around Weston-super-Mare. Memoir for 1:50000 geological sheet 279, New Series, with parts of sheets 263 and 295 (England and Wales)

By A. Whittaker and G. W. Green

Bibliographical reference: Whittaker, A. and Green, G. W. 1983. Geology of the country around Weston-super-Mare. Memoir Geological Survey of Great Britain, Sheet 279 with parts of 263 and 295.

Institute of Geological Sciences, Natural Environment Research Council, Geological Survey of Great Britain England and Wales. Production of this memoir was supported by the Department of the Environment. The views expressed in it are not necessarily those of that Department.

• Authors
• A. Whittaker and G. W. Green
• Contributors
• Geophysics J. D. Cornwell
• Offshore geology B. N. Fletcher
• Palaeontology D. T. Donovan H. C. Ivimey-Cook M. Mitchell M. J. Reynolds G. Warrington
• Petrography R. W. Sanderson

London Her Majesty's Stationery Office 1983. © Crown Copyright 1983 First published 1983. ISBN 0 11 884348 6. Produced in the UK for HMSO Dd 736191 C20 11/83

• Authors
• A. Whittaker, BSc, PhD and G. W. Green, MA Institute of Geological Sciences, Nicker Hill, Keyworth, Nottingham NG12 5GG
• Contributors
• J. D. Cornwell, MSc, PhD Institute of Geological Sciences, Keyworth
• B. N. Fletcher, BSc, PhD, M. Mitchell, MA,
• M. J. Reynolds, MPhil and G. Warrington, BSc, PhD Institute of Geological Sciences, Ring Road Halton, Leeds LS15 8TQ
• H. C. Ivimey-Cook, BSc, PhD and R.W. Sanderson, BSc Institute of Geological Sciences, London
• Professor D. T. Donovan, DSc Department of Geology, University College London, Gower Street, London WC1

(Front cover)

(Rear cover)

Preface

The geological map of the Weston-super-Mare district described in this memoir mainly comprises the New Series Weston-super-Mare (279) Sheet, but also encompasses small areas of the Cardiff (263) and Taunton (295) sheets, respectively in the north and south, and the islands of Steep Holm and Flat Holm. The oldest exposed rock in the district forms part of the Devonian of the Quantock Hills. Carboniferous Limestone (Dinantian) strata are well seen onshore in coastal and other sections in the northern part of the district and offshore in the islands in the Bristol Channel. Mesozoic rocks of the Triassic and Jurassic systems are magnificently exposed along the north Somerset coast, where they form the greater part of the type area for the Triassic Jurassic system boundary proposed by the Geological Society of London (Cope and others, 1980; Warrington and others, 1980). The Institute's boreholes at Brent Knoll in the centre of the district proved a thick and almost complete Lias sequence overlying Triassic strata with halite deposits, thus confirming the presence of a Somerset saltfield. Numerous boreholes in the wide alluvial tracts of the Somerset Levels have yielded much information on sedimentation during the last 10000 years.

The district forms part of the region covered by the Old Series one-inch Sheet 20, which was geologically surveyed on the one-inch scale by Sir Henry T. De la Beche and published in 1834, and revised in 1839. Some of the geological features were described in De la Beche's classic Report on the geology of Cornwall, Devon, and west Somerset, published in 1839. The part of Avon (then Somerset) lying to the north of Weston-super-Mare, which occupies the south-eastern corner of the New Series Cardiff (263) Sheet, was surveyed on the six-inch scale by Sir Aubrey Strahan in 1896 and the map was published in 1897, to be followed by a descriptive memoir in 1902. A revised edition of this map was published in 1902 and a second edition of the memoir in 1912. The Taunton (295) Sheet was resurveyed mainly by W. A. E. Ussher and published in 1907.

With Dr J. R. Earp as District Geologist, a primary six-inch survey of the northern part of the district was started by Mr G. W. Green in 1967, who also resurveyed the onshore English part of the Cardiff (263) Sheet and was joined by Dr B. Kelk for a short period. The remainder of the mainland area was completed by Dr A. Whittaker between 1968 and 1970, under the direction of Mr G. Bisson of the Institute's Exeter office. The offshore islands were mapped by Mr Green and Dr Kelk during visits in 1972 and 1975.

During the survey the collection of fossils additional to those found by the surveyors was undertaken by Messrs M. Mitchell and M. J. Reynolds (Carboniferous) and Dr H. C. Ivimey-Cook and Mr R. D. Clark (Mesozoic). The Carboniferous macrofossils were studied and named by Mr Mitchell, the conodont faunas by Mr Reynolds and the foraminifera by Dr W. H. C. Ramsbottom. Mr Mitchell has contributed to this memoir a chapter on the Carboniferous palaeontology. Dr Ivimey-Cook named the Mesozoic fossils. For specialist assistance in naming the ammonites, thanks are given to Professor D. T. Donovan (Lower Lias) and Dr M. K. Howarth (Middle and Upper Lias). Dr Ivimey-Cook and Professor Donovan have jointly supplied an Appendix on the Lower Lias ammonite faunas. Dr G. Warrington has contributed an Appendix on the Triassic and Jurassic palynology and other micropalaeontological work on the Mesozoic successions.

A special petrographical report on the Carboniferous volcanic rocks has been provided by Mr R. W. Sanderson, who re-examined material in the Institute's sliced rock collection and additionally studied rocks from his own collecting. He has drawn on unpublished work on the volcanic rocks of the Bristol area, including Weston-super-Mare, by Dr R. Dearnley, which included chemical analyses by Mr G. A. Sergeant. Other petrological descriptions were contributed by Dr J. R. Hawkes. A chapter on geophysical surveys in the area, written by Dr J. D. Cornwell, includes details of recent surveys by the Institute. The results of offshore surveys by the Institute's Marine Geology Unit have been provided by Dr B. N. Fletcher. Dr Whittaker has summarised the results of the Institute's boreholes at Brent Knoll, and Dr Ivimey-Cook and Dr Warrington have given additional details of these in their respective appendices. A fuller account of the boreholes is in preparation. The editing of the memoir was undertaken by Mr Bisson.

We gratefully acknowledge the courteous assistance of many individuals and firms and of officials of the Weston District Council in supplying information on boreholes, trial pits, quarries and temporary sections, and in giving facilities for their examination by the surveyors themselves. Professor C. Kidson kindly provided some scientific results from work being carried out by members of the Geography Department, University College of Wales, Aberystwyth.

Preparation of the memoir has been partially funded by the Department of the Environment.

G. M. Brown, Director, Institute of Geological Sciences, Nicker Hill, Keyworth, Nottingham NG12 5GG. 5 April 1983

Geology of the country around Weston-super-Mare–summary

This book is an authoritative account of the geology of the diverse country around Weston-super-Mare, including the coastal headlands of Middle Hope and Brean Down, the western end of the Mendip Hills, part of the Somerset Levels and the north Somerset coast between Hinkley and Watchet, as well as the islands of Steep Holm and Flat Holm. Miles of coastal cliffs provide magnificent exposures of the Carboniferous Limestone with some interbedded volcanic rocks in the north, and of Triassic and Lower Lias rocks in the south.

The Institute's boreholes at Brent Knoll, near the axis of the Central Somerset Basin, give a section through most of the Lower Jurassic strata and the greater part of the Triassic System, and numerous other boreholes in the Somerset Levels provide a picture of sedimentation around the Bristol Channel during the last 10 000 years. A summary of recent work by the Institute and others on the submarine geology of the upper reaches of the Bristol Channel is also included.

The memoir traces the geological history of the area, from the limestones of the Carboniferous formed in shallow tropical seas with abundant animal life, now preserved as fossils, through the barren red beds of the Permian and Triassic, deposited when a desert covered the area, to the richly ammonitiferous limestones and shales of the early Jurassic, which were laid down in a sea that inundated the district. On the flat lowlands of the Somerset Levels drift deposits conceal a varied undulating topography.

Geological sequence

The geological divisions shown on the Weston-super-Mare (279) Sheet and described in this memoir are listed below. Certain New Red Sandstone strata, not cropping out at the surface but proved in a borehole, are omitted from the sequence but are mentioned briefly in the text.

(Geological succession)

Superficial deposits (Drift)

Solid formations

Notes

• National Grid references are given in square brackets throughout the memoir. They all lie within the 100-km square ST (or 31).
• Numbers prefixed by the letter E refer to specimens and thin sections in the collections of the Petrology Unit, Institute of Geological Sciences.
• Numbers preceded by A refer to photographs in the Institute collections.

List of six-inch maps

The following is a list of the 6-inch geological maps included, wholly or in part, in the Weston-super-Mare Sheet with the date of survey of each map. The surveying officers are: E. A. Edmonds, G. W. Green, B. Kelk, J. C. Thackray, A. Whittaker, B. J. Williams.

Chapter 1 Introduction

The land area represented on the Weston-super-Mare (279) 1:50 000 geological sheet is situated in the counties of Avon and Somerset. A large part of the sheet area is offshore and forms the eastern extremity of the Bristol Channel known as Bridgwater Bay. The main towns, Weston-super-Mare, Burnham-on-Sea and Highbridge, besides being industrial centres, are holiday and tourist resorts and are served by excellent rail and road communications.

The landscape is for the most part low lying, but it is varied by more rugged terrain in parts of the north-east and by gently undulating country in the south-west. The higher parts of the north-east are the westerly continuation of the Mendip Hills, which become progressively lower westwards and form a series of hills or elongate east–west ridges which are, geologically speaking, inliers of Palaeozoic rocks (Figure 1). The 168 m OD height of Bleadon Hill is exceeded only in the south-west of the district, where the northern part of the Quantock Hills, composed of Devonian rocks, attains a height of about 240 m. The gently undulating country flanking the Quantock Hills averages perhaps 60 m OD and is underlain by Mesozoic rocks; it gives way to the lowest-lying part of the landscape, underlain by alluvial deposits with general surface about 6 m OD, which forms the coastal part of the Somerset Levels. Within the flat tract of the Somerset Levels, isolated hills composed of Jurassic strata relieve the visual monotony of the topography. The three physiographic divisions of high rugged ground, gently undulating country and flat-lying Levels, are thus seen to be intimately related to the outcrop of Palaeozoic, Mesozoic and Quaternary rocks, respectively. Fringing the west-facing coastal belt are blown sand deposits, seawards of which are extensive tidal flats up to 5 km wide. The islands of Flat Holm and Steep Holm at the mouth of the River Severn are the westward continuations of the Palaeozoic rocks (Carboniferous Limestone) of the onshore north-eastern part of the district.

The district is drained mainly by the rivers Parrett, Brue and Axe and by the man-made Huntspill River. The sluggish larger rivers meander over the flat-lying levels, although man's interference with the natural drainage has obliterated the original pattern, producing long, straight drains, which in many instances do not relate to the older watercourses at all. The levels are crossed by numerous ditches or rhynes which divide the land into rectangular fields. The huge tidal range (up to 12 m) experienced by the coastal strip is due to its position at the narrow end of the trumpet-shaped Bristol Channel. The height of high water of spring tides is about that of the general surface level of the alluvial flats, so that without interference by man there would be continual danger of extensive flooding. Indeed, much of the history of the district concerns the attempts made by man to control the environment, so making it habitable and productive.

Stone Age Man was probably active in the area, because Palaeolithic, Mesolithic and Neolithic remains are known from the limestone caves of the Mendip Hills. The burial mounds of the Mendip area suggest that a sizeable Bronze Age population settled on the low-lying levels and moors as well as the higher ground. A similar pattern of settlement persisted through the Iron Age, the period of the well-known lake villages of nearby Glastonbury and Meare. In the 1st century AD the occupying Romans were exploiting the mineral deposits of Mendip, and constructed artificial barriers along the coast to prevent incursions of the tide. Despite periodic catastrophic floods, the Saxons established villages behind the sea defences; the Domesday Book for the manor of Brentmarsh (Brent Knoll, East Brent, Lympsham and Berrow) shows that the number of cows was higher for this manor, and the ploughing activity less, than for comparable manors in adjoining areas, a pattern which has persisted to this day. After the Norman Conquest much of the low-lying area was brought under the control of the ecclesiastical authorities of Glastonbury and Wells, and it was possible to embark on more ambitious drainage projects which facilitated the construction of navigable waterways. The drainage problem hereabouts centres on the fact that inland of the coastal clay belt (surface about 6 m OD) lies an even lower-lying area of peat moors which are below the level of high spring tides. Drainage of the peat moors is further complicated by the tendency of the peat to shrink when dried, so that successful draining causes lowering of the land surface. The water in the rhynes empties into the main river channels via sluice-gates and after heavy rainfall the upland water is confined in high-level channels. This, together with large sluice-gates at the coastal end of the waterways to stop incursions of sea water, suffices to prevent excessive flooding, although under exceptional conditions flooding still occurs. The present drainage system, largely dating from the 18th and early 19th centuries, enabled reclamation and enclosure of large areas of peat moor.

Geological history

The Devonian rocks of west Somerset and north Devon, which impinge on to the area under discussion by way of the northernmost part of the Quantock Hills, comprise mixed fluviatile or deltaic deposits interbedded with fully marine sediments. They form an intermediate zone between the continental Old Red Sandstone facies of the area to the north and the Devonian marine facies to the south. In this sense the area under discussion can be considered a hinge zone during later Palaeozoic times.

The Hangman Grits, comprising grey quartzitic sandstones with plant fragments, siltstones and pebble beds and purple cross-bedded sandstones, were probably deposited under fluviatile conditions, the sediment being supplied from the north. They are sandwiched between marine deposits in the Quantock area. Within the overlying shallow-water marine Ilfracombe Slates is the Cockercombe Tuff, containing spilitic fragments and a fossil band indicating accumulation under marine conditions; the spilitic material was possibly derived from a submarine eruption to the south or east in Givetian times. The Morte Slates, the youngest Devonian strata exposed in the Quantock area, are shallow-water marine sediments and are probably of Frasnian age. Higher Devonian strata were proved in the nearby Knap Farm Borehole at Cannington Park (Whittaker and Scrivener, 1978, 1982), conformably underlying the Carboniferous sequence. These highest Devonian deposits of sandstones, shales and limestones (some of which are pebbly) contain marine fossils. In the northern part of the Weston-super-Mare district, but not exposed at the surface, the Upper Devonian is represented by the Portishead Beds, a thick sequence of sandstones with subordinate mudstones and pebbly sandstones, mainly red in colour, of Old Red Sandstone facies.

The relatively shallow-water sedimentation of the Upper Devonian gave way, in earliest Carboniferous times, to deeper water conditions in which the Lower Limestone Shale was deposited. Subsequently the seas shallowed to allow deposition of the main part of the Carboniferous Limestone sequence. Again the area under consideration was close to an important hinge zone, because the shallow-water shelf or platform carbonates give way to the south to basinal cherts, dark shales and limestones (including turbidites) of the Bampton district. The discovery of Waulsortian-type reef deposits in the Lower Carboniferous sequence of Cannington Park (Whittaker and Scrivener, 1978, 1982) provides evidence of a possible barrier or fringe between the platform carbonates to the north and the turbiditic sequence to the south. The situation is comparable with that in Ireland, where reef facies developed at or near the juxtaposition of 'Culm' beds and Carboniferous Limestone. Although no Upper Carboniferous rocks have been proved within the map area, evidence beyond the eastern margin of the map indicates that Upper Coal measures are probably present beneath the Mesozoic rocks in the north-eastern corner, between Middle Hope and Worle Hill. In this part of the district deposition was not continuous throughout Carboniferous times and it is probable that uplift and considerable erosion of the Carboniferous Limestone took place along the north-east to south-west trending Lower Severn Axis, in mid-Carboniferous times, prior to the laying down of the Upper Coal Measures. Within Upper Carboniferous times, the district appears again to have been a zone across which changes took place; the tract of land to the north was covered by the typical Upper Carboniferous deltas and swamps, whereas Upper Carboniferous sediments in the Cannington area are perhaps more akin to the basinal Culm measures farther south-west.

Towards the close of the Carboniferous period the area was convulsed by major earth movements during the Variscan orogeny. The district forms part of the Hercynian or Variscan fold belt, which can be traced from southern Ireland via south-west England into the mainland of Europe. Tremendous compressive forces from a general southerly direction folded and reverse-faulted the Palaeozoic rocks along dominantly east to west trending axes and possibly initiated an important system of northwest-trending strike-slip faults. The style of tectonics is at least partly, but probably considerably, related to the degree of competence of individual Palaeozoic formations and their tectonic level; this has given rise to complicated geological structures in the Palaeozoic rocks of the Variscan fold belt. In the northernmost part of the district a veering of the fold axes to a north-easterly direction marks the transition from the main Variscan fold belt to an area in which a Malvernoid (north–south) axial trend is dominant. There is evidence that a tensional phase of normal faulting occurred, possibly during or subsequent to uplift of the area towards the close of the Variscan orogeny.

The picture presented at the beginning of Permo-Triassic times, is of rugged, perhaps mountainous terrain, crossed by fault-bounded depressions receiving sediment from the active erosion of higher ground. One such depression, trending east-south-east, lies under much of the map area and continues westwards beneath the waters of the Bristol Channel and eastwards beneath the Glastonbury area. As New Red Sandstone deposition progressed, evidence suggests that normal faults were active, allowing the accumulation of greater thicknesses of sediment in the basinal areas but with successive overlap of younger over older sediments.

Thus the area of sedimentation was expanding as time went on. Deposits marginal to the eroded stumps of the Palaeozoic high ground, laid down during later New Red Sandstone times, are represented by the Dolomitic Conglomerate, which is contemporaneous with part of the red fine-grained basinal sediments of the Mercia Mudstone Group ('Keuper Marl'). Towards the close of Triassic times incursions of a sea from the west or the south gave rise to intertidal deposits and a sabkha environment was established prior to the major marine transgression recorded in the Penarth Group ('Rhaetic') sediments.

Throughout the Lower Jurassic much of the area was subjected to a fully marine environment. The youngest solid rocks present within the confines of the 1:50000 sheet comprise limestones of Middle Jurassic age which now form the top of Brent Knoll. Like earlier Mesozoic sediments, the Jurassic rocks were laid down over a region which sagged differentially, giving thicker deposits in the downwarped areas or troughs whose limits were established in preceding epochs. By analogy with adjacent regions, it is probable that marine sedimentation continued through most, if not all, of Jurassic and Cretaceous times and this was accompanied by some continued degree of extensional tectonics. Compressive earth movements in Tertiary times were probably particularly effective in areas with long-established zones of structural weakness, accentuating or complicating existing structures and possibly producing new ones.

Lack of information prevents the enumeration of detailed events during the Pleistocene. It is possible that the Bristol Channel and the Somerset lowlands were glaciated some time during the early part of the Quaternary. In later Pleistocene times the area suffered a periglacial climate, and extensive spreads of Head deposits were formed under freeze-thaw conditions. Also during Pleistocene times there were warmer periods when some of the water locked up in ice was released to raise sea level and give rise to beaches at higher levels than those of the present day, and possibly also to produce the sands and gravels now known as the Burtle Beds. The release of water, which had been locked up in ice during the last glaciation, produced the rise in sea level responsible for the so-called Flandrian transgression, which gave rise to the alluvial deposits which cover such a large part of the district. AW, GWG

Chapter 2 Devonian

Middle Devonian

Hangman Grits

Devonian rocks crop out in the south-western part of the district, to the west and south-west of the village of East Quantoxhead and make a strong topographic contrast with the surrounding gentle landscapes of the Mesozoic rocks. The main outcrop is the northernmost part of the Quantock Hills; smaller outcrops occur as inliers surrounded by Triassic rocks and comprise a partially exhumed Triassic topography.

The beds are all referable to the Hangman Grits, a thick sequence of arenaceous deposits of Old Red Sandstone type which is traceable from north Devon to west Somerset. The Devonian strata which underlie and overlie the Hangman Grits are not exposed within the district; they are marine sediments, and the Hangman beds comprise the earlier of two incursions of continental arenaceous material into the predominantly marine Devonian sequence of this region.

The 'latest detailed geological account of the Devonian rocks of the Quantock area is that by Webby (1965), and the following brief account is based largely on his work. Webby divided the Hangman Grits into the Triscombe Beds, a lower division comprising green, massive, quartzitic sandstones interbedded with subordinate grey, green and brown siltstones and green conglomerates, and the Hodder's Combe Beds, a higher division comprising maroon, cross-bedded sandstones, conglomerates and quartzitic sandstones. According to Webby, the Triscombe Beds are thicker than 457 m in this area and the Hodder's Combe Beds thicker than 304 m; respectively they may correspond stratigraphically with the Trentishoe Grits and the Rawn's Beds (and perhaps the Sherry Combe Beds) of north Devon. An estimated 385 m of these beds may be present within the present district. Both lithological divisions have yielded plant fragments, and the sedimentary features of the rocks suggest deposition under fluviatile conditions, with a source area possibly lying to the north. The beds are thought to be of Eifelian (Middle Devonian) age.

Details

Within the map area exposures are poor. Old quarries are present around the perimeter of the main outcrop but rarely reveal more than a few metres of rock. The workings [ST 1192 4256]; [ST 1217 4254] near Perry show purplish grey and green medium-grained and in places flaggy sandstones dipping to the north-east. Old Quarries [ST 1312 4234]; [ST 1315 4226] and sandstone pavements [ST 1335 4230]; [ST 1339 4222] near Smith's Combe reveal a few metres of mainly purple, medium-grained sandstone; similar rock types crop out in the small escarpment [ST 1389 4220] 500 m SSE of Town's End. Lying loose on the surface over some of the Devonian outcrop are small blocks of mainly grey and purple, in places gritty but mostly medium-grained sandstones, gritty pebbly sandstones similar to the Rawn's Beds of north Devon, and sandstones with poorly preserved fossil casts reminiscent of the Sherry Combe Beds of north Devon.

The Devonian inliers north of the main outcrop expose a little grey and purple medium-grained sandstone at Perry Gully [ST 1155 4321], Knowle Wood [ST 1266 4267] and near East Quantoxhead [ST 1270 4348]. AW

Upper Devonian

Portishead Beds

In the northern part of the district, Upper Devonian (Upper Old Red Sandstone) rocks, named the Portishead Beds, are inferred to incrop beneath the Mesozoic strata (see (Figure 18)) between Middle Hope and Flat Holm and south-west of Brean Down, whilst they crop out in the core of the main Mendip anticline (the Blackdown Pericline) a short distance to the east of the present district. These rocks comprise dull purplish-red feldspathic and commonly quartzitic sandstones, interbedded with green, red and reddish purple sandy shales and siltstones, micaceous fine-grained sandstones and, particularly in the lower part of the succession, with pebbly sandstones and sandy quartz-conglomerates. In the Blackdown Pericline (Green and Welch, 1965, p. 10) about 500 m of beds are present, with the base not seen. In the Portishead area (Kellaway and Welch, 1955, p. 6; Pick, 1964), immediately to the north of the district, they are about 250 m in thickness, with the Woodhill Bay Conglomerate (3 m thick) at the base unconformably overlying the Black Nore Sandstone (Lower Old Red Sandstone). The latter comprises some 450 m of beds (base not seen) essentially similar in lithology to the Upper Old Red Sandstone. GWG

Chapter 3 Carboniferous

Lower Carboniferous (Dinantian)

The Lower Carboniferous is represented by the Carboniferous Limestone which crops out on the mainland as discontinuous elongated easterly trending ridges or inliers surrounded by Mesozoic or younger rocks. Flat Holm and Steep Holm represent the continuation offshore of two of these ridges and, likewise, are surrounded by Mesozoic rocks (under the sea).

The sequence comprises a thick series of platform or shelf type limestones, of which the bottom and top are seen in the immediately adjacent areas of Broadfield Down and the Mendips to the east, though not in the Weston district itself. These rocks represent an important resource of high-grade limestone, although for environmental reasons quarrying is now much reduced in scale. Two minor volcanic episodes are represented within the sequence.

Brief history of research

The earliest Geological Survey maps of this and adjacent areas (Old Series Sheets 19 and 20; New Series Cardiff (263) Sheet) showed the Lower Carboniferous rocks divided by Strahan into the Lower Limestone Shale below and the Carboniferous or Mountain Limestone above. A major advance followed from Vaughan's (1905, 1906) classification of the Lower Carboniferous rocks in the Avon Gorge of Bristol into the Avonian Series, based on a succession of coral-brachiopod zones. Over the next two decades or so this zonal scheme was applied with enthusiasm by Vaughan himself and many other workers to the Lower Carboniferous rocks throughout most of southern Britain, though it suffered an increasing degree of modification in the process (for a summary see George and others, 1976). Application of Vaughan's zonal scheme to the rocks of the Weston district and adjacent areas was made by Sibly (1905; 1906), Reynolds and Vaughan (1911), Bamber (1924), Bush (1929) and Smith and Willan (1937). The Carboniferous volcanic rocks in the Weston–Bristol area received detailed attention in a number of papers between 1898 and 1917, although the determination of their precise stratigraphic positions waited on the completion of the present 6-inch survey.

The demise of Vaughan's zonal scheme was heralded by the publication by Ramsbottom (1973) of a suggested scheme of correlation of the Dinantian rocks of Britain based on the recognition of six widespread major transgressions, the latest of which coincided with the base of Vaughan's D₂ Zone. This hypothesis was expanded and emended by Ramsbottom (1979). Important parts of the succession, especially as compared with the succession in northern Britain, were shown to be missing in the Avon Gorge type succession (1973, p. 595, fig. 8). The transgressions approximately coincide with the bases of the six regional stages established by the Dinantian Working Party of the Stratigraphy Committee of the Geological Society of London (George and others, 1976) as a standard for correlation to replace Vaughan's zones. These stages as modified by Ramsbottom and Mitchell (1980) are used in this account, and their relationship to the lithological formations is given in (Table 1). The Weston sequence given in George and others (1976, fig. 4, column A) was based on the recent survey but has since been revised.

Stratigraphy

Within the individual inliers in the Weston district the Carboniferous Limestone is well exposed. However, owing to the combined effects of faulting, folding, pre-Triassic erosion and the widespread cover of later rocks, no complete sequence of the Carboniferous Limestone is seen in any one inlier. The overall Lower Carboniferous sequences (Table 1) have therefore to be pieced together by comparison both with adjoining inliers and with the adjacent districts, New Series Bristol (264) and Wells (280) geological sheets.

Two distinct, though related, sequences can be recognised. The northern sequence, the thinner and less complete of the two, is built up by combining the successions found in the Middle Hope and Weston–Worle inliers. Part of a rather similar succession can be seen on Flat Holm. The southern, thicker and more complete, sequence is that found in the main mass of the western Mendips (Green and Welch, 1965). Connecting links between them are provided by successions seen on Brean Down and Steep Holm. A summary is given in (Table 1), the succession at Burrington Combe being included for purposes of comparison.

The estimated overall thickness of the beds between the top of the Old Red Sandstone and the base of the Hotwells Limestone is of the order of 450 to 500 m for Broadfield Down and the Avon Gorge, 700 to 750 m for the Weston sequence (after making an allowance for the unexposed part of the sequence), 800 m for Burrington Combe and around 900 to 930 m for the western Mendips. This southward and westward increase in thickness accords well with current knowledge of Lower Carboniferous sedimentation in the south-western Dinantian province (e.g. George, 1972).

Black Rock Limestone and older rocks

Although the Lower Limestone Shale is not exposed in the Weston district there is no reason to suppose that it is not widely present at depth. The formation comprises grey-green to dark grey and nearly black shales with interbedded limestones. The thickness ranges from about 150 m in the western Mendips (Blackdown Pericline) to about 100 m in the Clevedon area. An intermediate thickness of 120 m is suggested for the Middle Hope area, in conformity with the behaviour of the overlying formations.

The Black Rock Limestone consists predominantly of grey to nearly black medium-grained to fine-grained fossiliferous limestones with abundant crinoidal debris. The colour varies according to the crinoid content, being lightest where this is most abundant and coarsest in grain, and vice versa. The most complete section is in the cliffs and foreshore of Middle Hope, where the formation comprises well-bedded limestones alternating, at intervals of roughly 10 m, with massive posts from 1.5 to 10 m thick, thus giving a rhythmic appearance to the sequence. A sequence 5 to 9 m thick with abundant chert nodules and silicified fossils forms a recognisable marker at about the middle of the succession. Volcanic rocks occur just above this chert level (see below). From faunal evidence (p. 21) it is unlikely that more than 20 to 30 m of Black Rock Limestone are present below the lowest exposed beds. The upper limit of the formation is taken at a prominent bedding plane marking the top of a 7-m-thick bed of grey crinoidal limestone. The uppermost 0.8 m of the bed is completely dolomitised.

There is no exposure of Black Rock Limestone between Middle Hope and the Mendips. In the western Mendips the higher part of the sequence, which is absent at Middle Hope (see below), shows a general downward passage from more massive paler crinoidal limestones into the very dark fine-grained well-bedded types, for example in the sequence at Uphill Quarry (Plate 2). Exposures in the lower beds are insufficient to indicate whether the rhythmic sequence noted at Middle Hope is present. The median chert is a feature of the succession throughout the western Mendips, where it attains some 25 to 30 m in thickness, with well-marked sheets of chert as well as numerous nodules and silicified fossils. Although the chert forms the lowest part of the Black Rock Group succession exposed in the western Mendips of the district, it is instructive to compare the whole succession, compiled from the adjacent areas on the Wells (280) Sheet, to the Middle Hope succession:

The main difference is in the limestones above the chert level, where faunal evidence (p. 22) indicates that an appreciable thickness of beds is missing at the top of the Middle Hope sequence. These beds above the main chert are approximately 30 to 40 m thicker in the western Mendips than at Burrington Combe.

Black Rock Dolomite

The name 'Black Rock Dolomite' denotes that the formation is equivalent to part of the Black Rock Limestone of neighbouring areas. The formation is completely exposed at Middle Hope, where it comprises medium-grained to fine-grained grey dolomite, well-bedded near the base and massive above, which weathers to a pale grey to off-white colour with a characteristic criss-cross pattern of cracks. Fossils are not uncommon and are mainly in the form of calcite-filled moulds. Detailed fossil collecting shows that the junction of the Caninophyllum patulum and Siphonophyllia cylindrica biozones (= Middle and Upper faunas) of the Black Rock Limestone (p. 22) lies at about the middle of the Black Rock Dolomite at the western end of the Middle Hope peninsula. Judging from the fossil remains, the dolomite appears to have been originally a fossiliferous crinoidal limestone similar to the coarser varieties of Black Rock Limestone. This suggestion gains support from the thickness variation across the peninsula—30 m in the west and 45 m in the east—which is inversely related to that of the underlying Black Rock Limestone. The upper surface of the formation is very sharp and relatively smooth. It is surprising that, in spite of the faunal evidence for a considerable discontinuity, there is no clear evidence of erosion at this contact.

In the Mendips, east of the longitude of Hutton, the Black Rock Dolomite is similar in thickness and facies to that of Middle Hope. However, large corals and brachiopods typical of the cylindrica Biozone of the Black Rock Limestone (p. 36) are locally preserved, showing that the Dolomite lies higher stratigraphically than it does at Middle Hope. West of the longitude of Hutton there is a massive increase in dolomitisation, which may extend upwards to some 25 m above the base of the Birnbeck Limestone, and locally it is impossible to distinguish the Black Rock Dolomite sensu stricto, which has therefore to be included with the overlying dolomites. The maximum thickness of dolomite at Bleadon is about 140 m, while it may be even more at Brean Down where the base is not seen. Mapping eastwards from the Fort at the western end of Brean Down, lateral passage of the Gully Oolite and the lower massive facies of the Birnbeck Limestone into dolomite can be proved. Thus, hereabouts, the top 55 m of the dolomite represent altered Clifton Down Group limestones and hence only the remaining 85 m or so can represent Black Rock Dolomite sensu stricto.

Gully Oolite

The Gully Oolite at the bottom of the Clifton Down Group is one of the most easily recognised formations in the succession. For fuller details and discussion of the lithological variations within the formation the reader is referred to the account by George (1978a) of its occurrence in Gower (there named the Caswell Bay Oolite), where its character appears to be fully comparable to that of the Weston district. In general terms its most distinctive features are the homogeneous texture, typically comprising evenly-graded rather coarse ooliths set in a sparry matrix, the pale grey colour weathering to conspicuous white at outcrop, the massive jointed habit with few or no bedding planes but with finely cross-bedded texture seen on weathered joint faces, and the sharp, well-defined base and top. It is only rarely fossiliferous. The base is exposed only in the western part of the Middle Hope peninsula, where it is very sharp and nearly flat. Hereabouts the lowest 1 to 3 m of the formation are fossiliferous and include many bands with abundant crinoidal debris similar to the 'Sub-oolite Bed' of the Bristol area (Kellaway and Welch, 1955, p. 11). At Brean Down, the southern limit of its known occurrence, the Gully Oolite includes lenses of well-sorted very coarse crinoid ossicle debris and scattered corals in the uppermost third of the formation. Lateral passage southwards into grey massive coarse crinoidal limestones occurs within the district and must be abrupt, but the details of the transition are obscured by extensive dolomitisation (p. 35).

The top of the Gully Oolite is well exposed at many places on Flat Holm and less spectacularly at Spring Cove, Weston, and at the western end of Brean Down. In all cases it shows a marked hiatus with the beds above. On Flat Holm, the uppermost part of the Gully Oolite is closely comparable with the Heatherslade Bed in the same position in Gower, which was described in detail by George (1978a), who considered it to represent an intertidal algal-mat beach deposit. On Flat Holm the bed consists of a dark grey to nearly black micrite 'crust' some 50 to 60 cm in thickness, with fenestral structures and scattered ooliths and gastropods. There is a passage downwards into 'normal' oolite. Micritisation extends downwards very irregularly and in places micrite fills anastomosing cracks in the oolite below. The uppermost part of the crust shows good evidence of penecontemporaneous brecciation. On Brean Down the uppermost part of the Gully Oolite comprises highly altered dolomite-mudstone or siltstone and may represent the altered remnant of a micrite crust similar to that on Flat Holm. At Spring Cove, the top 3 m or so of the Gully Oolite are strongly fissured, with dolomitic material apparently from the overlying beds filling the fissures or lining their walls. This may partly represent a penecontemporaneous karstic surface, though it has had a complex history, as shown by the extensive reddening and dolomitisation associated with many of the fissures and their fillings. The alteration may be associated with either the overlying Spring Cove volcanic rocks (p. 30) or a nearby Triassic land surface, or, most probably, both.

The complete thickness of the Gully Oolite is nowhere seen, though it may have been about 35 m at Brean Down, where the lowest part has been completely altered to dolomite (p. 35). At both Middle Hope and Flat Holm an incomplete thickness of 37 m was measured. The thickness at Broadfield Down, to the east of the Weston district, is 37 m.

Caswell Bay Mudstone

The name 'Caswell Bay Mudstone' is applied to a varied group of fine-grained sediments and is taken from the type locality in Gower (One-inch Swansea Sheet 247, 1972). At the time of the survey the rocks at Weston were called Clifton Down Mudstone by the author (George and others, 1976, fig. 4) or Lower Clifton Down Mudstone (Ramsbottom, 1977, fig. 2), but in view of the uncertainty of their relationship to the much thicker Clifton Down Mudstone of the type section in the Avon Gorge (Kellaway and Welch, 1955, p. 12) the new designation is now thought to be preferable. These beds are best exposed on Flat Holm, where they vary in thickness from 3.7 to 5.3 m, and the lower half of the sequence includes well-marked calcitemudstones (chinastones) with fenestral structures and desiccation cracks. The upper part of the sequence here comprises finely interbedded dolomitic mudstones, siltstones and shales with scattered laminae and thin lenses of bioclastic limestone containing crinoid debris and some ooliths. Many levels are strongly burrowed. On the mainland at Spring Cove they are dominantly composed of laminated algal dolomitic mudstones about 3 to 4 m in total thickness, but they attenuate southwards and are represented by only 1.5 m of dolomitic shale and mudstone at the western end of Brean Down. They are absent elsewhere in the western Mendips and apparently also on Steep Holm.

Birnbeck Limestone (new usage)

The term 'Birnbeck Limestone' is applied to grey to dark grey bioclastic limestones which lie between the top of the Caswell Bay Mudstone and the bottom of the Goblin Combe Oolite. The type section extends along the coast from Spring Cove to the foot of the steps of the Royal Pier Hotel near the Birnbeck Pier. Two main facies^‡1  are here recognised, a lower group of grey, very massive, finely cross-bedded, rather unfossiliferous limestones, commonly oolitic (beds B3 to B11, pp. 29–30), and an upper group of grey to dark grey, bioturbated, more fossiliferous, well-bedded limestones (beds B12 to B14; (Plate 3)). The individual beds in the lower part range up to over 8 m in thickness, and those of the upper part are from 0.3 m to rarely more than 1 m thick. The commonest fossils are Palaeosmilia murchisoni, Siphonophyllia garwoodi, Bellerophon sp., and brachiopods, notably Delepinea carinata. Thin dolomitic siltstone bands occur at intervals in the succession associated with both facies. The sequence in the type area includes a basaltic lava flow at Spring Cove (Plate 6) which extends eastwards for some distance inland. Reddening of the limestones for 12 m below the lava is due to their ash content (see pp. 15–16), and the reddening for 9 m above the lava may have the same cause. On the mainland the broad sequence of the two facies (Figure 2) is maintained consistently, with the thickness of the lower massive part ranging from 30 m in the Weston–Worle area to 20 m on Brean Down, and that of the upper well-bedded part from 43 to 45 m in the Weston–Worle area to 41 m at Brean Down and 37 m at Steep Holm and Bleadon. Preferential dolomitisation of the massive facies is widespread, particularly in the western Mendips and on Steep Holm. The relationships of the facies of the Birnbeck Limestone and the adjacent formations are discussed below (pp. 24–27). The formation has not been recorded from Broadfield Down.

The Birnbeck Limestone is recognised on Flat Holm, but there is a marked change as compared with the mainland both in terms of facies and in the order of superposition of the facies. Interbedded with alternations of both the massive and the well-bedded facies of the type section occur six discrete sequences of finely banded dolomitic mudstones, siltstones and shales of a facies like that displayed in the upper part of the Caswell Bay Mudstone (see above). The alternating limestone-mudstone succession, which measures 31 to 32 m in thickness, is here given the name Flat Holm Limestone Member (Plate 4). The beds above the member show a passage from the massive facies below to the well-bedded facies above. Detailed correlation of the Flat Holm sequence with the mainland is not possible, with the exception of a well-marked band containing distinctive large caninoid corals (bed D5, p. 32), which is also recognised in the type section of Flat Holm (bed B6, p. 30).

Goblin Combe Oolite

The Goblin Combe Oolite, named from the type area on Broadfield Down, consists of pale grey to grey, medium-grained to coarse-grained, oolitic-crinoidal limestones, individual beds varying from oolites with scattered crinoid debris to pure crinoidal limestones. The full sequence is exposed in the Weston–Milton area, where it measures about 70 m in thickness. Where it is exposed on the coast or in large quarries the lowest part of the formation is seen to consist of a single large post of limestone commonly from 8 m to as much as 18 m or more in thickness (as on Steep Holm for instance). The lower part of the basal bed comprises well-sorted coarse crinoidal limestone but this passes upwards into an increasingly oolitic facies. Under suitable conditions of weathering cross-bedding is conspicuous. Fossils are mainly restricted to scattered specimens of P. murchisoni. At Anchor Head, on the coast, an 18-m-thick variably dolomitised horizon is present in the middle of the succession.

Burrington Oolite

It has been noted above (p. 8) that in the Brean Down–Bleadon area dolomitisation has obliterated all the strati-graphical details of the succession beneath the Birnbeck Limestone. East of Hutton, however, dolomitisation on both limbs of the Mendip anticline is much reduced and appears to be confined to the Black Rock Dolomite sensu stricto but, perhaps partly owing to lack of exposure, it is not possible to recognise the Weston-type sequence and the beds are all grouped together as Turrington Oolite'. At Hutton Combe [ST 361 585], for instance, roadside exposures show that massive grey crinoidal limestones directly overlie the Black Rock Dolomite in the position occupied by the Gully Oolite in the rest of the Weston district. Higher beds comprise pale grey to grey crinoidal-oolitic limestone with appreciable thicknesses of pure oolites, much of which must correspond to the Goblin Combe Oolite. A similar situation holds in the southern limb of the anticline east of Shiplate, just beyond the eastern margin of the Weston district. The calculated thickness of the Burrington Oolite is of the order of 230 to 240 m on either side of the anticline. An indication of the presence of beds equivalent to the Birnbeck Limestone in the western part of the Wells (280) district is given by the presence of P. murchisoni and D. carinata (Welch in Green and Welch, 1965, p. 34) about 100 m above the base of the Burrington Oolite in the Crook Peak area.

Clifton Down Limestone and younger Carboniferous rocks

In the Weston–Worle area, the lower part of the Clifton Down Limestone is everywhere a group of thinly bedded laminated calcite- and dolomite-siltstones and mudstones with interbedded bioclastic and oolitic beds, which varies from 11.4 to 13.4 m in thickness. These beds are apparently in a similar stratigraphical position to the Rib Mudstone of Burrington. The rocks weather to a pale pinkish grey to fawnish grey colour and include strongly burrowed levels. Where fresh, as in Worle Hill Quarry, they are dark grey and black in colour and the individual beds tend to be nodular in form with alternating darker and lighter coloured layers or laminae; clear evidence of algal structure is lacking although Dr J. R. Hawkes, commenting on a thin section (E36614) of these rocks, stated that 'linear texture (arrangement of dolomite and calcite crystals) in the darker zones contrasts with the isotropic texture of the dolomite micrite bands suggesting the possibility of biological control'. These basal beds of the Clifton Down Limestone are closely comparable in lithology with both the upper part of the Caswell Bay Mudstone on Flat Holm and the micritic facies of the Flat Holm Limestone Member of the Birnbeck Limestone, much lower in the Carboniferous Limestone sequence (pp. 9–10). In the coastal sequences at Anchor Head and north of Weston Woods a higher group of similar thin-bedded dolomitic rocks (up to 11 m thick) also occurs, the intervening beds, comprising about 19 m of massive oolitic granular bioclastic limestones, are much dolomitised. Inland, the sequence above the basal micritic beds passes into a more oolitic facies and the higher horizon of fine-grained beds appears to be absent. The Clifton Down Limestone above this lowest 40 m or so of mixed facies passes up into more uniform massive grey to pale grey current-bedded oolites. These are 75 to 80 m thick, of which the top 12 to 18 m can be separately mapped because of the presence of interbedded pale grey calcite-mudstone bands and pellet beds. The highest exposed part of the Clifton Down Limestone comprises non-oolitic granular bioclastic limestones with Lithostrotion martini which, above the lowest 10 m, occurs as large masses, usually silicified and associated with chert nodules. Above these beds it may be assumed by analogy with the Mendips and the Bristol area that the remainder of the Clifton Down Limestone, if not removed by intra-Carboniferous erosion, comprises the calcite-mudstone facies (called 'Concretionary Beds' at Bristol) which everywhere marks its top. The nearest exposures of Carboniferous rocks to those of the Weston–Middle Hope area form part of the Clevedon–Portishead ridge, to the north-east of the Weston district, where much of the Carboniferous Limestone has been removed by intra-Carboniferous erosion along the line of the NE–SWtrending Lower Severn Axis (Kellaway and Welch, 1948, fig. 8). There is evidence (p. 95; (Figure 18)) that the southwesterly prolongation of the Lower Severn Axis lies between the Weston–Worle ridge and Middle Hope, and it is probable that Upper Coal Measures (Pennant Measures) are present beneath the Mesozoic cover here and that it oversteps northwards on to successively older formations of the Clifton Down Group. The Hotwells Limestone is presumed to incrop north of the Mendips and to be overstepped at about the latitude of the Weston–Worle ridge (Figure 18) and (Figure 19).

In the Mendip area of the district the greatest thickness of exposed Clifton Down Limestone is around Hutton, on the northern limb of the main (Blackdown) anticline. The basal beds, comprising pale grey splintery porcellanous oolites and subsidiary calcite-mudstones, apparently pass up into medium- to coarse-grained dolomites, the combined thickness (shown on the map as dolomite) being about 75 m [ST 359 591]. The dolomites, which are poorly exposed, probably represent secondarily altered granular and oolitic limestones comparable to those represented in the lower part of the Clifton Down Limestone farther north at Weston. These are overlain by massive, pale grey to white, coarse oolites, the lowest 30 m of which are exposed [ST 359 592], and which are comparable to those in the higher parts of the exposed Clifton Down Limestone sequence at Weston. The Lithostrotion beds are present at outcrop in the Mendips, a short distance to the east of the present district, and were also proved in a borehole at Weston Gasworks [ST 329 607] between the Mendips and the Weston–Worle inlier. The remaining part of the Clifton Down_. Limestone sequence and the succeeding Hotwells Limestone have not been proved in the Mendips of this district but are presumed to incrop beneath the Mesozoic cover in continuation of their occurrence farther to the east (Figure 18).

Volcanic rocks

History of research

The presence of igneous rocks in the Carboniferous Limestone of the Weston district has long been known (see historical summary by Morgan and Reynolds, 1904a, pp. 137–139). They were shown as intrusive 'traps' on the Old Series geological maps, but their volcanic origin, although previously suspected, was only firmly established in 1896 during the 6-inches-to-one-mile primary survey in the south-eastern corner of the Cardiff (263) Sheet. The first detailed field descriptions were given by Geikie and Strahan (1899) and these were followed by a fuller account, including petrographical descriptions, by Morgan and Reynolds (1904a). During the same year (1904) Boulton gave additional details of the lava at Spring Cove, and Morgan and Reynolds (1904b) provided a collation of all the field observations up to that time. In 1905, Sibly showed that the volcanic rocks occurred at two distinct horizons separated by a considerable thickness of sediments, and not at one level as had been thought hitherto. Reynolds's final work (1917) collected together observations made since the earlier papers and added new information gained from trial pits, together with a number of chemical analyses. The present survey has established the exact stratigraphical positions of the volcanic rocks. Petrographical work by Dr R. Dearnley, combined with chemical analyses by Mr G. A. Sergeant, has provided evidence for late-stage auto-metamorphism by the residual fluids of the lavas of the upper volcanic horizon at Weston and the approximately contemporaneous volcanic rocks of Goblin Combe (Broadfield Down). Finally, petrographical work by Mr R. W. Sanderson (pp. 15–18) has demonstrated that the reddening of the limestones at Spring Cove for 12 m below the lava is due to the weathering of contained tuffaceous material.

[Description]

Middle Hope

The volcanic beds at Middle Hope are seen in four exposures on the north coast of the peninsula, which were numbered 1 to 4 from west to east by Morgan and Reynolds (1904a, b). Their stratigraphical position is accurately fixed in relation to the main chert beds of the Black Rock Limestone which they directly overlie. The total thickness of the volcanic beds decreases from about 32 m at exposure 1 (Swallow Cliff), to 27.5 m at exposure 2, apparently rather less again at exposure 3 (total thickness not known), to at least 2 m but possibly as much as 8 m at exposure 4.

The succession is largely pyroclastic, with tuff beds of brown, red and green hues, usually with abundant lapilli of highly vesicular pumice which may be either ungraded or graded into coarse and fine layers locally showing cross-bedding. Extensive net-veining by calcite is characteristic (Plate 5). The succession is interrupted by fossiliferous limestone beds at intervals which range in thickness from 1 or 2 cm up to 3.7 m. The tuff beds themselves commonly contain marine fossils and show gradations through calcareous tuffs and nodular tuffaceous limestones into more normal limestones. Thinly interbedded tuffs and tuffaceous limestones are best seen in the lowest part of the sequence. Only exposure 1 shows lava, a much decomposed basaltic 'pillow lava' some 3 to 4 m in thickness. This fact, taken together with the general decrease in thickness of the volcanic rocks when traced eastwards, has led to the conclusion that the centre of volcanic activity lay somewhere to the west of this locality.

The volcanic beds are less resistant to weathering than the surrounding Carboniferous Limestone; thus, their outcrop is marked inland by a hollow and on the coast by bays or re-entrants between headlands or projections of the limestone.

Spring Cove, Weston-super-Mare

The Spring Cove basaltic lava is well exposed on the coast at Spring Cove (Plate 6). Inland, there are numerous small exposures and much surface brash can be seen, or has been recorded, in the Milton Hill area. Its stratigraphical position is clearly fixed on the coast and elsewhere in relation to well-defined mapping units (pp. 29–30). It is now calculated to lie some 145 m higher in the Carboniferous Limestone sequence than the Middle Hope Volcanic Beds–a thickness more than double previous estimates (e.g. Strahan and Cantrill, 1912, p.30). The thickness of the lava appears to be variable: on the coast it is about 15 m, but careful inspection on the ground over much of Weston Woods has failed to prove its presence and it must be very thin or absent. North of Milton (or Butt's) Quarry, and in the area of the cross-roads on Milton Hill, calculations based on width of outcrop and local dips suggest a figure of at least 30 to 35 m, whereas a borehole in the bottom of Milton Quarry [ST 336 627] proved a total thickness of 12 m only a short distance to the south.

The exposures at Spring Cove were fully described by Geikie and Strahan (1899) and Boulton (1904). In summary, the rock is a massive highly amygdaloidal pillow lava in which pieces of limestone and calcareous matter are commonly caught up and are in all stages of alteration and assimilation (Plate 6). The lower contact of the flow with the underlying limestones is gently undulating and cuts across the sediments in such a way as to show that the latter were not fully consolidated at the time of eruption. Boulton (1904) suggested that the rock mass consisted of lenticular sheets of lava and fluxion tuffs and agglomerates representing minor flows sloping from north to south, and from this deduced that the vent lay somewhere to the north. Geikie (in the discussion) doubted the explosive origin of any of the material in the flow. Re-examination of the rocks on the ground has shown a confused picture, in which Boulton's subdivisions are not clearly discernible.

The limestones are reddened for a vertical distance of some 12 m below the lava and Geikie and Strahan (1899) attributed this to the presence of 'fine volcanic dust'. Morgan and Reynolds (1904b, p. 202) found no clear evidence of the presence of ashy fragments in the limestone below the basalt but noted that ashy particles were plentiful above it to a height of 8 ft (2.4 m). They did not, however, offer any alternative explanation for the reddening, which is unknown from this part of the succession elsewhere in the district. Re-examination of thin-section material by Mr R. W. Sanderson has confirmed Geikie and Strahan's contention (p. 16).

The basaltic lavas and tuffs at Goblin Combe, Broadfield Down, and the basalt at Cadbury Camp and Tickenham (Kellaway and Welch, 1955, pp. 15–16) occur within the Clifton Down Mudstone approximately mid-way between the base of the Goblin Combe Oolite and the top of the Gully Oolite (Figure 4). Making due allowance for facies changes (p. 26), it is evident that these occurrences must have been approximately contemporaneous with the Spring Cove volcanic rocks. It follows that volcanic activity southwest of Bristol occurred over quite a wide area and that several small vents were involved.

Uphill

A small occurrence of amygdaloidal basalt or dolerite conformable with the enclosing sediments was discovered during the cutting of the (then) Great Western Railway line at Uphill [ST 326 579]. Reynolds (1917, p. 26) estimated the thickness of the bed as about 4 ft (1.2 m). The bed probably lies at about the top of the main chert horizon of the Black Rock Limestone, and corresponds to the horizon of the volcanic rock occurrences at Middle Hope. Unfortunately it was not determined whether the rock was intrusive or extrusive, and no exposure can now be seen. No trace of ash or lapilli was found in the limestones above or below the bed (Morgan and Reynolds, 1904a, p. 146; 1904b, p. 212). Mr R. W. Sanderson considers that the balance of evidence favours an intrusive origin for the rock (see pp. 16–17). GWG

Petrography of the (Carboniferous) volcanic rocks

Middle Hope

The dark greenish grey basalt from below Swallow Cliff, Middle Hope [ST 325 661] (exposure 1, p.27) is a sparsely amygdaloidal porphyritic olivine-basalt in which little, apart from the texture, remains of the original material. Thin sections (E16522), (E52330), (Plate 7), fig. 1) show yellowish, sericitised laths of albite measuring up to 0.27 mm in length, invested with and largely replaced by K-feldspar. Small discrete areas of pale clay-mineral represent original pyroxene grains. Opaque ore is abundant, occurring as minute (0.009 mm) octahedra, either solitary or commonly as skeletal and linear aggregates. Granular secondary sphene is also plentiful. Clay and calcite-chlorite pseudomorphs show shapes which suggest that olivine and pyroxene phenocrysts were originally present. Sparse ovoid amygdales are lined with sparry calcite or opaque ore and filled with pale green clay. The chilled margins of the amygdales are crowded with acicular crystallites of magnetite and sphene which exhibit areas of parallel growth.

The soft green tuffs below and above the lava ((E52306) and (E52305) respectively) are similar in composition, being accumulations of angular pyroclasts in an abundant clay matrix, the whole probably of pyrogenetic origin. These tuffs (beds A8 and A12, p. 27), possibly Geikie's beds 7 and 11 respectively (Geikie and Strahan, 1899), are greyish olive (near 10Y 4/2) to light olive (10Y 5/4) where the clay matrix is abundant (E52306). Argillisation has obliterated much of the structure of the clasts, the smaller (about 0.16 mm to 1.2 mm) being largely homogeneous aggregates of a pale green clay-mica. With increasing size, however, the original scoriaceous nature becomes apparent, with the structure picked out by accumulations of hematite and goethite dust and leucoxene'. Larger fragments (up to about 15 mm) were evidently highly vesicular microporphyritic lavas with a glassy groundmass containing crystallites, and some show evidence of autobrecciation. The matrix is usually a pale green clay or carbonate. The lower tuff ((Plate 7), fig. 2) has a matrix of spherulitic clay. X-ray diffraction analysis of the insoluble material from specimens (E52305) and (E52306) by Mr R. J. Merriman shows that the clay fraction in both cases is composed of a mixed-layer chlorite/ vermiculite mineral with subordinate amounts of illite. From microscopic evidence, the illite is largely in the clasts and the mixed-layer mineral in the matrix.

From a lower bed (?Geikie's bed 4, (E52307); ?bed A6, p. 28) a brown-tinted olive-grey (5Y 4/1) fine-grained rock shows some admixture of calcareous sediment and a matrix partly of sparry calcite and partly of clay with a granular appearance. Argillised pyroclasts may be partly replaced by carbonate, but the vesicular structure remains in evidence by the distribution of opaque dust. Shell fragments and crinoid ossicles show marginal replacement by chalcedonic silica and orthoclase. The latter occurs as euhedral crystals up to c. 0.2 mm in length, in which growth zones are marked by inclusions of residual calcite granules.

Samples of tuffaceous rocks ((E11422), (E16524), (E16525), (E16533), (E16534)) from the exposures along the shore eastwards towards Woodspring Bay show a reduction in the size of the pyroclasts and a greater admixture with sediment. The volcanic fragments also tend to be more finely vesicular or pumiceous than those in the rocks at Swallow Cliff.

Spring Cove

At Spring Cove [ST 309 625] a somewhat uneven or pillow-like flow of fine-grained porphyritic lava ((E3212), (E16531), (Plate 7), fig. 3) was described by Teall (in Strahan and Cantrill, 1912, p. 31) as 'a fine-grained, chocolate-coloured rock, composed of pseudomorphs after olivine and probably augite, in a groundmass showing microlitic structure. The phenocrysts are represented by pseudomorphs in carbonate. The microlitic feldspars of the groundmass are colourless, but they no longer show their characteristic optical properties. The groundmass is deeply stained with ferric oxide. Although all the minerals have been destroyed, the structure has been perfectly preserved and there can be no doubt whatever that the original rock was an olivine-basalt.' Further examination shows that the feldspars form laths, rarely longer than 0.8 mm, exhibiting a somewhat swirling, flow-induced orientation. Together with calcite, the pseudomorphs after ferromagnesian phenocrysts also usually contain patches of clay. Many of the phenocrysts exhibit slightly corroded euhedral outlines and some attain 2.9 mm in length.

The 'variolitic' basalt at Spring Cove ((E16517), (Plate 7), fig. 4) described and figured by Morgan and Reynolds (1904a) is a scoriaceous rock, pale red (5R 6/2) and finely mottled dark reddish brown (10R 3/4). The dark mottlings are ovoid, conjoining vacuoles filled with generally strongly iron-stained calcite in which float 'U'-shaped vermicular clay aggregates. No certain varioles are to be seen, but groups of angularly ovoid pseudomorphs of iron-stained clay after a ferromagnesian mineral ((Plate 7), fig. 4) may be what Reynolds (1917) described as '... a group of imperfect varioles'.

Associated with the lava at Spring Cove are tuffaceous limestones which are stained deep red. Specimens from a short distance above the lava ((E3204), (E16515), (E52304)) are composed of subrounded argillised fragments of fine-grained basalt set in a matrix of calcite spar, and probably represent erosion products from the lava. Below the lava the sediments are strongly iron-stained for some 12 m. They have an 'ashy' appearance in hand specimen but the tuff content reported by Geikie and Strahan (1899) was disputed by Morgan and Reynolds (1904a), Boulton (1904) and Reynolds (1917). A series of specimens ((E3199), (E32012), (E16514), (E16530), (E52242), (E52243), (E52244), (E52245), (E52303)) from the red rocks below the lava all contain iron-stained subangular to rounded pumice fragments ((Plate 7), fig. 5). The pumice is generally devitrificd, and replaced either by a clay mineral or carbonate where not opacified by iron oxides. Rarely the pumice is preserved in a green palagonitic material (E52244). Together with the pumice, abraded crinoid ossicles, micrite pellets and ooliths, and rare grains of feldspathic sandstone are present, all embedded in a sparry carbonate cement, commonly with patches of colourless clay. X-ray diffraction studies by Mr R.J. Merriman on the insoluble residue from specimen (E52242), 6.0 m below the basalt, indicate a vermiculite/chlorite mixed-layer mineral and illite with hematite and goethite. In contrast to this, a specimen (E52304) of tuffaceous limestone from 0.5 m above the lava contains only illite/smectite mixed-layer minerals. Thin oolitic coatings of carbonate on the clastic fragments indicate that the pumice at Spring Cove has suffered marine transport.

Autochthonous development of orthoclase crystals with the adularia habit ((E52242), (E52243), (E52244), (Plate 7), fig. 5), similar to those mentioned from the Middle Hope tuffs, is a noticeable feature, confined to micrite pellets and ooliths in rocks with a high pyrogenetic content. Growth of the crystals appears largely to have occurred before the development of the final oolitic coating in which orthoclase is rarely found. Crystallisation appears to have been spontaneous as clastic nuclei are not seen.

Recrystallisation of the carbonate matrix to aggregates of brown ankerite in well-formed rhombs, with commonly minor proportions of granular calcite, is apparent in all the rocks.

Milton Hill

Morgan and Reynolds (1904a) and Reynolds (1917) described poorly exposed occurrences of porphyritic basalt and tuff, the eastward continuation of the Spring Cove lava, from the vicinity of Milton Hill crossroads. Most of the localities are now built over, but specimens preserved in the Institute's collection show the rocks from near Florence Cottage [ST 3428 6323] (E11433), Milton Hill [ST 3377 6294] (E11429) and [ST 3355 6299] ((E11432), (E16540)), and Butt's Quarry [ST 3360 6270] ((E36607); see p. 30) to be autometasomatised, more or less vesicular, porphyritic olivine-basalts. At the eastern end of Weston Woods, a former exposure [ST 3345 6286] (Reynolds, 1917, locality 97, Furze Close) provided relatively unaltered material. The rocks ((E11431), (E16539); (Plate 7), fig. 6) are fine-grained, glomeroporphyritic olivine-basalts with andesitic affinities. Subhedral phenocrysts of fresh, neutral tinted augite (0.3 to 0.85 mm) and completely argillised commonly euhedral olivine (<2.2 mm) are set in a hyalocrystalline groundmass with an intersertal texture. The lath-shaped feldspars of the groundmass are quite clear and fresh, with compositions near An₅₀ (β′ = 1.560 ± 0.002). The laths are usually <0.23 mm in length, but uncommon microphenocrysts attain 0.5 mm with a proportionate increase in width. Granular augite is abundant in the groundmass but unevenly distributed and mainly associated with a felsic mesostasis. The grains average 0.02 mm in diameter and are aggregated in larger clumps. Interstices are filled either with pale green clay, representing original glass, or with turbid brown felsic material. Minute inclusions prevent certain optical identification of the latter, but, as its refractive index lies between those of Canada Balsam and the andesine-labradorite laths, it is probably andesine. Accessory opaque ore occurs as needle-like crystals or linear aggregates.

Slightly more altered samples ((E11419), (E16519)–(E16520)) exhibit replacement of pyroxene by carbonate and marginal K-feldspathisation of the plagioclase. Also present in these rocks are prismatic crystals of primary pale brown hornblende. Well-crystallised pseudomorphs of bastite after pyroxene are present in the rock from Milton Hill (E11429) and, to a lesser extent, also at Butt's Quarry (E36607).

A fragment of calcareous tuff (E16521) from between Milton Hill and Kewstoke [about ST 3366 6307] is composed of ?accretionary lapilli of ferruginous microspar and fragments of carbonated basalt set in a coarse calcite spar matrix. Uncommon angular grains of quartz occur in the lapilli and matrix.

Uphill

The basalt formerly exposed in the railway cutting near Uphill Station [ST 326 579] is a completely altered amygdaloidal olivine-basalt ((E672), (E1648), (E7672), (E11430), (E16528)–(E16529), (E16532)). In texture and mineralogy this rock appears to have been similar to that of Milton Hill, but is of coarser grain.

Reynolds's (1917) investigation suggests that the basalt is no more than 1.2 m thick. In combination, the recorded features of the occurrence-relatively coarse grain in relationship to thickness compared with similar rocks at Milton Hill and Middle Hope, and apparent absence of associated tuffaceous rocks-suggest that, although the basalt is strongly amygdaloidal, it may be intrusive rather than a lava flow.

Chemistry of the basalts

Complete chemical analyses of the variolitic basalt from Spring Cove and partial analyses (for silica and alkalis) of samples from the other major occurrences were published by Reynolds (1917). These figures are repeated in (Table 2).

In a review of the Carboniferous basalts in the Bristol area, which included the present occurrences, Dearnley (1960) interpreted their petrographical and chemical characteristics as being indicative of late-stage metasomatism of the lavas by residual fluids. It appears generally that in a series from the more massive least-altered material to vesicular highly altered rock the total alkali content increases, and the Na:K ratio decreases, leading to strongly potassic, orthoclase-bearing types.

Summary

The lavas from Weston-super-Mare are alkaline olivine-basalts which have undergone variable degrees of autometasomatism by alkali-rich residual fluids. There is no evidence to suggest that there was any marked difference in composition between the Middle Hope magma and the later magma which gave rise to the Spring Cove-Milton Hill lavas. Reynolds' (1917) arrangement of these igneous rocks into three groups (viz. normal olivine-basalts; olivine-bearing rocks with high potassium and low sodium; olivineorthoclase-basalts) most probably reflects their post-eruptive alteration by alkali-rich fluids rather than any original differences in magma composition or reaction with sea water. In extreme cases the feldspars are reduced to albite and/or orthoclase with abundant sericite inclusions.

Despite similarities in the volcanic sequences at Middle Hope and Spring Cove, where tuffs and tuffaceous limestones are found below and above the lava flows, there are differences in detail. These suggest that the former locality was probably proximal to the eruptive source, as the tuffs show little sign of having been reworked and mixed with sediment, and ash effusions continued after the outflow of lava. At Spring Cove the pyroclast-bearing rocks appear not to be primary ash deposits. Those below the lava show evidence of transport and resedimentation, and the beds above the lava may best be regarded as the results of comminution and erosion of the cindery lava.

The sediments and pyroclastic rocks associated with the lavas exhibit enrichment in iron and potash. These elements were probably largely derived from the fragments of lava and volcanic glass by leaching during an argillisation process after the eruption. Iron was combined in ankerite and in a goethite-rich dust responsible for the deep red colour of the sediments, while potash concentrations became sufficient for orthoclase crystals to form at low temperatures. It may be postulated that the pyrogenetic material was deposited in an enclosed basin where concentration of leached material could occur, and that the processes, at least in part, were akin to those investigated by Sheppard and Gude (1973) in saline lacustrine environments in the USA, where K-feldspar deposits were the ultimate product of the alteration of silicic tuffs. RWS

Stratigraphical palaeontology of the Dinantian strata

Introduction

Full fossil lists for the Dinantian of the Weston-super-Mare district have not been reproduced in this memoir, but are held on data cards by the IGS Palaeontology Unit. The more important fossil records are given in the text, with the details of the Lower Carboniferous stratigraphy. (Table 3) and (Table 4) list the significant fossils. Foraminifera have been identified by Dr W. H. C. Ramsbottom, conodonts by Mr M. J. Reynolds (who has carried out the bulk of the collecting work), and corals and brachiopods by Mr M. Mitchell. The authors of specific names are included in the Index of genera and species.

The Lower Carboniferous formations are related to the regional stages proposed in 1976 by George and others (Table 1). The Weston and Mendip successions are more complete than that in the Avon Gorge and some of the fossiliferous parts of these sequences cannot be correlated with any part of the original Vaughanian zonal scheme ((Figure 4); Ramsbottom, 1977, fig. 2). The faunas of the Weston area are particularly important for the understanding of the Chadian and Arundian biostratigraphy of the South-Western Province.

Black Rock Limestone and Black Rock Dolomite

The Black Rock Limestone and Dolomite have been taken together as one unit for the purposes of describing the faunas. The three faunal assemblages described by Mitchell and Green (1965, p. 182) from the Black Rock Limestone of Burrington Combe have been given formal assemblage biozone names by Ramsbottom and Mitchell (1980), and these can be used as a standard for comparison of the Black Rock faunas of other areas (Mitchell, 1981). The Zaphrentites delanouei Biozone (= Lower Fauna) is dominated by brachiopods and occasional zaphrentoids; the base of the Caninophyllum patulum Biozone (= Middle Fauna) is marked by the incoming of a rich coral fauna including Caninia cornucopiae, Caninophyllum patulum (C. patulum greeni in the lower part, and C. patulum patulum in the upper part), Cyathoclisia tabernaculum and Sychnoelasma konincki; and the Siphonophyllia cylindrica Biozone (= Upper Fauna) is characterised by Siphonophyllia cylindrica (sensu stricto).

At Middle Hope, all but the lowest beds of the Black Rock Limestone are exposed (p. 28) and the faunas are listed in (Table 3), which indicates the Assemblage biozones and the Belgian conodont biozones (Groessens, 1976). The faunas from beds A1(a) and A1(b) include Sychnoelasma clevedonensis and Zaphrentites delanouei and are referred to the delanouei Biozone. The lowest typical patulum Biozone coral assemblage is recorded from bed A1 (c). Canin phyllum patulum greeni is present from bed A1(c) to bed A16 and C. patulum patulum is doubtfully present in All but occurs from A18 to A22(a). The faunas are poor in the Black Rock Dolomite (A22), but the highest C. patulum recorded is 12.19 m above the base of the bed. The only macrofossils found in the upper part of the Dolomite are from the top bed of the formation, where the fauna, containing a poorly preserved coral identified as Siphonophyllia sp. sp. ?cylindrica group and Delepinea notata, indicates a correlation with the cylindrica Biozone. The base of the latter biozone is taken 16 m below the top of bed 22. MM

The conodont faunas recovered from the lowest part of the Middle Hope section (bed A1(a)) include no siphonodellids, but the presence of Patrognathus aariabilis is indicative of a Siphonodella Biozone age. The limits of the succeeding Polygnathus communis Carina Biozone are defined by the range of the eponymous subspecies from beds A1(b) to A15. Beds A16 to A22 (a) are taken to represent the Scaliognathus anchoralis Biozone; no S. anchoralis was found in the samples taken, but the conodont fauna as a whole compares well with the faunas of the anchoralis Biozone elsewhere. Forms characteristic of, but not necessarily restricted to, this biozone include Gnathodus delicatus, G. simplicatus, G. texanus, Polygnathus bischoffi and Pseudopolygnaths triangulus pinnatus. Bed A22(b) marks the first occurrence of Mestognathus beckmanni, indicating the base of the beckmanni Biozone. Polygnathus bischoffi is present in beds A22(a) and A22(b) and overlaps the junction between the anchoralis and beckmanni biozones. A comparison between the coral and conodont faunas shows that the delanouei–patulum biozonal boundary is above the base of the communis Carina Biozone, and that the patulum–cylindrica boundary coincides with the base of the beckmanni Biozone. M J R

The faunas from the outcrops at Brean Down, Uphill and Bleadon are noted in (Table 4) and are all from similar levels in the upper part of the Black Rock Limestone and Dolomite. They include Siphonophyllia cylindrica (sensu stricto) and Delepinea notata, and correlate with the cylindrica Biozone. No significant conodont faunas have been collected from these sections. Conodonts have been recovered, however, from a lower horizon in the quarry [ST 3265 5791] near Flat Roof Farm, the assemblage being comparable with those from the anchoralis Biozone of Middle Hope. The significant elements in this fauna are Gnathodus texanus, Polygnathus cf. mehli, P. nodomarginatus, Pseudopolygnathus cf. multistriatus and Ps. triangulus pinnatus.

The thicknesses of the Black Rock faunal units for the Weston area sections are summarised in (Table 5), where they are compared with those for Burrington Combe and the Avon Gorge. The Burrington Combe sequence is the most complete and thickest yet known in detail, and the Avon Gorge is the most incomplete, with the upper part of the patulum Biozone and the whole of the cylindrica Biozone not represented (Mitchell, 1972, p. 159). The Brean Down–Uphill succession is probably thicker than that at Burring-ton but the faunas from only the upper part have been studied. The Middle Hope section shows an intermediate situation, with only 16 m of the cylindrica Biozone present.

The base of the Chadian was drawn at the base of the beds containing the Upper Fauna (= cylindrica Biozone) at Burrington by George and others (1976, fig. 4; column B), and the faunas from the Black Rock Limestone and Dolomite of the Weston district are in agreement with this correlation. The Hastarian–Ivorian boundary is taken at the base of the Carina Biozone (see (Table 3); Ramsbottom and Mitchell, 1980).

Gully Oolite

The Gully Oolite is generally poorly fossiliferous but faunas of Chadian age have been recorded at two localities (Table 4). One of the richest Gully Oolite faunas from the Bristol–Mendip area was collected at Middle Hope (bed A23), from crinoidal and oolitic beds at the base of the formation (= Sub-oolite Bed of Mitchell, 1972, p. 156; George and others, 1976, fig. 4:4). This fauna contains Caninia cf. caninoides, Delepinea notata, Megachonetes cf. magna [cf. Vaughan, 1905, p1.26, fig. 3], Michelinia cf. megastoma and the first substantiated record from the South-Western Province of Levitusia humerosa, a species diagnostic of the Chadian. At Brean Down, lenses of crinoidal limestone within the Gully Oolite (bed F2) have yielded Koninckophyllum praecursor, M. cf. megastoma and Palaeosmilia cf. murchisoni [small form].

Caswell Bay Mudstone

No fauna is recorded but laminated algal beds are present in the Caswell Bay Mudstone, which is taken as the regressive phase of the Chadian (George and others, 1976, fig. 4: column A; = Lower Clifton Down Mudstone of Ramsbottom, 1977, fig.2).

Birnbeck Limestone

Much of the more massively bedded lower part of the Birnbeck Limestone is poorly fossiliferous but M. cf. megastoma and P. murchisoni have been collected. The richest fauna is from a coral bed which occurs 11.4 m above the base of the formation at Spring Cove, Weston (bed B6) and 10.5 m above the base on Flat Holm (bed D5). At both localities the corals are common in the basal 0.3 m of a massive post of limestone. This coral band was noted by George and others (1976, fig. 4:6) and contains an Arundian fauna with Caninia caninoides, C. subibicina, Siphonophyllia garwoodi and Konickophyllum praecursor. No conodont faunas have been recovered and the algae and foraminifera are poor from the lower part of the Birnbeck Limestone. However, Dr Ramsbottom has identified abundant double-walled Koninckopora, suggesting an early Arundian age, from the basal limestone (bed D3) of the formation on Flat Holm. The earliest record of P. murchisoni is from the base of the lowest bed of the Birnbeck Limestone (B3) of the coast section south of the Worle Hill Thrust, and this indicates that the base of the Arundian should be taken at the base of the Birnbeck Limestone throughout the district.

The thinner-bedded upper part of the Birnbeck Limestone contains a richer and more varied fauna than the lower part. This includes Clisiophyllum multiseptatum, M. cf. megastoma, P. murchisoni [common], S. garwoodi and Sychnoelasma kentensis, but the most important fossil is the diagnostic Arundian chonetoid Delepinea carinata. This assemblage is typical of the Arnside Fauna of the North-West Province (Garwood, 1913, pp. 467, 508; Mitchell, 1978b, p. 173) and is one of the most widespread and easily identified of the Dinantian faunas. It has not been recorded from Broadfield Down, or the Avon Gorge where the beds correlated with the Birnbeck Limestone are thin lagoonalphase deposits with three beds of crinoidal limestone. A rich fauna has been collected from the Birnbeck Limestone of Little Down Quarry, where in addition to P. murchisoni, S. garwoodi and D. carinata, the presence of Delepinea cf. destinezi suggests that beds older than those with the main D. carinata fauna are present. C. multiseptatum is recorded only from the top of the Birnbeck Limestone (bed E7) and the base of the Goblin Combe Oolite (bed B15), and this restricted mid-Arundian range corresponds with the occurrence of the species in the Furness succession (Mitchell, 1978a, p.34). Conodonts are rare in the upper part of the Birnbeck Limestone and the overlying Goblin Combe Oolite and Clifton down Limestone.

Goblin Combe Oolite

The fauna of the Goblin Combe Oolite is more restricted in species than that in the upper part of the Birnbeck Limestone. It includes C. multiseptatum, P. murchisoni [common], Linoprotonia sp. hemisphaerica group and L. cf. O, and is of Arundian age.

Algae and foraminifera indicate a late Arundian (V2a) age and are similar to those from the Goblin Combe Oolite of the Avon Gorge (George and others, 1976, fig. 4:10).

Clifton Down Limestone

Fossils are not common in the Clifton Down Limestone, except in some bands which are locally crowded with individuals of relatively few species, and this distribution suggests a slightly restricted environment. The common fossils are Axophyllum [Carcinophyllum] aaughani, Composita ficoidea, Lithostrotion aranea, L. martini, Linoprotonia corrugatohemispherica and Linoprotonia sp. hemisphaerica group. This assemblage is typical of the Holkerian. The characteristic brachiopod Davidsonina carbonaria, which was listed by Sibly (1905, p. 557) as 'abundant near the base of S₂', is represented in the present account by the records from the Monk's Hill section and the north side of the marine lake at Knightstone. All the cerioid Lithostrotion specimens collected from the Clifton Down Limestone are here referred to L. aranea, although in some specimens the columella is thin, suggesting comparison with L. basaltiforme bristoliense of Vaughan (1903, p. 106).

Sibly (1905, p.557) recorded Stenoscisma [Camarophoria] isorhyncha as 'abundant at the base of S₁' (equivalent to a late Arundian age), but the single specimen collected during the present survey was from the Clifton Down Limestone (bed C5) of the coast section north of the Worle Hill Thrust and was associated with a typical Holkerian fauna. MM

Sedimentation and correlation

The palaeogeographical framework within which the Carboniferous Limestone of the South-Western or Avonian Province was laid down is now well known and has been summarised by George (1972). A reconstruction (ibid., p. 227) shows 'the Avonian rocks to be for the most part shelf limestones on the southern flanks of a persistently emergent St George's Land [trending very approximately east–west] against which the rock formations gently wedge out... deposition of the limestones was never at any considerable depth–mostly it was on broad flats submerged to only a few metres–and minor changes in the rate of elevation to the north or subsidence to the south caused at times large areas of the marginal tracts to be exposed and eroded, at other times the back-shore flats to be flooded for many miles as the migrating floors encroached upon the slopes of St George's Land.' The present-day Weston–Broadfield Down–Mendip district lies entirely within this marine shelf region.

In a stimulating analysis of Dinantian stratigraphy Ramsbottom (1970; 1973; 1979) has categorised six main lithological and faunal facies which recur within the Dinantian and are believed to be indicative of deposition in different water depths (Figure 3). Although, as the author himself admitted, such a classification represents an oversimplification of the numerous Carboniferous Limestone facies and their interrelationships, it nevertheless forms a useful basis for discussion of the stratigraphy. It is important, however, not to interpret these facies as depth indicators in too rigid a manner. Ramsbottom (1973, p. 569), referring to modern work, suggested that the algal calcite or dolomitic mudstones (the highest, group 1 a) were formed 'mostly at supra-tidal or inter-tidal levels and that the oolitic limestones formed in water generally of less than three metres depth' and bioclastic limestones (type 3) in more open water 'no more than a few metres deep'. The depth ranges were thus so small that it is not surprising to commonly find an intimate and very variable intermixing of facies types within a very small thickness (sometimes only a few centimetres) of sediment, such as has been described in the Chadian rocks of Gower by George (1978a). In these cases depth of water can have been only one of the faciescontrolling factors and others, such as exposure and currents, may have been locally more important. Given this caveat, the rocks of the Weston district and adjacent areas can be broadly grouped into facies types as follows (Figure 4): (1a) calcite-mudstones and dolomite-mudstones with stromatolitic algae (lower part of the Caswell Bay Mudstone; uppermost part of the Clifton Down Limestone);(1b) calcite-mudstones and dolomite-mudstones (upper part of the Caswell Bay Mudstone; mudstone parts of the Flat Holm Limestone Member; basal part of the Clifton Down Limestone); (2) oolitic limestones (Gully Oolite; Goblin Combe Oolite except for the lowest part; Birnbeck Limestone massive facies; part of the Burrington Oolite and Clifton Down Limestone); (3) pale grey thick-bedded bioclastic limestones (Birnbeck Limestone, well-bedded facies, basal Goblin Combe Oolite; part of the Burrington Oolite; Hotwells Limestone); (4) dark grey to almost black, thin-bedded bioclastic limestones (parts of the Lower Limestone Shale; Black Rock Limestone); (5) calcareous mudstones (much of the Lower Limestone Shale); (6) non-calcareous mudstones (not represented in the present district). The Lithostrotion beds in the Clifton Down Limestone appear to correspond to facies type 3 on faunal grounds, although they range from massive paler limestones to thinner-bedded darker types, and the association with chert seems to characterise type 4 facies (cf. the Black Rock Limestone).

Using the facies as depth indicators, in conjunction with studies of the fauna, Ramsbottom suggested that the stratigraphy of the Dinantian is to be interpreted in terms of six major cycles or mesothems, each commencing with a major transgression owing to a eustatic rise in sea level, followed by regression due to a eustatic fall in sea level. He designated the mesothems D1, D2, etc., including substages (e.g. D1a, D1b) where appropriate, and pointed out that each mesothem in turn encompassed many minor cycles or cyclothems. The Geological Society Dinantian Working Party (George and others, 1976) equated these major cycles with named stages^‡2 . In general the regressive phases are marked by the type 1 lagoonal' facies and the transgressive phases by the deeper water bioclastic limestones.

The first major transgression (D1) took place at about the start of the Tournaisian. The base of the Stage is placed on palynological evidence (Dolby and Neves, 1970) a short distance below the dramatic change from the continental-type sedimentation of the Upper Old Red Sandstone to the marine Lower Limestone Shale (types 4 and 5 facies), which in turn is succeeded by dark limestones of facies type 4 (Black Rock Limestone). The second major transgression, which corresponds to the base of the Chadian Stage, coincides in the central and western Mendips with an important faunal change at the base of the Siphonophyllia cylindrica (s.s.) Biozone within the Black Rock Limestone (p. 21), but apparently not with any marked lithological change. In the Middle Hope area, in the north of the district, extensive secondary dolomitisation occurs in this part of the sequence and there is stratigraphical evidence for an appreciable non-sequence (see below).

The facies changes between the base of the Gully Oolite (D2b) and the base of the Hotwells Limestone (D5b) in the Mendip–Bristol district present a relatively complex picture of transgressive and regressive phases in which it is difficult to distinguish the relative importance of the different features for the purposes of regional correlation because of the effects of general overall subsidence within the area (see (Figure 4) and compare with Ramsbottom, 1977, fig. 2). There is little thickness change in the Dinantian rocks below the Hotwells Limestone as between Broadfield Down and the Avon Gorge, compared to a roughly estimated doubling of thickness between Broadfield Down and the western Mendips (Figure 4). The major cycles appear to be most clearly identifiable (George and others, 1976, fig. 4, column D) on Broadfield Down, where the lithological succession is well differentiated into regressive and transgressive facies. The thickening of the strata south of Broadfield Down is due to the increased proportion of grey oolitic and bioclastic limestones and, by contrast, the micritic 'lagoonal' strata attentuate and disappear southwards. As clear identification of the boundaries of the cycles in the field depends more than anything else on the presence of these contrasting facies, it follows that recognition of the limits of the cycles becomes increasingly difficult in the southern part of the district. The characteristic macrofaunal elements are rarely sufficiently abundant to enable the stage boundaries to be fixed within narrow limits independently of the facies changes, and their recognition depends on microfossil determination.

At Bristol the base of the Gully Oolite, which corresponds to the base of the Chadian there, is marked by an appreciable non-sequence (Mitchell, 1972). The faunal evidence (p. 22) indicates that some 70 to 80 m of strata (Black Rock Limestone) are missing at Middle Hope, as compared with the Mendips. A small part of this thickness difference may be due to reduced sedimentation, but this does not seem to be an important feature in Tournaisian–earliest Chadian times. Ramsbottom (1973, p. 575) suggested the possibility that erosion associated with the non-sequence resulted in the removal everywhere of typical regressive phase lagoonal sediments, of which the extensive dolomitisation of the underlying rock (Black Rock Dolomite) is now the sole remaining evidence. Quite apart from the unlikelihood of the complete removal without trace of any remanie debris from such a huge area (possibly some 5000 km² including South Wales), this hypothesis ignores the fact that the lowest part of the Gully Oolite is itself commonly strongly dolomitised (e.g. Reynolds, 1921, p. 233; Mitchell and Green, 1977, footnote p. 188, and on Brean Down, p. 35). It seems more likely that the conditions for deposition of lagoonal sediments were never established and that the dolomitisation occurred either when the Gully Oolite was forming or soon afterwards. Murray and Wright (1971, p. 258) have suggested that the Gully Oolite may have been partly subaerial in formation farther north; also in the Weston district the upper surface may show evidence of emergence (see also Mitchell and Green, 1977, footnote p. 188). In either case, it is probable that fresh water would have had ready access to the formation and this can be an important factor in the promotion of secondary dolomitisation (Folk and Land, 1975).

The general picture of sedimentation for the Chadian (D2b) given by Ramsbottom (1973) and the Geological Society Dinantian Working Party (George and others, 1976) for Bristol is a simple regressive sequence above the basal non-sequence, of crinoidal limestone ('Sub-oolite Bed')–Gully Oolite–(Lower) Clifton Down Mudstone. Although this sequence is seen at Bristol and on Broadfield Down, it becomes increasingly difficult to recognise as the facies are traced southwards and south-westwards. The basal crinoidal limestone is very poorly represented, the top of the Gully Oolite marks a widespread break (the 'Mid-Avonian Break' of Dixon and Vaughan, 1912, and later authors) leading to emergence in part, possibly all, of the Weston area, and the implied lateral equivalence of the Caswell Bay Mudstone to the whole of the (Lower) Clifton Down Mudstone is now known to be incorrect (see below). The conditions of formation of the Gully Oolite appear to have been unusually stable, for it is unique amongst Avonian oolites in its combination of huge areal extent, relative purity, and uniformity in thickness. Lateral passage southwards into massive coarse well-sorted crinoidal limestones seems proven (see above) and this appears to correspond to increasing depth of water downslope of the main inshore shoals where the oolite was forming. It is now generally agreed (e.g. compare Ramsbottom, 1973, with George, 1978b) that the Caswell Bay Mudstone was formed in an intertidal to supratidal environment. The presence of lenses of oolitic and bioclastic crinoidal limestones in intimate contact with the 'normal' micritic facies, however, provides a good illustration of why the relationship of water depth to rock type must be argued with caution (see George, 1978a, b, for detailed discussion).

The Arundian (D3) strata at Broadfield Down and Bristol were thought by Ramsbottom (1973) and George and others (1976) to consist of the Goblin Combe Oolite resting unconformably on the (Lower) Clifton Down Mudstone of Chadian age and succeeded by an apparently simple regressive sequence of calcite-mudstone facies which forms the upper part of the Clifton Down Mudstone and the basal part of the Clifton Down Limestone. New evidence (see below) from the Avon Gorge shows this correlation is over-simplified, and that the (Lower) Clifton Down Mudstone of Bristol, and by implication of Broadfield Down, is part Chadian and part Arundian. At Weston, the Birnbeck Limestone represents the basal transgressive phase of the Arundian. Although the Birnbeck Limestone succession of Brean Down is similar to that at Weston (Figure 2), the southward thickening of the lower massive cross-bedded unit, apparently at the expense of the underlying Caswell Bay Mudstone, and the increased proportion of the well-bedded upper facies at the expense of the massive cross-bedded facies, is probably due to increased water depth to the south. Examination of the Avon Gorge section by the author and Mr M. Mitchell has revealed that, of the 41.20 m of (Lower) Clifton Down Mudstone between the Goblin Combe Oolite and the Gully Oolite, the basal 20.25 m are dolomitised chinastones with stromatolites, mudstones and shales, very similar in lithology to the Caswell Bay Mudstone of Weston and they are taken to be its lateral equivalent; the uppermost 20.95 m include three beds of massive, cross-bedded, crinoidal or oolitic limestones interbedded with chinastones with some stromatolites, mudstones and shales. Dr W. H. C. Ramsbottom has identified Arundian algae and foraminifera from the highest of these limestones, and possible Chadian foraminifera from the lowest. Thus part of the (Lower) Clifton Down Mudstone can be correlated with the Birnbeck Limestone, indicating that one or more incursions of open sea conditions reached as far north as Bristol in late Chadian and early Arundian times. The correlation of the upper part of the (Lower) Clifton Down Mudstone at Broadfield Down with the Birnbeck Limestone is also suggested by the evidence of the contemporary volcanic rocks. A relatively extended period of local volcanicity is witnessed to by the presence of 20 m of tuffaceous limestones, with a median basalt submarine flow, in the basal part of the Birnbeck Limestone in the Weston area, and the tuffaceous debris provides evidence for weathering under subaerial conditions (pp. 15–18). The only other known local volcanic occurrences in the Chadian–Arundian succession are on Broadfield Down (basaltic lava and tuffs, 15 m thick) and east of Clevedon (three small outcrops of basalt along a strike section of 1.2 km; N. G. Ref.[ST 456 721] to [ST 466 725]), all of which occur at about the middle of the (Lower) Clifton Down Mudstone in an environment that includes the possibility of subaerial weathering. As both the Weston and the other occurrences are at comparable levels in thickness terms above the top of the Gully Oolite, and as volcanicity in the Dinantian rocks of the province is not common, it seems reasonable to conclude that the occurrences are interrelated and probably contemporaneous in lower Arundian rocks.

On Broadfield Down the top (regressive phase) of the Arundian Stage (D3) was taken within the Clifton Down Limestone above a basal group (37 m) of calcite-mudstones with oolite bands. The transgressive phase of the succeeding Holkerian Stage (D4), immediately above, comprises limestones with silicified Lithostrotion masses and some chert which then pass up through oolites (Brockley Combe Oolite) into a well-marked regressive algal-rich phase at the top of the Clifton Down Limestone (equivalent to the Concretionary Beds of the Avon Gorge). This general sequence is widespread in the Bristol area and has been well described in the Avon Gorge (Reynolds, 1921) and at Wick and Chipping Sodbury (Murray and Wright, 1971). Minor cyclicity representing both transgressive and regressive phases is noted in the above sections in the lower third of the Clifton Down Limestone. In an analysis of the facies in terms of modern analogues Murray and Wright concluded that the limestones (biomicrites, pelmicrites) with Lithostrotion masses or coralline debris represent an open shelf-sea environment, in contrast to the intertidal algal-rich (stromatolitic) phases and the tidal bars and shoals represented by the oolites. In the Weston–Mendip district, as in the areas to the north-east, the Clifton Down Limestone is divided into a middle Lithostrotion phase and an uppermost calcite-mudstone phase, and a very variable lower group representative of both transgressive (oolites; crinoidal limestones) and regressive (oolites; finely bedded calcite- and dolomite-mudstones) phases.

At Burrington Combe, which is equidistant from Broadfield Down and Weston, George and others (1976) have put the base of the Holkerian (D4) directly above a thin dolomitic mudstone with fenestral structures (named the Rib Mudstone = bed 13 of Mitchell and Green, 1965, p.187) within the Burrington Oolite. The beds above include the diagnostic Holkerian brachiopod Davidsonia carbonaria. The base ofthe Clifton Down Limestone in the Burrington area is taken at the base of a well-marked 'lagoonal phase', estimated to lie about 38 m above the Rib Mudstone (Green and Welch, 1965, p.25), and the Lithostrotion beds occur some 53 m higher still in the succession (Table 1). At Weston, the base of the Clifton Down Limestone is likewise distinguished by a well-marked 'lagoonal phase', with the diagnostic Holkerian brachiopod D. carbonaria being found in the overlying beds; the Lithostrotion beds occur about 120 m above the base of the formation. From this it would appear that the basal beds of the Clifton Down Limestone at Weston may be the approximate equivalent ofthe Rib Mudstone of Burrington^‡3 . The relationship of this general Weston–Mendip succession to that of Broadfield Down, where the base ofthe Lithostrotion beds marks the base of the Holkerian, is uncertain. Two alternative explanations are suggested–either there is a considerable non-sequence at the base ofthe Holkerian at Broadfield Down or else the sediments show an appreciable degree of diachronism where traced southwards towards the Weston–Mendip area. GWG

Details

In the following account the bed numbers and symbols refer to those used for the fossil collections.

Middle Hope

Coastal cliff and foreshore sections provide almost continuous exposures through the uppermost three-quarters of the Black Rock Limestone, the Black Rock Dolomite and much of the Gully Oolite. The Middle Hope Volcanic Beds are exposed at four different places along the coast.

The most complete and easily accessible sections occur in the cliffs in the western third of the peninsula. The highest beds (Gully Oolite) occur adjacent to the National Trust car park [ST 330 659] at the northern end of the Sand Bay road, and successively lower beds can be examined by walking round the coast in a clockwise direction from this point. The description of the volcanic rocks ['exposure 1': 325 661] is, with some modifications, based on the published accounts (p. 12)

Continuing eastwards along the coast, the volcanic rocks are next seen at 'exposure 2' [ST 3371 6643], where the total thickness is 29.5 m. The bed numbers in square brackets are those used by Morgan and Reynolds (1904b):

'Exposure 3' [ST 3392 6655] of the volcanic beds is a further 200 m east-north-east along the coast. The strata in this section are bent into a sharp syncline immediately adjacent to, and on the northern side of, a fault and there has been some associated repetition and distortion of the rocks which do not appear to have been allowed for in the published account (cp. Morgan and Reynolds, 1904b: bed 9 'somewhat nodular limestone' appears to be the same as bed 7, and bed 8 'argillaceous limestone' is a brown calcareous tuff with lapilli identical with their bed 10). The total thickness is not known because the highest exposed bed is the 'pipe-rock' (bed 16 or exposure) and because of the structural disturbance below, but it appears that the general character of the strata underneath this bed is comparable to that of 'exposure 2', although the beds may be rather reduced in thickness and more calcareous in composition.

The Black Rock Limestone below the volcanic beds is well exposed in the cliffs and foreshore eastwards to around the extremity of the peninsula at St Thomas's Head. The volcanic beds are next seen immediately south of the jetty of the marine research station at 'exposure 4' [ST 3489 6692]. If Morgan and Reynolds' record (1904b, p. 200) is added to the author's own observations, the sequence reads in descending order: [Black Rock Limestone]: shaly beds, 0.75 m; fine red tuffaceous limestone, 0.6 m; fairly coarse tuff, 0.6 m; not seen, estimated 2 m; red and yellow crinoidal limestone, 2.4 m; not seen, estimated 3 m to Black Rock Limestone of which the upper part is red-stained. The minimum thickness of the volcanic beds is thus 2 m, but the staining of the underlying limestones suggests that the thickness could be as much as 4 m, or even possibly twice as much as this. Almost complete coastal exposure occurs north and south of 'exposure 4' and the section is given by Bush (1929), though the thicknesses quoted must be received with caution, especially as he did not recognise the presence of a thrust fault repeating part of the sequence in the southern part of the section. The Gully Oolite is the least well exposed formation in the section and its contact with the Black Rock Dolomite cannot clearly be seen owing to structural disturbance. A summary of the thicknesses at St Thomas's Head is, in descending sequence: Gully Oolite, seen 37 m; Black Rock Dolomite, 45 m, with a well-marked level of zaphrentid and large caniniid corals in the lower part; Black Rock Limestone (a) 55 m to base of the chert (b) 85 m seen below the chert. As compared to the western end of the peninsula (pp. 2728), it will be noted that the Black Rock Dolomite has increased in thickness at the expense of the underlying Black Rock Limestone but that there is an overall thinning of some 12 m between the top of the Black Rock Dolomite and the chert.

Weston- Worle

The Weston-Worle inlier is structurally divided into two parts by the E–W-trending Worle Hill Thrust. The youngest exposed beds in the Carboniferous Limestone succession are to the north of the thrust and are well exposed on the coast north of Weston Woods and inland on the north side of the main ridge. The oldest beds occur to the south of the thrust and are best seen on the coast at Weston itself (Figure 5), with an overlap of about 120 m of beds common to both sides of the thrust.

South of the Worle Hill Thrust, commencing with the youngest beds, the main coast section starts in the low cliffs on the west side of Knightstone 'island' [ST 312 618]; it continues past Anchor Head and Birnbeck Pier and ends at the Worle Hill Thrust, which intersects the coast at a point 240 m ENE of the toll gate on the Weston Woods–Kewstoke Road [ST 3130 6270]. The section is as follows:

The Gully Oolite is strongly dolomitised for about 30 m westward along the coast from the position of the thrust.

Birnbeck Island [ST 305 626] is underlain by massive Goblin Combe Oolite, but at low water the northernmost reef is seen to be Birnbeck Limestone and just south of the end of the lifeboat slipway basal Clifton Down Limestone (bed B21) is exposed.

Inland, exposure of the higher parts of the Carboniferous Limestone succession is much better than that of the lower parts. Apart from the Gully Oolite, which is seen at intervals immediately adjacent to the Worle Hill Thrust, the details of the beds up to the Goblin Combe Oolite are difficult to make out in Weston Woods owing to indifferent exposure. No exposure of either the Caswell Bay Mudstone or the basalt was seen. Weathered basalt debris is common in the gardens on Milton Hill east of Weston Woods and the top of the flow was seen in the bottom of Butt's or Milton Quarry [ST 336 627], where it was reported from trial borings to be 12.2 m in thickness. No details can be added to those already published. The easternmost recorded exposure was of bedded tuff dipping southwards in an old pond [ST 3429 6327] just beyond the eastern end of Worlebury Golf Course, but nothing can now be seen there. The best inland exposure in the Carboniferous Limestone is in Butt's Quarry, where the succession from the top of the basalt to the basal part of the Goblin Combe Oolite (bed B15) almost exactly replicates that of the coast section. Immediately adjacent quarries to the north and east [ST 339 627]; [ST 340 626] also give good exposures of the Birnbeck Limestone and the basal Goblin Combe Oolite. These beds are also seen about 1.5 km to the west, in the Town Quarry [ST 321 624], where they are cut by a north–south fault which brings down the lowest 12 m of the Goblin Combe Oolite on the west side. On the east side of the quarry about 15 m of the well-bedded facies of the Birnbeck Limestone intervene between the Goblin Combe Oolite capping the southern rim of the quarry and a bed of the massive facies (6 m seen) at the foot of the northern wall. This massive bed must lie at a considerably higher level than similar beds on the coast (cp. pp.9–10).

The beds of the Goblin Combe Oolite above the basal coarse crinoidal bed are exposed at intervals in small quarries and natural outcrops in the northern part of Weston and, most clearly, on Worle Hill. The dominant rock types are pale grey massive crinoidal-oolitic limestones and purer oolites. The dolomite (bed B l8) noted on the coast cannot be traced inland. Scattered corals, including Lithostrotion martini and P. murchisoni, are the commonest fossils.

The Clifton Down Limestone as seen in numerous small exposures inland comprises a basal well-bedded micrite or calcite-mudstone ('chinastone') sequence, usually more or less dolomitised, as on the coast (bed B21), but not invariably so. The dolomitisation noted on the coast in the succeeding rocks (bed B22) is not consistently present inland, where the dominant rock types are more or less pure cross-bedded oolites, locally porcellanous in texture. These rocks are best seen in a road cutting [ST 3304 6187] and an old quarry [ST 3329 6192], both exposing 10 to 12 m of beds; the latter appears to be one (Milton Road Quarry) described by Sibly (1905, p. 560) as including a distinctive fauna, but this could not be confirmed during the recent survey. Between Ashcombe Park and Worle the Clifton Down Limestone becomes strongly dolomitised and it is not possible satisfactorily to separate the basal micrite from the overlying oolite beds.

The coastal section below and north of the Worle Hill Thrust exposes beds between the upper part of the Goblin Combe Oolite and the base of the Lithostrotion beds in the upper part of the Clifton Down Limestone (Figure 6). The Lithostrotion beds, which comprise the highest exposed part of the Clifton Down Limestone in the district, are well seen on Monk's Hill, where the underlying part of the sections duplicates that seen on the coast.

The coastal section between the thrust [ST 3132 6270] and a dip fault [ST 3143 6272] about 130 m farther along the coast to the east is cut by several small dip faults, but it is possible to match the beds across them and thus to make out a complete succession. The mapped line for the base of the banded mudstone-oolite beds is taken at the lowest calcite-mudstone band. The brachiopod Composita ficoidea commonly occurs as seams of abundant individuals in both the oolite and the fine-grained facies. The uppermost bed is transitional to the typical Lithostrotion beds (p.12). The section is as follows:

The section is continued on the east side of the dip fault mentioned above. The throw of this fault is not known but must be in excess of 4.5 m. The relationship of the bottom bed of the previous section to the top of the succeeding section cannot be proved but it seems probable from outcrop calculations that between 10 and 20 m of strata are unrepresented. Although the remainder of the coast section to the angle of Sand Bay is cut by a number of dip faults the strata can be matched across them and a complete succession can be established:

The easternmost 475 m of the coastal section, which extends to the angle of Sand Bay, provides discontinuous exposures in the basal Clifton Down Limestone (bed C3) and in the uppermost Goblin Combe Oolite, which is strongly shattered in the easternmost 50 m of exposure (see also p. 94).

A reasonably good match can be made between the coastal sections north and south of the Worle Hill Thrust. Beds (C3) to (C5) of the former correspond with beds (B21) to (B23) of the latter, and the oolites (bed B24) on Knightstone 'island' correspond approximately to the oolites beneath the banded oolitecalcite mudstone beds at the top of the Sand Bay section.

Inland, the highest exposed beds in the Clifton Down Limestone are best seen in the road-section on Monk's Hill [ST 337 632] to [ST 335 633] and in the small valley immediately to the west in which the Monk's Steps are situated. The two sections are complementary to one another and the following succession represents a composite section, in descending sequence: grey to dark grey well-bedded granular limestones with abundant L. martini masses, commonly silicified and with lines of chert nodules, seen 30 m; similar limestones, but with only scattered Lithostrotion, becoming slightly oolitic towards the base, 11 m; grey cross-bedded oolites with Composita and with horizons of calcite-mudstone 0.5 to 2 m thick at the top, middle and base, 9 m; light grey rather porcellanous oolite with Davidsonina carbonaria near the top, 5.2 m; light grey calcite mudstone, 3.2 m; grey mostly massive cross-bedded oolites, 12 m [the Monk's Steps join the road here]; oolites as before, about 20 m. A small strike fault is seen at the bottom of the section where the Monk's Hill road joins the main Kewstoke Road. The thickness of beds with calcite-mudstone is unusually great (17.4 m) and contrasts with the more normal thicknesses as seen on the coast.

Between Monk's Hill and the Worle Hill Quarry [ST 351 634], at the eastern end of the inlier, there are numerous exposures in the upper part of the Clifton Down Limestone in which the main subdivisions may be readily made out; the quarry itself provides good exposures in the lower beds. The strata in the quarry are folded into a syncline with the southern limb inverted adjacent to the Worle Hill Thrust, which is exposed at the southern end [ST 351 633] of the workings. The thrust has brought the Goblin Combe Oolite into juxtaposition with the basal member (bed C3 of Sand bay section) of the Clifton Down Limestone. The latter comprises 12.2 m (base not seen) of black thinly bedded nodular dolomitic mudstone with shaly partings (p. 11), and is overlain by massive limestones estimated to be 100 to 120 m in thickness. Some thick beds of granular crinoidal limestone with scattered Lithostrotion occur but the dominant rock types are oolites; no exact correlation can be made with the coast sections. The higher thin-bedded fine-grained dolomitic horizon (B23 and C5 pars, on coast) is not represented, and the main oolites (e.g. C4 to C14+) below the banded oolite-mudstone member are thicker than anywhere else in the inlier. Typical Lithostrotion beds are present on the northern edge of Worle Hill, eastwards from Hatley Rocks [ST 341 634], where they are separated from the Clifton Down Limestone in the Worle Hill quarry (noted above) by a northerly branch a the Worle Hill Thrust (p. 100).

Flat Holm

Good cliff sections provide complete exposure through the greater part of the Gully Oolite, the whole of the Caswell Bay Mudstone and much of the Birnbeck Limestone, including the type sections of the Flat Holm Limestone Member of the latter (see p. 10; (Plate 4); (Figure 2) and (Figure 7)). For descriptive purposes in the following account, the micritic bands of the Flat Holm Member are designated with letters A to F in ascending order. A standard sequence for the island is given below. The Gully Oolite is most fully exposed on the east coast, between Coal Beach and Dripping Cove; the mudstone-limestone sequences comprising the Flat Holm Member are best seen on the east coast between Dripping Cove and Lighthouse Point, and on the west coast from Bottles-well Point northwards for about 150 m. The upper part of the Birnbeck Limestone, with Palaeosmilia murchisoni and Delepinea carinata, is best seen on the north-west coast, as the succession to the south on both east and west coasts is truncated by a thrust fault in the southern part of the island. The goniatites mentioned by Strahan and Cantrill (1912, p. 26) appear to be gastropods (Bellerophon) in the Birnbeck Limestone.

The dark grey micrite 'crust' at the top of the Gully Oolite (p. 8) is present in all the sections and appears to be identical with the Heatherslade Bed of Gower described by George (1978a) from a similar position. The thickness is usually about 0.5 m and the upper part is much brecciated. Exceptionally, in a trench section [ST 2214 6464] west-south-west of the lighthouse, no less than 2.75 to 3 m of rubbly textured 'chinastone' are seen, with scattered 'floating' ooliths increasing downwards, and passing down without break into normal oolite.

The Caswell Bay Mudstone varies in thickness from 3.7 m to a maximum of 5.3 m in the trench [ST 2214 6464] mentioned above. The individual sections vary in detail from that given above. South of the lighthouse the basal layer locally comprises grey mudstone up to 0.4 m thick, apparently occupying hollows in a very hummocky surface on the underlying beds. Locally the overlying 'chinastones' show good algal laminae and desiccation cracks on their upper surfaces and are strongly dolomitised. The uppermost 'chinastone' is apparently widespread and is much burrowed in places. The top two-thirds of the sequence is dominantly composed of buff-weathering laminated and burrowed dolomitemicrites, in parts of varying thickness, and is apparently identical to the micritic bands in the overlying Birnbeck Limestone.

The Wolves

The Wolves are two small reefs which lie 1.4 km to the north-west of Flat Holm and are only uncovered for a short time at the lowest tides. They are composed of medium-grained oolite, apparently dipping at about 30° in a general south-easterly direction. Samples collected from the reefs by Dr R. A. Waters in 1979 yielded a microfaunal assemblage determined by Dr W. H. C Ramsbottom as of Arundian age, thereby indicating a correlation with the Goblin Combe Oolite.

Steep Holm

The inaccessibility of some of the cliffs on Steep Holm, combined with local but intense tectonic disturbance (Plate 3), has of necessity left some room for doubt in the interpretation of the geological section. However, the rocks range from the Black Rock Dolomite to the Goblin Combe Oolite (Figure 8) and are closely comparable to the development seen at Brean Down on the mainland. The cliffs at the eastern end of the island provide the most complete section:

Northwards from the base of the above section the rocks are strongly upturned to form a monoclinal fold, and the associated shattering and disturbance make it impossible to continue the detailed measured section. Examination of photographs of the cliffs taken from offshore indicates that some 50 m of massive strata intervene between the base of the measured section and well-bedded rocks at the foot of the cliffs on the north coast, exposed at extreme low tide but inaccessible for sampling. Sampling at several places along the coast in the lower part of the massive strata proved typical Gully Oolite; thus it seems likely that the bedded strata underneath represent the uppermost part of the Black Rock Dolomite and hence the combined thickness of Birnbeck Limestone and Gully Oolite is just under 102 m. There is no indication of the Caswell Bay Mudstone intervening between the lower massive beds of the Birnbeck Limestone and the Gully Oolite and it is assumed that this formation is absent. At the western end of the island, landings confirmed the presence of the Gully Oolite on the northern third of the coast; the middle third, also in massive limestone, was inaccessible but assumed to be in either Gully Oolite or basal Birnbeck Limestone; the southern third is in strongly folded well-bedded Birnbeck Limestone which continues along the full length of the south coast. Inland exposures are sparse but small quarries [ST 2265 6058]; [ST 229 606] prove that a capping of Goblin Combe Oolite is present.

Brean Down

The Brean Down peninsula is composed of Carboniferous Limestone ranging from the Black Rock Dolomite to the Goblin Combe Oolite. The oldest and youngest beds are exposed on the south and north coasts respectively, while full successions are seen at the eastern and western ends of the Down. With the exception of the Goblin Combe Oolite, which is exposed along much of the northern coast, the following section is taken from the western end of the Down, starting from the base of the Goblin Combe Oolite at Sprat Beach [ST 285 593], then extending past Howe Rock to the cliffs on the south coast:

The basal massive facies of the Birnbeck Limestone (beds F5 to 7) are best seen in the dry 'moat' on the east side of the Fort [ST 2804 5926]. Although the rocks have been entirely altered to dolomite they can be readily recognised by virtue of their massive nature and their position overlying the Gully Oolite. The Caswell Bay Mudstone is apparently represented by a single mudstone band, and the dolomitic mudstone bands in the lower part of the Birnbeck Limestone may correspond to some of the mudstone phases recognised on Flat Holm (Figure 2).

The dolomitic siltstone band (bed F3) may represent a highly altered remnant of the micrite capping to the Gully Oolite similar to that recognised on Flat Holm (pp. 32–33). The top two-thirds of the mapped Gully Oolite passes laterally into massive dolomite a short distance to the east of the Fort but the lower part can be followed for a distance of about 0.8 km. The lower massive facies of the Birnbeck Limestone can be traced for a similar distance by means of the slight depressions which mark the outcrops of its basal mudstone bands. Eastwards of this area the entire sequence below the well-bedded part of the Birnbeck Limestone cannot be mapped separately from the underlying massive Black Rock Dolomite, presumably because it has all passed into dolomite. Lack of exposure west of the site of the Roman Temple [ST 2932 5882] makes it uncertain, however, exactly where the transition occurs. East of the temple site dolomite is ubiquitous below the well-bedded part of the Birnbeck Limestone. At the east end of Brean Down there are good exposures of the basal beds of the Goblin Combe Oolite (on the north coast) and of the top two-thirds of the Birnbeck Limestone in a large disused quarry [ST 3025 5885] which shows a closely comparable sequence to that recorded above at the western end of the peninsula. The section continues for 14 m below this in massive dolomites with cross-bedding etched out by the sea in suitable situations, and this clearly represents the basal massive facies of the Birnbeck Limestone of the section at Howe Rock.

Black Rock is a reef 325 m E of Brean Down consisting of massive dolomite and strongly dolomitised crinoidal limestone with scattered P. murchisoni and Linoprotonia sp.–presumably in the basal massive facies of the Birnbeck Limestone.

Uphill

The small inlier at Uphill represents the western prolongation of the Black Down Pericline of the Mendips and is entirely in the Black Rock Limestone above the main chert. The cliff and quarry section (Uphill Quarry) below St Nicholas's Church shows a single bed of unfossiliferous grey partly dolomitised granular crinoidal limestone and dolomite up to 7.5 m thick resting on some 37 m of well-bedded grey to dark grey fossiliferous crinoidal granular limestone, which becomes fine-grained and almost black in the lower part of the sequence (Plate 2). Fossils collected from below the dolomite are characteristic of the Siphonophyllia cylindrica Biozone of the Black Rock Limestone: S. cylindrica (sensu stricto), Caninia sp. ?cornucopiae, Cyathoclisia tabernaculum, Fasciculophyllum densum, F. omaliusi, Michelinia cf. megastoma, Sychnoelasma konincki and Megachonetes cf. magna (cf. Vaughan, 1905, pl. 26, fig. 3). Scattered exposures of dark grey crinoidal limestones occur along the western edge of the inlier and the main periclinal axis is seen near the southern edge. It is estimated that some 150 m of Black Rock Limestones are represented in the exposures.

Bleadon–Hutton

The Bleadon–Hutton area covers the western end of the Black Down Pericline, where the formations range from the Black Rock Limestone to the Clifton Down Limestone. The oldest exposed beds comprise the main chert of the Black Rock Limestone (p. 7), which is seen in a number of small exposures in the core area of the pericline south-east of the (disused) Uphill and Bleadon Station [ST 3290 5762]; [ST 3275 5765] and in the eastern higher parts of Bleadon Hill [ST 354 576]; [ST 356 573]. Reynolds's record (1917, pp. 26–27) of cherty limestone associated with a thin basalt in the bottom of a small quarry [ST 326 579] on the west side of the main railway line is probably referable to the top of the main chert. Sampling from higher beds on the Black Rock Limestone in the now much overgrown quarry [327579] on the opposite (east) side of the railway line yielded abundant conodonts referable to the Scaliognathus anchoralis Biozone (p. 22). The best exposure of the higher beds in the Black Rock Limestone is seen in crags on the western side of Purn Hill. Here, a short distance below the base of the Black Rock Dolomite, dark grey medium-grained to fine-grained crinoidal limestones yielded Cravenia sp., Siphonophyllia cylindrica (sensu stricto), Sychnoelasma konincki, D. notata, M. cf. magna (cf. Vaughan, 1905, pl. 26, fig. 3)–an assemblage of the Siphonophyllia cylindrica Biozone (p. 21).

The Black Rock Dolomite is thickest at Bleadon, where, as on Brean Down (p. 8), dolomitisation has extended upwards to the base of the well-bedded facies of the Birnbeck Limestone (see below)–a total of some 135 to 140 m of beds–so that it is not possible to separate the Black Rock Dolomite sensu stricto from the overlying dolomites referable to the Clifton Down Group. Scattered fossils however, prove that a considerable thickness of dolomite belongs to the Black Rock Group (cylindrica Biozone), e.g. natural exposures north of Bleadon village [ST 3437 5708], estimated to lie about 18 m above the base, yielded S. cylindrica? and Megachonetes sp., and a roadside quarry [ST 3325 5705] at the southern end of Purn Hill, some 60 m above the base, showed Michelinia cf. megastoma and S. cylindrica (sensu stricto). On the northern limb of the pericline, the Black Rock Dolomite shows a marked local thickening in the extreme west of Oldmixon but typically the thickness varies between about 30 and 40 m. Debris from a pipe-trench [ST 3355 5809] yielded a cylindrica Biozone assemblage with S. cylindrica?, D. notata and M. cf. magna within 10 m of the top of the formation.

On the northern limb of the pericline no subdivision of the rocks between the top of the Black Rock Dolomite and the base of the Clifton Down Limestone was found possible and the whole sequence was designated as Burrington Oolite (p. 11). The lower part of the sequence is best exposed at Hutton or Canada Combe [ST 361 585], and the higher beds are very incompletely exposed in natural outcrops at intervals along the northern slopes of Bleadon Hill. On the southern limb of the pericline, all the exposed beds above the dolomites are seen at Little Down Quarry, Bleadon [ST 342 567], where the section, in ascending sequence, comprises: fawn very massive dolomite and crinoidal dolomitic limestones, seen 30 m; Birnbeck Limestone in typical, well-bedded, fossiliferous facies, 37 m; Goblin Combe Oolite comprising pale grey to grey coarse crinoidal oolitic limestone, usually in one bed, seen 30 m. The Birnbeck Limestone includes a distinctive band of fine-grained dolomite and dolomite-mudstone (0.5 m thick) about 9 m above the base, which may correspond to a similar Bed (F11) seen on Brean Down. The fauna of the Birnbeck Limestone includes K. praecursor, P. murchisoni, Siphonophyllia garwoodi, Sychnoelasma aff. kentensis, D. carinata and D. cf. destinezi.

The Clifton Down Limestone is exposed only on the northern limb of the pericline in the Hutton area. Its junction with the underlying beds is seen in a small quarry [ST 3498 5843], at the northern edge of Hutton Wood, where 2 m of white-weathering, grey, well-bedded calcite-mudstones with common ooliths overlie massive current-bedded oolites (Burrington Oolite). Similar rocks are seen in natural exposures for a further 300 m along the strike to the east. The incompletely exposed succession east of the village has been briefly described above. GWG

Chapter 4 Permo-Triassic

Introduction

The Permo-Triassic rocks of the Weston district are disposed within three main basins, which are separated by outcrops of Palaeozoic rocks but are probably intercommunicating. The basins are partly depositional in origin and partly due to post-Variscan earth movements (p. 98). They decrease in size and depth northwards. The Central Somerset Basin, which is limited to the south and north by the Quantock Hills and the Mendip Hills respectively, is the major Mesozoic structural and depositional feature of the district and contains in excess of 695 m of Permo-Triassic rocks. The smaller basin between the Mendips and the Weston–Worle ridge includes some 260 m of Triassic rocks, and that north of the Weston–Worle ridge accommodates 100 m or more of these beds. Towards the margins of the basins successively younger strata progressively overlap on to the Palaeozoic basement, and locally in the Mendips and the Middle Hope peninsula overlap of the whole succession by Lower Lias strata can be proved. The Permo-Triassic rocks buried a topography of considerable relief, with the development of scree slopes and fan conglomerates immediately adjacent to the Carboniferous Limestone. This Permo-Triassic topography is now being exhumed in the area of the present-day Carboniferous Limestone outcrops and in the north Quantock area.

The oldest Mesozoic rocks exposed at the surface within the confines of the Weston district belong to the Mercia Mudstone Group. However, older Permo-Triassic sandstones were proved in the Burton Row Borehole, Brent Knoll, and were also recorded from the Puriton Borehole (McMurtrie, 1912; Ussher, 1911). The Puriton Borehole [ST 3191 4086] was drilled to a depth of over 631 m. W. A. E. Ussher examined the cores and recorded 'Keuper Marl' to the depth of 388.63 m, overlying 'Upper Sandstones' ('Keuper' or 'Bunter') to 450.19 m. Below the 'Upper Sandstones' a pebble bed or conglomeratic series of rocks, proved to 454.46 m and tentatively equated with the Budleigh Salterton Pebble Beds, rested upon red, fine-grained 'marly' sandstones, referred to the 'Lower Sandstones' (Permian), to the final depth of 631.64 m. The Burton Row Borehole [ST 3356 5208] proved a generally similar sequence, with Mercia Mudstone Group (Grey Marl, Tea Green Marl and 'Keuper Marl') from 422.06 to 906.25 m, Sherwood Sandstone Group ('Upper Sandstones') to 953.44 m, and 'Lower Sandstones' to the final depth of 1105.17 m.

The nomenclature to be applied to the Triassic strata in the British Isles has long been debated. In (Table 6) the names adopted by Warrington and others (1980) are compared with the terms used on the published Weston-super-Mare 1:50000 map and in some other publications on this district. The divisions shown on the map are based upon lithological characteristics. In mapping inland it was found impracticable to trace a line between the Tea Green Marl and the Grey Marl and these strata were mapped as one unit; this is the Blue Anchor Formation of Warrington and others (1980), which includes the Sully Beds of Richardson (1911). For the purpose of mapping, the Westbury Formation and the Cotham Member (Cotham Beds of authors) were combined under the old name 'Rhaetic', though in the revised nomenclature the Cotham Member is placed in the calcareous Lilstock Formation of the Penarth Group. The higher member of the Lilstock Formation, the Langport Member, is made up of the Langport Beds plus the marly bed forming the lowest 0.25 m of Richardson's (1911) Watchet Beds. This member forms the basal part of the topographic feature associated with the Lower Lias, and in mapping it was possible to delineate its base but not its top; for the purpose of the map, the member was included with the Lower Lias. About 6 m of beds that are lithologically part of the Lower Lias are here described with the Jurassic rocks in Chapter 5, although on the basis of the current definition of the base of the Jurassic (Cope and others, 1980) they are of Triassic age.

Sherwood Sandstone Group

Strata of the Sherwood Sandstone Group are not present at outcrop but were proved in the Burton Row Borehole (above), where they comprised pink, purple and red sandstones with some pebbly beds (see Appendix 1).

Mercia Mudstone Group

The greatest known thickness of the Mercia Mudstone Group (formerly 'Keuper Marl') is the 484 m proved in the Burton Row Borehole (Appendix 1, p. 121). In that borehole the group comprised purplish and brownish red weakly calcareous and dolomitic siltstones with green bands, patches and spots. Anhydrite, mainly in the form of nodules, was common at various levels and rock salt was present between the depths of 693.76 and 742.40 m (Whittaker, 1972a; 1973a). At a similar stratigraphical level, the Puriton Borehole proved rock salt within the Mercia Mudstone between about 183 and 219.15 m from the surface.

The topmost 67 m or so of the Mercia Mudstone red marl are exposed in St Audrie's Bay. A low regional dip to the south-west hereabouts has superimposed upon it local small-scale basins and domes, which in Perry Gully are seen to be structures related to the presence of Devonian rocks near the surface. The red marl rests unconformably on Devonian rocks, which form a number of inliers in this area, so that a partly exhumed Permo-Triassic topography is present. Close to the Mendip Hills the marls pass laterally into conglomerates and breccias (the Dolomitic Conglomerate) which are fossil scree and fan deposits.

Towards the top of the red marl sequence hard green siltstone bands are increasingly common; these are calcareous or dolomitic and some display polygonal shrinkage cracks in their upper surfaces. The top of the red marl division is taken at the top of the highest prominent red silty mudstone band, despite the fact that there is a unit (perhaps 6 m thick) of Tea Green Marl and Grey Marl type below that level.

No horizon older than the Mercia Mudstone Group (including beds of equivalent age in the Dolomitic Conglomerate facies) appears to be present north of the Mendips. There the successions thicken rapidly away from the Palaeozoic outcrops towards the centres of the basins.

Knowledge of the thicker successions is confined to the Gasworks Borehole [ST 3290 6068] at Weston and a borehole [ST 364 593] at Locking, both of which may be compared to the Banwell Moor Borehole [ST 399 609] which lies 3 km east of the Weston district (Green and Welch, 1965, pp. 198199). Details of the boreholes are given in Appendix 2 (p. 125). The basal strata everywhere consist of ill-sorted red marly beds with abundant pebbles and fragments of Devonian and Carboniferous rocks (mainly Carboniferous Limestone), and vary from a few metres to about 30 m in thickness. The remainder of the succession, as seen in the boreholes, comprises red mudstones and siltstones, which become sandier and siltier towards the base; a poorly-defined belt ?20 to 45 m in thickness, including many gypsum nodules and fewer gypsum bands, occurs about one-third above the base, and above this, about midway in the sequence, a belt of strata some 30 to 45 m thick with many grey bands interbedded with the red beds is estimated to lie some 60 to 90 m below the Tea Green Marl (p.4.9).

Blue Anchor Formation

The Blue Anchor Formation includes the Tea Green Marl and the overlying Grey Marl. The Tea Green Marl proper is defined as the rocks between the highest prominent bed of red mudstone and the lowest dark grey mudstone bed of the overlying Grey Marl. At St Audrie's Bay it is only 5.18 m thick, and comprises green and dark greenish grey silty mudstones or siltstones. Junctions of green with red blocky mudstones are seldom sharp.

The Grey Marl is defined at its base by the lowest dark grey mudstone band above the Tea Green Marl and at its top by the base of the Westbury Formation. The unit is about 26 m thick at St Audrie's Bay and about 21 m at Lilstock. A great variety of lithologies is present but the sequence essentially comprises alternations of dark grey mudstones, some of which are shaly, and greenish grey or buffish grey silty mudstones or siltstones, which are dolomitic in part. Throughout the sequence occur veins and nodular masses of white, colourless and red-tinted gypsum. At two levels particularly abundant gypsum nodules strung out along the bedding and interbedded with alternating dark grey and greenish grey mudstones are present through 3 to 6 m of the succession. These evaporite-rich horizons are designated C and B in downward sequence in the sections given below (see (Figure 9) and (Figure 10)). Finely laminated mudstones displaying burrows are present in places, as are a few distorted horizons possibly caused by solution of evaporite minerals. Nodular evaporites are less common in the cliff sections of Grey Marl at St Audrie's Bay and Lilstock, but are present in the foreshore section at Watchet and in the Burton Row Borehole. Besides the evaporites, laterally widespread correlating levels are present about 3.0 to 4.5 m above the base of the Grey Marl, and comprise a lower, hard grey siltstone or silty sandstone bed with cavities (presumably once filled with evaporite minerals) and an overlying thin (up to 0.05 m) red marl band. This horizon is designated A (Figure 9).

The topmost few metres of the Grey Marl have yielded Rhaetian fossils, and because of this were classified as Rhaetic and equated with the Sully Beds of Glamorgan by Richardson (1911). The correlation is good but the classification of the upper part of the Grey Marl sequence as 'Rhaetic' is misleading. The topmost Grey Marl strata are lithologically akin to the lower Grey Marl strata, despite the occurrence of relatively thin dark grey shale bands. In the top few metres of the Grey Marl sequence in west Somerset ripple-marked siltstones, sun cracks, erosion surfaces and conglomeratic beds present a strong lithological contrast with the overlying Westbury Formation shales and limestones. The Westbury Formation is readily delineated when mapping inland and forms an excellent mapping unit commonly associated with a good feature. There is palynological evidence that the Grey Marl of south-west England is of Rhaetian age, and that some of the red mudstones underlying the Blue Anchor Formation are also referable to this stage (Appendix 4; Warrington, 1971).

In the northern part of the Weston district the Blue Anchor Formation does not exceed 12 to 14 m in thickness and is much less where the Triassic rocks overlap against Carboniferous Limestone. The exact stratigraphical relationship of these beds to the thick development of the formation in the Central Somerset Basin is uncertain. Rock types similar to those in the basin area are present but the general appearance of the succession is much nearer to that of the Tea Green Marl of that area. Kellaway and Oakley (1933) described exposures in road and railway cuttings at Uphill, adjacent to the Mendips (p. 52). Cored boreholes to the north of the Mendips, just beyond the eastern edge of the district, proved the darkest beds (olive-green to dark greenish grey) to be within the uppermost 4.5 m of the Blue Anchor Formation here approximately 12 to 13 m in thickness. These boreholes showed a red marl bed (0.8 to 1.0 m) about 1.5 m above the base of the Tea Green Marl. Although this bed invites comparison with a red marl widely recognised at a similar stratigraphical position south of the Mendips (see above; also Green and Welch, 1965, p. 68), it has not been recognised elsewhere in the Mendip area and the resemblance may be fortuitous. A discontinuous horizon of celestine. or calcite-celestine nodules is present at the base of the Tea Green Marl or within 1 m below it in the red marls within a short horizontal distance of the Carboniferous Limestone outcrops in the northern area. This may represent either one or both of the horizons of nodular evaporites in the area to the south (see above) or the east (Green and Welch, 1965, p.68, 167; Nickless and others, 1976, pp. 69–75). It is thought that the celestine is an early diagenetic replacement after gypsum or anhydrite. Later replacement of the celestine by calcite and quartz has also occurred. The presence of Carboniferous Limestone nearby appears to favour these forms of diagenesis, which have not occurred in the more central parts of the basins. The primary minerals may have formed in a sabkha-type environment (Nickless and others, 1975). Salt pseudomorphs were recorded by Bristow and Etheridge (1873) in a bed about one-third below the top of the Tea Green Marl–Grey Marl succession at Uphill. In view of the similarities between the northern and basinal successions and, particularly, in the absence of clear evidence for severe intra-formational erosion, it is suggested that any differences between them are due more to differences in the marginal and basinal environments, and in particular the rate of sedimentation, than to undetected major non-sequences or unconformities.

Differences of opinion exist about the environment of deposition of the Mercia Mudstone (Warrington, 1974a; Arthurton, 1980). One view sees its accumulation on desert plains in association with seasonal rainfall flooding into salt lakes, and with periods of desiccation produced by evaporation, resulting in the formation of evaporite minerals. Another view favours its accumulation in a hot, shallow inland sea with high salinity resulting from concentration by evaporation. The recognition by Elliott (1961) of sedimentary structures, such as micaceous bedding planes, mudstone pebbles, slip-layers, miniature ripples, slumped beds and pseudomorphs after halite in the Mercia Mudstone suggests deposition from water. Dumbleton and West (1966, p. 17) carried out mineralogical determinations and concluded that the minerals present are consistent with the origin of the Mercia Mudstone from 'erosion debris of surrounding mountains, which was deposited in a large inland basin having internal drainage and a hot climate in which dry periods alternated with heavy rains'.

The locally variable lithologies of the Tea Green Marl proper, with burrows, laminated beds, local intraforma­tional conglomerates and nodular sulphate minerals, herald the more extensive development of similar features in the overlying Grey Marl. The sediments suggest incursions of the sea across south-west England, possibly from the region of the present Bristol Channel or English Channel. The Grey Marl presents evidence, such as the occurrence of porphyroblastic sulphate nodules associated with car­bonaceous (?algal-mat) mudstones, suggestive of supratidal deposition in a sabkha environment (see also Stevenson, 1970; Sellwood and others, 1970; Stevenson and Warring­ton, 1971). Alternating with the supratidal deposits are various lithologies (including laminated dolomitised siltstones) which are compatible with deposition in a low-energy intertidal environment. Marine fossils are also present in these beds. The stratigraphy indicates numerous incursions and retreats of the sea and the gradual trans­formation to a fully marine environment near the base of the Penarth Group. The lowest beds of Tea Green Marl and Grey Marl type occur interbedded with red mudstones near the top of the red marl sequence, and represent the early establishment of sabkha conditions in the Mercia Mudstone of Somerset. However, these conditions did not persist long and were followed by re-establishment of red marl con­ditions. The presence of thin red bands in the Blue Anchor Formation attests to the brief recurrence of red marl conditions between dominant periods of sabkha-type de­position. Only late in the accumulation of the Grey Marl did more fully marine conditions become established prior to the main marine transgression. The stratigraphy illus­trates well the change from the 'continental' conditions of the earlier Triassic by repeated incursions of the sea. AW, GWG

Penarth Group

The Penarth Group is divisible into three units on the basis of lithology. In upward sequence, these are the Westbury Formation, the Cotham Member and the Langport Member. The last two make up the Lilstock Formation. Much of the Watchet Beds of Somerset, as defined by Richardson (1911), is—for reasons given below (p. 59)—regarded as part of the lowest Lias.

Pioneer work by Richardson, including detailed strati-graphical measurement of the 'Rhaetic' in south-west Britain, established a sound framework for subsequent researches. He introduced a set of stratigraphical terms which came into general use, substituting the term Westbury Beds (now Westbury Formation) for 'Black Shales', and dividing strata previously called White Lias into several parts, amongst which were the Cotham Beds and Langport Beds. However, for Somerset at least, some of Richardson's work required revision.

Although the Penarth Group forms a readily mappable unit in the northern area of the Weston district, the only detailed stratigraphical information available is from road and railway cuttings east of Uphill (see especially Kellaway and Oakley, 1933, but also Wright, 1860; Woodward, 1876; Bristow and Etheridge, 1873). Mapping evidence indicates that the succession in the Uphill area is probably typical for the area as a whole, though the thicknesses may be reduced north of the Weston–Worle ridge.

Westbury Formation

The Westbury Formation in the south comprises mainly very dark grey shaly mudstones with dark grey argillaceous limestones, commonly lenticular and associated with fibrous calcite ('beer') and subordinate rippled sandstone beds. The mudstones contain a bivalve fauna, including Rhaetavicula contorta, and vertebrate remains are locally relatively abundant in the sandstones. Several 'bone beds' are present, particularly in the lower part of the formation. The thickness of the Westbury Formation is just over 10 m at Lilstock, but varies considerably from place to place. Individual beds correlate well with Richardson's (1911) description of the sequence. The incompetent mudstones tend to act as a locus for strike faults and the contact with the overlying Cotham Member is commonly faulted.

In the Uphill sections the Westbury Formation varies in thickness between 5.62 and 7.09 m. Dark shales rest with a sharp junction on the Grey Marl. There is a single (Ceratodus) bone-bed, 0.13 to 0.2 m in thickness, at about 1 m, or slightly less, above the base of the formation. This succession contrasts with that farther north in the Bristol area, where the Ceratodus bone-bed normally marks the base of the formation, and also with the sequence, in the Central Somerset Basin, where several bone-beds are present in the lower part of the succession. Between Locking and Banwell, east of the district margin, the bone-bed occurs at the base, and the formation's smaller thickness (5.03 to 5.28 m) may have resulted from overlapping of the lowest shale bed (p. 58). Near the Carboniferous outcrop thickness variations appear to be related to distance from the shoreline. Various discontinuous argillaceous limestone bands have been recorded in the Westbury Formation.

Lilstock Formation

The Cotham Member comprises distinctive pale grey or greenish grey calcareous mudstones, limestones, siltstones and sandstones, which are 1.45 m thick at Lilstock (Plate 9). The lower part of the member contains contorted slumped calcareous siltstones, rippled beds and a laterally persistent horizon bearing a network of polygonal shrinkage cracks (Plate 10). In section, the polygonal cracks are seen as narrow vertical pipes which penetrate the contorted beds to various levels and which are filled with sediment similar to that occurring in overlying beds. The bed recognised by Richardson (1911) as the equivalent of the Cotham Marble at Lilstock, but not at St Audrie's, is taken to be the basal bed of the Langport Member (Whittaker, 1978).

The Cotham Member is 2.19 m thick in the Uphill road cutting, with the usual calcareous mudstone lithologies including a rather atypical representative of the Cotham Marble (0.25 m thick) at the top. The base is sharp and in the railway cutting Wright (1860) recorded a band (0.04 m thick) of 'dark grit containing scales and teeth of Fish' at this level, but the band is not present in the road cutting. Thicknesses of the Cotham Member in the Locking–Banwell area are from 1.6 m to about 2 m.

The Langport Member consists mainly of pale grey limestones with interbedded grey or bluish grey mudstones. The limestones are commonly lenticular, nodular or even globose, particularly in the lower part of the member, where some individual limestones are extremely fine-grained or porcellanous and in places laminated. A higher group of two or three limestone beds (with the 'Sun Bed' at the top) weathers to a distinctive creamy white colour and has an irregular base; the top is irregular in places and locally displays U-shaped burrows. The member is 1.37 m thick east of Lilstock, but its average thickness is about 1 m.

Above the Sun Bed there is grey hard and calcareous mudstone ('marl') which contains limestone lenticles up to 0.08 m thick. Up to 0.97 m thick at Blue Anchor, to the west of the present district, these beds are only 0.34 m thick within the district at St Audrie's Bay. They are the lowest part of the 'Watchet Beds' as defined by Richardson (1911); lithologically they are of Langport Member aspect, and they correlate with the similar but much thicker (2.18 m) marly beds at Lavernock (bed A of Richardson, 1905) on the opposite shore of the Bristol Channel. The shaly higher beds of Richardson's 'Watchet Beds' are regarded as the basal Blue Lias beds, which are transgressive (Whittaker, 1978).

The Langport Member is mainly absent in the northern area, although the Sun Bed has been mapped in a very small isolated outlier east of Hutton village. It is not known whether its absence is due to pre-Blue Lias erosion or to non-deposition.

At intervals between Hutton and Uphill, the Tea Green Marl and Penarth Group overlap on to the Carboniferous Limestone, and are in turn overlapped by the Blue Lias. At Middle Hope also, Blue Lias limestones and clays rest directly on the Carboniferous, although borehole evidence indicates that the exposures are virtually at the point of overlap (p. 49). There is no evidence of the marginal pebbly and sandy facies that is characteristic of such situations farther east (Green and Welch, 1965), and it is assumed that the contemporary Carboniferous Limestone outcrops were either submarine or of such low relief and limited extent that no significant erosion of them took place during the overlap phase. AW, GWG

Conditions of deposition

The bone beds of the basal Westbury Formation are associated with pebbly or gritty sandstones which in places yield rather angular pebbles of Grey Marl siltstone, and pebbles of chert and milky quartz, together with abundant bone and tooth fragments that are commonly worn. Small aggregates of pyrite are also present. The poor sorting and the stratigraphical position vis-a-vis the underlying sabkhatype deposits support the view that these beds were deposited in a transgressive littoral zone. However, locally (e.g. on the Watchet foreshore), the lower Westbury Formation sandstones show polygonal cracks on their upper surfaces; these are filled with overlying sediment and are presumably the result of brief emergence and desiccation. Fine wispy laminae of vivid green mudstone alternate with sandstones, siltstones and dark grey mudstones in the lowest 1 m or so, and show small-scale ripples suggestive of current action. The limestones of the lower part of the Westbury Formation are commonly associated with thin (about 10 to 20 mm), rippled, brown-weathering sandstones, and the disarticulated bivalve shells present on upper and lower limestone surfaces suggest the possibility of current-concentrated shell beds. Between these higher energy periods with local emergence occurred periods of quiet, possibly deeper-water sedimentation, represented by the dark grey shales which make up the bulk of the Westbury Formation. The shales are poorly bituminous, pyritous, and contain a fauna mostly of thin-shelled bivalves; they are associated with concretionary limestones which are thought to be of an early diagenetic origin. The conditions of deposition of broadly similar bituminous shales are discussed at length by Hallam (1975). Fossils tend to be concentrated in bands, and between these the water may have been more stagnant, anaerobic or possibly slightly oxygenated, but not necessarily deep. The repetition of similar lithologies throughout the whole thickness of the Westbury Formation, including some thin vivid green wisps or laminae, suggests the presence of three sedimentary cycles.

The Cotham Member in places rests irregularly on the Westbury Formation, the junction being commonly associated with calcitic 'beef'. A sharp change in depositional conditions, probably accompanied by at least local erosion, from a relatively low-energy environment to one of higher energy, is indicated by the sediments. These beds are calcareous, in contrast to most of the Westbury Formation and, broadly speaking, they fine upwards from sandy or silty to argillaceous. Ripple marks are common and sand-ripple lenticles occur in mudstones; contorted or deformed beds in the lower part of the Cotham Member are interpreted as slump or dewatering structures formed before the sediments were consolidated, possibly activated by contemporaneous fault movements. The polygonal cracks which affect these beds are interpreted as deep desiccation cracks, signifying emergence; they are filled with sandy material, identical to that in succeeding beds, which are rippled.

The porcellanous or laminated limestones in the succeeding basal Langport Member are in places similar to atypical limestone types found in Blue Lias beds of Caloceras johnstoni Subzone age (see p. 59). Local thickenings and thinnings of individual beds related to small-scale contemporary palaeofaults are observable at Lilstock. The environment of deposition is interpreted in terms of warm, very shallow-water, shelf lagoons. Tectonic movements resulted in the development of intraformational conglomerates and irregular bed junctions, and porcellanous limestone rims or edges to non-porcellanous limestone (cf. the Dorset White Lias in Hallam, 1960b) suggest the possibility of emergence and rapid recrystallisation of calcium carbonate to produce porcellanous rock. Periodic deepening of the water may account for the presence of interbedded grey mudstones, and the bored, sun-cracked, upper surface of the Sun Bed signifies emergence and desiccation. A brief return to deeper water but relatively high-energy conditions produced Langport Member silty mudstones (the basal part of the Watchet Beds in the sense of Richardson), and was followed by the deposition of the fissile shales of the transgressive basal Blue Lias.

Minor cycles are recognisable within all the three principal divisions of the Penarth Group, but particularly in the Westbury Formation and the Langport Member. The minor cycles are best illustrated by the Blue Lias sequence (see below), but those of the Penarth Group are essentially similar, probably differing (in limestone-shale ratios, etc.) only because these sediments were deposited in a different environment to the Lower Lias. Like the Blue Lias minor rhythms, the earlier cycles begin with shale and pass upwards through mudstone, limestone and mudstone again. The cycles are regarded as indicating shallowing of the depositional environment upwards to the limestone member, and then deepening through the higher mudstone to the sharp base of the next higher shale; however, not all the cycles are fully developed. The rhythm is observable on a large scale also; the upward passage from shaly Westbury Formation through marly Cotham Member to the more calcareous Langport Member (White Lias) and marly succeeding beds, parallels the minor cycles. The Penarth Group cycle is succeeded by the transgressive basal shale of the Blue Lias. Correlation of these sequences along the Somerset coast is illustrated in (Figure 11) and (Figure 12). AW

Details

Mercia Mudstone Group

Beds below the Blue Anchor Formation

Watchet

East of Watchet harbour, red marls with green siltstone bands dip north and are traceable to Helwell Bay [ST 0786 4336], where they are faulted against Lower Lias mudstones of semicostatum Zone age. Because of faulting on the foreshore, thicknesses are difficult to determine, but it is estimated that over 40 m of the topmost part of the Mercia Mudstone are present. A noteworthy feature of the stratigraphy hereabouts is the most westerly exposure of Grey Marl-type lithologies in the red Mercia Mudstone; in a sea cave [ST 0777 4340], west of the faulted point of Helwell Bay, occur very dark grey mudstones associated with green blocky mudstones and greyish green siltstones. The lithologies are similar to those in the Grey Marl and presumably record an early establishment of the Grey Marl environment, which quickly reverted to that of the red and green marl. This horizon has also been located at St Audrie's Bay and in the Burton Row Borehole (p. 123).

St Audrie's Bay

The best exposures of the red Mercia Mudstone in the district are those in St Audrie's Bay, where the cliffs west of the Blue Ben Fault [ST 1055 4312] to [ST 1200 4375] show about 67 m of south-west-dipping beds which pass up into the Tea Green Marl; the following section was measured below that unit (Figure 9):

It will be seen that the lower part of the section examined comprises typical red, blocky and silty mudstones with a few thin, green siltstones or mudstones interbedded. Above the level of bed A6 the green siltstone or mudstone bands become much more numerous, but they are still subordinate to the red bands. At about the level of bed A16 upwards for 3.5 m, green mudstones and siltstones become the dominant element in the stratigraphy with poor, thin red bands. Between beds A19 and A23 green and greyish green mudstones and siltstones are present, with no red bands, through 3.5 m of strata. In the middle of this unit are two dark grey marl bands identical with those from the Grey Marl sequence above. The dark grey mudstone bands are the lowest beds in this section to yield indigenous Triassic miospores (Warrington, 1974b), and they occur at the same stratigraphical level as similar beds within the red Mercia Mudstone at Watchet (see above) and in the Burton Row Borehole (p. 123). Thus, within the red marl sequence is a small-scale unit which shows that Tea Green Marl and Grey Marl conditions were established at an earlier date than has hitherto been recognised. The greyish green silty mudstones of bed A23 contain red patches, and strata above this level (to the base of the Tea Green Marl) comprise alternating beds of red and green mudstones and siltstones, showing that conditions reverted for a short time to those under which the red and green marls formed. The red marl-Tea Green Marl junction is taken at the top of the highest prominent red band in the section, although it has been noted that a thin, but laterally persistent, red band occurs near the bottom of the Grey Marl sequence at Watchet, St Audrie's and in the Burton Row Borehole.

Lilstock-Knighton–Benhole Farm–Stert Flats

Alternating red and green marls high in the red Mercia Mudstone succession occur south of the major fault at Lilstock Bay [ST 1771 4525]. The beds dip south at about 14° and hard silty beds form reefs that strike roughly east to west across the foreshore.

Inland, a small faulted inlier [ST 1804 4475] occurs east of Lilstock, and two elongate east-west strips of red Mercia Mudstone occur as faulted inliers and floor valleys between Knighton [ST 1933 4446] and Benhole Farm [ST 1977 4575]. The top few metres of the alternating red and green mudstones crop out along the valley sides beneath the Tea Green Marl. Members of the Geography Department, University College of Wales, Aberystwyth, pointed out to the writer an outcrop of undoubted Mercia Mudstone on the Stert Flats [about 2406 4603] north of Catsford Common; this outcrop comprises thin alternating bands of red and green mudstone, presumably from the upper part of the Mercia Mudstone sequence. AW

Uphill–Bleadon–Shiplate

In the Uphill area, west of the A370 road, the main outcrop of the red marls and the overlying Tea Green Marl occurs in the eastern part of the ridge immediately south of the village. Good exposures of red marl can still be seen in the side of the A370 road cutting [ST 326 584], where about 30 m of red marls were formerly exposed beneath the Tea Green Marl. Green mottling is noticeable in places and green bands occur at fairly regular intervals; Kellaway and Oakley (1933, p. 475) remarked that 'in the centre of each of these there is usually a nodular layer of 'box-calcite', hollow nodular masses lined with crystalline calcite and often filled with loose crystals of calcite… principally minute scalenohedra'. These nodules may be replacements of evaporite minerals (see p. 39). In a road cutting [ST 3231 5824] red marls just below the junction with the Tea Green Marl, show crystalline calcite nodules, probably after an evaporite mineral (cf. p. 53). Kellaway and Oakley (1933, p. 472) recorded that the red marls in part of the east bank of this cutting were thrown into asymmetrical folds with an amplitude of 1.8 m or more and approaching monoclines in form. These structures are now no longer exposed. Over much of the ridge, Tea Green Marl rests directly on the Carboniferous Limestone but small inliers of red marl were augered in three areas.

The south side of the Mendips, east of the Western Region main-line railway, represents a partially exhumed Triassic land surface formed of Carboniferous Limestone but with intervening valleys still infilled with Triassic rocks readily identified by their bright red soil and scattered outcrops. The Triassic rocks immediately adjacent to the Carboniferous Limestone comprise well-bedded to massive red marly conglomerates and breccias, pebbly marlstones and some interbedded marls, but these pass laterally within a short distance into red marl. Exposures in the Triassic rocks are confined to the harder marginal facies. The best exposures are on the south side of the disused Uphill Railway Quarry [ST 326 579], a road cutting leading north from Bleadon [ST 3405 5715], numerous natural outcrops on the hillside west of Shiplate Slait [about 352 574 eastwards to the district margin] and an old quarry at The Copse, Wonderstone [ST 3503 5678]. At the last locality over 3 m of well-bedded red and yellow, mostly fine-grained, conglomerate has been used locally as a building stone. The Carboniferous Limestone is in places traversed by joints infilled with red Triassic debris (p. 99); for example, at Little Down Quarry, Bleadon [ST 342 567] and on the south side of Purn Hill [ST 3325 5702], where a small roadside quarry exposes an infilled fissure, up to 2 m wide and striking at 027°, which may represent a pre-Triassic fault plane.

Oldmixon–Hutton–Locking

Eastwards from Oldmixon, the northern side of the Mendips largely represents an exhumed Triassic cliff-line in Carboniferous Limestone. Good exposures of the Dolomitic Conglomerate marginal facies can be seen south of Lower Canada [ST 3580 5850]; [ST 3590 5848]; [ST 3570 5830]. Here the higher–and stratigraphically younger–outcrops comprise yellow conglomerates and breccias, which give way to red sediments lower down, with a zone of mixed colours between. The rocks have been worked locally for minerals (possibly ochre) and probably also for building stone. The upward passage from red into yellow rocks is widespread in the Mendips farther east and is due to an increasing degree of secondary dolomitisation (Green and Welch, 1965, p. 65). The yellow facies is best seen in a small quarry [ST 3583 5923] at Ludwell Farm, where 4 to 5 m of yellow fine-grained dolomite are exposed, banked against Carboniferous Limestone. The dip, mainly depositional, is from 10° to 20° to the north. Dr J. R. Hawkes reports on a thin section (E36610) that 'the rock consists mainly of fine granular dolomite crystals averaging around 0.015 mm in size with some large rhombohedral dolomite crystals up to about 0.1 mm in size… also a few irregular patches of calcite which may represent original fossil debris, or perhaps sparry limestone.'

A borehole at Locking [ST 3637 5934] proved 145 m of Triassic deposits, mainly red marls, overlying Carboniferous Limestone. The base of the Tea Green Marl is estimated to have lain originally about 30 m above the top of the borehole (see p. 38 and Appendix 2). The sub-Triassic surface of the Carboniferous rocks, thus apparently slopes northwards at about 35° between the nearest limestone outcrop to the south and the borehole, an inclination appreciably in excess of the present-day nearby hillside slopes.

Weston-super-Mare–Kewstoke

A borehole at Weston Gasworks [ST 3290 6068] proved red rocks to a depth of 242 m, overlying Carboniferous Limestone. The base of the Tea Green Marl is estimated to have lain some 15 m above the top of the borehole (see p. 38 and Appendix 2). The Carboniferous Limestone of the Weston–Worle ridge represents an exhumed Triassic ridge, as shown by the presence of Triassic rocks round its edges, including many small partially exhumed outwardly radiating Triassic valleys. Boreholes for the Weston Technical College [ST 3175 6180], near the sea front, proved red marl beneath thick drift (p. 92).

Debris from graves in the town cemetery [ST 3285 6190] included fine-grained breccias and marlstones with 'potato stones' (quartz-lined geodes) adjacent to the contact with the Carboniferous Limestone. This contact is seen at Worle, in a small quarry [ST 3566 6318] and in a good roadside section [ST 3522 6343] adjacent to the Worle Hill Quarry; the Triassic rocks comprise yellow dolomite and yellow dolomitised conglomerates, respectively.

At Home Farm, Norton, a bank section [ST 3423 6372] exposed, immediately beneath the Tea Green Marl, 1.5 m of red marls with green patches associated with numerous nodules of crystalline calcite, probably after evaporite minerals (p. 39). Dr J. R. Hawkes reports that the nodules 'bear vughs and crystallisation seems to have concentrated around... and proceeded inwards to these centres' (i.e. the vughs).

Middle Hope

The only exposed Triassic rocks on the Middle Hope peninsula, apart from a small patch [ST 3452 6644] of yellow fine-grained dolomitised conglomerate of doubtful affinities between the top of the Carboniferous Limestone and the base of the Lower Lias, are infillings up to 2.4 m in width in prominent north-west-trending joints or small faults at St Thomas's Head. A borehole 400 m west-south-west of Woodspring Priory started in Lower Lias at about 6.1 m above OD and ended in red marls and conglomerates at 39.6 m below OD; the top of the Triassic was not recorded but over 25.7 m of red rocks were noted in the lowest part of the borehole. The contact between the Carboniferous Limestone and the Lower Lias crops out at 33 m above OD at a distance of 135 m to the north-north-west of the borehole. The southward gradient of the Carboniferous Limestone surface is thus more than 1 in 2. GWG

Blue Anchor Formation

Watchet

The following section was measured across the foreshore reefs at a place [ST 0770 4355] some 500 m E of Watchet harbour (Figure 10):

In places it was necessary to calculate thicknesses by measuring horizontally (normal to the strike) across the beds, and then noting the dip; 'control' measurements at places where normal vertical bed-by-bed measurement was possible showed that this technique did not greatly distort the thicknesses.

St Audrie's Bay

The Blue Anchor Formation is well exposed in the cliff immediately west of the Slip [ST 1053 4314] at St Audrie's Bay (Figure 9). Across the foreshore, hard bands protrude above the level of marine beach deposits and dip uniformly at about 12° to the west-south-west. It is noteworthy that the Grey Marl hereabouts lacks the bedded nodular gypsum so common and characteristic of these beds farther west. However, veins of gypsum are not uncommon at St Audrie's Bay. The cliff face is prone to sudden, large rockfalls from the Grey Marl; large scree mounds are frequently visible at the foot of the cliff but may be washed away completely by the next high tide. The following section was measured west of The Slip:

From the topmost 1.52 m of the Grey Marl sequence Bristow and Etheridge (1873) recorded fish scales, annelid burrows and 'Gervillia' praecursor.

At the eastern end of St Audrie's Bay several faulted patches of Tea Green Marl and Grey Marl [ST 1144 4376]; [ST 1146 4384]; [ST 1163 4376] are associated with the western closure of the Blue Ben Syncline and the Blue Ben Fault. In the cliff [ST 1204 4378], disturbed Tea Green Marl and Grey Marl are caught up along the major Blue Ben Fault plane.

Quantock's Head–Kilve–Lilstock

Two elongate and narrow fault-bounded patches [1355 4448] of Tea Green Marl and Grey Marl occur near low water mark on the foreshore between Quantock's Head and Kilve Pill.

At Lilstock Bay, Tea Green Marl and Grey Marl are preserved in a syncline which is exposed in the cliff. The beds are accessible on the eastern limb of the structure [1778 4518] but the succession is incomplete near the base because of faulting (Figure 10).

Nodular anhydrite has not been recorded from this locality. The topmost bed (bed No. 1) of Richardson's (1911, p. 30) 'Sully Beds' at Lilstock is a dark grey earthy limestone with an irregular base, and is best classified with the Westbury Formation on lithological grounds. The 'Sully Beds' yielded bivalves and fish remains to Richardson.

Blue Ben–Kilton

As on the coast, small faulted patches of Tea Green Marl and Grey Marl are caught up along the major Blue Ben Fault and associated faults [ST 1210 4374]; [ST 1243 4369]; [ST 1266 4381]; [ST 1288 4368]; in some places they form low, mound-like features, but in others they are revealed only by augering along the line of the fault. East of Court House [ST 1402 4361], the Grey Marl forms the face of a strong scarp which is capped by the Penarth Group and Lower Lias; the grey and green marls are faulted out against Lias clays on the western side of East Wood but can be traced as a strong scarp feature into Kilve. The Grey Marl-Penarth Group junction hereabouts is commonly faulted.

At Kilve [ST 1500 4336] the Tea Green Marl and Grey Marl crop out in the stream bed beneath gravels, and in the valley sides south of this place. At least 4.5 m of typical greenish and buffish grey marly siltstones are visible close to the old mill [ST 1507 4298].

Lilstock–Wick Moor

North-east of Lilstock, a strong ridge composed of Tea Green Marl and Grey Marl forms a faulted inlier [ST 1720 4527]. Eastwards from the exposure of Grey Marl at Lilstock Bay [ST 1780 4521], a fault-bounded strip of these beds occupies the northern and higher slopes of a valley as far as the vicinity of North Moor [ST 2075 4548]. Augering hereabouts suggests that the red mudstones pass up gradually through alternating red and green siltstone bands into typical Tea Green Marl and Grey Marl. However, the junction between Grey Marl and the overlying strata is faulted almost, throughout the length of the crop so that for 1.35 km [ST 1826 4534] to [ST 1961 4538] Grey Marl beds are in contact with the Pre-planorbis Beds of the Lias; eastwards from the last locality the junction between Grey Marl and the Westbury Formation appears to be normal, although the presence hereabouts of lithologies of Westbury Formation type in the highest part of the Grey Marl precludes the precise delineation of this boundary by augering. The eastward continuation of this outcrop of Blue Anchor Formation is terminated against the north-east-trending Hinkley Point Fault.

Motorway site investigation boreholes and exposures on the scarp face of the Polden Hills have shown the combined Tea Green Marl and Grey Marl hereabouts to be about 21 m thick. AW

Mendip Hills–Middle Hope

The road and railway cuttings at Uphill have provided the most detailed information on the Blue Anchor Formation succession (see especially Kellaway and Oakley, 1933; Bristow and Etheridge, 1873), but at the present time only part of the former section is visible. The combined thickness of Tea Green Marl and Grey Marl in the road cutting [ST 325 582] was 12.5 m, of which Kellaway and Oakley considered that the Grey Marl accounted for 2.4 m. A distinctive brecciated bed was thought to represent a non-sequence, but it may be a diagenetically altered evaporite-rich bed. The overlying strata comprised 'greenish-grey marls and marlstones alternating with thin seams of blackish shale', and the authors could find 'no clear evidence of non-sequence' at the top. The succession described from the road cutting is apparently strictly local because it cannot be clearly identified in the detailed published accounts (e.g. Bristow and Etheridge, 1873) of the nearby railway cutting section, where the total thickness was 11.0 m and the darkest beds (dark greenish grey, fissile in places) occurred between 8.7 and 4.8 m below the top. Kellaway (in Kellaway and Oakley, 1933) suggested that a conglomerate at 2.03 m from the top (Wright, 1860) or a 0.1-m 'tough siliceous limestone with nodular pyrites' 1.93 m from the top (Bristow and Etheridge, 1873), might represent the basal breccia bed of the Grey Marl in the road cutting. Up to about 20 greenish grey and buff indurated marlstone bands (0.5 to 0.25 m thick) were recorded within the sequence.

West of Hutton three outliers of Tea Green Marl form small whale-backed ridges. A cutting [ST 344 588] up to 6 m deep was made through one of these ridges during the 1939–45 war to provide access to a factory to the north. Although much grassed over, exposures of red and Tea Green Marl, dipping at 8° to the south and including a small strike fault, can still be seen. Nodules of celestine and some calcite occur at intervals at the base of the succession as seen in ditches and on ploughed ground (p. 39).

Between Hutton and Upper Canada the Tea Green Marl was augered beneath the Penarth Group. It is much reduced in thickness as compared with the areas to the west, and is apparently overstepped by the Penarth Group at the eastern end of the outlier. GWG

Penarth Group

Watchet–Doniford

The following section in the Westbury Formation was measured on the foreshore [ST 0765 4362] near Helwell Bay, just east of Watchet harbour. The bed numbers adopted are those used by Richardson (1911) in the adjoining cliff-section (Figure 11).

The lower parts of bed 8 and beds 10 to 12 are commonly disturbed and are vertical in places. To obtained measurements of thickness from the thicker shales it was necessary to record outcrop width and dip and then to calculate the thickness. This procedure may have resulted in slight errors in the recorded thicknesses.

In Doniford Bay [ST 0827 4345] the whole of the Penarth Group is exposed in foreshore reefs. The following section was measured across the foreshore; bed numbers are those of Richardson's (1911) St Audrie's section except that his Gotham Bed 1 is here placed in the Langport Member.

Thicknesses of the thicker shale beds were calculated from measurements of outcrop width and dip.

St Audrie's Bay

Penarth Group strata are present in the cliff at the western end of St Audrie's Bay [1044 4307] and can be traced across the foreshore [1022 4343] where they strike out to sea in a northwesterly direction. The following section is exposed at the foot of the cliff' [ST 1032 4327] west of St Audrie's Slip; the bed numbers being those of Richardson (1911) with the proviso noted on p. 53:

The lowest stratum in bed 3 of the Cotham Member displays flow folds and contorted structures in its upper part. Apparently these structures formed before consolidation of the sediments and they are attributed to slumping. What appear to be narrow neptunean dykes are visible in the cliff face at about this stratigraphical level. The pipes of sediment originate in the topmost sandstone of bed 3 (Cotham Member) and penetrate down to 0.46 m into the underlying beds. The pipes are commonly between 0.01 and 0.03 m wide, and lensoid in vertical section with points at top and bottom. On bedding planes the structures form a polygonal network (Plate 10). They are filled with sediment like that described from the topmost sandstone in bed 3, and are interpreted as deep desiccation cracks (Whittaker, 1978).

The section given below was measured in the cliff [ST 1043 4306] behind and to the south of a patch of landslipped strata visible at beach level. At this locality, as in places at the previous locality, the Cotham Member is brought into contact with the Westbury Formation by strike faults; consequently the Cotham Member may be thicker than recorded. Richardson's (1911) bed numbers are again used, subject to the amendment noted on p. 53.

Beds at this level were obscured by landslipped strata, but Richardson (1911) measured the whole section. The section below overlaps with that given above and is based upon Richardson's account and bed numbers (1911, p. 23), but with revised stratigraphic nomenclature in square brackets and metricated thicknesses

According to Richardson (1911), the Westbury Formation strata below bed 17 were poorly exposed, necessitating reliance on the earlier measurements of Bristow and Etheridge (1873).

At the eastern end of St Audrie's Bay [ST 1147 4376], small faulted patches of Penarth Group are associated with the Blue Ben Fault. The strata are similar to those already described from the western part of the bay.

Quantock's Head–Kilve

Two small faulted patches of Penarth Group are preserved on the foreshore [ST 1376 4446]; [ST 1384 4465] north-west of Kilve. Exposure is poor owing to gravel, mud and seaweed cover, but lithologies are like those to the west and thicknesses are similar.

Lilstock Bay

At Lilstock Bay the strata are folded into a syncline and Penarth Group beds are accessible near the axis of the fold. The following section was measured in a landslipped block [ST 1768 4512] on the eastern limb of the fold (Figure 11) and (Figure 12), and Richardson's bed numbers have been employed, except for the modification noted on p.53.

Bed 23 of the above section was numbered bed 1 of the 'Sully Beds' by Richardson. Lithologically it is more akin to the Westbury Formation limestones and is here classified with that group. Vertebrate remains are present in it and also in bed 15, which in places is very sandy.

A similar sequence is present east of the headland at the eastern end of Lilstock Bay, where the following section was measured in the cliff [ST 1797 4527]:

The easternmost exposure of the Penarth Group along the coastal section is in the cliff east of Lilstock Bay. Many east-westtrending faulted slices of ground are present in this area and Penarth Group beds are exposed on the foreshore dipping westwards in three places [ST 1788 4533]; [ST 1828 4544]; [ST 1838 4545], but are not well seen because of mud and seaweed cover. The cliff section [ST 1846 4542] (Plate 9) is given below:

To the west of this area the contact of the Cotham Member with the Westbury Formation is irregular and faulted in places.

Bedding planes exposed on the foreshore in this vicinity display polygonal markings in beds 2–3 of the Cotham Member similar to those noted from other localities above; these structures are interpreted as deep shrinkage cracks. Just to the west of the section given above, limestones in the lowest beds of the Langport Member (beds 4–11) thicken to 0.61 m in large lenticular spheroidal masses. It is mainly the middle bed which thickens to such exaggerated proportions; the basal limestone bed thickens and thins slightly but is fairly persistent along the outcrop. The basal limestone was correlated by Richardson (1911, p. 29) with the Cotham Marble, but it is indistinguishable from limestones at this horizon elsewhere along the coast at places where Richardson stated that the Cotham Marble equivalent was absent. Lithologically the bed is akin to the higher limestones of the Langport Member throughout the coast section and consequently it is here grouped with this unit.

Inland exposures

Inland exposures of the Penarth Group are poor. South-east of Court House [ST 1410 4347] a band of Westbury Formation black shales and limestones can be traced round the southern scarp slopes of a prominent hill. Faulted patches and inliers of Penarth Group are present hereabouts [ST 1438 4328] but the main crop is traceable into the valley at Kilve, where Westbury Formation shales are exposed in the bed of the stream [ST 1500 4339].

East and north-east of Lilstock are fault-bounded patches of Penarth Group [ST 1734 4521]; [ST 1790 4514]; [ST 1822 4520]; [ST 1857 4497] mainly proved by augering. Near the ridge top, south of Benhole Farm [ST 1968 4540], are faulted Westbury Formation beds; farther east [ST 2006 4538] to [ST 2073 4549] the Penarth Group is traceable as a continuous band, with Cotham Member limestones resting on black shales and dipping north or north-north-west. AW

Mendip Hills

The succession in the road and railway cuttings at Uphill was described by Kellaway and Oakley (1933) but at the present time only part of the road cutting is visible. Kellaway recorded a faunal list from these beds. In the road cutting the Westbury Formation measured 7.09 m on the east side and 6.50 m on the west side, the difference arising mainly from the decreased thickness of shales in the middle of the succession. This compares with a thickness of 5.62 m in the railway cutting where, however, the decreased thickness was due to an overall thinning of the succession. This thinning was probably related to the increasing nearness of the Rhaetian shoreline to the east and south-east.

The Cotham Member is 2.19 m thick in the Uphill road cutting, and comprises yellow marls with three interbedded blue-hearted splintery limestones or calcite-mudstones (each about 0.25 m thick) which are thinly laminated and show desiccation cracks in places. The uppermost bed–the presumed equivalent of the Cotham Marble–contained Modiolus sp., Protocardia sp., broken bivalve shells, Gyrolepis alberti, fish bones, scales and teeth. Kellaway (in Kellaway and Oakley, 1933, p. 481) also noted annelid borings in this bed. The overlying brown unfossiliferous shale (seen 0.02 m) is of doubtful affinity but possibly Lower Lias: the Langport Member (White Lias) is thus presumed to be absent here. In the railway cutting the Cotham Member was not separately distinguished from the overlying Lower Lias (Bristow and Etheridge, 1873).

At Upper Canada, the Penarth Group is overstepped by the Lower Lias within 50 to 100 m of the Carboniferous Limestone outcrop. A similar situation must also obtain at Middle Hope, underneath the Lower Lias which crops out in the southern part of the peninsula (p. 49).

Additional information on the Penarth Group is provided by boreholes sunk in 1966 for the M5 motorway route investigation in the Locking–Banwell area, about 1.7 km E of the district margin, and examined by Messrs G. A. Kellaway, R. J. Wyatt and I. H. S. Hall. The Westbury Formation [about 379 594] varies from 5.03 to 5.28 m in thickness, with a thin bone bed at its base (unlike Uphill, p. 42). Complete core recovery of the Cotham Member (1.63 m) was obtained in only one borehole [ST 3783 5996], which showed Cotham Marble (0.10 m) at the top. No Langport Member (White Lias) was present (but see p. 42). GWG

Chapter 5 Jurassic

Within the Weston district only Lower Jurassic rocks and a small patch of Middle Jurassic rocks are preserved. Excellent sections in Lower Lias rocks are present along the north-facing cliffs and foreshore between Watchet and Hinkley Point, but otherwise the Jurassic strata are poorly exposed. The extensive alluvial deposits of the Somerset Levels blanket much of the Jurassic outcrop within the district, but the stratigraphy of the underlying strata is known in some detail as a result of drilling. Borings in the vicinity of Brent Knoll have penetrated much of the Lower Jurassic sequence and link, stratigraphically, the coastal exposures of Lower Lias with the earliest Aalenian (Middle Jurassic) strata preserved at the summit of the hill.

The Lias sequence hereabouts is thick compared with the Lower Jurassic of the Mendip–Radstock area to the east. As a whole it is predominantly argillaceous and of offshore or basinal facies. Limestone-rich or calcareous beds are present at several levels and there is only one major development of silt-grade material in the higher part of the Upper Lias.

Lower Lias

The Lower Lias was proved to be 373 m thick in the Burton Row Borehole. This figure excludes the Langport Member, but does include strata referred by previous workers to the 'Watchet Beds', which have in the past been classified with the 'Rhaetic' (see p. 42). The term 'Watchet Beds' was introduced by Richardson (1911) for a few metres of grey, strongly fissile shales which overlie the White Lias (Lang-port Member) of the Somerset coast. In the writer's view (Whittaker, 1978) Richardson's lithological correlation of these shales with the silty marls above the 'White Lias' limestones at Lavernock in south Glamorgan is erroneous. These latter beds comprise mudstones or fine-grained siltstones which may in places be ripple-marked; the sediments are comparable to those in the lower part of the Langport Member and indicate a relatively high-energy environment. In contrast, the greater part of the strata grouped by Richardson as 'Watchet Beds' in west Somerset are typical Liassic grey shales, and bituminous in places; these sediments indicate formation in a low-energy environment. Therefore, most of Richardson's north Somerset 'Watchet Beds' are here regarded as basal Lower Lias deposits which are transgressive. The classification of the north Somerset strata adopted here is the same as that used by Bristow and Etheridge (1873) in the St Audrie's Slip vertical section; this placed the base of the Lower Lias at 0.25 m above the top of the Sun Bed, which is the highest limestone of the Langport Member.

The Lower Lias strata exposed at the surface within the district belong mainly to the Blue Lias and comprise alternating limestone and shale or mudstone (Plate 1) and (Plate 11), though there are outcrops of the lateral equivalent of the lower part of the Shales-with-Teef' of Dorset (Lang and others, 1923). The Blue Lias of Dorset is about 29 m thick, however, compared with that of the north Somerset coast, which is of the order of 175 m thick, the great expansion of thickness having taken place in the mudstones. The familiar description of the Blue Lias as limestone-shale' alternations is a simplification of a cyclic pattern of sedimentation, which, in its fullest expression, is represented in upward order as follows: shale (commonly with a sharp base), mudstone, limestone and mudstone. The shale unit may or may not be bituminous; characteristically it contains only ammonites, fish remains and, in places, a fauna of small, thin-shelled bivalves; pyrite is common; where bituminous, the shales consist of thin alternations of carbonaceous and clay laminae with fine calcite stringers parallel with, and at various angles to, the bedding. On weathering selenite is commonly deposited along the bedding and gives rise to the characteristic wafery-thin 'paper shale' appearance. The base is generally sharp but the top of the shale locally contains Chondrites mottles. Depositional conditions for these sediments have been much discussed (Hallam, 1975). It is clear that substrate conditions must have been anaerobic to explain the abundance of sulphides and the lack of burrowing organisms in finely laminated sediment. The deposits are thought to have been laid down in relatively deep water during periods when regular pulses of sedimentation took place.

The mudstones are non-fissile or poorly-fissile, of medium to dark grey colour, blocky and calcareous. Usually they contain a benthonic fauna and they also commonly show trace-fossil mottling, particularly near the junctions with other lithologies although not at the sharp bases of shales; however these characteristics are less well marked in the Central Somerset Basin than elsewhere.

The limestones are of two distinct types. The more common type is dark bluish grey in colour, uniform throughout, hard, compact and splintery; most of these limestones are normally persistent laterally but some are lenticular or even nodular in places; groups or individual beds of nodules or lenticles are also laterally persistent and traceable throughout the length of the coast section. The second type of limestone is an extremely fine-grained or porcellanous rock which may be laminated; it was described by Hallam (1960a) from certain levels in the Dorset Blue Lias, and occurs at a similar horizon (Caloceras johnstoni Subzone) in Somerset. These lithologies are reminiscent of some seen in the hard limestone bands of the White Lias (Langport Member). Individual limestone bands in the Blue Lias of the north Somerset coast maintain a uniform thickness over the whole length of the coastline and can be correlated with individual bands proved in the Burton Row Borehole. It is uncommon for the junctions of limestone beds to be sharp; usually there is a transitional zone (perhaps up to 0.03 m thick) at upper and lower surfaces of the limestone bands within which it is difficult to place a precise boundary. Within, or very close to, these transition zones it is not unusual for there to be abundant fossil remains, particularly of ammonites, which are much commoner on under and upper surfaces of limestones than within the limestone.

Whether or not these limestone-shale rhythms of the Blue Lias are of primary or secondary origin is a topic which has been much debated. The wide lateral extent of individual bands (but with great variation of mudstone thickness), the presence of trace-fossil mottling and the regular pattern of the lithological cycle, all argue in favour of a primary, or very early diagenetic, origin for most of the limestone bands, although it is possible that some of the nodular limestones are secondary. The shale with sharp base betokens a transgression, producing a relatively rapid deepening of water, followed by regular pulses of sediment up to the level of the succeeding mudstone, when conditions became aerated and continuous deposition occurred but with varying amounts of sediment being laid down from place to place. A less sudden change of conditions to shallow water allowed the limestone member to be deposited, and another change, to somewhat deeper water conditions, caused the topmost mudstone member of the cycle to be laid down. The shale to limestone part of the sequence can best be thought of as the waxing part of the cycle, with the upper mudstone member as the waning part.

The earliest detailed account of the Lower Lias stratigraphy is that of Woodward (1893), who noted that the coastal strata were referable to the ammonite zones of 'A. planorbis' to 'A. semicostatus' inclusive, with a possible representative of the 'A. turneri' Zone. Judging by Woodward's description of the Helwell Bay fossil localities (op. cit., p. 96), it seems that the ammonites referred by him to 'A. turneri' were in all probability Arnioceras bodleyi, a species common at this locality and which can be of similar appearance in a crushed state. Later work has shown that Woodward's thicknesses of the strata (about 50 m) are a considerable underestimate; between 160 and 200 m of Lower Lias strata are present in the thickest sections. However, Woodward divided the sequence into five units and was ahead of his time in attempting this finer classification.

More recently, Palmer (1972) divided the exposed Lower Lias into seven divisions, designated A to G, and proposed formal names for them. The work was carried out independently of the present survey and the stratigraphical results agree very closely with those here recorded (Table 7).

However, his study was limited to the area between Watchet and Lilstock and does not bring out the lateral variations in thickness which are so characteristic a feature of the Lias stratigraphy of this coast.

Despite the extensive faulting and small-scale folding which affect the Lower Lias of the north Somerset coast, it is possible to elucidate the stratigraphy in great detail. With the exception of certain mudstones^‡5 , the main outcrop of which is in Doniford Bay, the Lower Lias beds exposed are referable to the Blue Lias. Within the Blue Lias of the coast are several lithological or stratigraphical divisions which aid detailed mapping. Using aerial photographs as base maps localities on the foreshore can be pinpointed and groups of strata can be accurately delineated. Strike faulting precludes the possibility of mapping inland in such detail, although locally it is possible to recognise, for instance, important mudstone divisions which form valleys between higher ground of predominantly limestone groups. The divisions mapped on the coast are as follows:

Comparison of these divisions with the lithological units of Palmer (1972) is shown in (Table 7).

Individual beds and groups of strata are readily traced along the coast and correlate well with the section proved in the Burton Row Borehole. Mapping divisions 1, 2 and 3 are present in South Wales as Trueman's (1920) groups C, D (Lavernock Shales) and E respectively, and in the Keynsham–Radstock area as the 'basal limestones and shales', the Saltford Shales and the 'higher limestones and shales' respectively (Donovan, 1956). Division 2 is represented on the Dorset coast by Lang's (1924) beds H55 to H72.

Lower Lias beds above the level of division 5 were proved in the Burton Row Borehole (Appendix 1, p. 121). AW

In the northern part of the Weston district only the lowest part of the Lias succession has survived post-Liassic erosion. The thickest section known in detail is adjacent to the Mendips, in the railway cutting east of Uphill, where about 10 m of shales and limestones are referable to the Psiloceras planorbis Zone (including the Pre-planorbis Beds) and 3 m (seen) apparently to the Alsatites liasicus and Schlotheimia angulata zones (see p. 78). Higher beds are present in a shallow basin lying between Middle Hope and the Weston–Worle ridge but no details are known. The maximum thickness of the Lower Lias present here is estimated to be of the order of 40 to 50 m. Although the Lower Lias overlaps directly on to the Carboniferous Limestone in a number of places, the facies is typically basinal and there is no evidence for littoral or sublittoral facies similar to those known in the Mendips, at Broadfield Down or in Glamorgan. It is thought, therefore, that most of the area lay very close to, or below, sea level at the time of deposition. GWG

Middle and Upper Lias

Between the Polden Hills and Bleadon Hill the only major physiographic feature which relieves the monotony of the Somerset Levels is Brent Knoll, composed mainly of Middle and Upper Lias strata and reaching some 140 m above OD at its summit. The knoll proper, which is the south-eastern part of the hill, forms a relatively steep-sided, flat-topped cone, flanked on the north-west by a platform

The summit of the hill was shown on the original one-inch to one-mile geological map as Marlstone or Middle Lias, the lower parts of the hill being regarded as Lower Lias. Subsequent work, mainly by Woodward (1887; 1893), led to reinterpretation of the stratigraphy along the lines that the hill was capped by a thin representative of the Cotswold 'Cephalopod Bed' (see below). The occurrence of 'Ammonites' communis, and 'A.' bifrons in rubbly limestones, noted by Woodward (1887), also proved that the Upper Lias was present; the discovery of 'Middle Lias' ammonites, 'Ammonites margaritatus, A. loscombei', by Woodward, in spoil from a well near the foot of the western part of the hill, led to the conclusion that no Lower Lias was exposed at Brent Knoll. These latter ammonites would now be regarded as Amaltheus cf. stokesi, denoting the stokesi Subzone of the Amaltheus margaritatus Zone (Middle Lias), and Tragophylloceras sp., which ranges through the Pliensbachian. The present work has borne out the essential accuracy of Woodward's observations on the stratigraphy.

The sequence of strata from the top of Brent Knoll down to about Ordnance Datum can be summarised as follows, using data from both surface observation and the record of the Hill Lane Borehole (Appendix 2, p. 124).

In the Hill Lane Borehole the Amaltheus subnodosus and Amaltheus stokesi subzones of the margaritatus Zone could be established between 23.64 and 48.60 m and between 48.90 and 82.37 m respectively. The highest subzone, namely that of Amaltheus gibbosus, was not proved, although Amaltheus was present up to 4.06 m above the top of the subnodosus Subzone. Pleuroceras cf. spinatum, indicating the Pleuroceras spifiatum Zone, occurred in the borehole in hard calcareous silty and pebbly mudstones near the top of the Middle Lias.

The lowest 12 m of the Upper Lias sequence were seen in the Hill Lane Borehole. The beds are referable to the Harpoceras falciferum Zone, though the Dactylioceras tenuicostatum Zone may be present between 13.45 and 14.96 m in beds where the fauna recovered includes only indeterminate harpoceratids and Dactylioceras sp. down to 14.71 m. Above 8.50 m in the Hill Lane Borehole are abundant mottled, rubbly limestones comparable in position to those described by Woodward (1887). Precise biostratigraphical junctions are difficult to place above this level because exposure is poor and most of the fossils found during the present survey of Brent Knoll are not in situ. However collecting suggests the presence of the following zones and subzones: Harpoceras falciferum Subzone of the falciferum Zone, Peronoceras fibulatum Subzone and Catacoeloceras crassum Subzone of the Hildoceras bifrons Zone, Grammoceras striatulum Subzone of the Grammoceras thouarsense Zone and the Phlyseogrammoceras dispansum and Dumortieria moorei subzones of the Dumortieria levesquei Zone. The small amount of field evidence available suggests that the Upper Lias clays and limestones may range up to the Phlyseogrammoceras dispansum Subzone and that the succeeding Upper Lias silts are all referable to the higher parts of the levesquei Zone.

Inferior Oolite

Capping the Lias silts and clays of Brent Knoll and forming the summit are perhaps 5 to 10 m of calcareous beds of rather variable lithology ranging from brownish grey, hard, calcarenites to grey, moderately fine-grained limestones with ooliths. The beds are poorly exposed and there are numerous loose blocks which may derive from the Hill Fort at the summit of the hill.

Woodward (1876; 1887; 1893) referred to these beds as a thin capping of the so-called 'Cephalopoda-bed' on the evidence of rubble and loose blocks with rare Rhynchonella cynocephala and Serpula, then known from horizons below the Inferior Oolite near Beaminster and Crewkerne. However the range of Homoeorhynchia cynocephala is now known to extend upwards to rocks of Aalenian age. During the survey three specimens of Leioceras sp. were obtained from what could have been outcrops but may have been loose blocks of limestone, and Leioceras opalinum was found in a loose block, not far out of place, near the summit of Brent Knoll. These occurrences suggest an Aalenian age for these rocks. AW

Details

Lower Lias

Coastal sections

Doniford Bay

In Doniford Bay, gently flexured Lower Lias strata can be traced. North of the Doniford Bay Fault [ST 0870 4355] the beds dip at between 8° and 37° towards the north-north-east and the following section was measured across the offshore reefs (Figure 13):

South of the Doniford Bay Fault and dipping at about 18° to the north is a succession predominantly of shales and mudstones. These beds form a topographic low compared with the limestone-rich sequence farther offshore. Reefs of fissile shale and nodular limestone occupy an east-west belt of foreshore close to the beach and contain Arnioceras sp. and Coroniceras sp. Above these beds and commonly covered by a skin of mud is a group of mudstones containing a similar ammonite fauna. The strata are better exposed in the cliff [ST 0783 4325] where abundant less crushed Arnioceras sp. are present. The following section was measured on the foreshore:

The highest part of the section hereabouts can be measured in the cliff [ST 0783 4325] northwards as far as the Doniford Bay Fault exposed near the steps [ST 0780 4337] at Helwell Bay (Figure 14). Adjacent to this fault are at least six crumpled bands of hard calcareous mudstones, each about 0.15 m thick, in a disturbed (4.5-m thick) sequence of mudstones. The limestones contain Gryphaea sp. but no ammonites. They are considered to represent the highest Lower Lias beds exposed in Helwell Bay. The thicknesses given below, and measured in the western cliffs of Helwell Bay, are only approximate as numerous small dislocations affect these uniformly monotonous mudstones making the section difficult to measure.

In the mudstones on the foreshore close to the Doniford Bay Fault [ST 0832 4333] and [ST 0817 4337] two large thick-shelled ammonites identified as Coroniceras have been found. Calculations of thickness based upon dip and width of outcrop suggest that these specimens can be placed near the top of bed 257 in the cliff section given above. A small faulted inlier of semicostatum Zone strata (about the level of beds 242 to 247) occurs at the eastern end of Doniford Bay [ST 0866 4324] and yields species of Arnioceras and Coroniceras.

A faulted syncline [ST 0915 4355] offshore from Doniford preserves alternating limestones and shales which yield Vermiceras. These beds belong to the lowest part of the bucklandi Zone.

St Audrie's Bay

In St Audrie's Bay, Lower Lias beds dip at about 14° south-west. West of St Audrie's Slip are cliffs in which the alternating limestones and shales come successively to beach level, so that it is possible to make detailed vertical measurements of sections and to identify the biostratigraphic sequence. Strike faulting repeats the outcrop of the topmost part of the planorbis Zone on the foreshore [1013 4334] and small-scale low-angle thrust faulting repeats part of the succession in the cliff [1029 4319]. Starting in the cliff section [ST 0979 4332] near Doniford, and working eastwards, the following succession was measured (Figure 13):

Individual beds in the cliff section are recognised with ease on the foreshore. Ammonites are plentiful in the liasicus Zone strata, but they are fewer and less well-preserved in the planorbis Zone strata compared with foreshore exposures to the east and the west. This is considered to be due to the fact that the foreshore outcrop of these strata, at St Audrie's Bay, is roughly parallel to the direction of wave attack, allowing erosion and destruction of the delicate, thin shells of the ammonites.

Blue Ben–Quantock's Head

At the eastern end of St Audrie's Bay, Lower Lias strata are faulted against Mercia Mudstone and form the prominent headland of Blue Ben. The cliff section east of here is essentially a strike section complicated by east–west faulting (Plate 12). At beach level, in the cliff, are beds close to the angulata–bucklandi zonal junction. This is the Hettangian–Sinemurian stage boundary and can be located at a shale-mudstone contact (base of bed 146) on the eastern side of Blue Ben [ST 1241 4388], where Vermiceras solaroides is present in mudstones overlying shales with Schlotheimia. The alternating limestone-shale sequence of the conybeari Subzone is succeeded in the cliff by a predominantly shale-mudstone group. Although the latter beds are inaccessible along much of the cliff exposure, fallen blocks of shale at the foot of the cliff contain Coroniceras of the rotiforme group. Just west of Quantock's Head [ST 1300 4395], near the highest part (about 51 m OD) of the cliff, it is possible to examine the youngest beds exposed, hereabouts overlain by Head gravels. Thick bituminous shales are present and contain large (0.76-m-diameter) thick-shelled ammonites similar to those found in Doniford Bay, where they were referred to Corniceras lyra of the lyraSubzone of the semicostatum Zone. Also present are tuberculate coroniceratids and weathered Arnioceras?. This cliff section is of particular importance because it links undoubted conybeari Subzone beds in the lower part with strata containing ammonites indicative of the semicostatum Zone at the top. For definition of Lyra Subzone see p. 128.

On the foreshore, strata from the planorbis, liasicus, angulata and bucklandi zones are disposed in narrow, elongate, gentle, east–west folds which commonly show closure at both ends. The folds are faulted, mainly along east–west lines, with subsidiary fractures trending north-west. The planorbis Zone strata are not well exposed, particularly in the lower parts, but higher beds are like those at St Audrie's Bay, described above. Outcrops are present at three places [ST 1154 4375]; [ST 1165 4395]; [ST 1288 4424]. The liasicus Zone is well exposed but difficult to measure in the foreshore because of structural complications. Beds of the angulata Zone are well exposed but have thinned noticeably at Blue Ben from the St Audrie's Bay and Watchet areas to the west. All the main limestones and groups of limestones are present, and have similar thicknesses to those beds elsewhere; shales have thinned markedly but can still be correlated in detail with other sections. Several limestones have irregular junctions, at top or bottom, which suggests that slight local erosion took place at several levels during angulata Zone times. The Blue Ben locality is the closest that well-exposed Lower Lias strata come to the Quantock massif; two inliers of Hangman Grit [ST 1172 4322]; [ST 1278 4346], mantled by red mudstones of the Mercia Mudstone Group, are within 800 m and 600 m respectively of the Blue Ben exposures and essentially mark the position of the southern edge of the main Mesozoic structural and sedimentary basin (p. 98). The thinning of the angulata Zone hereabouts is ascribed to differential subsidence of the area in the vicinity of the Quantock Swell during these times.

The bucklandi Zone is well represented between Blue Ben and Quantock's Head. The Metophioceras Bed (bed 161) is easily recognisable by the occurrence of abundant large specimens of Vermiceras conybeari clustered on its under surface; it maintains this characteristic from the Blue Ben area eastwards to Hinkley Point. In the higher parts of the bucklandi Zone the beds are thinner than at Hinkley Point, presumably because of the proximity of the Quantock massif. In the following section, strata down to bed 166 were inaccessible and were measured from photographs of the cliff [ST 1300 4395]; strata below were measured either in the cliff or on the foreshore (Figure 13).

Quantock's Head–Kilve

The sheer cliff at Quantock's Head [ST 1335 4409] consists in its lowest part of conybeari Subzone limestones and shales; these are overlain by the predominantly argillaceous higher parts of the bucklandi Zone and the lower part of the semicostatum Zone. Huge fallen blocks of shale yield Coroniceras of the rotiforme group and the presence of the semicostatum Zone beds is inferred from the height of the cliff and the known stratal thicknesses in the vicinity. Between Quantock's Head and Kilve, strata of the lowest part of the semicostatum Zone and of the bucklandi Zone maintain similar thicknesses to corresponding beds in the cliff west of Quantock's Head. Strata of the rotiforme Subzone are well exposed on the foreshore [ST 1400 4431] and the lowest recorded ammonite of the C. rotiforme group occurs at this locality in the top 0.6 m of a shale (bed 164). In the middle parts of the rotiforme Subzone are discrete, thin (up to 0.05-m) shell beds containing ammonites, which maintain the same characters and stratigraphical position over many kilometres of coast. See p. 126 for discussion of revised zones and subzones about the bucklandi to semicostatum Zone boundary.

The thick massive limestone bed 204 occurs in the cliff [ST 1422 4437] immediately west of Kilve Pill and the succession down to bed 97 can be inspected between that locality and Quantock's Head. Strata between beds 204 and 161 at Kilve are similar in thickness to those west of Quantock's Head. However, the conybeari Subzone beds are slightly thicker in the area between Quantock's Head and Kilve than at Blue Ben. The greater increase in thickness takes place in the shale beds, although there is a measurable increase in limestone thicknesses also. Similarly, those parts of the angulata Zone measured between Quantock's Head and Kilve are thicker than their counterparts at Blue Ben; this increase takes place mainly in individual shale members.

On the foreshore, beds of bucklandi Zone age are well exposed [ST 1386 4437] in a synclinal structure which is intensely disturbed by strong, nearly east-west faults. To the north and east, the planorbis Zone strata, similar to those at Watchet, are seen on the seaward fringe of the intertidal area and are affected by folding and faulting. The beds of liasicus Zone age form a low-lying belt in the intertidal area [ST 1400 4452]; although the outcrops are usually obscured by a thin skin of mud, enough exposure is available to demonstrate that the lithologies and fossils are similar to those of the Watchet and St Audrie's areas.

Kilve–Lilstock Bay

Along the coast between Kilve and Lilstock Bay the strata dip generally towards the south or south-south-west and are cut by numerous east-west faults (Plate 13). Consequently the sections are repeated to some extent. The higher part of the planorbis Zone, with Caloceras, is exposed at low tide [ST 1538 4517] and close by in the cliff [ST 1571 4510]. Strata of liasicus Zone age are well exposed on the foreshore [ST 1490 4470]; [ST 1545 4506], as are those of angulata Zone age on the foreshore [ST 1510 4490] and in the cliff [ST 1540 4490]. North of Park Farm the higher beds of liasicus Zone age are exposed [ST 1667 4557], and limestones of angulata Zone age with low dips to the north form prominent ledge-like reefs close to low water mark [ST 1615 4560]. Between Kilve and Lilstock Bay four patches of strata of bucklandi Zone age are preserved [ST 1495 4461]; [ST 1555 4525]; [ST 1588 4538]; [ST 1656 4574]. The beds maintain their usual characteristics for this area, in that a lower group of limestones and shales yielding Vermiceras underlie a higher, thicker group mainly of shales containing Coroniceras of the rotiforme group. An interesting feature of these strata hereabouts is the occurrence of bell-shaped masses of limestone at the level of beds 166 to 188. Each mass has a central core of Shelly limestone containing uncrushed ammonites, surrounded by less fossiliferous grey limestone; surrounding the latter and displaying quaquaversal dips away from the bell-shaped limestone mounds are shales. The mounds are associated with concretionary limestones some of which display a septarian aspect, as exposed on the foreshore [ST 1510 4462]. Similar structures are present on the foreshore at Hinkley Point (see below). The highest parts of the Lower Lias hereabouts are present in the cliff [ST 1513 4464] and the sequence from bed 224 to bed 69 is exposed in the cliffs and on the foreshore to a place [ST 1470 4461] north-east of Kilve Pill (Figure 16).

Thicknesses of strata of bucklandi Zone age hereabouts are practically the same as those of the corresponding beds measured near Quantock's Head. The division 3 strata are about 7.6 m thicker east of Kilve beach than at Blue Ben. Individual beds can be correlated with confidence and the increase of thickness (eastwards from Blue Ben) is in individual shale members, the limestones maintaining similar thicknesses between the two sections. Two of the limestones with irregular bases in beds of angulata Zone age at Blue Ben maintain this characteristic to the east of Kilve.

The higher part of the Blue Lias sequence given below, down to the level of bed 152, was measured in the cliff [ST 1584 4521] 0.75 km WNW of Park Farm, and the lower part 160 m to the NE [ST 1571 4511] (Figure 15).

The division 3 strata (see (Table 7)) in the above section are thicker than those of the Blue Ben and Kilve sequences. Most of the thickness increase is taken up in the shales, but at various levels the limestones are also thicker. In contrast to the Blue Ben and Kilve sequences, no limestones with irregular junctions were observed in the section given above. The easterly thickening away from Blue Ben is interpreted as being due to increasing distance from the margin of the Mesozoic basin. The sequence of beds of liasicus Zone age is slightly thinner than that recorded from St Audrie's Bay, although all the individual beds are recognisable. Bed 55 is represented hereabouts by two layers of limestone nodules rather than one. The topmost beds of planorbis Zone age are similar to the corresponding beds recorded from St Audrie's Bay.

Lilstock Bay

On the foreshore of the western part of the bay at Lilstock abundant Lias limestone cobbles and boulders obscure the solid geology. To the east, strata of the lowest two zones of the Lower Lias are well exposed on the foreshore but are cut by numerous east–west faults into thin elongate blocks; cliff sections expose strata representing the Lower Lias ammonite zones up to and including the lower part of the angulata Zone. The following section was measured in the cliff [ST 1785 4527] (Figure 15).

The beds of planorbis Zone age are like those at St Audrie's Bay and are found from Lilstock Bay to the Benhole Farm Fault [ST 1872 4588]. In the cliff [ST 1887 4555] the liasicus-angulata zonal boundary is exposed and can be examined on the foreshore [ST 1852 4563] close by. Preserved in a trough between two faults are bucklandi Zone strata. Most of the shale outcrops are often covered by present day intertidal deposits but beds of Blue Lias type yield Vermiceras.

Benhole Farm–Stolford

Gently northward dipping strata of the higher parts of the angulata Zone form a low cliff between Benhole Farm and Hinkley Point Power Station. The angulata-bucklandi zonal junction can be traced on the foreshore from Benhole Farm to the power station. Forming strong reefs offshore from the power station are strata of the bucklandi Zone and the lowest part of the semicostatum Zone.

The section given below was measured across the offshore reefs at the power station (Figure 16); in many instances (mainly in the thicker shale bands) direct measurement was not possible and thicknesses had to be calculated.

The presence of the bell-shaped mounds of limestone at the level of beds 182 to 186 shows that these structures are widespread, since similar masses are present in the same beds near Kilve (p. 71). The structures are composed of fine-grained, splintery, dark bluish grey limestone with abundant small C. cf. rotiforme preserved in an uncrushed state. There are also many pale grey silty limestone masses of cylindrical shape which have no regular orientation and may be curled around the bases of the large mounds. In places the shales adjacent to the limestone masses are disposed quaquaversally about a central core of limestone.

Isolated exposures of strata of bucklandi Zone age are present near Stolford [ST 2310 4620]; the beds are folded and quite severely faulted in places. Members of the Geography Department, University College of Wales, Aberystwyth, pointed out to the writer in the field the presence of Lower Lias exposures beneath the mud deposits of the Stert Flats. At one place [ST 2385 4605] thin greyish blue limestones with interbedded shales contain large thick-shelled arietitids, referable to a position high in the bucklandi Zone or low in the semicostatum Zone. The beds strike at 170°, dip westwards at about 14° and are slightly faulted. Lias limestones containing Vermiceras were observed at a place about [ST 2393 4608] north of Catsford Common. These beds are probably referable to the conybeari Subzone of the bucklandi Zone and dip westwards at about 7°.

East–west-striking and northerly dipping Lias limestones and shales are visible at extreme low tide beneath thin mud deposits near the seaward edge of the mud flats [ST 2380 4701].

Inland sections

Blue Ben-Kilve

The Lower Lias is present inland north of the Blue Ben Fault. A bluff at the top of the valley side [ST 1266 4371] exposes 1 m of easterly-dipping thin limestones alternating with shales which probably belong to the Pre-planorbis Beds. North of this, east of Blue Ben and probably associated with a zone of faulting, the following sequence is exposed [ST 1273 4382]:

Near Court House [ST 1363 4368] soils have been derived from Lias clay, with some local patches of Lias limestone. East of Court House a northerly inclined dip-slope formed by the lowest Lias beds is traceable eastwards towards the village of Kilve and brashy limestone debris is strewn across its surface. Old quarries [ST 1430 4346] in a wood are mainly overgrown but expose Lias limestones and shales referable to the planorbis Zone.

The Kilve Stream exposes SE-dipping shales and limestones probably referable to the angulata Zone north-west of the Chantry [ST 1458 4413]. South-east of the Chantry several poor exposures of Lower Lias lie in the stream bed: north-west of a point [ST 1492 4373] 230 m N of Parkhouse Farm the Lias bedrock dips in a general southerly or south-easterly direction, whereas south of this locality dips are in a northerly direction; at the farm [ST 1497 4350], 0.30 m of dark grey limestone overlying 0.64 m of shale dip north-east and are probably referable to the planorbis Zone.

Kilton–Stockland Bristol

Old quarries [ST 1631 4477]; [ST 1653 4513] between Kilton and Lilstock are overgrown or filled in, but abundant limestone debris in the vicinities suggests that the Lower Lias was worked for lime hereabouts.

The Lower Lias ridge between Knighton and Wick Moor exposes a northward-dipping limestone pavement [ST 1932 4487], very close to old workings. Brown and grey clay soils with local brashy patches of limestone debris occupy much of the ground between Shurton and Stockland Bristol. Just east of Stogursey [ST 2033 4290], beds of planorbis Zone age form a strong, northerly-inclined dip slope. Two old quarries [ST 2261 4256]; [ST 2261 4287] near Cockwood expose up to 1.83 m of alternating limestones and shales and yielded specimens of P. planorbis. The field mapping has demonstrated a complicated pattern of east-west faults in this neighbourhood, particularly about the crop of the Penarth Group and lowest Lias beds.

Benhole Farm area

Old quarries [ST 1866 4533] at the crest of a strong east-west ridge show the following northward-dipping succession in Lower Lias, division 1.

South of Benhole Farm, and occupying the crest of the same east-west ridge, are old workings [ST 1956 4540] which yield limestone blocks like the Bottom Lias Bed (Pre-planorbis Beds) of the coast. A northward-dipping limestone pavement immediately to the east of the old workings is close to an exposure [ST 1961 4539] of paper shales and limestone which is probably referable to the lowest Lias beds.

At Hinkley Point Power Station, east of Benhole Farm, foundation excavations and site investigation boreholes have provided information on the Lias stratigraphy, spanning beds 60 to 204 of the coastal sequence described above (see (Figure 22)). The successions are closely comparable. Some individual calcareous mudstone bands pass laterally into limestone or mudstone bands, but they are usually thin, and major groups of strata as well as individual beds are commonly persistent throughout the area. The youngest beds on the main part of the site are probably of rotifarme Subzone age, corresponding to beds 159–161; they contain Vermiceras and overlie conybeari Subzone limestones and mudstones with V. solaroides and V. conybeari in a bed equivalent to bed 154 on the coast. The base of the bucklandi Zone again appears to lie in the thick mudstone below bed 147. The angulata Zone is indicated by ammonites at a depth of 7.16 m in one borehole [ST 2130 4613], at 1.60 m below the base of bed 147. The occurrence of Waehntroceras? (at a depth of 59.54 m), Psilophyllites (at 59.89 m) and abundant schlotheimiids (at 60.25 m) in beds 60 to 64 in this borehole indicates the presence of the liasicus Zone. As with the coastal sections, the angulata–liasicus zonal boundary is above the bottom of the group of strata with abundant limestone bands (base of division 3).

Stretcholt area

East of the River Parrett an outcrop of Lower Lias, surrounded by alluvial deposits, forms Pawlett Hill. The Hill rises to over 30 m OD and has gentle clay slopes with a little scattered Lias limestone debris. Capping the hill, and forming a rudimentary dip-slope inclined in a direction a little west of north, are Lias clays with limestone containing Gryphaea; a farm wall [ST 2866 4282] of Lias limestone blocks contained a large arietitid ammonite which was probably derived from the hill. These fossils, in association with a predominantly mudstone stratigraphy, suggest that the Pawlett strata may be referred to the late bucklandi Zone or early semicostatum Zone as seen, for instance, in the strata exposed south of the major fault in Doniford Bay. AW

Mendip Hills–Middle Hope

The fullest succession recorded in the Lower Lias of the northern part of the district was in the railway cutting east of Uphill [ST 327 581] where the beds are much faulted (p. 105). Wright (1860; 1878) gave the section in ascending sequence above the Westbury Formation as: dark grit containing scales and teeth of fish, 0.04 m; alternating shales or marls and limestones (mainly in top 3.7 m) with Ostrea liassica, 'Modiola'minima, etc., about 6 m; dark grey laminated shale, then limestone, with 'Aeg.'planorbis, 4.3 m; alternating beds of light coloured limestone and shale with 'Aeg.' angulatum and 'Aeg.' liasicum, much fractured, dip 40°, 3.05 m. It may be assumed that the Cotham Member accounts for the lowest 2 m of the alternating limestone and marl sequence, and that the uppermost 3 m include beds of liasicus Zone age and possibly part of the angulatum Zone.

In Hutton, evidence for the presence of typical Blue Lias limestones and shales with bivalves was found in excavations for a new housing estate [ST 355 589].

At Locking, apparently starting a short distance above the Penarth Group, the following ascending sequence was noted in excavations for a new housing estate [ST 361 601]: shaly marl, seen 0.76 m; laminated limestone, 0.13 m; shaly marl, 0.15 m; flaggy limestone with Modiolus minimus and Plagiostoma sp.,0.6 m; not exposed, about 1.5 to 1.8 m; shaly marl, 0.23 m; limestone, 0.3 m; shaly marl with soft limestone (0.10 m) towards base, seen 1.2 m; total thickness 5 m. The uppermost 1.5 m of beds contained Psiloceras planorbis, thus leaving more than 3.5 m assignable to the Pre-planorbis Beds. The limestones are hard textured and dark grey to blue-hearted. The shaly marls are highly fissile. The dip increases southwards from 7° to 10° south-south-west as an eastwest-trending fault is approached.

North of the Weston–Worle ridge the Lower Lias is folded into a gentle basin. The Blue Lias makes a low feature above the Penarth Group beds, and parts of the northern, eastern and southern rims of the basin–formed by inwardly directed dip-slopes of Blue Lias limestones–are seen above the alluvium of the moors. A small bankside exposure of southerly dipping shales with limestone bands containing bivalves is seen near Weston Crematorim [ST 358 639]. At Middle Hope, the Blue Lias forms a well-marked dip-slope inclined at about 4° to 5° to the south. Although the Lias rests directly on the Carboniferous Limestone along the northern edge of the outcrop there is no indication of littoral facies. A borehole west of Woodspring Priory (p. 49) proved Triassic rocks beneath the Lias. On the coast north of Wick St Lawrence [ST 363 666], near the eastern margin of the district, Lias (presumably Lower Lias) was proved beneath the alluvium (Murray & Hawkins, 1976, figs. 2 & 3). At Collum Farm [ST 347 655], towards the central part of the Lias basin, there is a low mound rising above the moor level; blue and brown shales with rock bands were augered in numerous places hereabouts and it is considered that these represent part of the Lower Lias succession. The topography suggests that the rocks are dipping gently to the south-south-west and that, therefore, the axis of the basin lies farther to the south. GWG

Middle and Upper Lias

Brent Knoll

At the surface the Middle Lias at Brent Knoll is obscured either by landslipped Upper Lias clays and limestones or by weathered Upper Lias hillwash material, and no exposures of in situ Middle Lias mudstones were found at the time of the survey.

The flattish top of the platform forming the north-western part of Brent Knoll is a regular easterly-inclined dip-slope in rubbly mottled limestones which rim this part of the hill. Heavy grey clay soils are present on the platform and grey mottled limestone fragments form a field brash in places. The platform is limited by steep slopes which locally display strong but laterally impersistent topographic features; some of these are probably due to limestone bands in the Middle and Upper Lias, but others may have been formed by landslipping and mudflow. Traces of conglomeratic grey limestones with pink limestone pebbles were seen in a scar [ST 3322 5123] below the rubbly limestone. Another exposure of conglomeratic grey limestone, without pink pebbles but at a similar stratigraphical level, was observed 600 m to the north [ST 3323 5183].

The limestone feature is poorly seen around the south-eastern part of the hill (the Knoll proper) but the occurrence of surface brash shows that the limestone beds are present throughout. They are well exposed in old workings [ST 3352 5120] and in a ditch [ST 3357 5084]. Mottled rubbly limestones forming the 'rim' to a rounded 'platform' feature north-west of Battleborough were well exposed in a temporary section [ST 3400 5046]. Downslope from the 'rim' is a strong feature which correlates with that described from the north-western part of the hill, but at an anomalously low level allowing for normal dip and strike. Geometrical considerations make faulting unlikely and the feature is interpreted as being due to rotational landslip (p.85).

Above the level of the mottled rubbly limestone beds, but still within the outcrop of the Upper Lias clays, are several strong but impersistent features [ST 3376 5079]; [ST 3393 5069]; [ST 3422 5062]; [ST 3433 5087]; [ST 3437 5081] produced by limestone bands or lenses which are not necessarily at one stratigraphical level.

Capping the Upper Lias clays are about 52 m of brown-weathering grey silty clays and silts, in places with a blocky texture. These beds were designated Midford Sands by Woodward (1893). AW

Chapter 6 Pleistocene and Recent

Introduction

Deposits formed during Pleistocene and Recent times cover large parts of the onshore area of the district. These 'drift' deposits have been mapped on a lithological basis; they are relatively thin and the stratigraphical record is patchy. The main morphological features of the district must have been established by late Tertiary or early Quaternary times, and much of Pleistocene history was indirectly concerned with the successive advances and retreats of ice-sheets. The present district, for much if not all of this time, lay to the south of the main glaciated regions, although there is evidence that ice was active in adjacent parts of south-west Britain and it seems possible that ice may have moved up the Bristol Channel and into the Somerset lowlands during the Wolstonian glaciation. The growth of major ice sheets carries with it the corollary of lower sea levels and it is likely that the valleys were eroded significantly during the glacial episodes. Conversely, the ice sheets retreated during warmer interglacial episodes, which were characterised by rises in sea level.

The earliest drift deposits of the area, the Burtle Beds, may have been deposited during one such interval. They occur as localised patches of sand and gravel. Traditionally they have been thought of as marine beach deposits but more recently they have been interpreted as outwash deposits from a Wolstonian ice-sheet. However, observations from the type locality of Catcott Burtle suggest that there may be sands of different ages in the area of the Somerset Levels and care is necessary in interpreting these patchily exposed deposits.

During the last (Devensian) glaciation the district was not glaciated but experienced a periglacial environment. These conditions gave rise to spreads of solifluction deposits mapped as Head. For the most part the Head deposits are sands and gravels derived as waste from the surrounding upland areas. In places they form extensive spreads and also occur as valley fills or as cappings to ridge tops. The valley-fill deposits in the lower reaches of streams are covered by alluvial deposits and eventually pass beneath the main alluvial spreads of the levels.

At the height of the Devensian glaciation, the sea level stood at more than 25 m below the present sea level. Melting of the ice-sheets caused a rapid rise in sea level so that the sea inundated the lower-lying parts of the coastal areas. This 'Flandrian Transgression' resulted in the deposition of the alluvial sediments now constituting the Somerset Levels. Curves illustrating the rise of the Holocene sea level in the Bristol Channel (Kidson and Heyworth, 1973, 1976; Hawkins, 1971)^‡7  show that it did not proceed at a constant rate but was much more rapid between 9000 and 6500 years ago than it was after that time. Evidence from the sediments suggests that a birch forest had grown in the region of the Somerset Levels and that the part which occupied the lower parts of valleys was killed off by the inundation and is now preserved as peat. As the rise in sea level continued, the peat was covered by clay. Some recent work (op. cit.) suggests that as the sea level rise gradually slowed, about 6000 years ago, the sea reached its maximum extent and began to retreat leaving a wide clay surface (now at about the level of Ordnance Datum) which was colonised by plants. In this coastal belt the vegetation was overtaken by further incursions of the sea and renewed deposition of clay occurred. The occurrence of a sudden marine transgression, the 'Romano-British Transgression', during the Roman occupation of Britain (e.g. Godwin, 1943), has been challenged by various workers in recent years. The clay deposits which overlie the peat with base at Ordnance Datum in the coastal parts of Somerset are now considered not to be related to major eustatic sea level variations but to minor coastal changes. AW

Sand and gravel of unknown origin

A patch of sand and gravel measuring about 200 m by up to 80 m occurs on the south side of Bleadon Hill [ST 349 573] at a height of about 80 m OD. The deposit has been described by Findlay and others (1972). The beds comprise buff sand, calcreted sand and reddish brown pebbly sandy loam, overlain by subrounded calcreted cobbly gravel of Carboniferous Limestone with little or no matrix, dipping at about 37° southward. The deposit rests on Triassic red marls in its eastern part and appears to overlap on to the Carboniferous Limestone at the extreme western end, where it is associated with a small platform and cliff-like feature, adjacent to the west, cut into the limestone. No complete section has been seen through the deposit but mapping indicated a total thickness of about 7 m, of which the gravel may account for 4 to 5 m or more.

Dr H. C. Ivimey-Cook found a sparse microfauna in the sands, which he considered might be of Liassic age. Following a re-examination of this material Mr I. P. Wilkinson noted that the specimens showed evidence of reworking; he concluded that sufficient material was present to allow a Lower Lias age to be given to the fauna, though not necessarily to the deposit, and that a Sinemurian age was most likely. No Pleistocene microfauna has been recorded from the deposit. The possible modes of origin of the deposit discussed by Findlay and others (1972) were as: (1) a Mesozoic beach deposit, (2) a high-level Pleistocene beach deposit, (3) a marginal fluvio-glacial gravel and (4) as a lacustrine pro-glacial beach deposit. They were inclined to favour the third possibility, drawing an analogy with similar deposits on the Failand Ridge. GWG

? Glacially derived drift on Flat Holm and Steep Holm

The drift occurrences on Flat Holm have been disturbed by extensive fortification works and must therefore be interpreted with caution. Much of the land surface is covered with a skin, probably less than 1 m in thickness, of reddish brown sandy loam with a variable content of angular and rounded rock debris. A coastal section [ST 2190 6475] near the north-western corner of an old blockhouse in the Bottles Well Bay area in the south-western part of the island shows Carboniferous Limestone overlain by 0.25 to 0.6 m of chestnut-coloured unbedded loamy gritty sand with scattered pebbles of Carboniferous limestone and dolomite, grey-brown and reddish brown hard sandstone (Old Red Sandstone), flint, and a few grey-brown Pennant-type carbonaceous sandstone (Upper Coal Measures), overlain in turn, with an abrupt junction, by unbedded similarly coloured sandy and gritty loam with scattered angular rock debris, 0.1 to 0.2 m, and peaty soil, 0.1 m. The upper sandy loam may not be in situ, as the coarse fraction almost entirely comprises calcite and ferruginous chips of vein material, including traces of galena, and scattered plant debris (p. 118). A quantitative analysis by Mr G. E. Strong of the lower gritty sand showed gravel (>2 mm) 92.6 per cent, sand (0.063 to 2 mm) 6.2 per cent and silt/clay (<0.063 mm) 1.2 per cent, thus demonstrating that the sample should strictly be termed slightly sandy gravel. He reported that 'the gravel fraction consists of granules, pebbles and cobbles of a near greyish orange (10YR7/4) silty sandstone with some included granules and small pebbles (mainly quartz). The sand fraction consists of silty sandstone fragments and grains of quartz, including pink, milky and clear quartz. The grain shapes range from angular to well-rounded. There is abundant biotic debris of roots, cellulose fibres, shell-fragments and debris, probably of contemporary origin.' The evidence of recent contamination, coupled with the lack of bedding, suggests that the deposit is not in its original sedimentary state and is probably derived.

The presence of erratic pebbles in the beach on the northeast side of Flat Holm has long been known. Buckland and Conybeare (1824, p. 235) recorded 'pebbles of chalk-flint, red sandstone, quartz, flinty slate and porphyry' from here. A thin rock slice (E17444) from a boulder collected in 1936 from this locality was described by Dr A. J. Macgregor as a pink leucocratic soda-rich granite; he was unable to suggest a provenance and considered that it might have been imported by man. Mr R. K. Harrison examined a suite of 100 pebbles collected by Dr G. A. Kellaway from East Beach [ST 221 652] in 1971. Thin sections of pebbles were described as: laminated cherty dolomitic siltstone (E40285); hornfelsed siltstone (E40286); spicular chert (E40288); dolomite pseudo-microbreccia (E40291); graphic microgranite ((E40289), (E40293), (E40294)); feldspar-porphyry (E40287); fluxioned feldspar-rich lava (E40290); and silicified welded crystal lithic tuff (E40292). The remaining rocks were broadly characterised in percentages of the total pebble population as follows: jaspers and other fine-grained quartzose rocks, dark grey to purple, 24 per cent; quartzitic sandstone, purple-brown (probable Old Red Sandstone), 15 per cent; flints (Cretaceous), 13 per cent; quartzites, fine-grained, indurated (?Precambrian), 11 per cent; limestone and dolomite (Carboniferous), 10 per cent; coarse and medium grits, 10 per cent; ?andesite, purple and green, <5 per cent; quartz-feldspar-porphyry, grey, <5 per cent; ?obsidian, trace; vein-quartz and other igneous rocks, trace. Mr Harrison reported that the pebbles could be divided into two categories: relatively local and far-travelled. The former were readily distinguished (limestones, dolomites, flints, purple-brown quartzitic sandstones). The latter were highly variable in lithology, and provenances could not be given with any certainty; some, e.g. vein-quartz and Precambrian? quartzites, might have been secondarily derived from conglomerates in the Old Red Sandstone (Heard and Davies, 1924, p.512; Wallis, 1928). The andesites and welded tuff might have stemmed from North Wales or the Lake District and the graphic microgranite possibly from northern England (e.g. Carrock Fell).

In conclusion, the presence on Flat Holm, both in the beach- material and in the surface loam, of non-local rock types, especially flints and igneous rocks, suggests derivation from a glacial deposit, probably in the bed of the estuary. In this respect, it is significant that remnants of glacial drifts are present onshore in the area to the east of Flat Holm (for full details, see Gilbertson and Hawkins, 1977a). Also that the gravel spit extending eastwards from the end of Steep Holm, although composed mainly of local rocks, includes not uncommon flints and (greensand) cherts: the flints are subrounded and covered with a thick yellow patina. GWG

Burtle Beds

Deposits which have been described as Burtle Beds occur in the Somerset Levels both to the north and to the south of the Polden Hills. Within the present district spreads of Burtle Beds occur to the south of Burnham-on-Sea and Highbridge, at Stretcholt and Huntspill.

The Burtle Beds comprise low mounds of sand or sand and gravel rising above the general level of the surrounding alluvial surface; individual patches cover areas of a few hectares or much larger tracts. The sands comprise mainly fine quartz grains and much comminuted shell material occurring in stratified layers. According to Bulleid and Jackson (1937), in places near the surface 'the sand sometimes appears to have been washed away, the laminations interrupted, and the depression filled with the wash of subsequent tides'. Commonly the shelly sands are calcreted and indurated. Gravel constituents include chalk flints, quartz, water-worn sandstone pebbles, chert, and rounded lumps of red marl, Tea Green Marl and Grey Marl; fragments of Liassic fossils are also present. The Burtle Beds contain marine, fresh-water and land shells, as well as remains of elephant, rhinoceros, horse, auroch, red deer, fallow deer, roe deer, hyaena and wolf. Bulleid and Jackson (1941) concluded that the Burtle Beds had been laid down in a marine environment at a time when sea level was much higher than at present. They suggested that the mammalian remains and non-marine shells were washed into the depositional basin from surrounding areas.

Kidson (1970) noted the occurrence of a clay member (the Burtle Clay) underlying the sandy deposits and quoted C'⁴ dates for shells from the Burtle Sand ranging from 29 000 ± 1000 BP at Greylake to 36 000 ± 88 BP at Cutley. Kidson (1971) remarked on the unreliability of shell-dates, which if taken at their face value would suggest that there was a mid-Devensian high sea level, and concluded that the Burtle Beds were best referred to the Ipswichian or Hoxnian, but probably the former.

A new element was introduced into the discussion by Kellaway (1971), who considered the Burtle Beds were 'residual masses of glacial sand and gravel' and 'outwash deposits' from a pre-Devensian ice sheet (Wolstonian) which moved from the Bristol Channel across the lowlands of Somerset. This interpretation was unacceptable to Kidson and Haynes (1972), who stressed the marine or estuarine character of the sediments. Kidson and Heyworth (1974) excavated a pit at Catcott Burtle to examine the stratigraphical sequence and found that sands (Burtle Sand) overlay clays (Burtle Clay), which overlay weathered Lias clay. At the approximate level of the junction between Burtle Clay and Lias clay was 'a branch, root or trunk of Corylus (Hazel)' which yielded a C¹⁴ date of 4280 ± 70 BP. This Holocene date suggests that the overlying sands and clays were deposited at a comparatively recent date and not during Pleistocene times. Kidson and others suggested that Pleistocene Burtle Beds might have been reworked, at least on the margins, by the Flandrian Sea to produce Holocene sand deposits. Clearly, it may be that other sands, previously described as 'Burtle Beds' in the Somerset Levels area, are of Holocene age and of marine origin; it is equally possible that there are older sand and gravel deposits of uncertain origin in this general area. Until more evidence is forthcoming the age and origin of the Burtle Beds remain uncertain.

Head

Deposits mapped as Head comprise unsorted or poorly sorted gravelly sands or gravelly and sandy loams and are interpreted as being solifluxion deposits derived in the main from surrounding Palaeozoic outcrops. The deposits occupy either valley sides and bottoms or ridge tops and, in this part of Somerset, mainly overlie Mesozoic formations. Head also occurs as flat-topped or gently shelving terrace-like spreads and attains thicknesses of up to 4.5 m locally. In the Quantock Hills the valleys are narrow and steep-sided and gravelly detritus has only accumulated in valley bottoms and on some valley sides. However, where the gravelly detritus deposits join ground underlain by Mesozoic rocks, extensive spreads of gravelly Head are locally developed. One such spread forms a marked, terrace-like, flattish-topped feature south-south-west of Shurton [ST 200 440]. The spread originates in valleys in the Quantock Hills and has a length of over 6 km and a width of up to 2 km; near Fairfield House it slopes north-east at about 1:70. This major gravel spread may formerly have been connected to patches of gravel at Wick, Stolford and Whitewick Farm (p.86). Similarly, some of the isolated patches of Head which fringe the northern parts of the Quantock Hills may be remnants of a once continuous sheet of Head gravels which has been dissected in post-Devensian times.

For the most part the gravelly material comprises fairly well-rounded pebbles of Devonian sandstones of types found in the Quantock Hills, but slates are also present. As estimated by eye, it appears that the pebbles become slightly less well rounded as the gravel spreads are traced towards their sources in the Palaeozoic uplands. On the basis of the mapping it seems clear that the gravel spreads originated from detritus derived from extensive weathering and freeze-thaw agencies, probably during the late-Devensian glaciation, when the district experienced a periglacial climate and solifluxion processes transported the detrital material downslope. Cryoturbation has affected some of the gravels and provides evidence that there were cold spells subsequent to their deposition.

The degree of roundness of the pebbles of hard Devonian sandstone is unexpected if they have been transported by solifluxion and at most only a few kilometres from their source. There is ample evidence from parts of west Somerset that a Permo-Triassic topography is being exhumed by the stripping of deposits of that age, and that Permo-Triassic conglomerates underlie and are overlapped by later Triassic deposits in the area east of the Quantock Hills. Thin, feather-edge, Permo-Triassic conglomerates which may once have been present in valleys which fringe the eastern part of the Quantock Hills could have provided the earliest rounded Devonian sandstone pebbles for incorporation in the Head gravels. However, crude bedding and channel-fill features in parts of the usually unsorted and unbedded Doniford gravels attest to the action of water in the deposition of at least some of the gravels. In the light of Kellaway's (1971) suggestion that ice may have moved along the Bristol Channel to invade the lowland areas of central Somerset in Wolstonian times, it is conceivable that some material may have been carried into this district by ice to be redistributed by solifluxion processes during the Devensian. This could also explain the presence of a few chalk and flint erratics in some of the Devensian Head deposits of this part of Somerset. On the other hand, these erratics could be the soliflucted remnants of thin Cretaceous deposits, which may once have been present in this area. AW

Raised beach deposits

Raised beach deposits have long been known from Spring or Birnbeck Cove, Weston-super-Mare (Day, 1866; Mackintosh, 1868) and Middle Hope (Sanders, 1841; Strahan and Cantrill, 1912). Gilbertson and Hawkins (1977b) considered that the raised beach deposits at Swallow Cliff (p. 86) accumulated under periglacial conditions and compared them to the glacial deposits of Wolstonian age at Kenn, 9 km to the east-north-east (Gilbertson and Hawkins, 1977a), which occur in a similar stratigraphical context, that is, underlying estuarine deposits with a comparable macrofauna to bed 8 at Swallow Cliff (see below). The age of the underlying rock platform is not known. Radiocarbon dates of around 39000 to 33000 years BP for the mollusca of the beach (Callow and Hassal, 1969) have been generally rejected as exceeding the range of reliable dating (Kidson, 1971; Stephens, 1973; Gilbertson and Hawkins, 1977b). However, mainly on the basis of sea levels deduced from the terraces of the River Severn, Donovan (1962) considered that the Swallow Cliff beach and associated wave-cut platform were mid-Devensian (Upton Warren Interstadial), an age which would accord with the radiocarbon dates.

On the south side of the Mendips opposite the Anchor Inn, Bleadon (see below), breccias with marine shells and a cold mammalian fauna are recorded at about the same level as the marine beaches to the north. It appears here that an earlier beach deposit may have been incorporated into a later (Devensian) solifluction scree, but the details are not clear.

Along the north side of Brean Down there is a well-marked platform at about 12 to 14 m above OD (ApSimon and others, 1961, p. 71), and this apparently corresponds to the wave-cut platforms farther north. Correlation of these various features has been confused in the past because various authors omitted to state the datums of their measurements (see Gilbertson and Hawkins, 1977b and cp. Stephens, 1973, p.44). Comparison of the heights of these features with those of other areas must take into account the very high upper tidal limit of the Severn estuary (up to about 6 m OD for spring tides), which results in the correspondingly great height of the present-day shingle beaches. GWG

Older blown sand and associated deposits

Banked against the limestone cliffs of Brean Down, on its southern or landward side, are thick deposits of blown sand with breccia horizons of Carboniferous Limestone debris. These sediments are exposed in cross-section at Sand Cliff [ST 296 587] at their western or seaward end. They have been excellently described and their chronology discussed in detail by ApSimon and others (1961; see p.87). The oldest layers contain an arctic fauna and are presumed to be late Devensian in age.

Comparable breccia deposits to those of Brean Down are found farther inland, banked in places against the foot of the limestone hills. Opposite the Anchor Inn, Bleadon [ST 3317 5710], a scree of angular dolomitic limestone debris in a red-brown sand and silt matrix can be seen plastered against, and filling fissures within, the old cliff of Carboniferous Limestone. This deposit was formerly much better exposed (Palmer, 1934, pp. 151–152) and is recorded as comprising a lower coarser breccia, with reindeer remains, Microtus ratticeps and marine shells, overlain by an upper sandier gravel or breccia which extended to a height of 9.1 m above road level [16 m above OD]. The latter contained remains of British or Romano-British occupation, including pottery and bones of domestic animals. GWG

Cave and fissure deposits

As a result of mining and quarrying in the Carboniferous Limestone, bone-bearing caves and fissures have been uncovered from time to time. The literature relating to these finds is widely scattered and the reader is referred to a bibliography of Quaternary sites in the Mendips and surrounding areas (Donovan, 1954, 1964), which has been brought up to date and reissued with a valuable chronological summary by Hawkins and Tratman (1977). Four sites within the Weston district are mentioned. The Hutton Bone Cave [ST 3610 5813], discovered by miners in about 1756, contained an extensive cold-climate (Devensian) mammalian fauna associated with Upper Palaeolithic flint implements. A single specimen (since destroyed) of Hippopotamus amphibius was reported, which may have come from beneath the cold fauna or from a separate cave. Quarrying through the years at Uphill Quarry [ST 315 584] has uncovered no less than 13 caves and numerous associated cave-earth filled fissures (see Harrison, 1977). The faunas were essentially cold climate (Devensian) and were associated with Middle and Upper Palaeolithic flint implements. The Bleadon Bone Cave [ST 341 567], also uncovered by quarrying, yielded a temperate to cold (Devensian) fauna, but Palaeoloxodon antiquus is also recorded, suggesting an earlier warm fauna (Ipswichian). Lastly, molars of Dicerorhinus kirchbergensis recorded from a fissure in a quarry at Milton, Weston-super-Mare [ST 3400 6268], indicate an earlier (?Hoxnian, or possibly Ipswichian) warm climate. GWG

Alluvial deposits and peat

Estuarine clays occupy a large surface area in the eastern part of the district. The clays form a remarkably flat surface which attains a height of between 4.6 and 6.7 m above OD. South of the Mendips the clays occupy a belt of ground from the present-day coastline up to about 9 km inland, and beyond this farther inland are the peat moors of central Somerset. The surface of the peat moors is generally at a lower level than that of the coastal belt of clay and this has given rise to drainage difficulties.

Many boreholes, sunk by various authorities but principally by the South Western Road Construction Unit of the Department of the Environment prior to the construction of the M5 motorway, have yielded information on the nature and thickness of the drift deposits. When collated, this information presents a useful, though incomplete, picture of the distribution of sediments and the shape of the buried topography, for although the present-day surface of the Somerset Levels is low, flat and monotonous, the drift deposits cover a varied 'solid-rock' landscape. Important features which emerge from the subsurface data (Figure 17) are: a relatively deep channel striking east-north-east north of Puriton, which may be regarded as a proto-River Brue; a deep channel between Brent Knoll and Badgworth, which must join a proto-River Axe north of Brent Knoll; and a proto-Lox Yeo tributary of the River Axe.

It is clear from the drift-free areas that locally the present-day topography is intimately related to the solid geology. Topographic depressions are associated with soft rock types (e.g. Mercia Mudstone Group, Westbury Formation, certain Lias clays) or with intensely faulted belts which produce zones of weak strata. This is evident not only inland but also in the intertidal areas of the west Somerset coast, where the strike of soft strata determines topographical depressions and vice versa. In several places boreholes through the drift show that alluvial deposits rest directly on Blue Lias limestones, which presumably form roughly east-south-east-trending small-scale dip and scarp sub-drift topographical features similar to those in the coastal exposures of the stone beds. A sub-drift 'lobe' of Lias clay projects east-north-eastward from the Lias exposures at Pawlett. Similarly, a sub-drift 'hillock' of Lias is present north-north-east of Pawlett and defines the southern slopes of the proto-Brue valley, which lies north of the present-day Brue channel. The buried proto-Brue valley may be about 30 m below OD just north of Highbridge (see below) and strikes east–west from the mouth of the present-day Brue to north of Highbridge, thence east-south-east to the Bason Bridge area. Evidence from cores in the solid Lias clay beneath the drift suggests the presence of important east–west or east-south-east-trending faults in the vicinity of this proto-Brue valley, and it is here suggested that belts of weakness along zones of major strike-faulting determined the position of the valley. The northern slopes of the proto-Brue valley are defined by 'highs' beneath Burnham-on-Sea and Brent Knoll, but north of these places data are scarce.

The Burton Row Borehole, sited at the foot of Brent Knoll's northern slopes, proved 8.4 m of drift deposits to a level about 0.5 m below OD. The presence of such a thick alluvial sequence, overlain by landslipped material and hillwash so close to the hill, suggests that when it was an island the slopes of the Knoll must have been steep and possibly cliff-like. The deep channel east-north-east of Brent Knoll was proved by borings; the Rooks Bridge Borehole (Richardson, 1928, and see p. 124) proved drift deposits to a depth of about 30 m below OD, similar to the depth of the proto-Brue valley. This channel presumably joins a proto-Axe valley in the area between Brent Knoll and Bleadon, Because the Mercia Mudstone Group marls are mainly soft and easily eroded, a proto-Axe valley would most likely be located within the incrop of these rocks between North Yeo Farm and the southern slopes of Bleadon Hill. The southern edge of the sub-drift crop of the Mercia Mudstone lies between the drift boundary at the foot of Bleadon Hill and a borehole [ST 371 543] which proved probable Grey Marl just east of the Weston-super-Mare district. The course of the buried valley west of Bleadon is unknown.

In regard to the alluvial sediments of the Somerset Levels, the basal peat is between 0.15 and 0.61 m thick where present, and where oldest it has been dated by radiocarbon techniques at between 8500 and 8000 years BP. In many places the basal peat is overlain by sand which is grey, uncemented and contains some silty clay. The sand is up to 16.8 m thick and in places is overlain by up to 14 m of grey silty clay which contains intercalations of fine sand (intertidal deposits), succeeded by between 1.83 and 7.93 m of grey silty clay with traces of peat or peaty clay (saltmarsh deposits). Next above comes the 'OD peat', commonly between 0.61 and 2.13 m thick, which has been radiocarbon dated at over 5000 years BP and, unlike the older peat, is horizontal with its base lying at about Ordnance Datum. The overlying and uppermost deposit comprises grey or brown mottled clay, which is slightly silty in places and commonly peaty near the base; this clay deposit (the 'Roman Clay') averages about 4.30 m in thickness. AW

The Holocene deposits within the Weston-super-Mare district north of the Mendips occupy over half of the land area and are comparable to those of the Somerset Levels on the south side of the Mendips. Boreholes penetrating to bedrock are largely confined to the Weston–Uphill area and to the line of the M5 motorway a short distance to the east of the district, but are very sparse elsewhere, so that the form of the buried topography is not known in detail. However it has been shown that a major westerly-draining channel, probably extending to 30 to 35 m below OD in the Weston area and entirely excavated in Triassic red marls, is present between the Mendips and the Weston–Worle ridge. Another channel is subparallel to the present-day line of the River Yeo east of the district (see Gilbertson and Hawkins, 1977b, fig. 36).

The Holocene succession, as far as it is known, appears to be very similar to that south of the Mendips, with a thin peat (less than 1 m) commonly occurring at the base, and an overlying sequence of grey sands (partly sub-tidal), thickest in the direction of the sea and within the central deepest parts of the channels (over 16 m proved), that is replaced laterally on either side from above downwards by laminated silts and silty clays (intertidal deposits) which extend upwards with generally decreasing grain-size to the level of the 'OD peat'. The latter is widespread; it is usually about 0.6 to 1.4 m thick but ranges from a few centimetres to 1.52 m, and rests on a surface of blue clay at between 0.5 m below and 1.8 m above OD. Finally, overlying the 'OD peat' and forming the surface of the levels is the so-called 'Roman Clay', which varies from 3.0 to 5.5 m in thickness.

Sedimentation studies involving the transport of marine foraminifera up the Severn estuary have enabled the local depositional environment to be reconstructed (Murray and Hawkins, 1976) and in accordance with the data from sea level curves no evidence was found for a period of marine transgression in the late Holocene. Whether peat or mud accumulated in any one area depended on the relative growth rate of the peat versus the sedimentation rate of the mud, itself dependent on the continuing sea level rise. In the area under discussion the major growth of the 'OD peat' coincided with a slowing down of the ecstatic rise in sea level around 5000 to 4500 years BP. The subsequent overwhelming of the peat appears to have been due to climatic deterioration after the Holocene Climatic Optimum; it would not have been synchronous over the whole area but would have taken the form of a 'creeping transgression', spread over an appreciable period of time. GWG

Blown sand

Thick deposits of blown sand fringe Sand Bay, Weston Bay and Bridgwater Bay, where they overlap landwards on to the edge of the flats formed of estuarine clay and pass laterally seawards into the modern beach sand. At the south-eastern angle of Sand Bay these deposits extend up into the lower part of Kewstoke Woods, where they overlap on to the Carboniferous Limestone. In the central part of Weston-super-Mare blown sand, 3 to 4.5 m in thickness, has been proved in numerous boreholes (pp. 92–93). The thickest proving is at the southern end of this spread, where the following section was recorded in a borehole [ST 3128 5899] on the dunes near the sea: dry brown beach sand, 3.0 m; on moist brown sand with silt layers and some shells, 2.0 m; grey silty clay and some silt, 0.68 m; grey-brown sand, 0.23 m; clay with some silt, 3.0 m; grey-brown sand, 0.15 m; on blue-grey silty clay. GWG

Intertidal deposits

Seawards of the blown sand deposits which skirt the coastal strip are extensive spreads of tidal flats. The sediments of the intertidal zone of Bridgwater Bay have been studied and surveyed by Professor C. Kidson and his co-workers at the University College of Wales, Aberystwyth; their map (Kidson and Heyworth, 1976) shows the dominant sediment type in Bridgwater Bay to be mud. However, sand deposits are also important, particularly on either side of the mouth of the River Parrett at low water. Other ephemeral deposits include shingle, which occurs as shingle ridges or storm beaches, particularly from Stolford to Fenning Island, and also as discrete patches on Stert Flats. Kidson and Heyworth reported that the shingle is derived mainly from Lias limestone and Pleistocene gravel outcrops. Areas underlain by Quaternary marine clays are present offshore from Burnham-on-Sea, while a submerged forest occurs north of Stolford. Kidson and Heyworth brought forward evidence for abandoned channels of the Parrett in the present intertidal zone, which reinforces historical records of changes in this area.

Similar but less extensive intertidal zone deposits occur throughout the length of the coastline. Deposits of mud (up to 1 m thick) are present at Doniford Bay and Lilstock Bay, and extensive spreads of locally-derived shingle lie offshore from Watchet Harbour, offshore from the army camp at Doniford Bay, and in Lilstock Bay. The Watchet and Doniford shingle spreads are composed mainly of Devonian sandstone pebbles originating from the Head deposits along the Washford River and Doniford Stream respectively, whereas the spread at Lilstock comprises mainly Lias limestone cobbles derived by longshore drift from extensive outcrops of Blue Lias beds occurring between Lilstock and Kilve. AW

Coastal morphology

Characteristic of the north (Atlantic-facing) coast of the south-west England peninsula are bays and indentations shaped like a mirror-image of the letter L. Amongst these are St Ives Bay, Watergate Bay, Port Isaac Bay, Bude Bay, Barnstaple Bay and Bridgwater Bay. The form contrasts with that of the south (English Channel-facing) coast of the peninsula, which is characterised by roughly semi-circular shaped bays. The shape of the north coast bays is determined by a combination of geological structure and geographical position with respect to prevailing wind and wave attack. Thus, ESE-trending geological structures offer resistence to wave attack from the prevailing direction, tending to produce N–S coastlines. However, ESE-trending geologically weak zones are also present, and prolonged wave attack and erosion of these produces E–W or ESE-trending coastlines.

Bridgwater Bay provides an outstanding example of a bay shaped in this manner. The long, straight coast of north Devon and west Somerset from Morte Point to Stert is determined by competent and resistant ESE-trending rocks of Devonian and early Jurassic age; these rocks abut on the north side against submarine, less competent, softer strata which are associated with ESE-trending lines of structural weakness. In the northern part of the Weston district are east–west ribs or ridges of competent, resistant Carboniferous Limestone which project seawards as points and delimit Bridgwater Bay, Weston Bay and Sand Bay.

The detailed shape of the southern shores of Bridgwater Bay, as with the inland area, is controlled largely by the solid geology. Within the Mesozoic sequence of the coastal area are three main incompetent, softer and less resistant groups of rocks, the red mudstones of the Mercia Mudstone Group, the Westbury Formation and the Lower Lias shales of lithological division 2 (p. 61). Enclosing these groups are more competent strata with much hard siltstone or limestone. St Audrie's Bay provides an excellent example. The most westerly point is formed by Lower Lias division 1 limestones and shales and gives way, eastwards, to a small bay (where the slipway is located) cut in Westbury Formation and the softer Grey Marl. Farther east still, at a waterfall, is a prominent point formed in the resistant red and green marl alternations (the green siltstones being hard and forming prominent reefs on the foreshore), which gives way eastwards to the main indentation of St Audrie's Bay, shaped in soft, red mudstones of the Mercia Mudstone Group. The bay is delimited on the east by Blue Ben point, formed by Blue Lias limestones and shales which are faulted against the red marls. Similarly, Lilstock Bay is formed in red marls and Westbury Formation strata. AW

Prominent wave-cut platforms of Carboniferous Limestone occur at the foot of the coastal Carboniferous Limestone headlands in the northern part of the district. At Brean Down such a platform, the Howe Rock platform, appears to pass beneath, that is to predate, the oldest breccia horizon of Sand Cliff, which is presumed to be of late Devensian age (p.82). The inner edge lies at, or just below, OD (near sea level) and the outer edge is believed to be lower than 6 m below OD. (ApSimon and others, 1961, pp. 70–71). Mean sea level at the time of the cutting of the platform was considered to have been about 5 to 7 m below present OD (Apsimon and others, 1961, p. 105; Donovan, 1962) and to correspond in date to the aggradation of the Worcester Terrace of the River Severn (late Devensian). Similar rock platforms cut in Carboniferous rocks occur offshore from Weston and Middle Hope.

Along much of the north side of Brean Down there is a well-marked platform cut into the Carboniferous Limestone at about 12 to 14 m above OD, which apparently corresponds to rock platforms associated with raised beach deposits recorded farther north. Sea level at the time these platforms were cut is estimated to have been about 8 m higher than at present, but the evidence from Swallow Cliff (p. 86) suggests that two separate episodes of high sea level were involved. Higher, but much less well marked, platforms here and elsewhere may be related to higher sea levels (ApSimon and others, 1961; Ford and Stanton, 1968). GWG

Coast erosion and landslips

H. B. Woodward (in Ussher, 1908) briefly considered the question of coast erosion in this neighbourhood and quoted historical evidence for considerable cliff wastage involving substantial retreat of the land in the Watchet area. J. H. Blake, who geologically surveyed the Watchet coast area in the 19th century, was likewise greatly concerned about coastal erosion and discussed it in correspondence with H. W. Bristow in 1871. However, from Blake's sections (Geological Survey archive material), drawn in 1871, it is clear that there has been little change in the coast of the Watchet area, due to erosion, over the last 100 years or so. This is partly because of small-scale man-made works but, presumably, mainly because the wave-cut platform has reached such a width that the force of wave action during any but the highest tides is largely spent before reaching the cliffs. Nevertheless, tides still reach the base of the cliffs and undercutting and subsequent rock-falls are a constant hazard. Areas particularly prone to rock-falls include St Audrie's Bay and the cliffs between Kilve and Lilstock Bay.

The cliffs of this coast are commonly of the 'slope-overwall' type and suffer various types of landslip or mass-movement from clay sludging or clay flow to rotational slip. The kind of slip developed depends upon the rock types present and the availability of established planes of movement (e.g. fault planes) which facilitate collapse. Superficial movement (landslip) along pre-existing fault lines as the sea encroaches is a common phenomenon of the coast and old landslipped material, to the seaward of existing cliffs, marks abandoned, older cliff positions in some areas (e.g. St Audrie's Bay). Undercutting of the Westbury Formation at sea level facilitates mass-movement of various sorts, producing a melange of jumbled rock material from that formation, the Lilstock Formation and the Lower Lias. This is visible at St Audrie's Bay and Lilstock Bay.

Inland, landslips are generally small and confined to moderately steep slopes in Penarth Group or Lower Lias strata. East and east-north-east of Kilve is a gravelly, Head-capped hill underlain by Lias shales and limestones; the steep sides of the hill, lubricated by water seeping from the Head–Lias junction at the hillcrest, have been unstable in the past, and earthflows and small-scale slides have formed a hummocky topography.

The most extensive landslips in the district are at Brent Knoll. Small hummocky scars have developed locally, particularly on the northern and western slopes in Liassic mudstones or clays which rise steeply from the alluvium to a height of over 60 m OD at the limestone-capped platform which forms a 'shoulder' on the north-west side of the Knoll proper. On these steep slopes it is impossible to determine whether some strong but discontinuous features, which run around the hill, are related to the solid geology or are due to small-scale, rotationally-slipped, back-tilted blocks from higher levels. North and north-east of the Hill Fort at the summit are channel-like scars associated with hummocky ground and composite slip (mostly of earthflow type), which originates at the head of the slip close to the junction of Upper Lias clays with overlying Upper Lias silts. Larger landslips are evident on the southern slopes of Brent Knoll.

The landslip [ST 344 506] immediately north or north-northeast of Battleborough is probably an old rotational slip which is now much degraded and eroded. A small rounded hillock, rising to above 30 m OD, rests on the lower slopes of the Knoll and comprises lithologies similar to those associated with prominent limestone features which crop out at about 61 m OD. Immediately above this is a scar or embayment of dimensions corresponding to those of the slipped mass.

West of Battleborough a strong feature is formed by limestone which is correlated on lithological and faunal grounds with that making the flattish platform which rims the north-west part of Brent Knoll. The Battleborough feature is at a lower level than might be expected with undisturbed dip and strike. The evidence favours interpretation of this feature as a large-scale, rotationally-slipped block whose toe is truncated and covered by alluvial deposits, implying that the movement occurred during Pleistocene or early Holocene times. During the early stages of the Flandrian Transgression the flanking slopes of Brent Knoll (see p. 83) were cliff-like and hence very vulnerable to undercutting by an encroaching and rising sea and prone to superficial mass movement. AW

Details

Burtle Beds

In the vicinity of Stretcholt a low mound, with gentle slopes and barely reaching above the level of the surrounding alluvial flats, consists predominantly of sandy material. The soils on the mound are brown and grey sandy foams with scattered fragments of calcreted sandstone and a few pebbles. An exposure in the side of a pond [ST 2858 4400] south-west of Yearsley Farm showed about 1 m of brown sandy loam and well-rounded sand, with pebbles of hard shelly sandstone and Lias limestone. The deposit is banked against the Lower Lias clays of Pawlett Hill, but forms a poor feature where it meets the alluvial clays at their feather edge.

Underlying the alluvial clays to the west and west-south-west of Yearsley Farm, Burtle Beds have been observed in temporary sections. Hard, shelly sandstone fragments were obtained from a ditch [ST 2850 4407] at the depth of 1.37 m; similarly, at a locality [ST 2840 4422] near the eastern bank of the River Parrett, brown sands with calcreted shelly sandstones, seen for 0.46 m, were overlain by some 1.22 m of grey clay. These records show that the sandy Burtle Beds have a greater extent beneath the Holocene clays than is obvious from their surface outcrop, and that they predate some, at least, of the Holocene clays.

Even more poorly exposed is the spread of sands in the vicinity of Huntspill [ST 3110 4580]. Hereabouts, a very slight rise, barely apparent in places, marks the outcrop of another batch of sandy deposits referred to the Burtle Beds. Brown sandy and clayey loam soils, in places with brashy fragments of calcreted sandstone, are visible in the fields and the only method of determining the alluvial clay-sand boundary in many places is by augering. The common presence of Lias limestone fragments suggests that the Lias may be present at very shallow depth hereabouts, not only beneath the Burtle Beds but also beneath the alluvial clays. The Alstone House Borehole [ST 3170 4678] proved Lower Lias beneath only about 3.8 m of drift and confirms the inference from field mapping that the lower Lias is present at shallow depth in this area. AW

Head

Blue Ben–Kilve

Occupying a valley [ST 1260 4367] east of Blue Ben are reddish brown sandy loamy with numerous Devonian sandstone pebbles. Exposed in the cliff [ST 1260 4387] where the valley meets the sea are 1.52 m of reddish purple pebbly loam overlying Lias shales. The north-south-trending ridge immediately to the east is capped by reddish brown pebbly loams, 1.22 m of which are exposed overlying Lias shales at the cliff edge [ST 1296 4392].

In the vicinity of Court House [ST 1364 4368], East Quantoxhead, are two spreads of Head gravels. The south-westerly spread is composed of red pebbly and loamy sands which are exposed to a depth of 0.91 m in a pond [ST 1346 4318], to a depth of 1.07 m in a ditch [ST 1350 4327], and to a depth of 1.22 m in a bank [ST 1333 4310]. The gravel spread north of Court House occupies a featureless and flat depression, with purplish red or brown sandy and pebbly loarns at the surface; 1.52 m of gravel and loam overlie Lias clay near the edge of the deposit in a ditch [ST 1389 4368]. The best exposure of this gravelly Head, however, is at the cliff face [ST 1365 4426] east-north-east of Quantock's Head, where up to 2.44 m of brown sandy loam and sand contain abundant Devonian sandstone pebbles.

The valley in which lies the Kilve Chantry [ST 1463 4400] is floored by up to 2.44 m of red gravel comprising Devonian sandstone pebbles in sandy loam. Numerous exposures are visible in the sides of the brook where gravels rest on Lower Lias clays and limestones. The stream bank [ST 1452 4422] north-west of the Chantry reveals dark bluish grey Lias mudstone, seen for 0.61 m, overlain by 1.22 m of Head gravel, comprising reddish brown sandy loam with abundant moderately rounded Devonian sandstone pebbles. The pebbles are ill-sorted, with no apparent orientation and no signs of having been laid down in water. Below the soil layer for about 0.3 m is a band of small pebbles (up to 0.05 m across) overlying larger cobbles and boulders of varying sizes. Similar pebbly sands and loamy are visible for 0.91 m resting on Lias clay at one locality [ST 1500 4341], for 0.91 m resting on Grey Marl at another [ST 1501 4337], and for 1.22 m at a third [ST 1497 4316].

East and south-east of Kilve Chantry spreads of Head gravel cap hills and ridges. Reddish brown sands and loamy with abundant Devonian sandstone pebbles rest on Lias clay [ST 1540 4340] 0.5 km to the east of Parkhouse Farm; in places nearby, augering suggests that the gravels are only 0.5 m thick. A tiny patch [ST 1555 4424] of brown loam with abundant Devonian sandstone pebbles, perhaps 0.61 m thick, caps a hillock south-west of Park Farm.

Burton–Shurton

On the spread of gravel south-west of Shurton abundant rounded Devonian sandstone pebbles are present in brown sandy loam and clay soils; locally a few pebbles of Lias limestone and chalk were observed. An old quarry [ST 1907 4382] west of Burton has been filled in, but gravelly and pebbly soil in the adjacent field suggests that the pit was worked for gravel.

Wick

Two spreads of gravel on either side of the stream near Wick [ST 2160 4451] may be connected beneath the alluvial deposits. The gravels occur just below 15.22 m OD, and give rise to brown sandy loam soil with abundant Devonian sandstone pebbles and a few chalk and fresh flint pebbles. Capping a low rise [ST 2255 4475] to the east are remnant patches of loamy soil with a few Devonian sandstone pebbles, an indication that Head deposits once covered this ground.

Stolford–Whitewick Farm

Spreads of Head gravel occur at Stolford [ST 2300 4570] and Whitewick Farm [ST 2353 4525], and a small inlying patch [ST 2383 4544] is surrounded by alluvial deposits. Shallow exposures in drains and field boundaries show the gravels to comprise mainly rounded Devonian sandstone pebbles in brown and reddish brown loam and sandy loam. Soils are very stony owing to the abundant Devonian sandstone pebbles and are known to local farmers as 'stonerush'. It is probable that the three spreads are continuous beneath the alluvium of the Chalcott and Crow Marsh Drain. The flattish top of the Stolford spread rises just above 7.62 m OD, whereas the Whitewick Farm spread barely reaches this height just east of Chalcott Farm [ST 2362 4475]. AW

Raised beach deposits

Birnbeck Cove

The raised beach at Birnbeck Cove [ST 310 625] is now very poorly exposed. The underlying rock platform is estimated to lie at about 13.5 m above OD and about 1.5 m of partially cemented bouldery beach deposit can be seen in the cliff top. Previous accounts (p. 81) have recorded a basal conglomerate 0.9 to 1.2 m thick, cemented with ochreous material and containing numerous shells similar to those recorded from Middle Hope (Swanton, 1911, p. xv), overlain by dune sand 1 m or more thick, cemented at the top in places and unconformably overlain by thick Head deposits. Mackintosh (1868, p. 282) described the basal bed as consisting of angular, subangular and rounded pebbles with numerous bones in the upper part, but Day (1866, p. 116) insisted that 'the shingle which composed it is as much rounded as that upon the present beach'. Day stated that the bones 'included, I remember, limb bones of ruminants and their teeth, also a few teeth of carnivores; but the most abundant remains were the teeth of a small species of horse' and he added a footnote that 'the only carnivorous teeth in my possession are one of Hyaena spelaea, the Cave Hyena, and one of Canis vulpes, the fox.' The combination of marine and terrestrial fauna is puzzling and may be due to contamination from a fissure deposit in the immediately adjacent limestone cliff.

Middle Hope

In the Middle Hope area the raised beach deposits are seen in two separate localities in the sea-cliffs on the north side of the peninsula, at Swallow Cliff [ST 3250 6608] and east of North Bay [ST 3370 6641]. Gilbertson and Hawkins (1977b) reinvestigated the section at Swallow Cliff and a summary of their measured sections (table 1; fig. 2), using their bed numbers, is as follows:

The section abuts against an old vertical cliff, its base at about 12.5 m above OD.

Gilbertson and Hawkins recorded 13 molluscan taxa from bed 8, of which Macoma balthica (79.7 per cent) and Littorina spp. (14.3 per cent) accounted for most of the individuals. The mollusca represent a mixture of hard- and soft-bottom living forms and marine and estuarine forms of temperate aspect. Foraminifera, mainly marine but some estuarine, were recorded from many of the beds, including bed 9, and were considered to have been derived by niveo-aeolian processes from outside the area. Bed 8 was considered to be Ipswichian (Interglacial)) in age. On the evidence of their bedding and the angularity of many of their contained rock fragments, beds 2 to 7 were thought to have accumulated under periglacial conditions, probably in early Ipswichian or late Wolstonian times.

The exposure near North Bay shows about 1 m of beach boulder material with shells, lying directly on Carboniferous rocks. The base is estimated to lie up to 0.5 m above the platform base at Swallow Cliff. Between the two exposures and elsewhere on the peninsula there are extensive remnants of a wave-cut platform that slopes gently seaward between upper and lower limits of about 13 and 9 m above OD. GWG

Older blown sand and associated deposits

Brean Down

The following summary account of the older blown sand deposits at Sand Cliff [ST 296 587] is drawn from ApSimon and others (1961) using their bed numbers:

The oldest layers occur at the foot of the limestone sea-cliffs, and successively younger layers are exposed vertically upwards in the face of Sand Cliff and horizontally outwards along the foreshore to the south. The ruling (depositional) dips are southwards, about 25° in the older layers (13 to 8) decreasing from 10° to nearly horizontal in the younger layers (7 to 1) which only occur, apart from layer 3, in the lower part of the section near the foot of the cliff.

The section at the north end extends from 5.2 m to about 36.6 m above OD. Layers 13 to 11 contain an arctic fauna named by Dr R. J. G. Savage, including Dicrostonyx gulielmi and D. henseli (lemmings), Microtus anglicus (an extinct vole), Lepus timidus (arctic hare), Alopex lagopus (arctic fox), Canis lupus (wolf), Mammuthus primigenius (mammoth), Megaloceros giganteus (giant deer), Rangifer tarandus (reindeer) and Bos/Bison sp. (auroch or bison), and are presumed to be late Devensian in age. The soil horizon at the top of layer 11 apparently represents an amelioration of the climate and may correspond to the Allerød Interstadial (ApSimon and others, 1961, pp. 107–108). No archaeological or faunal evidence was found in layers 10 to 8, apart from the mature red soil which marks the top of layer 8. ApSimon and others (1961, p. 106) considered that the change from terra fusca to brown earth formation might have resulted from the amelioration of the climate from the Boreal to the Atlantic periods at about 7000 BP and that the breccia layer below the soil was formed at a time of frost activity before the vegetation cover had been sufficiently established to prevent the downwards sludging of talus at about the beginning of the Flandrian Stage. GWG

Alluvial deposits and peat

Pawlett–Huntspill

Between the Lias 'highs' of Pawlett and Huntspill is a depression in the solid rock surface. Three boreholes along the course of the man-made Huntspill River, at Slow Way Bridge [ST 302 451], Bleak Bridge [ST 307 448] and Puriton Drove Bridge [ST 316 445], proved the Lias rockhead at 11.92, 12.49 and 17.66 m below OD respectively. The boreholes proved the alluvial deposits to consist mainly of soft clays, brown coloured in the top metre or two, blue below; peat was proved only in the Puriton Drove Bridge Borehole, where it was 0.12 m thick with base at 0.32 m above OD.

The Pillsmouth Borehole [ST 302 475], south of Pillsmouth Farm, provided a more typical sequence of the alluvial deposits in that sand was found, at depth, beneath the clay sequence. The following section was proved:

The peat has its base at 1.52 m below OD.

Puriton–Withy Farm–Hackness–Edithmead

The account of the area immediately north of the Polden Hills (Figure 17) is based principally upon information obtained from site investigation boreholes prior to construction of the M5 motorway, with the consent of the Director, South Western Road Construction Unit. A borehole [ST 317 423] situated 750 m NNW of Puriton proved Lias rockhead at 7.9 m below OD. Above this were grey silty alluvial clays, with traces of peat in places and with a discrete peat bed 0.15 m thick with its base at 4.9 m below OD. Traces of fine sand were present in the clay above the peat and in the lowest metre or so of alluvium, which rested on Lias clay and limestone. A cluster of six boreholes [ST 319 432] at Puriton Drove proved Lias rockhead at shallower levels (ranging from 0.21 m below to 1.37 m above OD), thus demonstrating a rise in Lias bedrock, probably in strata of bucklandi Zone age, and suggesting that a sub-drift extension of the Pawlett ridge is present hereabouts (Figure 17). Grey clay is the dominant drift lithology, with peat up to 1.37 m thick [ST 319 431] resting on Lias; the peat thins (to between 0.61 and 1.07 m) to the north [ST 320 432], where it is sandwiched between alluvial grey clays just above the Lias bedrock.

A borehole [ST 321 437] 800 m WSW of Withy Farm showed that the Lias bedrock (limestone at sub-drift surface) drops to a level of 13.25 m below OD. The following section was proved:

The thicker peat has its base at 1.00 m above OD. Also of interest are the basal sands and sandy clays which are present below the grey clays and peaty of this Withy Farm area. The sequence proved in a borehole [ST 323 443] near the Huntspill River (south bank) was similar, except that a basal peat 0.15 m thick was found resting on Lias clay at 18.87 m below OD; otherwise, the grey alluvial clays contained a 1.22-m-thick peat band, with base at 0.76 m above OD, and a 7.6-m-thick batch of sands with base at 16.73 m below OD. A slightly different sequence was proved in a closely adjacent borehole [ST 323 444] on the north bank of the Huntspill River: the basal peat (0.53 m thick, with base at 20.16 m below OD and resting on Lias strata of the angulata Zone) was overlain by 7.01 m of sand with peaty traces, overlain in turn by a grey clay sequence with peat horizons. The 'OD peat' in this borehole was 0.99 m thick, with base at 0.18 m above OD; a lower peat bed (resting on the sand deposit) was 0.38 m thick at 12.01 m below OD; 4.88 m of grey clay with partings of silt or fine sand up to 1 mm thick overlay the middle peat.

Grey silty clay with partings and lenses of grey silt and sand characterised the lower parts of alluvial sequences proved in boreholes [ST 323 447]; [ST 324 447] in an area 1 km NW of Withy Farm. Peaty clays and discrete peat bands were present at about the OD level, but only traces of peat occurred in the higher parts of both boreholes. The Lias bedrock was proved at levels of 19.81 and 19.66 m below OD respectively.

Two boreholes [ST 326 458]; [ST 327 457] situated 1.5 km E of Hunt-spill proved Lias bedrock at 19.75 and 20.42 m below OD respectively. Grey silty clay (9.60 and 10.52 m thick) in both boreholes contained the 'OD peat', which was 1.22 m thick (base at 0.15 m below OD) in the former borehole, and 0.76 m (base at 0.46 m above OD) in the latter. Below the clay in both boreholes was dense, grey, silty, fine sand with traces of clay. A basal peat (resting on Lias clay) was 0.84 m thick in the first borehole but was represented only by peaty traces near the bottom of the sand in the second.

Two boreholes [ST 327 462]; [ST 328 462] situated 500 m WNW of Hackness proved Lias bedrock at a level of 17.37 m below OD. The grey silty clay at the top of both drift sequences was 8.53 and 7.92 m thick respectively, and contained a discrete peat band 0.61 m thick with base at 0.30 m above OD in the former borehole and 1.07 m thick with base at the same level in the latter. Grey fine sand with intercalations of silty clay was present in the lower parts of both boreholes with traces of peat close to the base. The Hackness Sluice Borehole [ST 333 462], to the east, proved 14.02 m of grey and blue clay, with a 1.22-m-thick peat band at 0.30 m above OD. Below the clays were 11.28 m of sand, resting on Lias clay bedrock at 19.96 m below OD.

Two boreholes [ST 328 465]; [ST 329 465], respectively on the south and north banks of the River Brue, proved 7.92 and 8.68 m of grey clay, including a 1.5-m-thick peat band with base at 0.13 m below OD and a 1.68-m-thick peat with base at 0.30-m below OD. The underlying grey, silty, fine sand, in places with clay partings, lenses and peaty patches, was 18.29 m thick in the former borehole and 18.13 m in the latter; the junctions with the Lias bedrock were proved at 20.88 and 20.73 m below OD respectively.

Lias rockhead was proved 23.32, 22.25, 23.47 and 20.42 m below OD in boreholes [ST 327 467], [ST 328 466], [ST 329 465] and [ST 330 465] just north of the present-day Brue channel. The clay in the higher parts of the boreholes was proved to be up to 9.14 m thick and contained a peat from 1.22 to 2.44 m in thickness, its base ranging between 0.46 m above and 0.76 m below OD. Beneath the clay were grey sands, in places with thin clay bands and partings. Of interest are the occurrence of a lower peat band 0.61 m thick, separating clay from sand at 6.40 m below OD in one borehole [ST 327 467], and a basal 4.88 m of grey silty clay with traces of peat underlying the sand in another [ST 328 466].

Two closely spaced boreholes east of Highbridge penetrated peats below the 'OD peat'. One borehole [ST 330 470] proved the following sequence:

This borehole is thought to be close to the southern slopes of a buried valley of the proto-River Brue; the coarse sand and gravel at the base of the drift deposits may represent the remnants of gravels laid down by the proto-Brue, in and alongside its early channel. The 0.23-m-thick peat (with base at 20.03 m below OD) which rests upon the basal sand and gravel has yielded a radiocarbon date of 8365 ± 100 BP.

The adjacent borehole [ST 331 470] proved a slightly different sequence, at the lower levels, as follows:

Correlation of the lowest drift deposits with the sand and gravel of the previous borehole seems reasonable, as does the equivalence of the two lowest peats, but the logs of the two boreholes show how rapid the lateral changes in the succession below the 'OD peat' can be.

North of the proto-Brue valley, and about 1500 m NE of Highbridge, the Pillmore Lane Borehole [ST 336 480] proved a similar sequence. The higher grey silty clays were proved to a depth of 11.89 m and contained a peat bed 1.37 m thick with its base at 0.15 m below OD. Below the clays were dark grey, fine to medium-grained sands with traces of clay and peat to the depth of 21.18 m, underlain by clay and sand with traces of peat to the junction with Lias mudstone.

Some 1500 m SE of Edithmead, two boreholes [ST 3377 4854]; [ST 3383 4851] proved similar sequences. The former penetrated grey silty clays to 10.67 m; these clays contained a 2.13-m-thick peat band with base at 0.46 m above OD. The clays rested on grey, fine- and medium-grained sand, with traces of clay and peat near the top, to the depth of 27.43 m, which is the junction with Lias mudstone at 21.79 m below OD.

East Huntspill–Bason Bridge

The flat surface of the Somerset Levels coastal clay belt persists into the East Huntspill area. The Bason Bridge Borehole [ST 3444 4548] was made in 1896 to provide a water supply for the local schools (see p. 124). 'Fine calcareous silt', probably representing the grey silty clay of later boreholes, was proved to a depth of 19.20 m and included a peat (thickness given as 0.91 to 1.22 m in Richardson, 1928), with base at 0.46 m below OD, which is the lateral equivalent of the 'OD peat' of the area to the west. Underlying the clay sequence were sands with 'clay' and 'shale', and including a bed of 'shingle' near the base. The 'shingle' contained shells (including derived Lias fossils), pieces of Lias, Carboniferous Limestone and quartz, all of which could have been locally derived. The classification of part or the whole of the sandy deposits as Burtle Beds (Richardson, 1928) is here considered to be erroneous; comparison with later subsurface information shows the silty sand deposits of the Bason Bridge Borehole to be a normal feature of the alluvial sequence in this part of the Somerset Levels. The presence of marine or estuarine shells in some of the sand and shingle deposits suggests that the gravel is a marine or estuarine deposit or that it is of mixed origin.

Another borehole [ST 348 458] was drilled in 1912 to the depth of 24.84 m, abandoned and filled in. A second well was made in 1922 and deepened to 37.95 m. The section is given below, the surface level being 5.18 m above OD:

The classification of the strata given above is open to question in that the 'clay and stone' could be Lower Lias. However, it is possible that the 'clay and stone' is the lateral equivalent of the 'shingle' of the other Bason Bridge Borehole (see above) and of the gravel of boreholes to the north-west. If the classification is accepted, the drift-solid junction is at a depth of 33.68 or 28.50 m below OD. Even if the 'clay and stone' is classified as Lower Lias, the drift-solid boundary is at 27.13 m below OD, so that this old borehole proved the occurrence of the deep buried valley of the proto-Brue just north of the present-day channel of the River Brue at Bason Bridge. It will be seen that this borehole section is comparable with the generalised stratigraphy of other sections hereabouts; the upper clay deposit was proved to a depth of 14.33 m and contained the 'OD peat' (0.15 m thick) at 0.46 m above OD. The clay overlay sediments apparently similar to deposits at the same stratigraphical level elsewhere. The only important difference is in the depth of Lias bedrock.

Edithmead–East Brent–Edingworth

Trial borings for the M5 motorway indicated a deep buried valley to the east and north-east of Brent Knoll, which presumably joins a proto-Axe valley to the north. The stratigraphy and lithologies proved in the boreholes are broadly similar to those already described from the area to the south; grey, silty clays, generally containing peat traces or discrete bands of peat, overlie predominantly sandy sediments, which rest on bedrock.

East of Edithmead a borehole about [ST 3405 4928], with surface level 5.5 m above OD, proved alluvial deposits to a depth of at least 28.35 m, including a 1.22-m-thick peat with base at about 0.91 m above OD. Another borehole [about 354 504], east of Battleborough, proved grey or bluish grey silty clay (with two discrete bands of peat) to a depth of 9.90 m; underlying the clay was grey sandy or clayey silt (with peat traces and a discrete peat band) to the final depth of 18.75 m; bedrock was not reached.

A borehole [ST 358 517] with surface level of 5.67 m above OD, situated east of East Brent, proved the following sequence: