Content and licensingview original scan buy a printed copy

Geology of the country around Newton Abbot. Memoir for 1:50 000 geological sheet 339, New Series

By E. B. Selwood R. A. Edwards S. Simpson J. A. Chesher R. J. O. Hamblin M. R. Henson B. W. Riddolls R. A. Waters

Bibliographical reference:Selwood, E. B., and others. 1984. Geology of the country around Newton Abbot. Memoir British Geological Survey, Sheet 339.

London: Her Majesty’s Stationery Office 1984 © Crown copyright 1984 ISBN 0 11 884274 9 Printed in England for Her Majesty's Stationery Office Dd 736172 K16

• Authors
• E. B. Selwood, R. A. Edwards, S. Simpson, J. A. Chesher, R. J. O. Hamblin, M. R. Henson, B. W. Riddolls, R. A. Waters
• Contributors
• Dartmoor Granite J. R. Hawkes, G. Bisson
• Economic geology R. C. Scrivener
• Geophysics G. P. Riddler
• Palaeontology S. C. Matthews, W. H. C. Ramsbottom, C. J. Wood
• Water supply E. A. Edmonds
• Research commissioned by the Natural Environment Research Council from the University Of Exeter
• Directors and supervisors of research E. B. Selwood and S. Simpson University of Exeter G. Bisson, District Geologist, Institute of Geological Sciences
• Research assistants J. A. Chesher, R. A. Edwards, R. J. O. Hamblin, M. R. Henson, B. W. Riddolls, R. A. Waters
• Authors and Contributors
• The late S. Simpson, M. A., Dr. rer. nat., E. B. Selwood, BSc, PhD, R. A. Edwards BSc, PhD, J. A. Chesher, BSc, PhD, R. J. O. Hamblin, BSc, PhD, M. R. Henson, BSc, PhD, B. W. Riddolls, BSc, PhD and R. A. Waters, BSc, PhD Department of Geology, University of Exeter, North Park Road, Exeter EX4 4QE.
• S. C. Matthews, BSc, PhD Department of Geology, University of Bristol, Queen's Building, University Walk, Bristol BS8 1TR
• G. Bisson, BSc, BSc (Eng.), ARSM, E. A. Edmonds, MSc and R. C. Scrivener, BSc British Geological Survey, St Just, 30 Pennsylvania Road, Exeter EX4 6BX
• J. R. Hawkes. BSc, PhD, G. P. Riddler. BSc, PhD and C. J. Wood, BSc British Geological Survey, London
• W. H. C. Ramsbottom, MA, PhD British Geological Survey Ring Road Halton, Leeds LS15 8TQ

(Front cover)

(Rear cover)

Other publications of the Survey dealing with this and adjoining districts

Books

• British Regional Geology
• South-West England, 4th Edition, 1975
• Memoirs
• Geology of the country around Okehampton, 1968
• Mineral Dossier
• Ball Clay, 1975

Maps

• 1:625 000
• Sheet 2 Geological Sheet 2 Quaternary Sheet 2 Aeromagnetic
• 1: 1 584 000
• Tectonic map of Great Britain and Northern Ireland
• 1:50 000
• Sheet 324 (Okehampton)
• Sheet 325 (Exeter)
• Sheet 326 and 340 (Sidmouth)
• Sheet 338 (Dartmoor Forest)
• Sheet 339 (Newton Abbot)
• Sheet 349 (Ivybridge)
• Sheet 350 (Torquay)
• Bouguer Gravity Anomaly Map (Provisional Edition), Portland
• Aeromagnetic Anomaly Map, Portland

Six-inch maps

The following is a list of six-inch (1: 10 560) geological National Grid sheets included wholly or in part in the area of 1: 50 000 Sheet 339, with the names of the surveyors and dates of survey. The surveyors were: G. Bisson, J. A. Chesher, R. A. Edwards, R. J. O. Hamblin, M. R. Henson, B. W. Riddolls, E. B. Selwood and R. A. Waters.

Preface

The original geological survey of the Newton Abbot district, by Sir Henry T. De la Beche, relied heavily on the earlier work of R. A. Godwin-Austen, and was published in 1834 on the Old Series one-inch Sheets 22 and 25. A hand-coloured edition of the New Series Sheet 339 was issued in 1899, following revision by H. B. Woodward, W. A. E. Ussher and C. Reid between 1874 and 1887; a colour-printed edition, incorporating further revision by Ussher on the six-inch scale, was published in 1913, as was the first edition of this Memoir.

The revision of the geology of the district presented in this second edition is the outcome of a co-operative venture between the University of Exeter and the Institute of Geological Sciences. The district was surveyed on the six-inch scale between 1966 and 1971 under a Natural Environment Research Council research contract by six research assistants employed by the University, working under the supervision of Professor Scott Simpson and Dr E. B. Selwood (for the University), and of Mr G. Bisson, Mr E. A. Edmonds and Mr J. E. Wright (for the Institute). Dr. J. A. Chesher mapped the Devonian and Carboniferous rocks of the Middle Teign Valley; rocks of similar age were mapped by Dr B. W. Riddolls west and south of Newton Abbot, and by Dr R. A. Waters west of the Bovey Basin and between Chudleigh and the Teign Estuary. Most of the Permo-Triassic rocks of the district were mapped by Dr M. R. Henson, those around the Haldon Hills by Dr R. J. O. Hamblin, and those south of the Teign Estuary by Dr Henson, Dr Selwood and Dr Waters. The Cretaceous and Tertiary beds on the Haldon Hills were mapped by Dr Hamblin, and Dr R. A. Edwards mapped rocks of these ages in the Bovey Basin. Mr Bisson surveyed that part of the Dartmoor Granite lying within the district, and Dr J. R. Hawkes and Mr J. Dangerfield studied the petrology of the granite as part of a wider investigation of the granites of south-west England.

The compilation of this Memoir, using the research assistants' PhD theses as a basis, has largely been the work of Dr Selwood, Dr Edwards and Professor Simpson, aided by the research assistants as necessary. Professor Simpson wrote the introductory chapter of the Memoir and, with Dr Henson, the account of the Permo-Triassic rocks (Chapter 7); Mr R. K. Harrison described metalliferous nodules in the Littleham Mudstone. Dr Selwood was responsible for compiling Chapters 2 to 5, dealing with the stratigraphy and structure of the Devonian and Carboniferous rocks, from the work of Dr Chesher, Dr Riddolls and Dr Waters. The description of the Dartmoor Granite in Chapter 6 was written by Dr Hawkes, with a contribution from Mr Bisson; other parts of the chapter on igneous rocks and metamorphism were compiled by Dr Selwood from the work of Dr Chesher (in particular), Dr Riddolls and Dr Waters. Dr Edwards and Dr Hamblin wrote Chapters 8 and 9 dealing with Cretaceous and Palaeogene rocks, drawing on their theses and later work by themselves and Mr C. J. Wood of the IGS, Palaeontology Department. Mr Wood contributed to the account of the Cretaceous stratigraphy and provided lists of the Cretaceous fossils of the district (Appendices 3 and 4). Dr Edwards wrote Chapter 10 on the Upper Palaeogene. The accounts of the post-Variscan structure and of the Pleistocene and Recent deposits (Chapters 11 and 12) were compiled by Dr Edwards, from the notes of all the surveyors, Chapter 12 incorporating work on the river estuaries by Mr R. S. Nunny and Dr J. M. Thomas of Exeter University. Chapter 13 summarises the results of geophysical surveys, including a geophysical investigation of the Permo-Triassic basin by Mr G. P. Riddler and geophysical work in the Haldon Hills by Mr E. M. Durrance, of Exeter University, and Dr Hamblin. The final chapter, describing the economic geology of the area, was written by Mr R. C. Scrivener and Dr Edwards, with a contribution on water supply by Mr Edmonds.

Fossils were collected by the surveyors, and by Dr D. E. Butler, Dr A. J. Gooday (formerly of Exeter University), Mr A. A. Morter, Dr Selwood, Mr E. P. Smith, Dr D. E. White and Mr Wood. Dr Selwood was generally responsible for the identification of the Devonian faunas; Professor M. R. House, of Hull University, identified the Frasnian goniatites, and Dr C. T. Scrutton, of Newcastle-upon-Tyne University, the Devonian corals; Dr S. C. Matthews, of Bristol University, and Dr Riddolls named the conodonts and Dr Matthews provided Appendix 1; Dr Gooday and Dr Riddolls named the ostracods. The Carboniferous fossils were identified by Dr W. H. C. Ramsbottom, who contributed Appendix 2. Samples from the Permo-Triassic rocks were examined for spores by Dr G. Warrington, but yielded only derived specimens which were named by Dr B. Owens. Mr Wood was responsible for describing the Cretaceous palaeontology.

Grateful acknowledgement is made to numerous individuals and organisations for their help and co-operation; in particular to E.C.C. Ball Clays Ltd, E.C.C. Quarries Ltd, Kingston Minerals Ltd, Colesville Ltd, Watts Blake Bearne and Co. plc, and Mr J. H. Walbeoffe-Wilson, for information generously provided.

This memoir has been edited by Mr Bisson, Mr Edmonds and Mr W. B. Evans.

G. M. Brown Director. Institute of Geological Sciences, Exhibition Road, South Kensington, London SW7 2DE 31 July 1981

Notes

• The word 'district' used in this memoir means the area included in the 1:50 000 Geological Sheet 339 (Newton Abbot).
• In this memoir National Grid references with eastings greater than 7894 lie within the 100-kilometre square SX (or 20), while those with eastings less than 0707 are within the square SY (or 30).
• Dips are given in the form 50°/287°. The first number is the angle of dip in degrees and the second is its full-circle bearing measured clockwise from True North.
• Numbers preceded by A refer to photographs in the Survey's collections.
• Letters preceding specimen numbers refer to Survey collections as follows:
• E English sliced rocks
• MR Museum Reserve collection
• SAL Palynological preparations housed at the Leeds Office of the Survey.
• Locality numbers (e.g. loc. 28) used in the Devonian passages refer to Appendix 1.

Geology of the country around Newton Abbot—summary

The Newton Abbot district is mainly rural country of considerable scenic beauty, encompassing a range of landscapes which reflect the wide variety of the underlying rocks. The Dartmoor granite uplands in the north-west are bordered by a charming tract of dissected Devonian and Carboniferous rocks in the middle valley of the River Teign, which is flanked by plateaus capped by Cretaceous sands and Tertiary gravels — the Haldon Hills — reminiscent more of east than south Devon. The Permian rocks form mainly lowlands with warm red soils, bounded to the east by the scarp of the Triassic Budleigh Salterton Pebble Beds; these New Red Sandstone rocks give rise to fine coastal scenery. A wide valley north of Newton Abbot is underlain by Tertiary clays of the Bovey Basin; south of there and around Chudleigh the relief is governed by the varying resistance to erosion of Devonian and Carboniferous slates, limestones and igneous rocks.

This, the second edition of a memoir originally published in 1913, has been completely rewritten, and incorporates general and detailed accounts of all the geological formations in the district. The oldest rocks, of Devonian and Carboniferous ages, are complexly deformed, but it has proved possible to show in detail how different formations have been juxtaposed by great thrust faults associated with the Variscan earth movements. A wealth of palaeontological data supports the Devonian and Carboniferous accounts. The Dartmoor Granite and other igneous rocks are fully described, as are the Permo-Triassic rocks which formed on alluvial fans and in river basins adjacent to the Variscan mountains. The Upper Greensand chapter, which includes an account of the hitherto little-known greensands of the Bovey Basin, contains much new palaeontological and stratigraphical information. The chapters on the Tertiary contain an account of the flint gravels of the Haldon Hills and their relationship to those around Newton Abbot; and a description of the Bovey Formation strata, which were deposited in rivers and lakes in a deep fault-controlled basin.

Other chapters deal with Quaternary rocks, geophysical investigations, and economic geology.

Geological succession

(Geological succession)

The rock formations of the Newton Abbot district are summarised below. Some of the units listed are equivalent, wholly or in part, so that the formations are not necessarily in stratigraphical order. Generalised thicknesses are given where possible.

Chapter 1 Introduction

Position and relief

The Newton Abbot district (Figure 1) is bounded in the west by Dartmoor and in the east by the coast from just north of Torquay to Budleigh Salterton. Only three-fifths of the sheet area is land, but the limited area contains several contrasted landscape types. In the east a fertile lowland of deep red soils underlain by Permian and Triassic rocks drains to the River Exe (Figure 2). It is bounded to the east by a west-facing escarpment of pebble beds and sandstones dipping eastwards off the siltstones and mudstones of the lowland. From the top of the scarp, an area of heath and coniferous woodland, there are magnificent views westward to the flat-topped skyline of the Haldon Hills and the farther silhouette of Dartmoor.

West of the lowlands the Haldon Hills form a north–south plateau at about 240 m above sea level, which is the interfluve between the River Exe and the River Teign. They constitute an outlier of typical east Devon landscape, expressing the underlying structure of flat-lying resistant gravels of the Eocene on soft Upper Greensand, capping gently eastward-dipping red rocks.

The Dartmoor upland extends into the north-west corner of the district as an expanse of granite country rising to over 300 m above sea level. Between the granite and the Haldon Hills is a tract of charming hilly country about 5 km wide, with characteristically steep gradients but no great elevations. It is underlain by mainly Carboniferous shales and sandstones whose structural trend is roughly east–west. The larger folds are distinct in the Teign Valley, making obvious features in the landscape. In general the oldest (Upper Devonian) strata crop out nearest the granite and there is a regional dip away from it. Similar countryside lies west of the Bovey Basin, around Ilsington.

In the south-west corner of the district, west of Newton Abbot, highly variable topography reflects the contrasting resistances to denudation of the massive limestones and soft slates of the Middle and Upper Devonian. The slates have been reduced to a low plain broken by a numerous isolated hillocks marking igneous intrusions and lavas. The limestone outcrops are bounded by steep convoluted escarpments, an effect of the generally flat-lying attitude of the strata. Additional complications of the topographic pattern are due to interbedded volcanic ashes and infaulted Upper Carboniferous sandstones and conglomerates. A similar landscape lies east of the Bovey Basin, between Chudleigh and the estuary of the Teign. The Devonian rocks of these two areas have been carried northwards and westwards over the Upper Carboniferous along a major thrust fault or thrust complex.

The Bovey Basin extends south-eastwards from the edge of the granite 3 km north-west of Bovey Tracey to Aller, south of Newton Abbot. The greater part of this area is a well-defined plain floored by clays, sands and lignites of Tertiary age. Much of the ground is low-lying moor or heath with silver birch and fairly extensive coniferous woodland. The Bovey Basin has developed along the Sticklepath–Lustleigh Fault Zone, one of the great north-west–south-east dextral wrench faults of Cornubia.

Geological history

The oldest strata that crop out in the district (Figure 2) are of early Middle Devonian (Eifelian) age; in the country to the south these succeed Lower Devonian strata conformably. The Lower and Middle Devonian sediments accumulated in a crustal downwarp that commenced during the early Devonian and persisted into Carboniferous times, and extended from Cornwall across northern Europe. The Eifelian (Figure 6) is most extensively represented by dark grey slates (Nordon Slate), which locally contain a rich shelly fauna, and thin lenticular limestones yielding stromatoporoids and corals. These beds accumulated in deeper water than the predominantly sandy Lower Devonian. In some areas the Eifelian is represented by a carbonate facies consisting mainly of shallow-water crinoidal limestones (Denbury Crinoidal Limestone). The slates, and to a lesser extent the limestones, are associated with submarine and subaerial volcanic rocks and it appears that crinoidal and stromatoporoidal/coral limestones accumulated on volcanic piles within areas of mud deposition.

From Givetian to early Frasnian times the same two facies persisted. The Nordon Slate again includes important horizons of volcanic rocks, though carbonate turbidites also appear, presumably derived from areas of shallow-water carbonate deposition. The carbonate facies is made up of a variety of limestone types (Pulsford Limestone, Chercombe Bridge Limestone and East Ogwell Limestone); generally the limestone sequence becomes paler and more massive upwards. The youngest horizons are particularly fossiliferous and yield a rich benthonic fauna in which stromatoporoids and corals predominate. These sequences probably represent shallow-water carbonate sediments with numerous reefs; spasmodic volcanism gave rise to the Kingsteignton Volcanic Group and the Foxley Tuff.

In the mid Frasnian a major change in sedimentation occurred, possibly the result of diastrophic events elsewhere. This change involved rapid subsidence of the south Devon area, so that deposition of massive limestones ceased and condensed sequences of cephalopod-rich nodular limestones and nodule-bearing slates (Luxton Nodular Limestone) were laid down. The deepening of the sea resulted in the virtual disappearance of the shelly benthos and its replacement by a planktonic fauna, dominated by ostracods, in a thick sequence of purple, green and grey slates (Gurrington Slate, Whiteway Slate, etc.) with submarine spilitic lavas. By the end of the Devonian this ostracod slate facies had invaded the whole district.

Lower Carboniferous rocks follow the Upper Devonian without a break. In the Teign Valley the ostracod faunas disappeared and a black slate sequence (Combe Shale) was deposited, followed by siliceous lithologies, including radiolarian cherts (Teign Chert), accompanied by volcanic rocks. Shallow water persisted at Chudleigh and in it was deposited a condensed sequence of slates and siliceous strata (Winstow Chert), generally free from volcanic rocks.

By the beginning of the Upper Carboniferous, uplift and erosion of Carboniferous, Devonian and possibly earlier rocks to the south were resulting in deposition of submarine fans now represented by the coarse sediments, including conglomerates, of the Ugbrooke Sandstone. Meanwhile in the north, shales and thin turbidite sandstones accumulated which are continuous with the widespread Crackington Formation of central Devon.

The earth movements became more intense and eventually initiated the main Variscan orogeny. Chief of the resulting tectonic structures within the district is a series of great thrust sheets, piled up one upon another, that brought close together areas once widely separated, and so produced striking differences between adjacent local successions.

The newly formed post-Variscan uplands were strongly attacked by desert weathering. Screes mantled the slopes, debris was swept across the depressions and sand dunes built up. Permian and Triassic breccias, aeolian sands and flood-plain sediments buried the planed-off uplands of a continent.

A break in the sedimentary record of the district was followed by a major marine transgression with deposition of neritic Cretaceous sands on Permo-Triassic and older rocks, followed by Eocene gravels. Bovey Formation clays and sands of probable Palaeogene age were deposited in the Bovey Basin. These sediments were formerly considered to be of lacustrine origin, but it now seems likely that they are mostly alluvial, and that sedimentation continued as the basin deepened along the north-west-trending Sticklepath–Lustleigh Fault Zone.

History of research

Important early publications on the district were written by De la Beche (1839) and Godwin-Austen (1842), the latter being a unified account of a series of short papers from 1834 onwards. Palaeontological advances were made by Phillips (1841), Milne-Edwards and Haime (1853) on corals, Davidson (1864–1865) on brachiopods, and Whidborne (1889–1907). Jones (1890) described ostracods, and Hinde and Fox (1895) radiolaria.

The complex and erroneous succession of the Devonian and Carboniferous proposed by Godwin-Austen was improved on by Champernowne (1881), and major advances stemmed from Ussher's (1890) resurvey. The 1890 meeting of the International Geological Congress led to detailed correlation of the Devonshire Upper Devonian with its equivalents in classical Continental successions (House and Butcher, 1973).

Ussher (1913), in the Newton Abbot Memoir, demons trated the presence of major thrusting in the Chudleigh area, subsequently verified by Anniss (1933). But he did not attribute all the thinning of the Upper Devonian and Lower Carboniferous in the area to faulting, as some did, nor did he think, as Godwin-Austen had done, that an important unconformity separated the Devonian from the Carboniferous. He believed (1913, p. 31) that the effects might be due to greatly attenuated sedimentation.

The work of House (1963), and House and Butcher (1973), has demonstrated that a schwelle persisted for a long time during the Upper Devonian and Lower Carboniferous in the Chudleigh region. Their results have been based mainly on the ammonoid succession, but they are now supported by conodont (Matthews, 1969; Tucker and van Straaten, 1970) as well as ostracod zonation.

The nature of the Permo-Triassic rocks was recognised by Conybeare (1823), and they were described by De la Beche (1822) and Godwin-Austen (1842). Pengelly (1862a) believed that the breccias between Torquay and Dawlish were not older than Keuper. Murchison (1867) and Whitaker (1869) thought them the counterpart of the Lower Permian (Rotliegende) of Germany. Woodward (1874) and Ussher (1875) maintained that the whole New Red Sandstone of Devon was Triassic. Irving (1888) agreed with Murchison about the Lower Permian age of the breccias, and regarded the Lower Marls (Exmouth Sandstone-and-Mudstone and Littleham Mudstone) as Upper Permian (Zechstein) and the Pebble Beds as Triassic. In 1892 Ussher recognised the Permian age of the breccias but placed the system boundary at the base of the Lower Marls. By 1913, in the first edition of this Memoir, he had accepted Irving's view and firmly placed the Lower Marls in the Permian.

Pengelly (1863) drew attention to the exotic pebbles of the Budleigh Salterton Pebble Beds and considered the possible derivation of the quartzites from the same source as the Gorran Haven quartzite of Cornwall. Fossils within the quartzites were described by Vicary and Salter (1864) and Davidson (1870, 1880), who dealt exclusively with the brachiopods. In 1880 Davidson compared the brachiopods of the Gres de May of Brittany with those of the pebbles. Shrubsole (1903) supported the view that the quartzite pebbles derived ultimately from the Armorican massif. H. H. Thomas (1902, 1909) considered that the staurolite of the Pebble Beds could have come from Brittany. Worth (1890) studied the petrography of breccia fragments from Dawlish to Exmouth. Shannon (1927) examined the heavy minerals of the different formations.

Vicary (1867) first drew attention to the large, flesh-coloured orthoclase (so-called murchisonite) crystals amongst the breccia fragments; he compared them with the orthoclase phenocrysts of the Dartmoor Granite from which, he suggested, they might have been derived. Later authors, except for Groves (1931), have accepted this idea, and Dangerfield and Hawkes (1969) have argued for the early unroofing of the granite.

De la Beche (1839) and Godwin-Austen (1842) speculated that the Permo-Triassic strata were mainly littoral deposits of a large inland sea. Scrivenor (1948) agreed, emphatically rejecting lacustrine conditions and alluvial fans. Blackwelder (1949) was the first contributor to have extensive first-hand knowledge of the kind of arid region possibly involved. He showed that the breccias were identical to the alluvial fan sheet-flood deposits of the Sierra Nevada. Only with the use of modern sedimentological methods has the formation of these sediments been understood (Laming, 1954, 1966, 1968, 1969; Hamblin, 1969; Henson, 1970, 1971, 1973).

The non-marine sand and sandstone recorded by De Luc (1811) on the Haldon Hills were identified as Greensand by De la Beche (1822). Later (1839) he recognised the Bovey Greensand as a "continuation of the Haldon green sand into an adjoining depression on the west" in an account largely based on work of Godwin-Austen which was not published until 1842. This Greensand was shown on Old Series Geological Sheet 22 (1834) and by Pengelly (1862b). However, when H. B. Woodward resurveyed the Bovey Basin in 1874–75 he was unable to confirm (1876) many of Godwin-Austen's Upper Greensand localities, and their existence has been denied by all subsequent workers until the recent survey (Edwards, 1969). Jukes-Browne and Hill (1900) gave a detailed description of the Haldon Greensand that was to form the basis of the account given in the first edition of this Memoir. Durrance and Hamblin (1969) concluded that variation in thickness of the Greensand on the Haldon Hills resulted from penecontemporaneous folding and suggested that sediments of Cenomanian as well as Albian age were present. This view was confirmed by Hart (1971), who dated Orbitolina occurrences in the Cretaceous of Devon as Lower Cenomanian.

The origin and age of the Haldon Gravels and the gravels developed in the Bovey Basin are contentious. Buckland and De la Beche (1836) considered the high-level gravels of Haldon and east Devon to be of Tertiary age. From Haldon, De la Beche (1839) recognised unabraded flints as derived from in situ solution of Chalk, and equated them with similar gravels in the Bovey Basin. Jukes-Browne and Hill later (1904) listed fossils from the flints on Haldon which suggest the former presence of all zones of the Chalk up to, but excluding, Actinocamax quadratus and Belemnitella mucronata. Mackintosh (1867) believed the Haldon Gravels to be marine equivalents of the gravels of the Blackdown Hills, whilst Belt (1876) thought that all the high-level gravels of Devon and Cornwall had dropped from floating icebergs. Ussher too (1878) first regarded the Haldon Gravels as Pleistocene. Woodward (1876) correlated flint gravels of the Bovey Basin, including some previously mapped as Upper Greensand, with those of the Haldon Hills. Both Reid and Jukes-Brown concurred. Reid (1898) referred the gravels and the Bovey clays to the Bagshot Beds, but Jukes-Browne (1907) envisaged the gravels being carried eastwards by Eocene streams and spread out on the sea floor in Bournemouth Beds times.

A composite origin for the gravels was also put forward by Clayden (1906), who suggested that they were deposited as beach shingle during the retreat of the late Cretaceous sea, and were then resorted by streams. In the Newton Abbot Memoir (Reid in Ussher, 1913) the flint gravels of the Bovey Basin were described as Oligocene marginal lacustrine deposits largely derived from the Haldon Gravels, the latter being considered Eocene and correlated with the Bagshot Beds. The Haldon Gravels were thought to be possibly delta-fan deposits spread out at the foot of Dartmoor; a view later found to be consistent with heavy mineral studies (Boswell, 1923; Groves, 1931) which show the Dartmoor Granite to be an important source. Vachell (1963) queried the correlation of the Haldon Gravels with the flint gravels at Aller, which he thought were derived from the Upper Greensand and gravels of the Haldon Hills. This view has not been generally accepted and the precise equivalence of the Aller Gravel with the various members now recognised within the Haldon Gravels is disputed (Edwards, 1969, 1973; Hamblin, 1973a, 1974).

The economic importance of the ball clays of the Bovey Basin is reflected in an extensive literature dating back to the early 18th century (see Key, 1862; Scott, 1929). the lignite occurring near Bovey Tracey, which formerly had some importance as fuel, figures prominently in early works. More recently the origin and age of the deposits have attracted attention. Most workers have concluded that the clays were derived directly from the Dartmoor Granite; however, Maw (1867) considered that the white Lower Tertiary clays of Hampshire, Devon, Dorset and the Isle of Wight represented the residue after solution of the calcareous part of the Chalk. Hooker (1855) was the first to describe fossil plants from the basin, and these were later monographed by Heer (1862) and referred to the Aquitanian Stage of the Lower Miocene, a view followed in the first edition of this Memoir. Chandler's (1957) re-examination of the flora suggested an Oligocene, possibly Middle Oligocene, age. Others have suggested an Eocene age (Gardner, 1879–1882; Jukes-Browne, 1907, 1911, 1914). C. Reid (in Ussher, 1913) concluded that the Bovey clays were lacustrine and deposited in a rift-valley related to north-west–south-east faulting. Best and Fookes were later (1971) to deduce from the liquidity indices of the clays that they were heavily over-consolidated. They suggested that the clays were possibly deposited in temporary lakes on a flood-plain, and were desiccated soon afterwards. A gravity survey by Fasham (1971) confirmed the presence of a major fault structure along the western side of the basin and indicated a maximum thickness of 1245 m for the contained sediments; this depth was reduced to 1066 m by Vincent (1974).

Chapter 2 Variscan structure

Introduction

The Carboniferous rocks in central Devon have been deformed into a series of east–west trending, generally upright open chevron folds, which are disposed in a number of anticlinorial and synclinorial structures (Freshney and Taylor, 1971). Between Okehampton and the Boscastle–Bude coast these structures pass southwards into a zone showing progressive overturning of folds towards the south, to give a belt of close, recumbent, south-facing folds. These folds are believed to occupy the overturned portion of a major overfold facing south, which was generated by the upright fold belt or suprastructure overriding a more intensely folded infrastructure to the south. The base of the suprastructure, on which movement occurred, is thought to be the Rusey Thrust, which can be traced from the north Cornish coast towards Launceston and is regarded as a plane of decollement.

Deformation north of the Rusey Thrust is characteristic of high structural levels, but to the south a lower level is indicated; low-grade metamorphism appears locally in the Tintagel area, and south-facing isoclinal to close recumbent folds persist as far south as Polzeath, where they confront recumbent north-facing folds.

The belt of southerly overturned folds north of the Rusey Thrust can be traced eastwards through Meldon to the area immediately north of the present district, but the major overfold identified on the coast cannot be recognised. Rather the structure is continued into the anticlinorium represented in the Teign Valley. Selwood and McCourt (1973) demonstrated that this anticlinorium had been thrust northwards on the south-dipping Bridford Thrust over rocks showing Meldon-style folds, and suggested that the thrust had overriden a synclinorial structure to give a minor confrontation of facing directions.

Farther south the Palaeozoic rocks of the Newton Abbot district are traversed by low-angle faults and thrusts, which have brought contrasting Upper Palaeozoic successions (see Chapter 3) into juxtaposition and divide the region into a number of structural units as follows (Figure 3):

These structures were initiated by uplift to the south of this district which generated a land mass from which the Ugbrooke Sandstone was derived in early Namurian times (Figure 4). Deformation, with extensive northward thrusting, migrated northwards in mid or late Namurian times, producing north-facing structures in the southern successions, the Kate Brook Succession and the Liverton Succession. The structures in the Teign Valley Unit evidently resulted from late Carboniferous movement, since the Teign Valley Unit includes a sequence which in central Devon has been shown to pass up without break into Westphalian rocks (Edmonds and others, 1969). The apparent continuity of deformation, producing north-facing recumbent structures which were moved northwards over the high-level structures in the Teign Valley Unit, suggests that the latter could have been generated in the frontal area of the advancing thrust sheets, b.ut this view is at variance with the structural picture determined on the coast. Resolution of the problem is hindered by the difficulty in recognising the equivalent of the Rusey Thrust east of Dartmoor.

In the Teign Valley the separation of high-level and low-level structures lies at the south end of the anticlinorial structure. To the east of the Bovey Basin the line corresponds to the Holne Thrust, and on the west of the basin to the Narracombe Thrust. The Holne Thrust introduced into the area north-facing recumbent folds and is clearly not the equivalent of the Rusey Thrust. South-facing recumbent folding is thus unknown east of the Bovey Basin. If such structures existed, they have been overriden by thrusting which brings north-facing recumbent folds against the essentially upright structures of the Teign Valley Unit. In terms of the coastal structures this involves closing the structural style south of Polzeath [SW 9380 7890] against that seen north of the Rusey Thrust.

West of the Bovey Basin the facing direction of the Ilsington Unit is unknown, so that the northern limit of north-facing folds could lie at the Narracombe Thrust or the Silverbrook Thrust. If this unit faces north then the structural situation is similar to that seen on the east side of the Basin. On the other hand the presence of south-facing recumbent folds could indicate a continuation of the south-facing folds at the southern end of the Teign Valley anticlinorium. Relative to the Teign Valley Unit the Ilsington Unit is not exotic and the break between these units across the Narracombe Thrust may not be great. The abrupt change in structural level across the Narracombe Thrust is emphasised by late vertical movements which have largely cut out the low-angle thrust. The attitude of this thrust is unknown but an equation with the Rusey Thrust would demand that it dipped northwards.

If the folds in the Ilsington Unit face south, then the important confrontation between north-facing and south-facing recumbent folds identified on the coast at Polzeath occurs at the Silverbrook Thrust. The interpretation of the Polzeath confrontation (Roberts and Sanderson, 1971) involving the formation of north-facing folds in late Visean times and south-facing folds in Westphalian times can be applied in principle east of Dartmoor only if the generation of north-facing recumbent structures was maintained at least until early Namurian times and the thrusting described here operated at a late stage in the deformation following the development of the major south-facing structures that are at least post-Westphalian A in age. The late stage of at least part of the northwards thrust movements is evidenced by the Bridford Thrust (Selwood and McCourt, 1973) and in a number of units involving the southern successions, particularly those at higher tectonic levels. These show no broad structure, and considerable shuffling of stratigraphical units probably took place on low-angle thrust planes prior to their arrival in their present position.

Prominent north–south block faulting, possibly associated with broad warping on similar lines, is of late Variscan origin and predates the north-west–south-east faults.

Northern units

Teign Valley Unit

The broad structure of the Teign Valley Unit is an anticlinorium in which axial planes are upright in the middle and fan over progressively towards north and south, being less steeply inclined in the south. The folds increase in tightness as they become more overturned. All plunge at 30° to 50° eastwards off the granite. Fold axes strike between 078° and 106°. The anticlinorium is limited immediately north of the district by the Bridford Thrust. The southern limit is the Narracombe Thrust in the Ilsington [SX 785 762] area, and the Holne Thrust in the Chudleigh [SX 868 796] area.

The Crackington Formation crops out extensively west of the Lustleigh Fault Zone and north of the Narracombe Thrust. Bedding dips generally 45° to 50° north-east, but local steep dips to the south-east suggest asymmetric, but not overturned, minor folds with north-easterly inclined axial planes. Immediately north of the Narracombe Thrust the Teign Chert is brought to the surface in the Narracombe Anticline, which closes in the Brimley [SX 800 771] area. The southern limb is cut by the fault but the northern limb dips at about 40° north-east. North of this anticline the beds are believed to be disposed in a major syncline, and the prevalence of north-east dips suggests that both folds are moderately overturned to the south-west.

Between the Bovey Tracey and Lustleigh wrench faults, from Parke [SX 805 787] to Lustleigh [SX 785 813], the Pullabrook Anticline is picked out by the outcrop of Teign Chert and associated dolerite sills. It plunges south-east away from the granite margin. The north-eastern limb dips at about 40° north-east but the south-western limb is largely hidden by Tertiary deposits. Generally low dips suggest a moderately inclined axial plane with overturning to the south-west.

The major structures of the Teign Valley Unit are best displayed by the Lower Carboniferous rocks of the Teign Valley east of the wrench fault zone (Figure 5). The computed data for each fold are given on page 11 in order from north to south.

The enveloping surface of the major folds is disposed in two regional easterly-plunging anticlinal structures, one centred about Great Leigh [SX 845 855] in the north and the other about Trusham [SX 854 822] in the south; the connecting synclinal fold is centred about George Teign Barton. These regional folds have an approximate wavelength of 3 km. A congruous relationship exists between major and regional folds. The Ashton Anticline occurs at the crest of the Great Leigh Regional Anticline, which plunges at 46°/080°. The northern limb of this anticline dips at 88°/358°, the southern limb at 60°/135° and the axial plane at 70°/140°. The Trusham Regional Anticline plunges at 39°/092°; the northern limb dips at 70°/190°, the southern limb at 70°/138° and the axial plane is vertical and strikes at 090°. The trough of the George Teign Barton Regional Syncline is occupied by the Rydon Syncline; the former fold plunges at 42°/086° and its axial plane dips at 85°/170°.

The folds between the Whetcombe Anticline and the Rydon Syncline are all upright structures, but farther north all folds are overturned in a northerly direction. The steep, inverted southern limb of the Court Barton Syncline is attenuated and displaced to the north-east by the Scanniclift Stretch Thrust, beyond which the limb continues in the Doddiscombsleigh Anticline. The southern limb of the Ashton Anticline is similarly affected by the Lower Ashton Stretch Thrust. These stretch thrusts, which trend northnorth-east–south-south-west and dip at about 70° east-south-east, were developed with the folding and predate the low-angle Bridford Thrust. Shales of the Crackington Formation cut by them are commonly highly cleaved, and some beds of sandstone have become drawn out into boudins [SX 8591 8460].

Cleavage is axial planar. It appears as a steep fracture cleavage particularly in fold cores in argillaceous Upper Devonian and basal Lower Carboniferous rocks, but is weak or absent at higher stratigraphical levels. The thicker sequences of shale within the Crackington Formation locally show pencil cleavage, probably reflecting the intersection of closely spaced fracture cleavage with bedding [SX 8545 8006]; [SX 8560 8467].

Ilsington Unit

The Ilsington Unit comprises a block of recumbently folded strata of the Teign Valley Succession in normal succession plunging eastwards off the Dartmoor Granite between the Narracombe Thrust and the Silverbrook Thrust. Major folds appear to be present but only one minor fold was seen. Facing direction could not be determined. The unit is cut by a late north–south normal fault (the Ilsington Fault) which throws Crackington Formation down to the west.

The Trusham Shale and Combe Shale show slaty cleavage dipping about 20° in various directions. Bedding is usually parallel or sub-parallel to the cleavage, suggesting tight or isoclinal, gently inclined or recumbent minor folds. Generally the bedding in the Teign Chert dips at about 20° north-east; probably most of the exposures lie on the gently dipping limbs of long limb/short limb recumbent or gently inclined open folds. The Crackington Formation west of the Ilsington Fault is particularly argillaceous and exhibits a slaty cleavage dipping at about 30° between east and north-east.

Liverton Unit

The Liverton Unit contains the Liverton Succession and lies west of the Bovey Basin, between the Silverbrook Thrust and the Bickington and Holne thrusts. Recumbent to gently inclined folds are characteristic, but facing directions have not been determined in the Newton Abbot district^‡1 . A number of north-east–south-west, south-dipping low-angle reverse faults repeat the succession in a series of fault slices. The Ilsington Fault continues southwards from the Ilsington Unit and has acted as a late normal fault throwing down mainly Crackington Formation to the west against Upper Devonian and Lower Carboniferous successions to the east.

Except towards the top, the Rora Slate exhibits a penetrative slaty cleavage dipping at about 15° to 20°, probably axial planar to gently inclined or recumbent tight to isoclinal folds. Cleavage generally dips south or southeast, but north of Willis's Cross [SX 799 753] it is inclined east or north-east; this suggests a broad open antiform. In the uppermost part of the Rora Slate, fracture cleavage cuts the bedding of calcareous silts tones and associated lithologies at an angle which suggests that the folding is more open, though axial planes are still gently inclined.

Minor folds in the Mount Ararat Chert and the Crackington Formation are gently inclined to recumbent, with the degree of tightness governed by lithology; in general they are more open than those of the Rora Slate. Monoclinal upright open ruck folds north-west of Mount Ararat [SX 798 748] and open upright folds in the Rora Quarries [SX 8036 7436] probably represent incompetent folding in the cherts rather than a further fold episode.

East of the Ilsington Fault, between the southern margin of the Rora–Langmead [SX 8015 7432] to [SX 7950 7342] outcrop of Rora Slate and the Silverbrook Thrust, Mount Ararat Chert occurs on the high ground horizontally overlying Rora Slate. The junction is conformable, though there may have been local low-angle movement. The area seems to be a recumbently folded block in which facing directions cannot be determined. Axial planar cleavage in the Rora Slate dips mainly south-eastwards but might have been tilted during the emplacement of the granite. Axial planes of minor folds dip both north and south and may also have been tilted.

South-east of the Rora–Langmead Rora Slate outcrop, north-east–south-west faults are common and repeat the Mount Ararat Chert and Crackington Formation in slices dipping east or south-east and presumably recumbently folded. Two lesser faults of the same trend bring in slices of Rora Slate to the north-east of Bickington [SX 797 730].

West of the Ilsington Fault Crackington Formation predominates. Bedding dips mostly south-east and seems to be the right way up. The minor fold style probably consists of long limb/short limb open recumbent folds. Thrust slices of Mount Ararat Chert at Penpark [SX 782 724] and Yeo Farm [SX 793 733] are limited by north-east–south-west faults which at Yeo are south-east-dipping low-angle thrust faults. Mount Ararat Chert also forms structural outliers capping hilltops. Three such outliers occur in the Higher Sigford–Lower Lounston [SX 7816 7441] to [SX 7845 7500] area in a zone of strong north–south late normal faulting probably associated with the Ilsington Fault. Where not affected by faults the junctions with the Crackington Formation are horizontal or very gently inclined; only in the Higher Sigford outlier are the Posidonia beds present below the Mount Ararat Chert, suggesting an inverted sequence, but probably the contact is nowhere conformable. The faults beneath the cherts are likely to resemble those at Yeo Farm, rather than to be part of a widespread thrust sheet.

^‡2 

Kate Brook Unit

The Kate Brook Unit lies between the Holne Thrust and the Bickington Thrust and includes Kate Brook Slate, slices of ?Lower Carboniferous chert along the line of the Holne Thrust, and a large area of recumbently folded Crackington Formation immediately south-west of Chudleigh.

The main structural feature of the Kate Brook Slate is the dominant slaty cleavage that dips at fairly low angles between north-east and south, suggesting recumbent or gently inclined folds. Bedding is rarely visible except where calcareous sandy lenticles lie parallel to the cleavage.

The Crackington Formation is assigned to the Kate Brook Unit because of its structural position and fold style. It was examined in borehole cores and temporary sections associated with the building of the Chudleigh by-pass near Bellamarsh Copse [SX 856 778]. A temporary section at [SX 8568 7787] exposed a close, inclined to recumbent northwest-facing syncline with a south-east-dipping axial plane. Boreholes in Bellamarsh Copse showed dips of 20° to 30° and numerous faults, many parallel or sub-parallel to the bedding.

Southern units

The southern units lie south of the Bickington Thrust. Before deformation they consisted of a thick succession of argillaceous Middle Devonian to basal Lower Carboniferous rocks which acted as an envelope to lenses of Middle Devonian and early Upper Devonian limestones and were capped by a fairly uniform Lower and Upper Carboniferous succession. Now, gently inclined to recumbent north-to north-west-facing folds show varying degrees of tightness. In the thick argillaceous envelope the slates exhibit a penetrative slaty cleavage, usually parallel or sub-parallel to bedding, and the folds are generally tight to close. Major recumbent folds are recognised in the Ugbrooke Unit; elsewhere, numerous low-angle faults and thrusts, most of which reflect northward movement along south-dipping or horizontal planes, have sliced up and repeated the succession, so that major folds are not recognisable. The thick Devonian limestone lenses exhibit few minor folds; they acted as semi-rigid bodies within the argillaceous envelope, and during recumbent folding the hinge zones of major folds ruptured so that the rocks became sliced into discrete lenses.

Ugbrooke Unit

The Ugbrooke Unit is bounded by the Bickington Thrust in the north and by the Ware Barton Thrust in the south, where it is overridden by the Kingsteignton Unit. It is doubtfully extended west of the Bovey Basin between the Bickington and Goodstone thrusts.

East of the Bovey Basin, gentle dips to north-east, east, south-east and south with virtually no westward dipping beds, together with inversions, point to the presence of close to tight, gently inclined, overturned, north- to north-west-facing folds with south-easterly dipping axial planes. Two major fold closures have been mapped and a third inferred. The major folds are the Combe Holdridge Syncline, the Well Anticline and the Ugbrooke Park Syncline. In the eastern and central part of the unit all three folds plunge south-west, but there is evidence in the west that the Combe Holdridge Syncline may plunge north-east beneath the post-Carboniferous cover. Near the Upper Greensand the south-east limb of this fold strikes south-west, and immediately to the north the Ugbrooke Sandstone strikes north–south; this suggests a gentle north-east-plunging closure under the Upper Greensand. If the Well Anticline and Ugbrooke Park Syncline also plunge north-east in the west, then the preservation of the Carboniferous in the Ugbrooke Unit is due to a plunge depression or basin resulting from later folding of the major folds.

No minor folds are exposed; rather it seems that movement parallel or sub-parallel to bedding and involving considerable tectonic thinning has taken place. Low-angle, south-dipping faults or thrusts roughly parallel to the bedding are common on both a minor and a major scale and are probably associated with the recumbent folding as well as with later sliding.

In the generation of the major recumbent folds and associated thrusts the relatively incompetent and thin Upper Devonian and Lower Carboniferous strata acted as a zone of movement between the stronger Ugbrooke Sandstone and the massive Devonian limestones, the movement occurring on numerous slip planes parallel to bedding.

Slaty cleavage dipping gently east or south-east parallel to the bedding is characteristic of the argillaceous parts of the succession. It is commonly weak where the argillaceous sequence is condensed.

Kingsteignton Unit

Palaeozoic rocks about the Teign Estuary, thrust over the inverted limb of the Ugbrooke Park Syncline to the north and unconformably overlain by the New Red Sandstone to the south and east, are dominantly argillaceous with a penetrative slaty cleavage dipping at up to 30° between south-east and south-west. Cleavage is usually parallel to bedding. Minor fold axes generally trend east–west or north-east–south-west, or locally north–south. The slaty cleavage is axial planar to these minor folds, which are recumbent to gently inclined and of varying tightness.

No major folds can be demonstrated. However, the Luxton Nodular Limestone on the north side of the estuary seems to overlie Whiteway Slate in inverted succession and may well be overlain by the Whiteway Slate to the south of the estuary; thus a major recumbent structure orientated about 100° may exist. Since no Luxton Nodular Limestone occurs at the western end of the estuary, and the spilite in the Whiteway Slate strikes north-west–south-east across the estuary and dips south-west, such a structure would plunge gently westward.

Highweek Unit

In the Highweek Unit, Gurrington Slate is thrust northwards on the Goodstone Thrust, over Chercombe Bridge Limestone that may be a remnant of the Ugbrooke Unit. No major folds, and only a few minor ones, have been detected.

Silty horizons in the slates near Bickington are disposed in small tight folds overturned towards the north-north-west and plunging at 45° east; cleavage dips at 20°/166° [SX 7960 7195]. Bedding is rarely discernible but may be responsible for a lineation on the slaty cleavage plunging at 10° to 38° between east and south-east.

Cleavage dipping gently between east and south-east suggests tight recumbent folding, a supposition in accord with the minor folds recorded. However, repetitions of the slate sequence are complex, and facies changes and thrusts are probably also present. The slate belt forms an arc, with cleavage strike swinging from 045° at Woodland [SX 791 688] to 110° near Newton Abbot. It may have been affected by drag movements along Tertiary wrench faults associated with the Bovey Basin, and possibly also by late Variscan flexuring on north–south axes.

Kink bands occur on cleavage surfaces in a few places and probably represent the last phase of deformation; they vary in trend from east to south.

Denbury Unit

The Beacon Hill Succession and the massive limestones of Newton Abbot lying west of the Ipplepen Fault have been carried over the Gurrington Slate of the Highweek Unit on the Forder Green Thrust between [SX 7895 6725], west of Forder Green, and Highweek [SX 8442 7136]. In addition, small areas of Nordon Slate at Dainton and Nordon Slate with limestone 2 km farther east are tentatively referred to the Denbury Unit.

Relations between the successions within the unit have not been fully resolved, but it seems that some movement of the base of the massive limestones has taken place over the Nordon Slate. Minor thrusting is indicated east of Denbury [SX 8235 6890], where East Ogwell Limestone of the Torbryan Succession has overridden Foxley Tuffof the Ugbrooke–East Ogwell Succession. Outliers of East Ogwell Limestone lying on the tuff north of the thrust are believed to represent klippen, and abundant tuff debris developed west of Rydon Hill [SX 8320 6890] suggests a window of Foxley Tuff in the surrounding limestone.

The limestones are not visibly folded, except for a broad anticlinal warping associated with the east–west Torbryan Fault and an open anticlinal fold in Chercombe Bridge Limestone in a fault-bounded block south-west of Halwell [SX 8235 6794]. The latter fold plunges at 22°/075°, has a northern limb dipping at 20°/043° and a southern limb dipping at 25°/141°. Small-scale warping, apparently attributable to differential movement along bedding planes and sub-parallel fractures, can be seen in the Denbury Crinoidal Limestone [SX 8203 6912], in the Pulsford Limestone [SX 8187 6684] and in the East Ogwell Limestone [SX 8132 6755].

The folding of the thick Nordon Slate succession is similar to that of the Gurrington Slate to the north; gentle south-east-dipping cleavage prevails which commonly approximates to the bedding. Interbedded limestones locally show tight recumbent folds overturned to the north [SX 8051 6649]. Close recumbent folds with a wavelength of 1.8 m plunging at 22°/102° are exposed in a disused quarry [SX 8753 6698] east of Dainton, and in Dainton Quarry [SX 8510 6600] east–west-trending isoclines show limbs dipping at 30° south and south-dipping axial planes.

In the south, late normal faults of north–south and north-west–south-east trend break up the limestone complexes of the Denbury Unit into discrete blocks.

Abbotskerswell Unit

The Abbotskerswell Unit consists of all of the Whiteway Slate and Luxton Nodular Limestone south of Newton Abbot, between the Ipplepen Fault in the west and the New Red Sandstone in the east. It has been overthrust by a variety of formations that constitute the East Ogwell Unit. Surface outcrops yield little structural information, the beds being cut by a flat-lying slaty cleavage that commonly approximates to bedding. Minor folds are rarely observed in the field, but polished slate specimens reveal small-scale tight recumbent folding. A tight over-fold at East Ogwell, plunge 15° to south-south-west with related minor folds, was temporarily exposed at [SX 8397 7010]; it may have originated during the overthrusting of the adjacent limestones of the East Ogwell Unit.

The Rydon Ball Farm Borehole (p. 183) passed through Upper Carboniferous rocks of the East Ogwell Unit into 30.96 m of Whiteway Slate of the Abbotskerswell Unit. The slates overlie a low-angle fault, below which occur seven slices of strata bounded by further low-angle faults. Luxton Nodular Limestone in the succession exhibits folds which are recumbent to gently inclined and close to tight. Riddolls (1970a, b) showed how movement on the axial plane slaty cleavage in the cores of such folds had broken up thin limestone bands into lenses. The seven slices may represent alternate upper and lower limits of recumbent folds, for younging in adjacent slices is commonly in opposite directions. However the occurrence near the base of the slices of 4.52 m of Upper Carboniferous strata shows that movement on some of the low-angle faults was considerable. The East Ogwell Limestone lies below another low-angle fault.

East Ogwell Unit

Higher ground within the area of the Abbotskerswell Unit is formed by faulted klippen composed of East Ogwell Limestone with local developments of Chercombe Bridge Limestone, tuffs and Ugbrooke Sandstone, which together constitute the East Ogwell Unit.

Where the limestone is massive, as at Stoneycombe Quarry [SX 8620 6685], numerous sub-horizontal fractures are present; these are probably related to low-angle thrust faults which pervade the unit. Even within normal successions, as in the klippen [SX 8520 6780]; [SX 8520 6810] east of Two Mile Oak Cross, where East Ogwell Limestone overlies tuff, movement has commonly taken place along stratigraphical contacts. Dolomitisation and recrystallisation of the limestone are characteristically associated with the thrust faulting.

No broad structure has been detected and it seems that the original succession was sliced up by low-angle faults prior to arriving in its present positions. Warping and gentle flexures are common in the massive East Ogwell Limestone, but tight folding is seldom observed; however a 0.6-m limestone at Grange Quarry [SX 8520 6860], west of Abbotskerswell, shows a tight recumbent fold with a north–south axis.

Faults

Thrust faults

Narracombe Thrust

The Narracombe Thrust trends east-south-east from Smallacombe [SX 776 772] to Woodhouse Cross [SX 795 765]. It throws Trusham Shale in the core of the Narracombe Anticline against Teign Chert east of the Ilsington Fault and against Crackington Formation west of it.

The fault separates two different tectonic styles within the Teign Valley Succession, a contrast which may reflect different original tectonic levels. Waters (1970) suggested that the Ilsington Unit may have been thrust north over the Teign Valley Unit. This implies a north-facing sequence in the Ilsington Unit, where facing direction is unknown. An alternative is that the Ilsington Unit forms part of the south-facing end of the anticlinorium developed in the Teign Valley Unit, but now separated from it by the Narracombe Thrust. Truncation of the north–south Ilsington Fault by the Narracombe Thrust attests later movements on the latter, possibly as a branch of the Lustleigh wrench fault system.

Silverbrook Thrust

The low-angle Silverbrook Thrust runs east-south-east from south of Honeywell [SX 775 753] to the Ilsington Fault. Farther east the Liverton and Ilsington units are separated by a normal fault, but the thrust fault can be traced from Silverbrook Mine [SX 789 758] to north-east of Lenda [SX 793 757], where the junction is a normal fault. The probable southward inclination of the thrust fault and of the north-east–south-west faults which have shuffled the Liver-ton Unit is consistent with northward movement.

Holne Thrust

The Holne Thrust was described by Reid and others (1912) as having carried Upper Devonian slates (Kate Brook Slate) northwards over Carboniferous beds. It is well exposed in the River Dart at Holne Bridge [SX 731 706]. Farther east it is overridden from the south by the Bickington Thrust, but just south of Blackpool [SX 813 741] it is brought to the surface by a north-east–south-west fault. East of the Bovey Basin, in the Chudleigh area, it separates the tight recumbent folds of the Crackington Formation and Kate Brook Slate of the Kate Brook Unit from the open upright folds of the Teign Valley Unit. East of Chudleigh Rocks it runs along the Kate Brook to Crammers Bridge [SX 881 809] and thence northwards to the Whiteway area, where it passes beneath the Bickington Thrust. Fault-bounded slices of chert are caught up in the thrust zone at Heightley [SX 862 786], Old Kate Bridge [SX 875 797] and north-west of Hams Barton [SX 877 805].

The Kate Brook Unit has been brought against the Liverton Unit west of the Bovey Basin and against the Teign Valley Unit near Chudleigh. Possibly in the latter area it has overridden the Ilsington and Liverton units. The thrust may well have been displaced by normal faults trending north-east–south-west and north–south.

Bickington Thrust

The Bickington Thrust is the most striking low-angle fracture in the district. North-west of Ashburton [SX 758 699] it has carried the southern units over the Kate Brook Unit, but within the present district it has overriden the Liverton Unit (Plate 9). South of Blackpool, where the Holne Thrust emerges from beneath the Bickington Thrust, the latter once more separates the southern units from the Kate Brook Unit. Chercombe Bridge Limestone overlies Kate Brook Slate hereabouts [SX 8094 7348] to [SX 8133 7366] but farther east spilites of the Gurrington Slate are brought over Kate Brook Slate. Pods of Chercombe Bridge Limestone are caught up in the thrust zone near Staplehill [SX 825 737].

On the eastern side of the Bovey Basin the junction between the Ugbrooke Unit and the Crackington Formation of the Kate Brook Unit seems to be a fault rather than a thrust. However, east of a north–south normal fault running from Heightley to Winstow Cottages [SX 8633 7804], Chercombe Bridge Limestone has been thrust over Kate Brook Slate; in The Glen [SX 866 786] the Kate Brook has cut through the limestone into the slates. The thrust trends north-east along the base of the Chudleigh escarpment and a klippe of Chercombe Bridge Limestone occurs west of Waddon [SX 885 797]. At Waddon the thrust swings north towards Harcombe [SX 8872 8161]. South of Whiteway House [SX 8768 8295] and on the north side of Chudleigh Rocks the southern units are in contact with the Teign Valley Unit.

Ware Barton Thrust

The Ware Barton Thrust separates the Ugbrooke and Kingsteignton units. It can be traced from St Oswald's [SX 878 734] to north-west of Bishopsteignton [SX 904 745] and is cut by numerous north-west–south-east normal faults. Near St Oswald's, Whiteway Slate of the Kingsteignton Unit is thrust over Chercombe Bridge Limestone of the Ugbrooke Unit; the thrust was exposed during construction of the Kingsteignton [SX 870 735] by-pass. From near Ware Barton [SX 881 727] north-eastwards to Murley Grange [SX 903 738] Nordon Slate of the Ugbrooke Unit has been overridden by Whiteway Slate of the Kingsteignton Unit. The irregular patch of Nordon Slate immediately south-west of Ware Barton marks a window through the overlying Whiteway Slate. North-east of Murley Grange the Whiteway Slate is either overthrust or faulted against the Kingsteignton Volcanic Group.

Goodstone Thrust

The Goodstone Thrust trends north-east from Mead [SX 779 709] towards Bickington and carries tuffs at the base of the Gurrington Slate over the Chercombe Bridge Limestone of the Ugbrooke Unit. Just west of Telegraph Hill [SX 804 730] it completely overrides the limestone and merges with the Bickington Thrust.

Forder Green Thrust

The Forder Green Thrust has carried the Denbury Unit over the Highweek Unit, but possibly not far. It is not exposed, but the nearby slates are much disturbed. The thrust traces an arc from the south-west corner of the district to near Highweek, roughly in accord with the swing of the cleavage in the Highweek Unit.

Stretch thrusts

Two stretch thrusts (Hills, 1972), high-angle reverse faults originating from the stretching and eventual rupture of the overturned steep limbs of folds, are present in the Teign Valley Unit. They trend north-north-east–south-south-west and dip steeply east-south-east at approximately 70°. The Lower Ashton Stretch Thrust extends from Pigeon Copse [SX 8483 8428] south-south-west across the River Teign to the east of Exmouth Cottages [SX 8400 8300], producing contortion of Trusham Shale green shales [SX 8417 8317] to the north-east of Exmouth Cottages. The Scanniclift Stretch Thrust extends from the northern margin of Scanniclift Copse [SX 8445 8645] south-south-west to Bennah [SX 8359 8460], causing the abrupt termination of the chert ridges at both extremities.

Normal faults

Normal north–south faults, and a subsidiary set trending east–west, break up the Palaeozoic rocks, particularly in the southern units, into a series of fault-bounded blocks. They post-date the folding and thrusting of the main deformation and are probably of late Variscan age. Mineralised north–south and east–west faults in the Teign. Valley Unit also cut the granite. The displacement of the Permo-Triassic rocks by faults trending around east–west suggests some reactivation of the fractures.

Teign Valley Unit

Two faults trending approximately north–south near Doddiscombsleigh [SX 8574 8662] to [SX 8552 8561] and Christow [SX 8310 8618] to [SX 8280 8525] are probably normal and cause only minor displacements.

The Teign Valley mineral lodes have been emplaced along a south-trending fracture zone 200 m wide along the edge of the granite's metamorphic aureole. No displacements have been mapped, but the sediments are weakened and broken and quarries show much argillisation and brecciation. This zone of fracture has been reduced by preferential erosion to a low-level gently sloping surface that runs along the foot of the scarp formed by the aureole rocks.

The eastern margin of the granite is characterised by deep V-shaped valleys running east or east-north-east. The northernmost valley [SX 8190 8605] is coincident with an east-north-east-trending fault which displaces both granite and aureole rocks. Farther south the valleys at Bowden [SX 8238 8450], Combe [SX 8365 8400], Canonteign [SX 8332 8242] and Hyner [SX 8330 8170] also probably owe their origin to erosion along minor fractures. The east–west to east-northeast–west-south-west micaceous hematite lodes in the granite near Hennock [SX 830 808] are displaced by minor north–south faults. At Lower Bowden [SX 8200 8030], south-west of Hennock, east-north-east–west-south-west faults displace the granite and adjacent strata.

Ilsington and Liverton units

The Ilsington Fault appears to be normal, throwing down to the west. It runs south from immediately north of Ilsington, where it is truncated by the Narracombe Thrust, to Combe [SX 789 740], cutting the Silverbrook Thrust en route. South of Combe it swings south-east and cuts the Bickington Thrust. In the Liverton Unit north–south and north-west–south-east normal faults occuron both sides of the Ilsington Fault and are associated with it. One of them forms the western limit to the structural slices of Mount Ararat Chert south of Lower Lounston and west of Conthe and may run south to a fault following the Lemon Valley and offsetting the Bickington Thrust. The Ilsington Fault and its associated faults are late Variscan or younger.

Southern units

The Ipplepen Fault is the most important of a series of north–south normal faults which break up the southern units into a series of north–south elongated blocks. It may be traced northwards from [SX 8416 6748] along the eastern margin of Stallage Common through East Ogwell to Highweek, and forms the western limit of the Abbotskerswell and East Ogwell units. Middle Devonian limestones are represented on either side of the fault. On the eastern side the limestones form klippen resting on a predominantly Upper Devonian succession which at the base of the Rydon Ball Farm Borehole (p. 183) rests tectonically on massive limestone; to the west the limestones appear to be in conformable contact with underlying Nordon Slate. It seems likely that the Ipplepen Fault throws down to the east and that the massive limestones at the base of the borehole are the same as those in the Denbury Unit.

Where the east–west faults approximate to the regional dip, as do the Torbryan and Denbury faults of the Denbury Unit, they form prominent features, but in general they are subsidiary to the north–south dislocations.

Wrench faults

As in much of south-west England, north-west–south-east dextral wrench faults cut most other structures. The major Sticklepath–Lustleigh Fault Zone shows important Tertiary movements, but Dearman (1963) has suggested that these could represent a reactivation of Palaeozoic fractures. Minor faults away from the Bovey Basin which show similar trends are almost certainly genetically related, but Tertiary movements cannot be demonstrated. Two such faults cut the Ipplepen Fault [SX 8385 6955]; [SX 8390 7050]. The more important may be traced from Abbotskerswell to West Ogwell Cross [SX 8590 6865] to [SX 8300 6995]. Similar faults within the East Ogwell Limestone north-east of Dainton [SX 8515 6670] commonly carry much coarse calcite.

Details

Teign Valley Unit

Teign Chert

Roadside exposures between Hill Copse [SX 841 844] and Exmouth Cottages [SX 843 841] show asymmetrical anticlines and synclines overturned to the north and plunging at 30° to 60° east. The axial planes dip at 70° to 80° south-east and the steep limbs of these folds are commonly sheared by faults.

A cutting for an electricity pylon [SX 8505 8145] near Crocombe Bridge showed a minor anticline overturned to the south-east on the northern limb of the Whetcombe Anticline. The fold plunges at 34°/160°, the northern limb dips at 60°/048° and the southern limb at 84°/063°. A small-scale anticline is present on the north-east limb of this anticline. In the south-west part of the excavation the anticline passes into a poorly exposed syncline.

Combe Shale

An old railway cutting [SX 8429 8250] 1.5 km west of Trusham exposes minor east-plunging open flexures on beds dipping steeply south. Kink bands occur in the south-west corner of the second level of Crockham Quarry [SX 8476 8068], 2.4 m below a 1.5–m dolerite sill.

Crackington Formation

A stream section [SX 8560 8472] to [SX 8580 8487] at Higher Ashton exposes a series of east-plunging minor folds. In France Brook [SX 8585 8455] to [SX 8600 8470] tight asymmetrical anticlines and synclines overturned towards the north plunge east and show axial planes dipping at 80° south. A roadside exposure [SX 8581 8448] shows an anticline plunging at 30° east. Alongside a forestry track in Kiddens Plantation [SX 8751 8433] minor east-plunging open folds are present on south-dipping beds. The banks of a forestry track [SX 8689 8233] in Whiteway Wood show folds with long steep north-dipping limbs and short gently inclined south-dipping limbs, plunge at 50° east.

Ilsington Unit

Teign Chert

Lenda Wood Quarry [SX 7920 7584] shows an open recumbent syncline closing southwards. Axial-planar fracture cleavage is evident in the hinge zone. The fold axis trends 140°.

Liverton Unit

Mount Ararat Chert

A track section [SX 7933 7477] to [SX 7972 7482] exposes open to close, gently inclined to recumbent folds with axes trending east or north-east. The fold limbs are of unequal length and axial planes dip gently south-east or south. At the south-west end of the section a north-dipping long limb passes into open monoclinal rucks trending 025° and is faulted against Ararat Slate. Mount Ararat Quarry [SX 7976 7474] contains a tight recumbent fold with axis trending 103° and limbs of unequal length.

Ramshorn Down Quarry [SX 7902 7402] exposes open to tight gently inclined to recumbent folds with axes orientated 102° and limbs of unequal length. One recumbent synform exhibits a tight angular core but becomes open along the axial plane. Axial planes dip gently north and south and are paralleled by a weak cleavage in the shales of the fold cores. Two faults dip at 50° northward.

At Rora Quarries [SX 8036 7436] the roadside quarry exhibits a tight to close slightly overturned syncline/anticline pair; fold axes trend 065° and axial planes dip north. In the larger quarry to the north an open upright syncline/anticline pair shows axes plunging at 24°/060°.

Near Penpark, crags [SX 7801 7240] show an open recumbent fold with south-dipping axial plane and axis plunging gently to 259°.

Crackington Formation

At Wilsworthy [SX 8013 7647] a close to tight recumbent fold shows axis trending 165° and axial plane dipping at about 10° east. Alongside a track [SX 7855 7362] north of the River Lemon an asymmetric close anticline with a south-east-dipping axial plane exhibits a sub-vertical steep limb; the fold axis plunges at 40°/085°.

Ugbrooke Unit

Using the Lower Carboniferous as a marker, the south-east limb of the Combe Holdridge Syncline may be traced north-east from Lindridge Hill [SX 874 745] through Whiteway Barton and Larcombe Bridge to the closure of the fold at the northern end of Hamblecombe Lane [SX 886 774]. Inversion has been proved by trenching near Whiteway Barton (p. 57). This limb of the fold is cut by low-angle fractures and later strike faults. A north-east–south-west strike fault running from Ideford [SX 894 775] to Whiteway Barton throws down Permian to the south-east against Winstow Chert; near Whiteway Barton it throws Posidonia beds against Whiteway Slate, and it may continue south-westwards to throw Posidonia beds and Winstow Chert against Luxton Nodular Limestone.

The axis of the Combe Holdridge Syncline runs north-east–south-west, passing between the two conglomerate outcrops in the Ugbrook Sandstone between Holdridge [SX 8740 7578] and Whiteway Barton. The north-west limb of the fold shows a normal succession and passes into the Well Anticline (below); it is locally thrust northwards over the inverted limb of the latter. At Sedgewell House [SY 13750 7658] Ugbrooke Sandstone is thrust over Winstow Chert, and in the Well area Winstow Chert is thrust over Luxton Nodular Limestone. It was probably thrusting of this type that placed the tectonic outlier of Lower Carboniferous and Ugbrooke Sandstone on the Ugbrooke Sandstone of the Ugbrooke Park Syncline south of Lower Dunscombe [SX 8857 7893].

The Well Anticline is in part thrust northwards over the Ugbrooke Park Syncline, and has itself been cut by low-angle fractures. The core of this major fold consists of the Luxton Nodular Limestone and the East Ogwell Limestone exposed in the Well area. The north-west limb is inverted.

The axis of the Ugbrooke Park Syncline runs south-westwards from Higher Dunscombe [SX 8927 7908]. It passes beneath an overthrust patch of Winstow Chert and Ugbrooke Sandstone and beneath Permian rocks near Ugbrooke House [SX 8757 7806] and Upper Greensand near Babcombe Copse [SX 868 765]. The northwest limb is normal.

On the Chudleigh escarpment north-eastwards from [SX 871 787] the Luxton Nodular Limestone, the Whiteway Slate and probably part of the Winstow Chert are cut out by a low-angle fault or thrust. The isolated upfaulted block of Luxton Nodular Limestone on the dip slope of the escarpment at Lower Dunscombe [SX 885 792] represents a large pod of the succession caught up on the thrust plane.

Kingsteignton Unit

Minor folds are exposed in the low cliffs on the south side of the Teign Estuary east of Netherton Point. Siliceous nodules and seams in the Whiteway Slate [SX 894 723] are disposed in close to open recumbent or moderately inclined folds. Thin calcareous bands and laminae in the slates [SX 899 724] are microcrenulated, the crenulations forming minute open, and locally tight, gently inclined to recumbent folds. The bands are commonly offset by the axial plane cleavage and some are thick in the hinge zones but peter out along the limbs. At the same locality a slightly larger open fold in slates with microcrenulated laminae and bands trends 064°; its axial plane dips gently southward.

Chapter 3 Distribution of Devonian and Carboniferous rocks

Introduction

A comprehensive but generalised view of the Devonian and Carboniferous stratigraphy of the district has been available since the publication of the first edition of this Memoir in 1913. Subsequent work has revealed the presence of a number of synchronous but distinctive successions (Figure 6), the boundaries between which are commonly tectonic (Figure 3) and (Figure 7). They may be divided into two groups–the northern successions and the southern successions–separated by the Bickington Thrust.

Of the northern successions, only the Teign Valley Succession, in the middle Teign Valley, is thought to be autochthonous; the rest occur in thrust sheets that represent progressively higher positions in the structural pile when traced towards the south. South-east of the Bickington Thrust, the southern successions lie in an area of facies change, complex tectonics, poor exposure and incomplete palaeontological control. In some places the successions merge through lateral facies changes and in others they are separated by tectonic breaks. Facies changes are most marked in the Middle Devonian, but by early Carboniferous times a fairly uniform pattern of sedimentation had been established. The Middle and earliest Upper Devonian rocks comprise limestones set in a slate succession and range from Eifelian to early Frasnian. The Eifelian to Frasnian Beacon Hill Succession is mainly Nordon Slate, though with local limestone horizons, and probably passes up into the Frasnian to ?early Tournaisian Gurrington Slate of the Bickington Succession, to form a dominantly argillaceous sequence, but the junction between the two is faulted. In contrast, the Ugbrooke–East Ogwell and Torbryan successions show Nordon Slate at the base, overlain by limestones. Thus, south-east of the Forder Green Thrust the Nordon Slate youngs westwards into the Beacon Hill Succession and eastwards into the carbonate successions. Upper Devonian and basal Lower Carboniferous rocks of the Ugbrooke–East Ogwell Succession comprise slates, nodular limestones and slates with nodules, while the corresponding but much thicker Bickington Succession consists of slates with volcanic rocks.

Northern successions

Teign Valley Succession

The Teign Valley Succession, characteristically developed between Christow and Bovey Tracey [SX 816 786], is a conformable sequence from late Famennian to Namurian. Palaeontological control is good in the late Famennian, early Tournaisian and late Visean, but the greater part of the Lower Carboniferous has proved unfossiliferous. Dolerite sills and tuff bands provide useful marker horizons. The succession is as follows:

Similar beds occur between the Lustleigh Fault and the Bovey Tracey Fault but they are absent south of the Silverbrook Thrust.

Liverton Succession

The Liverton Succession ranges from late Famennian to Namurian:

The north–south Ilsington Fault separates areas at two structural levels. On the eastern side the complete normal succession crops out, the high ground being formed by Mount Ararat Chert and Crackington Formation. To the west the ground consists dominantly of Crackington Formation with rare outliers of Mount Ararat Chert capping the hills; these represent the remnants of a thrust slice, as do structural inliers in the low ground north of Bickington, where erosion has revealed the Posidonia beds beneath the Crackington Formation. Intrusive dolerites are few and confined to the Mount Ararat Chert on the eastern side of the Ilsington Fault.

Kate Brook Succession

The Kate Brook Slate is exposed in the Kate Brook valley near Chudleigh in the Kate Brook Unit. West of the Bovey Basin it is cut out by the overriding Bickington Thrust. The slates are at least in part of Upper Devonian. age. Lower Carboniferous cherts in thin faulted slices along the line of the Holne Thrust, and Crackington Formation in a faulted wedge south-south-west of Chudleigh, complete the Kate Brook Succession.

Southern successions

Bickington–Beacon Hill Succession

The Beacon Hill Succession appears only in the Denbury Unit. It consists of thick Nordon Slate with local developments of limestone and volcanic rocks, and ranges in age from Eifelian to early Frasnian. Nordon Slate in the north-eastern part of the Denbury Unit forms the base of the Ugbrooke–East Ogwell Succession. Because the Nordon Slate in these successions cannot be separated in the field, all of the Nordon Slate outcrop north of the Pulsford Fault is described under the Ugbrooke–East Ogwell Succession. This area is, however, known to include an undelimited area, near West Ogwell, belonging to the Beacon Hill Succession (see p. 44).

South of the Pulsford Fault the Beacon Hill and Torbryan successions share the Nordon Slate as a common formation, but the greater part of it is arbitrarily referred to the Beacon Hill Succession. Only the Nordon Slate running east–west through Torbryan [SX 820 669], conformably beneath the Denbury Crinoidal Limestone, is referred to the Torbryan Succession. The marked contrast between these two successions suggests the possible presence of low-angle thrusting. The Nordon Slate south-west of Kingskerswell belongs to the Beacon Hill Succession, but its structural position is obscure.

The Gurrington Slate which constitutes the whole of the Bickington Succession covers an extensive area between Highweek and Woodland and ranges in age from mid Frasnian to ?early Tournaisian. It contains local develop ments of spilite lavas, some showing pillow structures (Plate 2). It lies between Middle Devonian limestones to the north-west and Middle Devonian slates and limestones to the south-east, a disposition which led Ussher (1913) and Middleton (1954, 1960) to suggest the presence of a faulted syncline, but this has not been confirmed by recent detailed palaeontological observations. The Gurrington Slate rests on Chercombe Bridge Limestone between Staplehill and Bickington; this contact is interpreted as a thrust, the limestone belonging to an eroded sheet (?Ugbrooke Unit) of the Ugbrooke–East Ogwell Succession.

Ugbrooke–East Ogwell Succession

The Ugbrooke–East Ogwell Succession occurs in all the Southern Units except the Highweek Unit. A sequence from Eifelian to ?Namurian in the Ugbrooke Unit shows little of the tectonic complication seen in the other units. It includes the Kingsteignton Volcanic Group near the Eifelian–Givetian boundary. The Middle Devonian limestones show complex facies changes and are succeeded by Upper Devonian nodular limestones and slates that accord well with the concept developed by Schmidt (1926) of becken and schwelle (basin and rise) sedimentation.

In the Denbury Unit a south-westward passage into the Beacon Hill Succession is observed. The Middle Devonian part of the succession exposed in the Lemon Valley, west of Newton Abbot, shows notable changes from that of the Ugbrooke Unit, including thick volcanic rocks (Foxley Tuff) between the Chercombe Bridge Limestone and the East Ogwell Limestone. Near Chercombe Bridge this succession closely resembles the Torbryan Succession. The latter is contained in a tectonic slice overthrust on the Ugbrooke–East Ogwell Succession in the southern part of the Denbury Unit.

The Kingsteignton, Abbotskerswell and East Ogwell units contain parts of the Ugbrooke–East Ogwell Succession in tectonically complicated areas. The East Ogwell Unit shows klippen of various ages resting mainly on Upper Devonian of the Abbotskerswell Unit.

The upper part of the Ugbrooke–East Ogwell Succession is as follows:

These beds overlie the following succession in the Ugbrooke Unit between Chudleigh and the Teign Estuary:

Torbryan Succession

In the southern part of the Denbury Unit strata, ranging from late Givetian or Eifelian to Frasnian, crop out in a north-dipping sequence around Torbryan. West of the village the strike swings north-westward and the Denbury

Crinoidal Limestone peters out. A northerly passage into the Ugbrooke–East Ogwell Succession of the Lemon Valley is postulated; to the north-west the Torbryan Succession is terminated by the Pulsford Fault, and to the north-east it is locally thrust over Foxley Tuff. Two isolated areas of East Ogwell Limestone resting on the tuff south of East Ogwell are considered to represent klippen associated with this thrusting.

The succession is:

Chapter 4 Devonian

General account

Ussher (1890, 1900) explored the possibility that there were major facies changes in the Devonian rocks of the Newton Abbot district, and recent studies of the ammonoids (Plate 4), ostracods (Plate 5), trilobites and conodonts (Plate 6) and (Plate 7) have led to precise correlations (Figure 8). Other groups, particularly the Middle Devonian coelenterate-brachiopod faunas, have not been fully investigated. (Table 1) shows the zonal scheme, after House (1975), together with the ostracod chronology after Rabien (1970). The ranges of stratigraphically important ostracod species are given in (Table 2). (Table 3) shows the ages of the conodont faunas listed in full in Appendix 1.

Middle Devonian

Known only in the southern successions, the Middle Devonian is represented by a group of limestones and an argillaceous equivalent, the Nordon Slate. The oldest faunas recorded, of early Eifelian age, appear in the Kingsteignton Volcanic Group of the Ugbrooke–East Ogwell Succession. In this succession and in the Torbryan Succession the Middle Devonian–Upper Devonian boundary lies near the top of the limestone group and in the Beacon Hill Succession near the top of the Nordon Slate, but as these lithostratigraphical units are essentially Middle Devonian they are described under that heading. Volcanic rocks, less extensive than in other parts of Devon, are restricted to specific horizons in particular successions.

Nordon Slate

The Nordon Slate of the Beacon Hill Succession is characteristically developed in the south-west part of the Denbury Unit, where it is represented by an unknown thickness of greenish-weathering, black or bluish grey slates. Dark grey and black limestones locally interbedded with slates occur as discrete bundles that map out as lenses. Exposure is poor and cleavage generally obscures the bedding, though thin silty laminations and sporadic calcareous siltstones up to 50 mm thick define bedding in places. Spilites, tuffs and small intrusions of dolerite occur locally. Rocks belonging to the spilitic suite were described by Middleton (1960) and include porphyritic spilitic andesite and keratophyre. The type locality for the Nordon Slate is 823 m at 297° from Denbury Village (p. 44).

Lithologically the Nordon Slate of the Torbryan Succession and Of the Ugbrooke–East Ogwell Succession in the Chercombe Bridge area is indistinguishable from that of the Beacon Hill Succession, but in the Teign Estuary area the slates are commonly faintly micaceous and locally show some decalcified partings a few millimetres thick.

Shelly fossils are abundant in places within the slates, but deformation is such that faunas are indeterminate. In contrast, the associated limestones have yielded varied conodont faunas which have proved invaluable for correlation. Within the Beacon Hill Succession the oldest recorded fauna comes from limestones near Coppa Dolla [SX 8137 6699] (locality 5^‡3 ), where the kockelianus (unrestricted) to late ensensis Zone is indicated. Other conodont faunas indicate that the beds range upwards to the early Frasnian (locs. 1, 3, 4 and 7). Just south of the district [SX 8051 6649] (loc. 2), a thin limestone contains conodonts referable to the uppermost hermanni-cristatus or lowermost asymmetricus Zone. This last locality yields Palmatolepis? cf. disparalvea, a rare form in the European Devonian.

The Nordon Slate south of Kingskerswell belongs to the Beacon Hill Succession and possibly to the Denbury Unit. Conodonts from a thin limestone [SX 8753 6698] (loc. 9) probably indicate the varcus Zone. A similar age is given at Dainton Quarry [SX 85106597] (loc. 8).

With the exception of the Nordon Slate occurring in the Ugbrooke–East Ogwell Succession in the Teign Estuary area, the faunas of the slates throughout the different successions are dominated by benthonic elements including crinoid debris, bryozoans, brachiopods and specimens of the trilobite Phacops schlotheimi, which suggest a mid-Eifelian age. In the Teign Estuary area a predominantly pelagic fauna is dominated by tentaculitoideans, including Nowakia cf. holynensis, Styliolina minuta and S. fissurella; these can be correlated with the early Eifelian. Benthonic elements include solitary rugose corals, Pleurodictyum sp., crinoid debris, bryozoans, Murchisonia? and subspecies of P. schlotheimi. This last fossil suggests that the beds may range up to the mid Eifelian.

Kingsteignton Volcanic Group

Spilitic lavas, tuffs, limestones and some slates occur between the Nordon Slate and the Chercombe Bridge Limestone of the Ugbrooke–East Ogwell Succession in the Teign Estuary area in the Ugbrooke Unit only, and form a prominent north-east–south-west ridge on the north side of the Teign Estuary. Some 9 m of interbedded tuffs and slates with a few thin spilite lavas are overlain by about 90 m of tuffs with subordinate lavas and thin slates and with lenses of limestone up to 30 m thick. The limestones in the middle and upper parts of the sequence are mainly crinoidal. In the central and eastern part of the outcrop the tuffs with limestone lenses appear to be overlain by about 105 m of spilitic lavas with rare tuff horizons; the lavas die out westwards. The beds dip at about 35° south-east and are inverted; they young north-westwards and are structurally overlain by the older Nordon Slate.

Slates low and high in the Kingsteignton Volcanic Group yield the tentaculitoideans N. cf. holynensis, S. cf. fissurella, S. cf. minuta and phacopids including P. schlotheimi, a fauna identical to that of the Nordon Slate. A contrasting fauna has been obtained from a thin sandstone at [SX 9028 7413], comprising the tabulate coral Cladochonus?, large brachiopods and phacopids, but no tentaculitoideans or other pelagic elements.

Crinoidal limestones within the tuffs commonly contain massive and laminar stromatoporoids associated with rare coral fragments. The two main limestone lenses sampled seem to be at about the same horizon and probably span the Eifelian–Givetian boundary. The conodonts Tortodus kockelianus and Polygnathus linguiformis from a limestone lens north of Wear Farm [SX 8871 7315] (loc. 10) indicate the late Eifelian kockelianus (unrestricted) Zone; conodonts from north of Forder Lane [SX 8968 7375] (loc. 11) suggest a horizon within the upper patulus to C. costatus Zone, the earliest date recognised in the district. Thus the lenses of crinoidal limestone could be in part equivalent to the Denbury Crinoidal Limestone and the basal part of the Chercombe Bridge Limestone in the Denbury Unit, and the spilite lavas in the upper part of the volcanic group probably correlate with the lower beds of this Chercombe Bridge Limestone. The northern boundary of the group is largely faulted, and the age of the youngest beds problematical.

The tuffs are uniform, chlorite-rich, spilitic crystal tuffs with a rude cleavage that is generally subparallel to the bedding. They are dark green where fresh, brownish green or buff where weathered, and typically coarse grained though locally fine. Calcite occurs as small stringers and as irregular blebs up to 5 mm across. Rounded bombs and fragments of spilite lava up to 0.15 m in diameter are found. The tuffs are about 90 m thick in the east and 135 m in the west. In thin section the tuffs [SX 9018 7388]; [SX 9032 7407] show extensive recrystallisation and development of secondary minerals.

Altered, rounded or broken laths of rarely twinned plagioclase feldspar up to 0.05 mm long predominate, with some isolated crystals up to 1 mm across. Together with a little sub-rounded quartz they constitute about 40 per cent of the rock. The rare lithic fragments are mainly of spilite lava. Chlorite abounds as irregular patches or wispy threads, commonly associated with extremely fine-grained unidentifiable material. Calcite replaces large plagioclase crystals and -fills irregular voids. Epidote and sericite are alteration products of plagioclase, and small euhedral tablets and laths of secondary unaltered albite occur pseudomorphing plagioclase. Hematite is abundant in some samples.

Exposure is generally good, in crags and lane sections. Beds immediately above the Nordon Slate, tuffs and thin spilite lavas with some slates, are well exposed at Ware Barton and Wear Farm. Spilite flows form small crags [SX 8930 7346] and features [SX 8851 7302]. The sequence of spilite lavas that overlies the tuffs contains sporadic thin tuff horizons, as south of Colway Cross [SX 8984 7420].

Impersistent limestones in the lower part of the Kingsteignton Volcanic Group can be picked out by lines of old quarries. Three lenses have been mapped which interfinger with tuffs. The two largest [SX 887 733]; [SX 897 738] are 0.5 km long and up to 30 m thick; they lie at much the same stratigraphical horizon, though because of facies changes that horizon is near the top of the tuffs in the east and in the central part of the tuffs in the west. The smallest lens mapped [SX 895 735] is 90 m long and up to 6 m thick, and seems to lie low in the tuff sequence.

The limestones are dominantly crinoidal but with some dark grey, thinly bedded fine-grained limestones and massive stromatoporoid limestones. The central part of each limestone body shows [SX 8968 7375]; [SX 8983 7381] pinkish grey, poorly bedded to massive, coarse, crinoidal limestone with fragments of stromatoporoid and thin partings of crinoidal limestone and shale, together with medium to dark grey massive limestone, commonly dominantly composed of stromatoporoid, and alternations of beds 75 to 100 mm thick of coarse crinoidal limestones and laterally persistent horizons of laminar stromatoporoid. At the margins of the lenses, dark grey, thinly bedded, fine-grained limestones are commonly interbedded with tuffs.

Lavas in the lower part of the group occur as flows generally less than 0.3 m thick. In thin section they consist of a mass of tiny twinned albite laths and rods, commonly grouped in stellate to subvariolitic clusters, set in a fine groundmass of chlorite, calcite and a little glass. Scattered square or stumpy phenocrysts ofalbite are partly replaced by alteration products or calcite. Calcite also occurs as interstitial plates. The vesicles are usually irregular and filled with chlorite, calcite and chalcedony. Lavas at the top of the group comprise about 105 m of vesicular and amygdaloidal spilites faulted against tuffs and limestones to the south and Chercombe Bridge Limestone to the north. They are poorly exposed and deeply weathered and were shown as dolerite on earlier editions of the map. No pillow lavas have been observed. Thin sections reveal two rock types: normal spilite, consisting of a mass of radiating to randomly orientated feldspar microliths set in chlorite and opaque ore with chalcedony and chlorite-filled amygdales, and albitised basalt, consisting of closely packed randomly orientated long albite laths set in a rather glassy groundmass with granular opaque ore. Some specimens exhibit chlorite-filled amygdales, others show olivine and pyroxene pseudomorphs.

Denbury Crinoidal Limestone

The Denbury Crinoidal Limestone occurs in the Ugbrooke–East Ogwell Succession in the Denbury Unit, conformably between the Nordon Slate and the Chercombe Bridge Limestone. It is of variable thickness and shows a regional dip of about 20° south-east. The main outcrop around Denbury contains over 200 m of strata and is truncated to the south by a north-west–south-east normal fault. North of the village the formation forms an irregular linear feature at the base of the Chercombe Bridge Limestone. It is about 19 m thick immediately north of Denbury and thins out completely south-west of West Ogwell; other thicknesses are 54 m in Deer Park, 26 m south-west of Chercombe Bridge and 80 m north-east of Chercombe Bridge.

Denbury Crinoidal Limestone in the Torbryan Succession is developed locally between the Nordon Slate and the Pulsford Limestone. North of Torbryan, northerly dips give way eastwards to north-easterly dips as the outcrop swings round to abut against the Torbryan Fault. Eastward increase in width of outcrop reflects both the shallowing of the angle of dip around the nose of a faulted anticline and increasing thickness from west to east. The formation is 38 m thick at Torbryan church and 62 m thick farther east. It dies out northwards, possibly being cut out by a low-angle thrust fault. Such a fault would help to explain the presence of the dissimilar but equivalent Beacon Hill Succession immediately to the west.

The characteristic lithology, crinoid debris in a shaly matrix or fine-grained calcareous cement, is best seen in an old quarry [SX 8203 6912] 365 m north-west of Denbury, which is the type locality. To the north of Denbury coarsely crystalline crinoidal limestone predominates. Thin sections of rocks near West Ogwell [SX 8240 7040] show crinoid columnals averaging 1.5 mm in diameter and rare fragments of tabulate corals. Intergranular spaces are filled with sparry calcite, partly as syntaxial overgrowths on the columns. Pressure solution and stylolite growth have caused truncation of the columnals. Staining reveals that much of the sparry calcite contains iron. Some of the columnals have undergone degrading recrystallisation to micrite.

At Chercombe Bridge both shaly and crystalline crinoidal limestones are exposed, the latter thinning out northwards into Broadridge Wood. Crinoid debris in a fine-grained calcirudite from near Chercombe Bridge [SX 8331 7113] is seen in thin section to have undergone degrading recrystallisation. Dolomite occurs as irregular grains and patches, and ferroan calcite as veins. Fossils other than crinoid debris are uncommon. Coral fragments occur in a few places; Acanthophyllum sp.and Thamnophyllum sp.are recorded at [SX 8330 7104] Conodonts are not present in sufficient numbers for accurate age determinations. Selwood (1965) showed that the Eifelian–Givetian boundary occurs in the overlying Chercombe Bridge Limestone. At Torbryan the formation underlies the Pulsford Limestone, which is dated as late Givetian.

Pulsford Limestone

The Pulsford Limestone consists of medium to dark grey thinly bedded limestones about 30 m in total thickness. Individual limestones are characteristically lenticular and interbedded with thin shale. The type locality is a disused quarry [SX 8077 6795] south-east of Pulsford, where the limestones contain some small stromatoporoids and are succeeded by a local development of crinoidal limestone at the base of the East Ogwell Limestone. Farther south-east exposure is poor, and mapping depends largely on surface debris, but a number of outcrops occur near Torbryan.

A thin section of a specimen from [SX 8074 6795] shows extensive recrystallisation and a high clay content. Apart from rare crinoid debris the constituent grains average 0.15 mm in diameter and the rock is a fine-grained calcarenite. Irregular to ovoid cloudy patches of micrite may be pellets. Rare quartz grains occur up to 0.025 mm in diameter. There has been considerable development of pseudosparite, but prior to alteration the rock may have been a clayey pelmicrite or biomicrite. A similar rock from the type locality [SX 8077 6795] shows replacement of crinoid debris by pseudosparite, infilling of shell fragments with sparry calcite, and veins of ferroan calcite; crinoid fragments are sufficiently abundant for this rock to be described as a biomicrite. Near Coppa Dolla [SX 8132 6738] the limestone contains up to 40 per cent pellets but crinoid fragments are rare. To the south-east [SX 8149 6727] crinoid debris, rare coral fragments and stromatoporoids are present and the rock is traversed by stylolites which are lined with mud.

Macrofossils are rare except for a few corals and stromatoporoids, which include Acanthophyllum heterophyllum, Stringophyllum sp., Caliapora battersbyi and Syringaxon sp.at [SX 8192 6685]. Conodont faunas everywhere indicate the varcus Zone (locs. 16–20).

Chercombe Bridge Limestone

The Chercombe Bridge Limestone is present in the Ugbrooke, Denbury and East Ogwell units. Its thickness is up to 245 m. It is typically developed in the northern part of the Denbury Unit in quarries and sections alongside the River Lemon, where easterly dipping dark to medium grey well-bedded fine-grained limestones are distinct from the underlying Denbury Crinoidal Limestone and the overlying Foxley Tuff, and reach a maximum thickness of 150 m. Broadridge Wood Quarry [SX 8390 7112] (Plate 3) has been chosen as the type locality (p. 46). The limestone beds, which rarely exceed 1.2 m in thickness, are separated by thin shale partings, and range from medium-grained calcilutites to fine- or medium-grained calcarenites with little terrigenous material. Patchy and lamellar dolomitisation occurs. Similar limestones are developed locally within the overlying Foxley Tuff. In the Ugbrooke Unit, between Chudleigh and the Teign Estuary, the limestones of the Chudleigh escarpment and most of those south of Whiteway Barton have been referred to the Chercombe Bridge Limestone, although they include thickly bedded limestones with small bioherms up to 6 m thick, and some massive limestones and pale grey poorly bedded to massive stromatoporoidal limestones.

Stromatoporoids dominate the Chercombe Bridge Limestone fauna; dendroid forms occur in many outcrops and Amphipora is common. Tabulate corals predominate over rugose corals. Brachiopods, gastropods and rare trilobites are present locally. Ussher's (1913, pp. 19–28) faunas have not been revised. The formation is mainly of Givetian age; in the Chercombe Bridge area it ranges down into the late Eifelian; in the Denbury Unit it is overlain by the Givetian Foxley Tuff; in the Ugbrooke Unit the top of the formation is diachronous, mid Frasnian at Chudleigh, in the north, but possibly Givetian near Whiteway Barton in the south. The overlying East Ogwell Limestone in the south passes laterally into upper Chercombe Bridge Limestone.

Foxley Tuff

Tuffaceous rocks probably of mid and late Givetian age lie between the Chercombe Bridge Limestone and the East Ogwell Limestone in the Denbury and East Ogwell units. They are green and chloritic, weathering brown, and generally fissile. Thin fine-grained tuffs in Bickleyball Quarry, low in the Chercombe Bridge Limestone of the Ugbrooke Unit, may be their lateral equivalents.

The Foxley Tuff along the River Lemon contains thin limestones resembling Chercombe Bridge Limestone, is 80 m thick and dips south-east. North of the river it terminates against the Ipplepen Fault, and a broader outcrop to the south has been overridden by East Ogwell Limestone of the Torbryan Succession. The Emblett Hill Borehole (p. 184) penetrated graded and small-scale cross-bedded green chloritic fine-grained and lapilli tuffs with white calcite particularly common in the coarser types. Interbedded limestones occur locally in upward-coarsening cycles of limestone–fine tuff–lapilli tuff.

Poorly exposed tuffs west of Wolborough [SX 8516 7030]; [SX 8499 7020] and east of Two Mile Oak Cross [SX 8509 6791] are of uncertain stratigraphical position. The former are limited by block faults, but the latter appear to be thrust over Whiteway Slate of the Abbotskerswell Unit. The junction of the tuffs with overlying East Ogwell Limestone is probably tectonic in each case.

Rqgose and tabulate corals and rare brachiopods have been found in calcareous bands in tuffs alongside the River Lemon and corals in the Emblett Hill Borehole. In the River Lemon succession the basal beds of the overlying East Ogwell Limestone are probably late Givetian.

East Ogwell Limestone

The type locality of the East Ogwell Limestone is the disused Ransley Quarry [SX 8443 7018] near East Ogwell. The formation is thickest in a series of klippen in the East Ogwell

Unit. Thinner sequences occur in the Denbury and Abbotskerswell units. Between Chudleigh and the Teign Estuary the formation is exposed in the Ugbrooke Unit and has been penetrated by a borehole in the Kingsteignton Unit; in these areas it shows irregular development, is commonly difficult to distinguish from the Chercombe Bridge Limestone, and thins northwards. Typically the formation consists of pale to medium grey or greyish pink, poorly bedded limestones ranging from calcilutites to fine-grained calcirudites. North-west of East Ogwell, in the Denbury Unit, the rocks are rich in bioclastic material and unusually coarse.

The East Ogwell Limestone of the Torbryan Succession consists of pale grey to pink massive limestones similar to the corresponding rocks of the Chercombe Bridge area. Delimitation of the successions is arbitrary in places, and junctions have been drawn at tectonic features. The lower beds are thickest south of the Pulsford Fault, where the base is marked by the first massive limestone above the thinly bedded Pulsford Limestone, the top of the formation is not exposed, and about 200 m of strata are present.

The East Ogwell Limestone overlies the Foxley Tuffin the Chercombe Bridge area and the Chercombe Bridge Limestone between Chudleigh and the Teign Estuary. In each case it lies at the top of the Newton Abbot limestone group and is succeeded by the Upper Devonian Luxton Nodular Limestone.

Fossils are abundant and diverse in the East Ogwell Limestone, and notable collections of corals and stromatoporoids were made by earlier workers. The formation is here dated principally by reference to conodont faunas, which indicate that deposition of East Ogwell Limestone ceased everywhere at about the same time in the mid Frasnian. In the Ugbrooke Unit the East Ogwell Limestone probably ranges from Givetian to early Frasnian, but datable localities all lie high in the formation. In the Kingsteignton Unit Rhodes and Dineley (1957) obtained four Middle Devonian conodont faunas from the Bishopsteignton Borehole [SX 910 733]: Polygnathus linguiformis and P. varcus suggest the varcus Zone. One of the samples appears to have been obtained from the Luxton Nodular Limestone (p. 56), the others probably from the East Ogwell Limestone. North of Torbryan, in the Denbury Unit, coral faunas suggest a horizon well above the Eifelian–Givetian boundary.

Upper Devonian

Upper Devonian rocks are widespread in the Newton Abbot district but a full sequence occurs only in the Ugbrooke–East Ogwell Succession. North of the Holne Thrust, only late Famennian rocks are represented. The Upper Devonian strata are largely argillaceous except for the condensed thinly bedded and nodular limestones which succeed the Newton Abbot limestone group. Volcanic activity, attested by tuffs and spilites, was less than in Middle Devonian times. Neither top nor bottom of the Upper Devonian, as defined palaeontologically, coincides exactly with lithological boundaries (Figure 6).

Hyner Shale

Outside the metamorphic aureole of the Dartmoor Granite the Hyner Shale of the Teign Valley Unit consists of hard, dark blue or bluish green, generally mottled, shales and mudstones. It passes conformably up into grey and green micaceous shales of the Trusham Shale. Calcareous bands and nodules, and calcareous shales and mudstones, occur near the top of the formation and a calcareous siltstone forms a small ridge well seen at Hyner [SX 8404 8214] and east of Christow [SX 8407 8488]. This siltstone locally yields rich late Famennian faunas. Entomozoid ostracods including Maternella dichotoma and Richterina (R.) costata indicate the presence of the hemisphaerica-dichotoma Zone. The trilobites Archegonus (Waribole) cf. warsteinensis, Cyrtosymbole (Calybole) cf. ussheri and Phacops granulatus more specifically suggest the Wocklumeria Stufe at the type section [SX 8405 8482] to [SX 8408 8487] (p. 51) and at [SX 8410 8168]. The shales immediately above this horizon yield a Lower Carboniferous fauna of the Gattendorfia Stufe including Richterina (R.) latior at [SX 8424 8123].

Inside the aureole calc-flintas are common [SX 8300 8175] to [SX 8355 8170]; [SX 8300 8245] to [SX 8340 8244] and siliceous nodules are particularly characteristic of the lower beds of the formation. Between the Lustleigh and Bovey Tracey faults, and to the west of the fault zone, strata of Hyner Shale and Trusham Shale cannot be separated and are described under Trusham Shale.

Rora Slate

The fossiliferous Upper Devonian inliers north of Bickington were described by Ussher (1890, 1913) as conformable with the overlying Lower Culm cherts on the northern slopes of Ramshorn Down, but elsewhere mostly faulted against the Culm Measures. They constitute the Rora Slate, variously coloured slates with a prominent horizon of calcareous laminated siltstone at the top and conformably overlain by the Mount Ararat Chert. The formation is known only within the Liverton Unit, where it is exposed on the eastern side of the Ilsington Fault. The inliers, commonly partly fault bounded, occupy the lower slopes of valleys. Although isoclinal to close recumbent folds must give rise to local repetition, there is an overall younging from the valley bottoms to the cherts of the hilltops.

The type locality for the Rora Slate is South Rora Down Quarry (p. 52). Composite successions of the Rora Slate present in the inliers near Lounston and in the area around Ramshorn Down and Rora Down are as follows:

The lower Rora Slate fauna is sporadic in occurrence and usually poorly preserved. It is essentially pelagic, with an irregularly distributed benthos in the form of scattered solitary rugose corals, rare horizons rich in the tabulate coral Cladochonus?, a few disarticulated crinoid ossicles, indeterminate juvenile brachiopods and rare trilobites. Mainly juvenile bivalves and ammonoids occur throughout the sequence and ostracods near the top. The ammonoid fauna suggests an upper Clymenia–Wocklumeria Stufen age. Kosmoclymenia sp.occurs throughout the sequence, as at Pool Farm [SX 7987 7540] and Liverton [SX 8010 7523] to [SX 8024 7516]; crags below Mount Ararat [SX 7952 7472] and [SX 7953 7470] yielded Kosmoclymenia undulata. Other ammonoids include Clymenia sp.from [SX 7978 7492], and the long-ranging genera Imitoceras sp.and Sporadoceras sp.from nodules north of Pool Farm [SX 7974 7562]. Trilobites from limestones in the slates below Mount Ararat include Dianops anophthalmus [SX 7965 7462] and Phacops cf. granulatus [SX 7952 7472].

A hemisphaerica-dichotoma Zone ostracod fauna at the top of the lower Rora Slate includes Maternella hemisphaerica, Richterina (R.) costata, R. (R.) striatula and R. (Fossirichterina) semen. The occurrence of R. (F.) semen at this and younger horizons means that the two-fold division of the hemisphaericadichotoma Zone on the Continent, in which Fossirichterina appears only in the lower division, cannot be recognised in south-west England.

The benthonic fauna of the upper Rora Slate is similar to that of the lower beds but less plentiful. At Rora Farm [SX 8030 7436], in the middle of the formation, occur horizons rich in Cladochonus? and with crinoid stems up to 40 mm long. Juvenile chonetid and orthoid brachiopods occur together with isolated valves of lingulids; possibly the lingulids were epiplanktonic, living attached to 'soft' materials in the surface waters (Craig, 1952; Rudwick, 1965). Trilobites are the commonest benthonic element, including Dianops anophthalmus, D. aff. griffithides, Phacops granulatus, P. aff. wedekindi, P. cf. wocklumeriae, Pseudowaribole (Pseudowaribole) cf. conifera, P. (P.) aff. octofera and Typhloproetus subcarintiacus. Phacopid moults have been found in a quarry [SX 7907 7388] near Combe Farm. A single bellerophontid gastropod was obtained from a nodule in buff slates immediately north of Ramshorn Down [SX 7909 7440]. Bivalves occur in the slates and siltstones as scattered single valves, or more commonly with both valves close together. They form shell beds a few millimetres thick in silty mudstone at [SX 7953 7347]. Some may be referred to Paracyclas?, Guerichia [Karadjalia] venusta, and Sanguinolites? ellipticus; most are juveniles less than 5 mm long. Possibly the pectiniform bivalves were epipelagic (Craig, in discussion, 1954).

The pelagic fauna of these upper beds is characterised by abundant ostracods in the slates, but not in the siltstones. Entomozoids predominate over chilobolbinids, eurychilinids and leperditellids; they indicate the hemisphaericadichotoma Zone. Ammonoids are rare and occur in decalcified nodules or as flattened impressions in the slates; a Wocklumeria Stufe age is indicated. Conodonts have been found only as degraded moulds.

Greenish grey silty and siliceous slates at the top of the formation in the Mount Ararat track section [SX 7972 7478] contain bands of orange decalcified nodules that yield Cymaclymenia cordata and Imitoceras sp.This suggests upper

Wocklumeria Stufe ranging up into the Gattendorfia Stufe; conodont evidence from the overlying Ararat Slate indicates that it cannot be younger than the Ammonellipsites Stufe.

Kate Brook Slate

Ussher (1913) recorded some Lower Culm lithologies in the outcrops below Chudleigh Rocks, but regarded as Upper Devonian the rest of the green slates within the inlier that occupies the valley of the Kate Brook. Anniss (1933) referred greyish blue calcareous shales and thin crinoid-bearing limestones between Chudleigh Rocks and Black Rock [SX 8665 7858] to the late Frasnian, and the remainder of the slates in the valley to the late Famennian. All these slates constitute the Kate Brook Slate. They contain rare calcareous sandy lenses and beds of green very fine-grained sandstone. They are typically exposed in the Kate Brook gorge [SX 8645 7858]. The formation is entirely fault bounded and is of unknown thickness.

In the northern part of the Chudleigh inlier, around Harcombe, dark purple slates with rare thin sandy limestones occur within the green slates. This could be a lateral facies change and might indicate a passage into one of the ostracod slate formations, perhaps the Rora Slate.

The Kate Brook Slate crops out from Whiteway House to Chudleigh, and also between Blackpool and Staplehill, where the Holne Thrust swings northwards away from the Bickington Thrust. In the Kate Brook gorge, between the Riding Parks and Chudleigh Rocks, the formation is overthrust by Chercombe Bridge Limestone, but the junction is obscured by slipped blocks of the limestone.

Apart from scattered crinoid debris the fauna in the Kate Brook Slate is sparse and poorly preserved. Spiriferoid and rhynchonelloid brachiopods are present but mostly indeterminate. Cyrtospirifer sp., which is dominantly Upper Devonian in age, but may occur earlier and is known later, occurs in the calcareous sandy lenses in the Kate Brook gorge [SX 8664 7867]. Anniss (1933) reported "Styliolae" from Brimley, presumably from the road cutting [SX 8813 7977], which means that the slates there are unlikely to be younger than mid Famennian. At the northern end of the Chudleigh inlier, in the A38 road cutting [SX 8823 8193] to [SX 8823 8173], juvenile posidoniid bivalves and ostracods occurring with the brachiopods add a pelagic element to the fauna.

Gurrington Slate

The Gurrington Slate comprises an unknown thickness of slates, with a uniform south-easterly dipping cleavage, that stretches in an arcuate outcrop north-eastwards from the south-west corner of the district to Newton Abbot. Purple and green slates in the central part of the outcrop are flanked by grey slates to north-west and south-east (Figure 9). Rare ferruginous or siliceous nodules occur throughout the formation. Horizons of bluish grey calcareous siltstone occur within the greyish green slates between Bremridge and Farlacombe [SX 792 711], and buff siltstones south-east of Bickington. The purple and green slates show silty lamination in some fresh exposures. The type locality for the formation is in Gurrington Quarry (p. 53). Neither top nor bottom boundary is known.

Spilite lavas are restricted to the grey slates and are indistinguishable from those in the Nordon Slate. Southwest of Telegraph Hill [SX 8063 7317] grey slates are underlain by tuffs with interbedded spilites that structurally overlie the Chercombe Bridge Limestone. The stratigraphical position of the volcanic rocks suggests a Frasnian age, and Middleton (1960) correlated these tuffs with the Frasnian Dartington Tuffs.

Spilite with well-defined pillow structures is exposed in the disused Chipley quarries (Plate 2). The eastern quarry [SX 8085 7215] shows some cherts between the pillows. A thin section of spilite from the centre of one pillow shows that up to 50 per cent of the rock is composed of amygdales about 1 mm across set in a felted groundmass of plagioclase and opaque minerals. Most of the amygdales are of calcite rimmed by chlorite but some are of quartz or chlorite alone. Chlorite and calcite are also common in the groundmass. An analysis of the Chipley pillow lavas was given by Ussher (1913). Vallance (1965) showed that sodium is more abundant in the cores than in the selvages of pillows.

Spilitic tuffs occur at many horizons within the grey slates, but are mappable only at the base of the northerly belt of grey slates between Mead and Telegraph Hill. They are greenish brown, highly sheared and altered and show little trace of original texture. Numerous cognate lapilli of spilitic lava are present.

The fauna of the Gurrington Slate is irregularly distributed and commonly poorly preserved. It is dominated by juvenile bivalves and planktonic ostracods ranging in age from mid Frasnian possibly to early Tournaisian. Styliolinids and ammonoids are rare. Trilobites provide the only evidence of benthos; many of the phacopids from Knowles Hill Quarry [SX 858 718] show Salterian moults. (Figure 9) relates ostracod age determinations to the disposition of the principal belts of coloured slates. Close juxtaposition of rocks of contrasting ages indicates the presence of strike faults or thrusts. The northern unit ranges from the volcanic rocks of probable Frasnian age, through the Famennian northern grey slates belt into purple and green slates of latest Famennian and possibly earliest Tournaisian age. The central unit is entirely within purple and green slates and ranges from late Frasnian in the north to mid Famennian in the south. The southern unit is of grey slates, in part known to be of mid-Frasnian age.

Luxton Nodular Limestone

House and Butcher (1962, 1973) have shown that in the Ugbrooke–East Ogwell Succession the massive limestones of the Chudleigh escarpment are overlain by a condensed Upper Devonian sequence. The succession they described is given below, to the right of the classification adopted in this Memoir:

The Luxton Nodular Limestone occurs in the Ugbrooke, Kingsteignton, Denbury, Abbotskerswell and East Ogwell units. Its type section is an outcrop of thinly bedded red nodular limestones and slates with abundant calcareous nodules on the foreshore of the Teign Estuary, west of Luxton's Steps [SX 9000 7288]. The formation comprises red, grey and green thinly bedded, fine-grained nodular limestones, slates with limestone nodules, slates and rare siliceous horizons.

The base of the formation is abrupt at the top of both the East Ogwell Limestone and the Chercombe Bridge Limestone. Two basal members are differentiated, but neither is mappable. The Kiln Wood beds, which succeed the Chercombe Bridge Limestone, consist of interbedded thin limestones and shales, which may be siliceous, and local thin tuff horizons. The Lower Dunscombe Goniatite bed, which succeeds the East Ogwell Limestone, consists of thin irregularly bedded, fine-grained limestones. These two basal members were thought by House and Butcher (1962, 1973) to be in succession (see table above), but conodont evidence (Tucker and van Straaten, 1970) suggests that they are broadly equivalent, a fact further brought out by their areal distribution.

Beds above the basal members are made up of nodular limestones, and slates with limestone nodules, which range in age from early to mid Famennian. In the Ugbrooke Unit these strata consist of condensed nodular limestones, but in the Kingsteignton and Abbotskerswell units they are thicker and more argillaceous and appear to represent a transition into early Famennian sequences of the Gurrington Slate seen in the Highweek Unit. The top of the formation is taken above the uppermost nodular limestone in condensed sequences, and elsewhere arbitrarily above the top horizon of carbonate nodules or nodular limestones. Part b of House and Butcher's Mount Pleasant Group is transitional between the dominantly carbonate and argillaceous parts of the sequence and has been placed arbitrarily in the Whiteway Slate.

The Kiln Wood beds contain a pelagic fauna with rare stunted benthos. The grey shales yield brachiopods, scattered tentaculitoideans, bivalves, orthoconic nautiloids, goniatites and ostracods, all commonly pyritised. Calcareous horizons contain a richer fauna, including foraminifers, trilobites and conodonts. The beds are generally mid and late Frasnian, but range up into the Cheiloceras Stufe in the Rydon Ball Farm Borehole. The Lower Dunscombe Goniatite bed contains a dominantly pelagic fauna, with restricted benthos, comprising rare corals, crinoid debris, brachiopods, bivalves, ammonoids, trilobites and conodonts. This fauna indicates some lateral equivalence to the Kiln Wood beds. The nodular limestones and slates above the Goniatite bed yield a pelagic fauna of brachiopods, bivalves, ammonoids, and conodonts, plus a few ostracods in the slates. Benthos is currently represented by a single rugose coral, rare scattered crinoid debris, and a single trilobite. The nodular limestones range upwards from the lower Cheiloceras Stufe into the Platyclymenia Stufe at Mount Pleasant and into the mid Clymenia Stufe at Well. Mixed conodont elements involving the reworking of marginifera and velifer Zone forms at Well, suggest a non-sequence, perhaps due to penecontemporaneous carbonate disolution accompanied by local erosion.

Whiteway Slate

Whiteway Slate is present in the Ugbrooke, Kingsteignton and Abbotskerswell units and possibly in the East Ogwell Unit. It consists predominantly of grey-black, green and purple slates, locally calcareous and siliceous, and rarely sideritic. The type section is in Whiteway Barton farmyard [SX 8844 7525]. The junction with the underlying Luxton Nodular Limestone generally lies within the Platyclymenia Stufe, but ranges up into the Clymenia Stufe at Well. The top of the formation known in the Ugbrooke Unit is taken below the lowest black sooty slates of the Winstow Slate and lies in the upper Gattendorfia Stufe.

Thickness variations in the Whiteway Slate parallel those in the Luxton Nodular Limestone, so that reduced successions occur in roughly the same areas in both formations. The Whiteway Slate of the Ugbrooke Unit is 26 m thick in the Chudleigh escarpment and 45 m at Whiteway Barton. It is thicker in the Kingsteignton and Abbotskerswell units, and probably transitional to the Gurrington Slate of the Highweek Unit. Three spilitic lavas are present in the formation in the Kingsteignton Unit.

The Whiteway Slate contains a rich planktonic fauna of ammonoids, ostracods and conodonts. The benthos includes crinoid debris, rare brachiopods including lingulids, bivalves, trilobites and extremely rare solitary corals and gastropods. The nodule-rich facies of the Chudleigh escarpment yields abundant ammonoids and ostracods, accompanied by a locally rich molluscan benthos of mostly immature gastropods and bivalves. Elsewhere ammonoids are rare. Entomozoid ostracods are ubiquitous but most abundant in the thinner successions, where the accompanying bivalves, mainly immature and less than 10 mm long, are commonly found with both valves still articulated and unbroken. Many of the trilobites from Whiteway Barton are also juveniles.

Correlation

The faunas allow correlation of the principal successions (Figure 6). The Teign Valley and Liverton successions both contain calcareous laminated siltstones disposed about the Devonian–Carboniferous boundary. In the carbonate successions to the south, faunas indicate rapid lateral changes of facies. Within the Ugbrooke–East Ogwell Succession carbonate deposition began in late Eifelian times in the Chercombe Bridge–East Ogwell area but in early to mid Givetian times between Chudleigh and the Teign Estuary. Early Givetian Chercombe Bridge Limestone is widely developed, but the overlying East Ogwell Limestone is diachronous. Thus, in the Ugbrooke Unit, south of Whiteway Barton the upper parts of the East Ogwell Limestone contain conodonts indicating a position close to the Givetian–Frasnian boundary and the Chercombe Bridge Limestone is wholly Givetian, whereas on the Chudleigh escarpment the latter limestone ranges up into early Frasnian. In the Denbury Unit the Chercombe Bridge Limestone includes mid Givetian and perhaps late Givetian strata, while in the East Ogwell Unit it seems to be mainly younger than mid Givetian, though one early Givetian locality is known.

In the Torbryan Succession the Nordon Slate may range well up into the Givetian, the overlying Pulsford Limestone being dated as late Givetian. Either carbonate deposition began much later than in the Ugbrooke–East Ogwell Succession or the succession has been tectonically thinned. The local presence of crinoidal limestones at the base of the carbonates of the Torbryan Succession invites comparison with the Denbury Crinoidal Limestone in the basal part of the Ugbrooke–East Ogwell carbonate strata. The Pulsford Limestone appears to be a condensed equivalent of part of the Chercombe Bridge Limestone. The East Ogwell Limestone, which overlies the Pulsford Limestone, is probably late Givetian to Frasnian in age. Uniform conditions were established at this time over the whole area represented by the Denbury Unit.

The end of carbonate deposition in the Ugbrooke–East Ogwell Succession is represented by the Luxton Nodular Limestone, whose base is not older than late mid-Frasnian. An irregular topography inherited from the massive limestones controlled facies variation in this formation and in the overlying Whiteway Slate. Faunas indicate that thicknesses increase away from submarine rises or schwellen, and that there are lateral facies changes at the base of the Luxton Nodular Limestone, where the Lower Dunscombe Goniatite bed and the Kiln Wood beds are broadly laterally equivalent.

The junction of the Luxton Nodular Limestone with the Whiteway Slate seems to be about middle Platyclymenia Stufe in age, except at Well, where the Luxton Nodular Limestone persisted into the Clymenia Stufe.

Ostracods and conodonts show that the Beacon Hill and Bickington successions together constitute an argillaceous equivalent of the Ugbrooke–East Ogwell and Torbryan successions. Possibly some of the unfossiliferous greyish black slates of the Gurrington Slate may be of Carboniferous age and represent thick equivalents of the Winstow Slate. The contact between the Nordon Slate of the Beacon Hill Succession and the Gurrington Slate of the Bickington Succession is the Forder Green Thrust that separates the Highweek Unit from the Denbury Unit, but the amount of northward translation is probably not great.

Volcanicity in Eifelian and Givetian times seems to have been localised. There is no evidence of the Kingsteignton Volcanic Group in the Denbury Unit. In the Ugbrooke Unit there is no mappable representative of the Foxley Tuff, although thin tuffs occur low in the Chercombe Bridge Limestone in Bickleyball Quarry [SX 8850 7400]. Tuffs and spilite lavas of probable Frasnian age occur at the base of the Gurrington Slate within the Bickington–Beacon Hill Succession in the north of the Highweek Unit. They are not represented in the Frasnian of the southern part of this unit, nor in the Ugbrooke–East Ogwell Succession except for extremely thin and local tuffaceous horizons in the Kiln Wood beds on the Chudleigh escarpment. The Whiteway Slate contains three thin spilitic lavas within the relatively thick sequence which forms the lateral transition into the Gurrington Slate.

Facies changes

The facies relationships in the Middle Devonian to Upper Carboniferous rocks of the Newton Abbot district are shown in (Figure 8).

Nordon Slate facies

The late Eifelian to early Frasnian Nordon Slate of the Beacon Hill Succession, although largely barren, contains a fauna of similar character to the Nordon Slate underlying the carbonate complexes to the east. It includes thick spilitic lavas and local thin beds of tuff and limestone. Some beds of the limestone are parallel laminated, locally graded, contain rolled stromatoporoids, corals and crinoid debris, and are interpreted as redeposited limestone, possibly representing carbonate turbidites derived from the unstable margins of the carbonate banks.

The Nordon Slate at the base of the carbonate complexes accumulated in a shallow marine shelf environment, but faunal differences suggest that conditions were not uniform. In the Denbury Unit these slates range up at least into the mid Eifelian and contain a locally abundant shelly fauna including solitary corals, bryozoa, crinoid debris, coarse-ribbed brachiopods and trilobites. A thin limestone in the West Ogwell area contains stromatoporoids and probable late Givetian conodonts, suggesting that here the argillaceous facies persisted well into the Givetian (p. 44). The Nordon Slate underlying the carbonates of the Ugbrooke Unit is of late Eifelian age and contains a dominantly pelagic fauna; styliolinids and nowakiids are randomly orientated, indicating little or no current activity. The benthos is sparse, and coarse-ribbed brachiopods are absent. The slates represented in the Ugbrooke Unit may have formed in deeper water than those in the Denbury Unit, and in this deeper water a volcanic pile represented by the Kingsteignton Volcanic Group may have provided sites for crinoid and stromatoporoid growth that gave rise to local carbonate bank developments.

Carbonate complexes

Carbonate deposition in the Denbury Unit began with the Denbury Crinoidal Limestone, which is absent in the northern part of the area in which the Torbryan Succession occurs. Active circulation is suggested by well-sorted crinoidal limestones with sparry cement. A shallow shelf environment is suggested. The Pulsford Limestone contains some crinoidal debris, together with a few corals and stromatoporoids, and probably represents a local area of more restricted circulation.

The Kingsteignton Volcanic Group contains crinoidal and stromatoporoid limestones interbedded on two scales. Firstly, massive beds up to several metres thick of crinoidal limestone and limestone with stromatoporoids in growth positions, and secondly, beds up to 100 mm thick of crinoidal limestone and laminar stromatoporoid limestone, the former having sharp bases. The periodic initiation and growth of stromatoporoid colonies on top of crinoid debris, and their subsequent rapid burial under current-borne crinoid debris, is possibly linked to strong currents active during storm conditions.

Above the crinoidal limestones in the Ugbrooke–East Ogwell Succession in the Denbury Unit and the spilite lavas of the Kingsteignton Volcanic Group in the Ugbrooke Unit, there is a rapid passage into Chercombe Bridge Limestone that seems locally to represent an environment of minimal circulation, perhaps a protected back reef. The limestones are biopelmicrites, but the gross lithology depends on the fossil content. Stromatoporoids are abundant, particularly Amphipora and other dendroid forms, and tabulate and solitary rugose corals are also common at certain levels.

Much of the Chercombe Bridge Limestone in the Ugbrooke Unit consists of lime mud with rolled stromatoporoids and corals at various levels, as at Bickleyball Quarry and near Chudleigh Rocks [SX 8637 7875]. The well-bedded nature of the limestone and lack of in situ faunas argue against a reef structure, but small bioherms are present, as in Rydon Quarry [SX 874 941] (also known as Kingsteignton Quarry). Elsewhere, as at Ridge Quarry [SX 8894 8018] on the Chudleigh escarpment, massive stromatoporoids are also locally developed. The Chercombe Bridge Limestone probably formed in several sub-environments including in situ stromatoporoid banks and areas of lime mud deposition with currents occasionally reworking and concentrating bioclastic material.

Two principal types of East Ogwell Limestone are pale grey calcilutite with some bioclastic material, which probably formed in open shelf conditions, and, more commonly, calcirudite with abundant bioclastic material including corals, brachiopods, trilobites and crinoids. The latter formed in an area of active circulation, possibly a bank. In the Emblett Hill Borehole coarse calcarenites and calcirudites are interbedded with thin layers of stromatoporoid or tabulate corals in growth position; this points to periods of quiescence.

A review of the carbonate complexes of south Devon was given by Scrutton (1977).

Luxton Nodular Limestone facies

The Luxton Nodular Limestone marks a change in sedimentation from the underlying shallow-water elastic carbonates. Fauna and lithology indicate reduced successions, which are best explained in terms of sea-floor rises or schwellen recognised in the Hercynian facies of the German Devonian (Schmidt, 1926). The sudden onset of this facies was caused by rapid subsidence which may have been related to Frasnian tectonic activity to the south of the present area. The Luxton Nodular Limestone shows gradations from rise deposits in the Ugbrooke Unit to transitional or rise-slope deposits in the Kingsteignton and Abbotskerswell units.

The basal beds show striking facies changes. It appears that the formation originated on an irregular floor on top of the earlier carbonate deposits. The thin Lower Dunscombe Goniatite bed, consisting almost entirely of carbonate, lies above the shallow-water East Ogwell Limestone and bears a rich fauna with a significant benthonic element, whereas the approximately equivalent Kiln Wood beds consist mainly of greyish black shales with a restricted, often pyritised, pelagic fauna, with little benthos, and overlie the fine-grained Chercombe Bridge Limestone that probably formed in deeper water.

Dark green and greenish grey silty shales with scattered limestone nodules and thin nodular limestones at Well and Whiteway Barton represent an environment transitional between those of the Kiln Wood beds and the Lower Dunscombe Goniatite bed. Compared with the former they contain more carbonate and a richer fauna, but the shales are slightly paler and the fossils are not pyritised; apparently the bottom conditions were not completely anaerobic. The structurally isolated exposure of Kiln Wood beds at Conitor [SX 8508 7059], and the developments in the Rydon Ball Farm Borehole, are thicker than any occurrence in the Ugbrooke Unit; possibly they formed in a separate basin.

The formation of limestone nodules and nodular limestones in shales above the basal beds of the formation at the end of the Frasnian marked continuing subsidence and further restriction of the benthonic fauna, similar to that discussed by Griindel and Rosier (1963).

The mixed mid and late Famennian conodont faunas at Well, and perhaps also at Winstow Cottages, pose a problem. It is difficult to envisage a mechanism involving erosion of earlier horizons, survival of conodonts and their transport and incorporation into a later deposit on a schwelle (rise). A possible interpretation seems to be solution of carbonate on the sea floor, leading to the concentration of conodonts from a number of zones.

Ostracod slate facies

The typical becken facies consists of slates with a dominantly pelagic fauna. These slates pass laterally into limestones thought to have accumulated on schwellen. The earliest such basinal deposit in the present district is the mid Frasnian lower part of the Gurrington Slate, which equates with part of the Luxton Nodular Limestone. As a result of progressive subsidence, ostracod shales extended over the schwellen as a condensed deposit, the Whiteway Slate. During Wocklumeria Stufe times the ostracod slate facies extended northwards beyond the becken on to the northern shelf, where they are represented in the basal formations of the Teign Valley and Liverton successions.

The Gurrington Slate represents the deepest water environment of the ostracod slate facies and appears to reflect deepening of the Nordon Slate basin; this deepening led to the disappearance of the shelly benthos and the end of limestone turbidite deposition in the Nordon Slate succession. The Gurrington Slate contains scattered thin beds of calcareous siltstone which are interpreted as distal turbidites, but these are confined to the South Knighton [SX 811 725] area in the northern belt of grey slates. The rare nodules in the formation are ferruginous or siliceous rather than calcareous.

The Whiteway Slate contains some carbonate nodules and thin siliceous limestones. Lateral passage towards the Gurrington Slate is marked by thicker deposits with rare siliceous and sideritic nodules.

A northern sub-facies consists of the Rora Slate of the Liverton Succession and the Hyner Shale and Trusham Shale of the Teign Valley Succession. It is distinguished by absence of volcanic rocks and the presence of silty lithologies around the Devonian–Carboniferous boundary. Fauna and fine-grained lithology point to a quiet environment. Perhaps deposition took place in a becken less deep and nearer land than that of the Gurrington Slate, for not only is the benthos more abundant and varied but the pelagic elements are locally abundant. The influx of silts about Devonian–Carboniferous boundary times was not accompanied by any change in the benthos. These laminated calcareous siltstones may be distal turbidites. They are absent from the southern successions except for one local incursion in the Famennian, and it is assumed that the Liverton and Teign Valley successions were nearer land at that time. The upper part of the Rora Slate is condensed compared to its correlatives in the Teign Valley Succession, but the cause is unknown.

Kate Brook Slate facies

Thin sandstones locally present within the Kate Brook Slate are commonly roughly graded and parallel laminated and may be turbidites. The coarse-ribbed spiriferoids and rhynchonelloids that dominate the fauna are absent from the ostracod slate facies. A distinctive ostracod fauna is known from only a single locality. The brachiopod fauna suggests a shelf environment, and the absence of a varied benthos indicates a position some distance seaward on the shelf.

Details

Middle Devonian

Nordon Slate

Bickaton to Chercombe Bridge [SX 8331 7107]

Near Waytown Cross, cleavage dips at 10°/062° [SX 7918 6764] in slates in contact with vesicular spilite. Typical pale to medium green spilites with porphyritic feldspars up to 5 mm long crop out at Torcorn Hill [SX 7883 6682]. Phenocrysts up to 15 mm long are present at Waytown [SX 7903 6723], and a thin section of rock from there shows albitic plagioclase as phenocrysts and microlites of the groundmass extensively replaced by calcite and chlorite. A few calcite-filled amygdales are present. Fine-grained opaque minerals in the groundmass include some pyrite. Rare secondary quartz occurs in patches. A spilite from Knowle Hill [SX 7981 6770] shows amygdales 1 to 3 mm across, rimmed by chlorite and filled with calcite and rare quartz. Plagioclase phenocrysts, probably albite, are extensively altered to calcite. The groundmass has a slight trachytic texture, with needles and laths of plagioclase. Much cryptocrystalline material is present, together with fine-grained opaque minerals.

Slates in Beacon Lane [SX 8062 6668] to [SX 8050 6688] are apparently strongly folded, showing polished cleavage and quartz veining. Within the Nordon Slate, lenticles and beds of dark grey and black limestone up to 0.3 m thick are separated by shaly partings. A thin section of limestone from an exposure at Waterford Cross [SX 8083 6682] shows that all the calcite is recrystallised and that pyrite is a common authigenic constituent; the limestone at this locality is horizontal and yields conodonts (Appendix 1, loc. 3) indicating the early Frasnian asymmetricus Zone. Black thinly bedded limestones dip at 20°/113° [SX 7989 6669] and yield conodonts (loc. 1) suggesting an early Givetian age, at 20°/054° [SX 8137 6699] and yield conodonts (loc. 5) of the kockelianus (unrestricted) to late ensensis Zone, and at 20°/028° [SX 8102 6746] and yield conodonts (loc. 4) probably of varcus age. Thin decalcified bands occur within slates in the bed of the Am Brook, north of Coppa Dolla; grey or black slates show cleavage dipping at 28° to 62° eastward and yielded Phacops schlotheimi (mid Eifelian) from a decalcified bed at Collacombe Bridge [SX 8105 6751]. Some slates near Levaton [SX 8050 6780] to [SX 8076 6778] are slightly tuffaceous or contain spilite fragments, cleavage orientation varying from 20°/068° to 15°/107°. On the south side of Beacon Hill [SX 8020 6732] laminated tuffaceous siltstone contains rare volcanic clasts, and similar rocks occur near Yeatt [SX 809 687], Denbury Camp [SX 817 685] and Heathfield [SX 816 693].

Occurrences of Nordon Slate north of the Pulsford Fault are arbitrarily referred to the Ugbrooke–East Ogwell Succession. The slates pass eastwards beneath Denbury Crinoidal Limestone and to the west their outcrop is terminated by the Forder Green Thrust. Slates in the lane leading north-east from Pulsford are locally tuffaceous [SX 8074 6823]; [SX 8078 6825]. Denbury Down is underlain by vesicular and porphyritic spilite, green, fine-grained, with a few feldspar phenocrysts up to 5 mm diameter and chlorite-filled amygdales [SX 8111 6858]. Thin sections show the rock to be composed mainly of albite laths up to 0.3 mm long, with chlorite and opaque minerals in the groundmass. Rare amygdales are filled with chlorite, calcite, and some quartz. Middleton (1960) reported vesicular 'bombs' at or near the base of the lava on the north side of Denbury Down, and related them to centres within the lava flow around which there was a release of volatiles. Greenish grey slates in the lane north of Denbury Camp [SX 8150 6885]; [SX 8160 6886] to [SX 8169 6887] dip at 30° east and contain bands of deformed brachiopod moulds and other organic fragments, chiefly crinoid debris.

Near Heathfield Farm, hedgebank exposures [SX 8160 6925] to [SX 8160 6927] are taken as the type locality for the Nordon Slate. They show greenish brown slates with lenticles and beds of limestone up to 0.23 m thick. Some of the thinner beds may be complete stromatoporoids; others contain conodont fragments, but no precise age is indicated. Soft rust-coloured decalcified layers up to 76 mm thick contain moulds of brachiopods, crinoid debris and rare trilobites. Surface brash and poor exposures of dark grey limestone near West Ogwell suggest the presence of a lenticular limestone. The slates have yielded no determinable fossils apart from mid-Eifelian trilobites from West Ogwell (Ussher, 1913, p. 17), but conodonts from the limestone [SX 8185 6993] south-south-west of West Ogwell Convent indicate a Givetian age (loc. 7). Thus the limestone and possibly some of the associated slates are younger than the nearby Eifelian Denbury Crinoidal Limestone, and they may be part of an undefined area of either the Beacon Hill or the Torbryan succession north of the Pulsford Fault. Weathered crinoidal slates crop out in Winterlears Plantation [SX 8314 7099] immediately west of Chercombe Bridge.

East Ogwell and Dainton

In the East Ogwell Unit Ussher (1913, p. 17) recorded possible Devonian slates in roadside exposures south-west of Wolborough, and buff slates in a roadside cutting [SX 8509 6973], which are referred to the Nordon Slate, have yielded an Eifelian fauna including Phacops schlotheimi. These slates cannot be delimited because of poor exposure, but appear to be associated with the adjoining klippe of East Ogwell Limestone which is thrust over Whiteway Slate of the Abbotskerswell Unit.

An old quarry [SX 8753 6698] east of Dainton shows dark grey limestone beds up to 305 mm thick and interbedded shales. The limestones contain probable varcus Zone conodonts (loc. 9). Thin sections show a mosaic of calcite grains up to 1 mm in diameter with patches of dolomite rhombs and micrite or microspar and much, recrystallised fossil debris. Quartz is rare, occurring as irregular blebs, probably of secondary origin. Lack of close packing of remnant skeletal grains suggests that an original biomicrite has been altered by neomorphic processes.

Ware Barton to Murley Grange

Ussher (1913) noted Nordon Slate immediately north of the Teign Estuary between Ware Barton [SX 884 729] and Murley Grange [SX 902 737], but the outcrops were not shown on the one-inch map. To the south, Nordon Slate is in structural contact with the Whiteway Slate of the Kingsteignton Unit, while to the north it overlies basal tuffs of the Kingsteignton Volcanic Group in an inverted sequence. The slates are grey where fresh, greenish grey and brown where weathered. They are commonly faintly micaceous, and some show rare decalcified partings a few millimetres thick. Randomly orientated tentaculitoideans occur on many bedding planes. Bedding and cleavage usually coincide and dip south or south-east at low angles. These slates crop out at the base of a ridge formed by the volcanic rocks. The junction is exposed in inverted succession immediately west of Murley Grange, where some 4.6 mn of tuffs in a disused quarry [SX 8993 7372] pass up into micaceous slates; bedding dip is 45°/190°. The fauna includes a varied but stunted benthos with solitary rugose corals, Pleurodictyum sp., rare crinoid fragments, bryozoans, Murchisonia sp.and occasional P. schlotheimi. A richer planktonic element is dominated by Nowakia cf. holynensis, Styliolina fissurella and S. cf. minuta, whose local abundance and random orientation suggests little current activity. Orthoconic nautiloids and indeterminate ammonoids are also present. The tentaculitoidean fauna places the Nordon Slate of the Ugbrooke Unit at the top of the early Eifelian.

In the floor of a cart track [SX 8811 7273] grey, yellowish-greenweathered, micaceous slates yielded solitary rugose corals, crinoid debris, S. minuta and a phacopid.

A bank 83 m south-west of Wear Farm [SX 8888 7310] contains greyish green to buff weathered slates with thin decalcified partings. Cleavage dips at 10°/018°. Bedding planes are commonly covered with randomly orientated styliolinids; bryozoans, Nowakia sp., S. cf. fissurella, S. minuta, small bivalves and orthoconic nautiloids and an ammonoid have been recorded. Immediately to the north-east [SX 8893 7314] greyish green slates with cleavage inclined at 35°/142° crop out beneath tuffs; they yield Pleurodictyum sp.and Phacops sp.In the farmyard [SX 8900 7317] a cutting exposes 4.6 m of greyish green and grey slates, which contain irregular pyrite nodules up to 12 mm in diameter, show cleavage dipping at 22°/150° and have yielded S. minuta.

Greyish green slates at [SX 8924 7333], with cleavage dip 32°/075°, yielded solitary rugose corals, gastropods, S. cf. fissurella, S. minuta, small bivalves, juvenile ammonoids and P. cf. schlotheimi.

Kingsteignton Volcanic Group

In the Ware Barton farmyard [SX 8838 7294] brown-weathered coarse tuffs contain 2.4 m of greyish green micaceous slates which have yielded scattered styliolinids, including S. cf. fissurella and S. cf. minuta, P. cf. schlotheimi, and rare crinoid debris. Bedding dips at 20°/285°. In a bank by the entrance to the farm medium- and coarse-grained weathered tuffs contain 0.15 m of vesicular spilite lava; the cleavage in the tuffs dips gently south or west. An old pit [SX 8865 7333] shows 1.8 m of thinly bedded dark grey fine-grained limestones, dip 30°/130°; a 0.15-m band of red clay and weathered fine-grained tuffaceous material near the middle of the section contains lenses or thin beds of limestone. Immediately to the south [SX 8863 7329] occur crags of thinly bedded bluish black limestone rich in Amphipora. An old pit [SX 8871 7315] shows 1.8 m of pinkish grey coarse crinoidal limestone, locally recrystallised, dip 38°/207°, faulted against weathered tuffs; conodonts (loc. 10) from the limestone include Tortodus kockelianus, Polygnathus linguiformis and bar forms of the kockelianus (unrestricted) Zone.

Greyish green slates with tuffs and limestones crop out at [SX 8943 7368], the slates yielding rare crinoid fragments, N. cf. holynensis, S. cf. minuta and P. schlotheimi, and pale grey, coarse, crinoidal limestone with fine-grained bluish black stromatoporoid limestone occurs at [SX 8951 7364].

The following sections occur in old quarries; at [SX 8964 7374] 4.6 m of poorly bedded coarse crinoidal limestone with bands of laminar stromatoporoid; at [SX 8968 7375] 4.6 m of poorly bedded stromatoporoid and crinoidal limestone, overlain by interbedded laminar stromatoporoid and crinoidal limestones, dip 35°/168°, with a condont fauna (loc. 11) of lower Eifelian age; at [SX 8986 7365] 1.2 m of weathered tuffs with rare small spilite fragments, cleavage dip 22°/143°; at [SX 898 738] 4.5 m of dolomitised red crinoidal limestone; at [SX 8983 7381] a 12-m face of massive to well-bedded limestone with laminar and massive stromatoporoid and with bands and lenses of coarse crinoidal limestone, dip 42°/175°; at [SX 8993 7372] 4.5 m of weathered tuffs, cleavage dip 45°/200°, beneath Nordon Slate; at [SX 9017 7388] 1.5 m of coarse purple tuffs, the colour due to disseminated hematite, cleavage dip 42°/195°.

A lane section [SX 9028 7413] north-west of Clanage Cross shows fine-grained decalcified sandstone yielding Cladochonus?, solitary rugose corals, brachiopods and Phacops sp.Another [SX 9032 7407] shows 1.5 m of green coarse tuffs with cleavage orientated 32°/252°.

Denbury Crinoidal Limestone

Crinoidal limestone exposed north-east of Torbryan church [SX 8208 6691] is 7.4 m thick and comprises beds up to 0.2 m thick dipping at about 26° north; some beds contain a matrix of red shale, others are strongly cemented by calcite. Conodonts (loc. 13) from this locality give no precise age.

Red shaly crinoidal limestone [SX 8203 6911] to [SX 8204 6913] dips at about 20°/087°–120°; the thinner beds are the more shaly and in places are eroded out; bedding is slightly folded and cut by minor steeply dipping north–south fractures. Conodonts (loc. 12) from the type locality [SX 8203 6912] give no age determination; bedding is slightly folded and cut by minor steeply dipping north–south fractures.

Coarsely crystalline crinoidal limestone in overgrown outcrops south-east of West Ogwell dips at 26°/128° [SX 8226 6975]. It forms massive beds dipping at 38°/115° [SX 8237 7007] and at 20° to 41° east [SX 8234 7029] to [SX 8243 7043]. Conodonts from [SX 8240 7040] (loc. 14) and [SX 8331 7113] (loc. 15) give no precise age determination. In road cuttings west of Chercombe Bridge shaly crinoidal limestones dipping at 35°/118° [SX 8324 7105] pass into massive crystalline limestone [SX 8331 7105].

Pulsford Limestone

Some 9 m of limestone are exposed in a disused quarry [SX 8077 6795]. The rock is dark grey, fine to medium grained, in lenticular beds from 50 mm to 150 mm thick with shaly partings between. The dip of 40°/050° is disturbed alongside minor north–south normal faults. Conodonts (loc. 16) suggest the beds belong to the late Givetian varcus Zone. Stratigraphically lower beds beside a footpath [SX 8074 6793] show well-bedded, dark grey limestone underlain by lenticular limestones with intercalated shales, the junction being disrupted by small north–south faults. Conodonts (loc. 17) indicate the varcus Zone at [SX 8078 6790]. Outcrops in woods north-east of Coppa Dolla [SX 8132 6738]; [SX 8134 6733]; [SX 8149 6727] comprise lenticles of dark grey limestones containing small stromatoporoids, crinoid debris and conodonts and dipping 20°–36°/035°–080°. Conodonts from [SX 8132 6738] (loc. 18) and [SX 8149 6727] (loc. 19) probably indicate the varcus Zone.

Up to 3 m of lenticular grey limestone in an old pit [SX 8158 6690] west of Torbryan church, dip at 30°/067°. The rock is medium grained and contains rare rugose corals, crinoid debris, and conodonts (loc. 20) which possibly indicate the varcus Zone. The north-easterly limit of the Pulsford Limestone is defined by numerous outcrops in Clennonpark Wood. At [SX 8243 6723] shaly lenticular beds near the base of the formation dip at 28°/351°, and by a disused kiln [SX 8258 6725] dark grey limestone with Favosites sp.dips at 15°/013°.

Chercombe Bridge Limestone

Goodstone to Staplehill

Around Bickington, Chercombe Bridge Limestone probably belonging to the Ugbrooke Unit is thrust over Carboniferous rocks and itself overthrust by Gurrington Slate, tuffs and slates with some spilites. Outcrops near the Bickington Thrust show recrystallisation and dolomitisation. Chert breccia north of Telegraph Hill [SX 8082 7338] is probably of tectonic origin. In a disused quarry [SX 7852 7173] at Goodstone coarse-to fine-grained limestone contains orange clay streaks which may be traces of the umber worked formerly at this locality. At Rentor Quarry [SX 7974 7236] well-bedded limestones dipping at 50°/139'and containing tabulate and rugose corals are cut by a fault dipping at 60° eastward.

Hard brown , chert breccias [SX 8077 7332] to [SX 8087 7357] appear to have replaced limestone. Quarries north-east of Telegraph Hill are in limestone that appears to have been thrust over Upper Devonian slate; at [SX 8115 7337] fine-grained partly dolomitised limestones dip at 24°/047° and contain flattened remains of the stromatoporoid Amphipora; at [SX 8106 7350] and [SX 8103 7353] dolomitisation is common. A small patch of recrystallised limestone at [SX 8179 7369] is bounded by thrust faults; so is the limestone south of Higher Staplehill, which dips at 36°/045° [SX 8260 7335].

Halwell to Bradley Manor

A quarry [SX 8235 6794] near Halwell shows 12.2 m of well-bedded limestone, fine to medium grained and in beds 200 mm to 250 mm thick with shale partings. The strata form an anticline, the northern limb dipping at 20°/043° and the southern limb at 25°/141°.

A quarry at West Ogwell Cross [SX 8296 6998] shows fine-grained limestone with Amphipora sp.dipping at 22° south-east; a vertical fault trends south-east and the adjacent limestone shows brecciation and recrystallisation. Coarse-grained variants of the Chercombe Bridge Limestone are uncommon; that at Piebald Copse [SX 8308 7029] is a calcarenite in which recrystallised skeletal grains occur as ghosts in a matrix of sparry calcite that may be recrystallised micrite. Farther east fine-grained limestones dip at 25°/121° [SX 8315 7022] and at 13°/060° [SX 8303 7037]; beds range from 25 mm to 300 mm in thickness and are commonly separated by thin red shales.

Limestones south of Chercombe Bridge contain shaly partings and dendroid stromatoporoids. Thin limestones at [SX 8321 7089] yielded conodonts (loc. 21) possibly belonging to the kockelianus (unrestricted) Zone. A 12-m face in a large disused quarry [SX 8322 7089] shows limestones with large calcite crystals and some soft red shaly limestones dipping at 30°/120°. Stringocephalus burtini from West Hill [SX 8338 7090] indicates a Givetian age. The coral fauna from [SX 8338 7096] includes Acanthophyllum sp., Alveolites sp., Grypophyllum sp. cf. G. convolutum and Heliolites porosus and lies close to the Eifelian–Givetian boundary. Immediately above this horizon, rich Givetian coral faunas including Ceratophyllum sp. cf. C. soetenicum, Cystiphylloides sp., Dendrostella trigemme, Dohmophyllum helianthoides, Grypophyllum?, Favosites sp., Heliophyllum sp., Stringophyllum normale, S. isactis, Thamnophyllum sp.and Thamnopora sp., together with the stromatoporoid Stachyodes caespitosa have been obtained from exposures on West Hill [SX 8339 7091] to [SX 8353 7088]. Trilobites from near Chercombe Bridge include Scutellum (S.) flabelliferum, a characteristic Eifelian form (Selwood, 1965). A late Eifelian, possibly kockelianus (unrestricted) Zone, conodont fauna is represented south of Chercombe Bridge [SX 8321 7089] (loc. 21), close to the base of the formation.

Notable fossiliferous localities also occur in the higher part of the formation on the south bank of the River Lemon [SX 8372 7102] to [SX 8388 7111], where the following corals have been recorded: Acanthophyllum sp., Alveolites sp., Amplexocarinia tortuosa, Coenites sp., Cystiphylloides secundum, Grypophyllum denckmanni, Heliophyllum sp., Stringophyllum buchelense, S. normale, Thamnophyllum germanicum, and Thamnopora sp.The stromatoporoid Stachyodes sp.is also present. Cores from the Emblett Hill Borehole (p. 184), which penetrated the upper part of the formation, showed calcarenites, calcilutites and rare calcirudites, with stromatoporoids and corals.

Slopes rising steeply to 60 m above the River Lemon show many exposures. Limestones dipping at 20°/128° are displaced by steep north-south faults at [SX 8343 7108]. Up to 12.2 m of limestone in a large disused quarry [SX 8350 7107] dip at 26°/150°; interbedded thin red shales contain Amphipora sp., and shell fragments and traces of stromatoporoids are present. Ages close to the Eifelian-Givetian boundary are indicated by Cystiphylloides sp.and Thamnophyllum sp.from [SX 8359 7138], by Acanthophyllum sp., C. secundum, Grypophyllum sp. cf. G. primum and Stringophyllum sp.from [SX 8365 7140] and by Acanthophyllum sp. cf. A. concavum, Caliapora sp. Cystiphylloides sp., G. convolutum and Thamnophyllum sp.from [SX 8413 7126].

Limestone faces up to 30 m high in and around the large disused Broadridge Wood Quarry [SX 8390 7112] (Plate 3) show dips of 25°/112°; red shales occur up to 50 mm thick. Thin sections of a limestone from this quarry revealed a calcilutite composed of very fine-grained calcite with patches of microspar. Recrystallised skeletal fragments were present, some filled with sparry calcite; pellets about 0.1 mm in diameter comprised up to 20 per cent of the rock, which before alteration may have been a biopelmicrite; prominent bands of Amphipora sp.are exposed and Caliapora sp., Dendrostella?, Stringophyllum isactis and Temnophyllum sp.have also been recorded.

In the East Ogwell Unit, the only outcrop of Chercombe Bridge Limestone is in the Lemon Valley immediately east of the Ipplepen Fault, where it occurs as a faulted klippe, thrust over Whiteway Slate and itself overthrust by East Ogwell Limestone. The limestone resembles that of the Denbury Unit and yields locally abundant coelenterate faunas. Middleton (1959) noted rugose corals, recording Columnaria rhenana and Thamnophyllum hoernesi trigemme which suggest a Givetian age. No conodonts are recorded.

Bradley Pit [SX 8442 7088] shows limestone beds up to 1.2 m thick containing abundant Amphipora and dipping at 20°/110°. In Berry's Wood fine-grained limestones with tabulate corals dip at 22°/125° [SX 8449 7103]. Fine-grained limestones with tabulate corals or small stromatoporoids dip at 15° south-south-east [SX 8452 7110] to 20° south-south-east [SX 8466 7112]; in a small pit at [SX 8496 7108] they occur in beds up to 250 mm thick dipping at 53°/200° and contain patches of coarse calcite crystals.

Kingsteignton to Luton

In its main outcrop, south of Whiteway Barton, the Chercombe Bridge Limestone dips south-east and overlies the East Ogwell Limestone in inverted succession; its contact with the Kingsteignton Volcanic Group north of the Teign Estuary is almost entirely faulted. Typical Chercombe Bridge Limestone occurs at the western end of the outcrop, and is well exposed in a strike section in Bickleyball Quarry [SX 8850 7400], dip 20°/142°, as follows:

The limestone in the south face of the quarry is dolomitised, and many outcrops in this area show patchy dolomitisation. The north face contains two large fissures penetrating to the quarry floor and filled by slipped Upper Greensand and flint gravel.

From south of Colladown [SX 884 737] to Ashwell [SX 899 744], where the formation is faulted against the Kingsteignton Volcanic Group, paler grey and coarser limestones are present. They are typical of the basal part of the succession and pass northwards into normal Chercombe Bridge Limestone. A quarry [SX 8791 7347] at St Oswald's shows 3.05 m of bluish grey, well-bedded, fine-grained limestone, dolomitised in the upper part. Bedding dips at 37°/162°. A small quarry [SX 8785 7395] adjacent to the Kingsteignton–Humber road exposes 1.85 m of fine-grained limestone rich in stromatoporoids dipping at 15°/135°. An old quarry [SX 8818 7352] exposes 2.45 m of bluish grey fine-grained limestone, well bedded at the base and passing up into thinly bedded limestones with reddish yellow shale partings. Bedding dips at 20°/153°. Similar limestones containing coral and stromatoporoid debris form crags to the north [SX 8812 7366] and dip at 45°/113°. In Combesend Wood an overgrown quarry [SX 8828 7367] contains 6.1 m of such limestone, dolomitised and fissured and dipping at 40°/180°. North of the wood a large quarry [SX 8825 7375] exposes 20 m of fine-grained limestone, well bedded below and thinner bedded above, the thicker beds locally rich in colonial corals and Amphipora. Dolomitisation is extensive in the southern half of the quarry, and the beds dip at 34°/115°.

South of Colladown Wood a small pit [SX 8834 7316] exposes 1.2 m of thinly bedded and lenticular bluish black fine-grained limestone with shaly partings, dip 32°/094°. Limestone crags [SX 8852 7333] to the north-east contain crinoid debris and horizons of coarse crinoidal limestone and pale grey stromatoporoid limestone. At [SX 8880 7383] 1.85 m of thinly bedded dolomitised limestone with abundant stromatoporoids dip at 28°/153°. An overgrown quarry [SX 8881 7385] to the north exposes 3.05 m of dolomitised limestone with abundant stromatoporoids and bands rich in colonial corals, dip 20°/128°. A quarry [SX 8903 7384] at the west side of Wood House exposes 9.1 m of fine-grained limestone with horizons rich in crinoid debris, dip 34°/134°; conodonts (loc. 22) indicate the varcus Zone. Stanning's Quarry [SX 8960 7425] shows some 6 m of limestone, massive, shelly and sparry and rich in tabulate corals, crinoid debris and brachiopods.

On both sides of the valley through Melland's Copse [SX 890 743] thinly bedded to massive fine-grained limestones dip at 30°/40° south-east. They contain horizons rich in crinoidal debris, and small massive stromatoporoids occur sporadically in the thinner beds. Dolomitisation is common. At the north-east end of Melland's Copse a pit [SX 8910 7429] exposes 3.05 m of massive to well-bedded, partly dolomitised limestone dipping at 30°/149°. Crags at the entrance of Kiln Copse Quarry [SX 8910 7448] show slightly dolomitised stromatoporoid-rich thickly bedded limestone dipping at 40°/135°. A lane section [SX 8950 7460] shows well-bedded south-dipping limestones with much calcite veining, while in an adjacent quarry [SX 8950 7457] 0.9 m of fine-grained massive limestone contain abundant stromatoporoids.

In Rydon Quarry [SX 874 741] about 30 m of bluish black poorly fossiliferous well-bedded fine-grained limestones are exposed in beds up to 1.8 m thick. Scattered horizons of pale grey thickly bedded limestones pass laterally into small bioherms up to 6 m thick composed of stromatoporoid and tabulate corals, especially alveolitid colonies. In the northern part of the quarry muchrecrystallised very fine-grained paler limestones contain small lenses of sparry calcite ('bird's eyes') up to 15 mm across, and large areas of stromatoporoid. Similar variants higher in the sequence point to a tendency for the limestones to become thicker bedded and paler in colour upwards. Bedding dips at 38°/175°.

From Rydon Quarry eastwards through Whitelands [SX 883 743] to Lindridge [SX 897 758], dark and medium grey, well-bedded, fine-grained limestones prevail. Some of the smaller outcrops show massive but probably laterally impersistent horizons up to 6 m thick. Crags on the wooded slopes to the north-east of Rydon Quarry show fine-grained, fairly massive limestones, like those in the northern part of the quarry.

South of Whiteway Barton [SX 8844 7518] a Givetian age is suggested for the Chercombe Bridge Limestone by the occurrence of the Eifelian–Givetian boundary within limestones of the underlying Kingsteignton Volcanic Group and by Ussher's (1913) record of Stringocephalus burtini in the basal part of the formation around Coombsend [SX 881 735].

An overgrown quarry [SX 8887 7499] in King's Wood exposes 6 m of bluish grey, well-bedded, fine-grained limestones, which pass up into East Ogwell Limestone. In the middle of the section 1.2 m of thinly bedded limestone contain small massive stromatoporoids and rare thin-shelled brachiopods. Bedding dips at 20°/124°.

A disused quarry [SX 8939 7522] contains 3.05 m of dolomitised well-bedded to massive limestone dipping at 30°/108°. In a small copse [SX 8915 7565] limestones in beds up to 1.52 m thick dip at 38°/114° and contain stromatoporoids; dolomitisation is extensive in the northern part of the copse. In a quarry [SX 8980 7678] a 4.6-m face in fine-grained limestone shows a dip of 42°/145°.

The junction with the East Ogwell Limestone in the northern part of the main outcrop between Lindridge Hill and Lindridge House is transitional, paler grey thicker bedded and massive Chercombe Bridge Limestone grading up into fine-grained pale grey East Ogwell Limestone rich in stromatoporoid. Elsewhere the latter limestone is a massive, coarse-grained bioclastic rock and the junction is easily recognised, but it is less distinct in some of the small inliers to the north.

Chudleigh to Whiteway House

The limestones of the Chudleigh escarpment are separated by the Bickington Thrust from the underlying Kate Brook Slate. They are overlain by the Luxton Nodular Limestone in the south-west, by the Winstow Chert in the central part of the escarpment, and unconformably by Permian breccias in the north-east. From Kiln Wood [SX 861 779] to the Riding Parks [SX 867 786], and including the virtually isolated klippe that forms Chudleigh Rocks [SX 864 787], the limestones are mostly dark grey, well bedded and very fine grained, but massive limestones occur in places.

In the east side of Kiln Wood Quarry [SX 8611 7793] the Kiln Wood beds of the Luxton Nodular Limestone overlie 1.52 m of massive limestones containing conodonts which range from upper asymmetricus Zone to lowermost gigas Zone (Tucker and van Straaten, 1970). The basal part of the overlying Luxton Nodular Limestone contains conodonts ranging from upper asymmetricus to upper gigas and goniatites of high cordatum age (House and Butcher, 1973). Thus the top of the formation here is not younger than mid Frasnian.

In Lawelldown Wood a small quarry [SX 8642 7827] exposes 3.05 m of thinly bedded limestone rich in stromatoporoids, dipping at 30°/060°. Another quarry and surrounding crags [SX 8649 7845] show locally lenticular fine-grained limestones, with rare stromatoporoids, dipping at 10°/101°.

At Palace Quarry [SX 8675 7870] about 12 m of fine-grained limestone in beds up to 1.52 m thick are exposed. Scattered rugose corals and, at the top of the quarry, thinner beds rich in Amphipora are common (Scrutton, 1969). A shallow syncline occupies the main face, which is cut by numerous faults. Dolomitisation is widespread in the south-western part of the quarry.

In Riding Parks [SX 8688 7875] limestone crags yield scattered rugose and tabulate corals, stromatoporoids and a few brachiopods. Scrutton (1969) recorded Alveolites suborbicularis, Amphipora sp. Disphyllum sp.aff. D. caespitosum, Macgeea sp. nov., Thamnophyllum sp. cf. T. kozlowskii and Thamnophyllum boloniensis, of Frasnian age, from 4.6 to 7.6 m below the top of the Chercombe Bridge Limestone. At [SX 8698 7875], immediately below outcrops of Luxton Nodular Limestone, thickly bedded limestones dip at 20°/142°.

Chudleigh Rocks, a virtually isolated block of Chercombe Bridge Limestone, show low dips to east or south-east. On the north side large quarries [SX 8641 7872] to [SX 8657 7874] expose a strike section of fine-grained limestones. The beds range from 0.15 m to 4.6 m in thickness, and the thicker beds contain a few stromatoporoids. Similar rocks crop out in the south-facing cliffs above the Kate Brook Gorge. At [SX 8637 7875] a bed rich in rolled massive stromatoporoids up to 0.45 m in diameter is overlain by stromatoporoid and coral debris associated with Amphipora.

In old quarries north-west of Castle Dyke massive and thickly bedded fine-grained limestones dip at 10°/210° [SX 8727 7886] and at 20°/150° [SX 8728 7892]. Similar limestones at [SX 8716 7890] contain abundant stromatoporoids including Amphipora and rare rugose corals about 19 m below the top of the formation; Scrutton (1969) recorded Amphipora rudis, Hermatostroma episcopale, Marisastrum sp.aft M. marmini, Stachyodes sp.aff. S. caespitosa, Stromatopora sp Thamnopora sp.and Trupetostroma sp., suggesting an age close to the Givetian–Frasnian boundary.

In Grealy Quarry [SX 8804 7922] some 9.15 m of massive limestone are heavily veined by calcite. At the northern end of Burrows Wood an old quarry [SX 8802 7948] exposes 6 m of limestone in beds up to 1.83 m thick dipping at 15°/1149; at the southern end [SX 8812 7932] 7.6 m of massive partly dolomitised limestone contain scattered colonial corals. To the south [SX 8824 7919] coarse limestone dips at 28°/174° and at [SX 8830 7943] dolomitised fine-grained limestone contains abundant stromatoporoid and horizons rich in crinoid debris.

The orchard behind Waddon Barton [SX 8845 7960] contains crags of massive, fine-grained limestone which locally [SX 8860 7965] contains abundant stromatoporoids and scattered rugose corals. The quarry behind Ridge House [SX 8894 8018] contains 9.15 m of massive limestone mostly composed of stromatoporoids. In the woods north of Amberley [SX 8865 8025] 9.15 m of fine-grained limestones in beds less than 0.9 m thick dip at 8°/080°. Immediately to the north the limestones contain vertical fissures filled with limestone rubble and orange silty sand.

Whidborne's (1889–1907) brachiopods and gastropods from Chudleigh probably came from massive bluish black limestone at the north-eastern end of the escarpment, where Ussher (1913) and Anniss (1933) listed gastropods from Lower Upcott Quarry [SX 8860 8045], Kerswell Quarry [SX 8846 8071] and Harcombe Quarry [SX 8839 8190]. Anniss (1933) considered that the limestones were probably early Frasnian, but House (1963) identified the Givetian Stringocephalus sp.in Holman's Wood Quarry [SX 8830 8116]. The face in this quarry shows 9.15 m of massive rubbly fine-grained limestones. They are poorly fossiliferous except in the top north-west corner of the quarry, where there are horizons rich in stromatoporoids with some tabulate and rugose corals. Scrutton (1969) recorded Alaiophyllum?, Amphipora rudis, Caliapora battersbyi, Grypophyllum denckmanni, Heliolites porosus, Scoliopora sp., Stachyodes caespitosa, S. radiata and Trupetostroma sp.Dineley and Rhodes (1956) obtained conodonts from the base of the upper level in the south-east corner of the quarry. Lower Upcott Quarry shows 7.6 m of horizontal thinly bedded fine-grained limestone, and Kerswell Quarry about 9 m of intensely fissured massive limestone. In Harcombe Quarry massive calcite-veined limestone 1.5 m thick is unconformably overlain by New Red Sandstone breccias.

A quarry at [SX 8848 8115] shows about 9 m of massive limestones consisting essentially of stromatoporoids, with patches of coarser limestone rich in crinoid debris.

Foxley Tuff

Green fine-grained tuff in an unsurfaced lane [SX 8350 6950] dips at 30°/118°. In the roadside [SX 8382 6998] opposite East Ogwell Post Office, cleavage in buff fine-grained tuff dips at 25°/120°. Calcareous lapilli tuff in the roadside [SX 8393 7002] dips at 30° eastward; a thin section of a specimen from a nearby building site consisted of clasts up to 20 mm in diameter, composed of spherulitic quartz and calcite in a fine indeterminate brown matrix, set in calcite cement. Along the lane from East Ogwell to Chercombe Bridge weathered coarse tuff dips at 30°/138° [SX 8352 7041], and coarse lithic tuff in Mill Copse [SX 8380 7087] dips at 39°/142°. West of Ogwell Mill tuffaceous rocks in a trench ranged from laminated calcareous tuff dipping at 32°/123° [SX 8292 7095] to lapilli tuff dipping at 45°/097° [SX 8397 7093]; some calcareous beds contained tabulate and rugose corals and rare brachiopods. A thin section of grey tuffaceous limestone from here showed round volcanic clasts up to 20 mm in diameter composed of plagioclase microlites in a groundmass of chlorite and calcite. Albite is present in laths up to 1.5 mm long. Calcite occurs as irregular patches throughout. The tuff is overlain by East Ogwell Limestone. On the north side of the River Lemon fine tuff in the roadside [SX 8420 7106] dips at 30°/163°.

Calcareous tuff at [SX 8509 6791] south of Abbotskerswell dips at 30° north-east. Shaly tuff is exposed beneath limestone at [SX 8513 6806].

East Ogwell Limestone

In the Denbury Unit most exposures of East Ogwell Limestone south of the Torbryan Fault show easterly dipping grey to pink limestones, commonly dolomitised. From Torbryan northwards to Tornewton [SX 816 679] the limestone is predominantly fine grained, relatively unfossiliferous and dips at around 20° eastward. East of Halwell greyish pink limestone is commonly dolomitised near faults, particularly between Stallage Common and Rydon Hill. The northern boundary of the limestones is everywhere a thrust, the formation having overridden rocks of the Ugbrooke–East Ogwell Succession. South of Denbury the underlying rocks are Denbury Crinoidal Limestone; to the east they are Foxley Tuff. Immediately north of East Ogwell the East Ogwell Limestone succeeds the Foxley Tuff in a narrow east-dipping belt, which is bounded to the east by the Ipplepen Fault. The limestone–tuff junction is faulted at surface. North-west of East Ogwell massive limestone overlying tuff is richer in bioclastic debris than at most other localities and dolomitisation is common, particularly close to faults.

Torbryan to Emblett Hill

In a small disused quarry [SX 8214 6671] south-east of Torbryan church limestones dip at 45°/082° and contain scattered rugose corals and stromatoporoids. Massive limestones with thin shales are common along wooded slopes north of Torbryan, generally dipping at 20° to 30° between north and north-west. Locally they contain abundant rugose and tabulate corals with a few stromatoporoids. Acanthophyllum concavum, Caliopora battersbyi, Grypophyllum denckmanni, Heliophyllum sp.and Stringophyllum sp.occur at [SX 8192 6691]; Acanthophyllum sp. cf. A. heterophyllum and Alveolites sp.at [SX 8197 6723]; and G. denckmanni and Stringophyllum sp. cf. S. normale at [SX 8188 6686]. At Torbryan caves [SX 8173 6738] the beds dip at 20°/020°.

In Pulsford Quarry [SX 8125 6790] strata are flat or dip at 5° to 10° in various directions; sub-horizontal fractures occur and vertical joints strike at 012°. In Pulsfordhill Wood [SX 8082 6799] a fine-grained calcirudite contains much red shale; crinoid debris, comprising only 50 per cent of the rock, lies in a matrix of calcite grains less than 1.5 m in diameter, and rare quartz grains. Closely spaced stylolites truncate skeletal grains, and shale is concentrated along them. Degrading recrystallisation has occurred in some columnals, which may also contain patches of dolomite. This rock, unusual within the East Ogwell Limestone, is lithologically similar to the Denbury Crinoidal Limestone, and may indicate some lateral equivalence of the two formations.

East of Halwell Farm limestones with red shaly partings crop out sporadically; dolomitisation in the neighbourhood of the farm is associated with faulting. Fine-grained limestones south-east of Denbury are commonly rust-coloured and dolomitised. In Clennonpark Wood limestones with tabulate corals dip at 16° northward [SX 8249 6732]; similar rocks dip at 20°/126° at [SX 8308 6747]. In an old quarry [SX 8332 6759] near Clennon medium-grained limestone containing Thamnopora is overlain by reddish purple shaly limestone dipping at 35°/080°. On the north side of Stallage Common outcrops are commonly rust-coloured and dolomitised. Limestones to the east [SX 8422 6785] contain tabulate corals, dip at 26°/032°, and are cut by joints dipping at 70°/212°. Near Blackrock Copse porous dolomitised limestones dip at 12°/063° [SX 8398 6795]. On Rydon Hill limestones dip at 20° eastward [SX 8393 6855] and rust-coloured dolomitised limestones dip at 10° northward [SX 8366 6854]. Near Rydon Cross they contain tabulate corals and dip at 30°/193° [SX 8406 6895].

Near East Ogwell coarse-grained limestones with corals and red shaly partings dip at 35°/060° [SX 8383 7028]. South-west of Ogwell Mill [SX 8377 7059] to [SX 8401 7085] a late Givetian coral fauna has been recorded immediately above the Foxley Tuff, including Alveolites sp., Disphyllum caespitosum, Hexagonaria sp., Stachyodes sp., Thamnophyllum sp.and Thamnopora sp.At [SX 8398 7084], trilobite fragments obtained from limestones dipping at 28° south-east include Scutellum (S.) costatum whidbornei? of late Givetian age. In a small quarry [SX 8397 7062] south-west of Emblett Hill, massive limestones containing the corals Acanthophyllum concavum, Alveolites sp., Cyathophyllum (C.) sp., Hexagonaria cf. marmini, Macgeea sp., Metriophyllum bouchardi and Thamnopora sp.probably lie near the Givetian–Frasnian boundary. Numerous joints cause the rock to break into large blocks; bedding is difficult to discern but may dip at about 15° eastward.

The Emblett Hill Borehole (p. 184) penetrated 21.92 m of East Ogwell Limestone. Coarse calcarenites and calcirudites, with abundant skeletal fragments and with layers up to 75 mm thick formed by stromatoporoids or tabulate coral colonies, overlay fine calcarenites and calcilutites which contained skeletal remains well preserved in limy mud, together with stromatoporoidal and algal deposits. The few coarse-grained limestones among these lower beds contained abundant well-sorted coralline skeletal grains in a matrix of sparry calcite. The upper beds (loc. 23) yielded an asymmetricus Zone fauna. Approximately the same horizon is exposed west of East Ogwell, where the fauna includes Alveolites sp., H. cf. marmini, Tabulophyllum sp.and Thamnopora sp.at [SX 8387 7035], and Alveolites sp., D. caespitosum, Macgeea sp., Peneckiella? and Thamnophyllum sp.at [SX 8386 7020]. If the record of Peneckiella at the latter locality is confirmed a Frasnian age is indicated.

Ipplepen–Wolborough–Kingskerswell

The East Ogwell Limestone of the East Ogwell Unit occurs east of the Ipplepen Fault in isolated masses which are interpreted as the remains of a thrust sheet. Near the thrust and other faults the limestone is commonly recrystallised and dolomitised. In most places the formation rests directly on Upper Devonian slates of the Abbotskerswell Unit. However, south-east of Emblett Hill it has overriden a klippe of Chercombe Bridge Limestone included in the East Ogwell Unit. It overlies Ugbrooke Sandstone at [SX 850 706] and tuffs immediately west of Wolborough [SX 847 701], east of Two Mile Oak Cross [SX 851 681]; [SX 852 678] and south-east of Dainton Hill [SX 866 666]. This contact with tuffs is probably tectonic, but south of the district it appears to be conformable. The top of the formation is not exposed but high horizons are present in Ransley Quarry and Conitor Quarry [SX 8497 6952], where representatives of the Luxton Nodular Limestone are infaulted. In East Ogwell, immediately above the basal thrust and stratigraphically below the beds of Ransley Quarry, shaly and crystalline limestones crop out which are lithologically similar to the Denbury Crinoidal Limestone in the Denbury Unit and to the crinoidal limestone at the base of the East Ogwell Limestone near Pulsford in the Torbryan Succession.

The two outliers of limestone south of Stallage Common appear to be klippen, underlain by Whiteway Slate. A disused kiln [SX 8344 6743] adjoins red shaly limestones dipping at 25°/160° and faulted against thickly bedded limestones inclined at 15°/125°; the fault plane strikes 110° and dips at 85° southward. Limestones between Denbury Cross [SX 8355 6737] and Dornafield Cross [SX 8395 6735] show much recrystallisation, probably associated with thrust faulting. A medium-grained calcarenite from a disused quarry in Gotemhill Wood [SX 8560 6755] shows alteration in thin section, but skeletal grains are not closely packed and the fine-grained matrix is probably original; dolomite rhombs are common and ferroan calcite forms thin veins; conodonts (loc. 30) indicates the varcus Zone.

At the southern end of Stoneycombe Quarry [SX 8620 6685] mainly massive fine-grained limestone with stromatoporoids and rugose corals dips at 70°/210°. Corals include Stringophyllum (S.) sp.and T. caespitosum caespitosum, of Givetian age. About 183 m to the north folded thinly bedded coarse crinoidal limestones with interbedded shales dip at 30° to 60° northward [SX 8620 6704] and have yielded conodonts (loc. 32) of late ensensis Zone age. This early Givetian fauna is noteworthy in including the relatively rare forms Polygnathus latus, P. linguiformis cooperi (probably derived) and Tortodus variabilis. Quarry faces north of the railway [SX 8614.6743] show massive and thickly bedded limestones with tabulate corals dipping at 72°/250° and cut by widely spaced joints inclined at 50°/006°. The northernmost face is cut by several sub-horizontal fractures. The west face of the quarry [SX 8590 6740] shows limestone with stromatoporoids and tabulate corals.

A coarse calcarenite by the railway [SX 8642 6750] is composed chiefly of recrystallised crinoid debris and contains some secondary quartz. Kerswell Down Quarries [SX 8718 6760] contain massive coarse limestone with traces of stromatoporoids. Conodonts from [SX 8735 6768] (loc. 33) indicate that the limestones belong to the varcus Zone. Near Kingskerswell Station [SX 8770 6770] the limestone is locally yellow and dolomitised.

Thinly bedded and shaly limestone containing conodonts (loc. 31) dips at 20° northward in Slade Lane [SX 8585 6835], near Abbotskerswell; the varcus Zone is probably indicated. At [SX 8520 6860] limestone beds up to 0.6 m thick with shaly intercalations form a tight fold with horizontal axis and have yielded conodonts (loc. 28) of probable varcus Zone age. Similar rocks to the north-east contain a few rugose corals and dip at 18°/032° [SX 8530 6877].

Near Two Mile Oak Cross limestone at [SX 8432 6841] dips at 30°/100° and contains tabulate corals. Dolomitised limestones dip at 40° east [SX 8432 6832] and 38°/075° [SX 8468 6857]. Limestone with tabulate corals at [SX 8520 6809] dips at 16°/111°. These outliers may be remnants of a thrust block.

East Ogwell Limestone was penetrated from 199.64 m to 209.34 m at the bottom of the Rydon Ball Farm Borehole (p. 183). The limestones were fine calcarenites and calcilutites with sporadic skeletal grains, the top 4.11 m being extensively dolomitised and cavernous. Stromatoporoidal or alveolitid layers were common and dolomite was finely disseminated. Conodonts indicated a Frasnian age.

Several small patches of thrust and faulted limestone occur north-west of Abbotskerswell: at [SX 8518 6971] shaly limestones with conodonts (loc. 27) are probably of varcus Zone age. In and around the disused Conitor Quarry [SX 849 696] occur thickly bedded and massive limestones with few organic constituents.

An old quarry at East Ogwell [SX 8413 7002] shows 9 m of crinoidal limestone in beds up to 180 mm thick dipping at 10° eastward. The rocks are locally dolomitised and stained yellow. Conodonts (loc. 24) indicate the late Givetian, probably the varcus Zone. Similar rocks in the roadside [SX 8407 7008] yielded Acanthophyllum concavum and in an old pit [SX 8411 7012] yielded Sociophyllum sp. cf. S. sociale, Alveolites sp., Favosites sp.and Thamnopora sp.Farther east coarse-grained limestone at the roadside [SX 8442 7005] contains corals and conodonts and dips at 70°/135°.

Massive limestone from the road cutting [SX 8441 7006] near Ransley Quarry is seen in thin section to be a very fine calcarenite, composed of irregular calcite grains about 0.15 mm across and recrystallised skeletal grains in a microspar matrix; the matrix probably formed by the alteration of limy mud and the rock may originally have been a biomicrite. In Ransley Quarry [SX 8443 7018] a 14-m face shows poorly defined bedding inclined at 50°/132°. The limestone contains red muddy layers up to 50 mm thick. The quarry and the road cutting have yielded (Scrutton, 1968) Frechastraea pentagona pentagona, F. pentagona minima, F. carinata, F. goldfussi, F. bowerbanki, Haplothecia ogwellensis, Phillipsastrea hennahi ussheri, P. ananas and P. rozkowskae. Conodonts from a composite sample including pink micritic limestone (loc. 25) indicated a Frasnian age within the range from the base of the triangularis Zone to lower gigas Zone. Shannon (1921) obtained Agoniatites sp.and Manticoceras cf. cordatum from the quarry and House (1963) concluded that these ammonoids indicated that the Givetian and Frasnian were represented. The pink micritic limestone is believed to represent the Luxton Nodular Limestone. The East Ogwell Limestone may not therefore be significantly younger at this locality than elsewhere in the district.

Thickly bedded and massive limestones crop out sporadically northwards from Ransley Quarry; at [SX 8448 7036] they dip at 45°/115° and yield conodonts (loc. 26) of lower gigas age. Disphyllum caespitosum has been obtained from [SX 8435 7032] north-west of Ransley Quarry. Immediately east of Ransley a small mass of fine-grained calcirudite near Canada Farm [SX 8478 7010] has been thrust over tuff. Some of the skeletal grains are deformed and all are closely packed; the matrix may have formed as sparry calcite and been altered by degrading recrystallisation. Similarly in a fine-grained calcirudite from Lang's Copse [SX 8496 7060] crinoid fragments, which constitute most of the skeletal grains, have been altered to micrite.

Coarse, locally dolomitised, limestones with red shaly layers exposed in Wolborough Quarry [SX 8522 7047] are massive and fractured by widely spaced joints. Cavities are filled by sparry calcite. The rocks yielded (Scrutton, 1968) Billingsastraea? battersbyi, Phillipsastraea devoniensis and P. hennahi hennahi. Stringophyllum sp.and Thamnophyllum caespitosum and the presence of several maenioceratids enabled House (1963) to identify the molarium Zone.

Conodonts (loc. 29) obtained from the foot of the main face [SX 8524 7041] include Ozarkodina brevis, and probably indicate the varcus Zone. In temporary exposures [SX 8513 7045] for a drain west of the main quarry the limestones were seen to be thrust over purple slates.

Barton area

Barton Quarry contains coarse-grained dolomitised bioclastic limestones which, at the northern end of the quarry [SX 9125 6720], dip at 28°/155°. The limestones have yielded a late Givetian coral fauna including Acanthophyllum concavum, Haplothecia pengellyi, Macgeea sp., Phillipsastrea hennahi hennahi, Sociophyllum sp., Alveolites sp., Syringopora sp.and Thamnopora sp., and the goniatites Tornoceras? and Wedekindella brilonense (House, 1963). The coarse texture of limestones in the northern part of Lummaton Quarry is attributable to the abundance of organic debris. Stromatoporoids and tabulate corals are common [SX 9125 6662] and dolomitisation is widespread.

Teign Estuary to Chudleigh

In the Kingsteignton Unit, boreholes have proved probable East Ogwell Limestone beneath terrace gravels. A borehole [SX 9084 7328] at Great Park Nurseries, near Bishopsteignton, is recorded as having passed through 6.1 m of terrace gravel and 22.9 m of limestone, into 16.8 m of red slate. Rhodes and Dineley (1957), possibly referring to the same borehole, noted that massive limestone passed down into red limestone and that the lower half of the sequence consisted of red slates with thin limestone and nodules. Probably the massive limestone is either East Ogwell Limestone or Chercombe Bridge Limestone and the underlying thin limestones and slates belong to the Luxton Nodular Limestone, the succession being inverted. However, late Givetian conodonts have been obtained from both formations, whereas outcrops of the Luxton Nodular Limestone are nowhere known to be older than mid Frasnian.

In the Ugbrooke Unit between the Teign Estuary and Chudleigh, coarse limestone reminiscent of that in the Denbury Unit occurs together with fine-grained types. Stromatoporoids, corals, crinoid debris, brachiopods and rare trilobites are present. South of Whiteway Barton the formation lies in an inverted succession and is up to 75 m thick. To the north-west it is faulted against Whiteway Slate. South-east of the Whiteway Barton valley the limestones are dominantly fine to medium grained, pale grey, massive and commonly rich in stromatoporoids. On the north side of the valley, beds near the top of the formation comprise pale grey or pinkish grey, coarse, massive limestones rich in bioclastic debris, with some finer-grained stromatoporoid-rich limestones.

A quarry [SX 8780 7460] south-west of Whiteway Barton shows massive and thickly bedded limestones commonly rich in stromatoporoids with some coarser limestones high in the quarry containing crinoid debris and brachiopods; bedding dips at 35°/120°; Ussher (1913) named this locality Buckley Wood Quarry and gave a list of stromatoporoids. Immediately south of Torhill Cottages [SX 8777 7473] 2.45 m of thickly bedded and massive medium- to coarse-grained limestones dip at 40°/130°; stromatoporoids, rare rugose corals, crinoid debris and brachiopods are present together with a conodont fauna (loc. 34) of probable early Frasnian age. Massive limestone in a quarry [SX 8784 7473] to the east yielded only one bar conodont fragment; limestones on the eastern side of this quarry dip at 50°/127° and are completely dolomitised.

Coarse massive East Ogwell Limestone in a quarry [SX 8820 7500] on the northern side of the valley south-west of Whiteway Barton is faulted against Luxton Nodular Limestone; Ussher (1913) collected Scutellum (S.) sp.The massive limestone also yields rare rugose corals, crinoid debris and brachiopods, and a probable P. asymmetricus Zone conodont fauna (loc. 35). In Strongs Cover [SX 8880 7519] limestones are faulted against thinly bedded limestones of the Luxton Nodular Limestone and yielded conodonts (loc. 36) including Polygnathus webbi.

Near Ideford the formation is overlain by the Chercombe Bridge Limestone in inverted succession, and faulted against Winstow Chert. It is thicker than it is to the south and the dominant lithology is pale grey to pinkish grey, coarse, massive bioclastic limestone rich in corals and stromatoporoids with some brachiopods. Dolomitisation is locally extensive. An old quarry [SX 8942 7687] near Ideford, partly obscured by tufa, shows thickly bedded coarse limestone dipping at 38°/119°. Scattered tabulate corals and brachiopods occur, and much of the limestone is dolomitised. Another old quarry [SX 8928 7699] exposes 6 m of coarse sparry limestone, generally massive but apparently dipping at 48°/011°. Dolomitisation is extensive and much of the face is obscured by tufa deposits.

At Well [SX 880 770] the limestones form part of an upfaulted block which extends north into Ugbrooke Park. They are faulted against Luxton Nodular Limestone to the south-west and Winstow Chert to the south-east, and to the north their outcrop is largely obscured by sand and gravel of unknown age. The rocks are pale grey to pink, massive fine-grained limestones and are apparently near the top of the formation, for both Ussher (1913) and Aniss (1933) recorded thinly bedded red 'goniatite limestone' (Luxton Nodular Limestone) above the massive limestone in an orchard immediately behind an old quarry [SX 8800 7702], a section no longer visible.

Within Ugbrooke Park the formation is faulted against Ugbrooke Sandstone; bedding is rarely seen but seems to dip eastwards. Crags [SX 8720 7804] at the western end of Higher Water Lake show massive fine-grained limestone with tabulate corals. Partly dolomitised medium- to coarse-grained limestones are also exposed [SX 8720 7799]; [SX 8739 7765]; [SX 8739 7759].

Lower Dunscombe Quarry [SX 8855 7907] exposes 4.55 m of massive and thickly bedded coarse limestones that dip at 19°/240°. These limestones are part of an upfaulted block of East Ogwell Limestone on the dip slope of the Chudleigh escarpment. At the top of the quarry massive limestones pass up into thinly bedded limestones of the Lower Dunscombe Goniatite bed (p. 55). The massive limestones contain crinoid debris and brachiopods. Scrutton (1969) noted that corals were not common, and recorded Alveolites sp., Catactotoechus sp. nov., Ceratophyllum? and Macgeea sp.aff. M. recta. He also identified broken fenestellid networks, Styliolina sp., fragments of echinoderm spines and trilobites. Conodont faunas were listed by Dineley and Rhodes (1956) and Tucker and van Straaten (1970), and range from middle asymmetricus Zone to lowermost gigas Zone, that is up into mid Frasnian, though south of Whiteway Barton no fossils younger than early Frasnian have been recorded.

Upper Devonian

Hyner Shale

Small field exposures have yielded: Petraia sp., Guerichia venusta, Sanguinolites?, indeterminate ammonoids, Richterina (R.) striatula [SX 8415 8137]; Maternella hemisphaerica, Richterina (R.) costata, R. (R.) striatula [SX 8418 8132]; and R. (R.) latior [SX 8424 8123]. A small pit [SX 8410 8168] contains 1.5 m of calcareous shales, mudstones and siltstones, cleavage dip 80°/135°, bedding dip 50°/140°. Calcareous silty mudstone 0.3 m above the base of these beds yielded the following fauna indicative of toVI: Chonetes sp., G. venusta, Sanguinolites? ellipticus, Archegonus (Waribole) cf. warsteinensis and Cyrtosymbole (Calybole) cf. ussheri. In the roadside at [SX 8418 8176] 7 m of shales and mudstones, with lenticular dark blue mottling aligned parallel to bedding, dip at 50°/140° and show cleavage dipping at 80°/135°. In a copse north of Hyner Bridge [SX 8372 8179] 27.4m of shales dipping at 70°/135° contain scattered, brown, probably decalcified, lenticles 25 to 50 mm long aligned parallel to the bedding.

In Hyner Bottom massive quartz-biotite-cordierite-hornfels is exposed within the aureole at Great Rock Mine [SX 8280 8154]. It passes eastward into calc-flintas which form massive crags cut by joints dipping at 86°/278° and 86°/181° and contain sporadic siliceous nodules. East of the drying sheds shales and siltstones with small siliceous lenticles [SX 8355 8168] dip at 80°/145°.

A section along the B3193 road [SX 8410 8197] to [SX 8402 8221] exposes 244 m of shales with small calcareous nodules aligned parallel to the bedding. These beds lie in the core of the Ranscombe Anticline, dip at 70°/170° and show minor easterly plunging flexures. Teign Lane [SX 8430 8225] to [SX 8460 8229] contains a strike section in shales and mudstones also in the core of the Ranscombe Anticline. The strata are vertical, strike east–west [SX 8458 8228], and pass into grey micaceous shales of the overlying Trusham Shale [SX 8476 8227].

Massive quartz-biotite-cordierite-hornfels is exposed [SX 8259 8280] [SX 8275 8260] along a water Teat from Shuttamoor Mine. It carries small veins and stringers of micaceous hematite and passes eastward into the calc-flintas of Canonteign waterfall [SX 8320 8244]. Small siliceous nodules are visible on some weathered surfaces. In a field near Lower Ashton [SX 8478 8400] 6 m of shales showing a dark blue mottling dip at 80°/146°; joints dip at 88°/243° and 60°/169°.

Exposures alongside the B3193 road [SX 8405 8482] to [SX 8408 8487] show the following succession, which is taken as the type section for the Hyner Shale:

Rora Slate

Inliers north-west of Bickington have been mapped by reference to debris of slates yielding crinoid debris, G. venusta, M. dichotoma and Richterina (R.) costata (hemisphaerica-dichotoma Zone) [SX 7974 7300] to [SX 7972 7289], and of decalcified siltstones and slates, the latter yielding M. dichotoma and R. (R.) costata (hemisphaerica-dichotoma Zone) [SX 7966 7314] to [SX 7963 7307].

A road section [SX 7953 7347] south-east of Ramshorn Down shows decalcified micaceous silty mudstones whose bedding planes are commonly covered with juvenile brachiopods and bivalves. Juvenile globular ammonoids, Archegonus (Waribole) sp.and Typhloproetus sp.are also present and the mudstones are referable to toV–VI. A few metres to the south bioturbated laminated silty mudstones with solitary corals and bivalves dip steeply and are cut by a horizontal fracture cleavage.

A track [SX 7924 7378] on Ramshorn Down cuts through weathered decalcified siltstones with black mudstone laminae yielding S.? cf ellipticus and small brachiopods. Bedding dips at 80°/300° and a fracture cleavage is sub-horizontal. To the east-south-east [SX 7933 7374] laminated micaceous calcareous siltstones exhibit a fracture cleavage dipping at 37°/l67° that gives rise to a prominent bedding/cleavage lineation. Other exposures hereabouts are as follows: [SX 7945 7365] to [SX 7952 7365] micaceous laminated silty mudstones show a steep dip and a cleavage inclined at 18°/132°; [SX 7935 7379] to [SX 7945 7384] slates and siliceous slates show cleavage and bedding dipping at 12°/153° and yield crinoid debris, small brachiopods, fragmentary bivalves, including Paracyclas?, Kosmoclymenia?, Archegonus (Waribole) sp., P. cf. granulatus, Chilobolbina aff. rhenana, M. dichotoma, M. hemisphaerica, R. (R.) costata, R. (R.) striatula, and a conodont mould; the ostracod and trilobite fauna is referable to toV–VI. A stream section [SX 7963 7374] in slates with cleavage dipping at 30°/129° yielded sparse fragmentary bivalves and trilobites, M. hemisphaerica, R. ( R.) costata and R. (R.) striatula (hemisphaerica-dichotoma Zone).

Slates in a field [SX 8003 7422] yielded an indeterminate ammonoid and R. (R.) costata. A track [SX 8010 7438] to [SX 8013 7431] leads north from Rora through silty micaceous slates whose cleavage and bedding dip at 30°/095°; the fauna including juvenile bivalves (Paracyclas? and G. venusta?), lingulid brachiopods, rare flattened ammonoids, fragments of proetids, M. dichotoma, M. hemisphaerica and R. (R.) striatula, is referable to the hemisphaerica-dichotoma Zone. Slates in a gateway [SX 8013 7436] yielded Cladochonus?, orthoid brachiopods and flattened ammonoids including Kosmoclymenia sp.; others in a field [SX 8012 7433] contained crinoid debris, orthoid and lingulid brachiopods and phacopids.

A track [SX 8015 7424] south of Rora exposes slates with rare decalcified nodules up to 15 mm across; cleavage and bedding dip at 25°/093°. The slates contain rare Cladochonus?, crinoid debris, juvenile brachiopods, S.? ellipticus, a flattened ammonoid, Archegonus (Waribole) sp., Dianops cf. griffithides, Entomoprimitia (Reptiprimitia) rabieni, M. dichotoma, M. hemisphaerica, Richterina (Fossirichterina) semen, R. (R.) striatula and conodont fragments referable to the hemisphaerica-dichotoma Zone. A temporary trench [SX 8024 7437] to [SX 8030 7436] east of Rora cut through silty slates with Paracyclas?, S.? ellipticus, orthoid and lingulid brachiopods, an indeterminate flattened ammonoid, Dianops sp., M. dichotoma, M. hemisphaerica, R. (R.) costata, R. (R.) striatula and R. (F.) semen. At the east end of the trench horizons of streaky greyish green mudstone are rich in Cladochonus?, crinoid debris, and R. (R.) striatula; cleavage dips at 60°/120°.

A roadside quarry [SX 7907 7388] exposes 3 m of bioturbated slaty mudstones. Black stringers of manganese oxide lie parallel to the slaty cleavage that dips at 10°/194°. A sparse hemisphaerica–dichotoma Zone fauna includes solitary corals, Cladochonus?, juvenile brachiopods and bivalves, phacopids, M. hemisphaerica, R. (R.) costata and R. (F.) semen. Weathered silty slates [SX 7895 7408] with horizons of reddish purple slate yielded rare crinoid debris, an orthid brachiopod, indeterminate juvenile bivalves, R. (R.) cf. striatula and R. (F.) semen; a rough cleavage dips at 40°/170°. Laminated calcareous siltstones dip at 28°/052° in a stream [SX 7927 7417]. The following hedge-bank exposures occur: at [SX 7909 7440] slates with cleavage and bedding dipping at 25°/135° yield crinoid debris, a bellerophontid gastropod, M. dichotoma, R. (R.) costata and R. (R.) striatula, referable to the hemisphaericadichotoma Zone; at [SX 7918 7470] slates show cleavage and bedding dipping at 25°/079°; nearby [SX 7924 7481] slates with abundant small decalcified lenses show cleavage and bedding dipping at 13°/057° and yield fragmentary bivalves, juvenile orthoid brachiopods, Parawocklumeria?, Dianops sp., R. (R.) costata, a single specimen of R. (R.) latior and R. (R.) striatula, probably referable to upper toVI and the latior Zone.

South Rora Down Quarry [SX 7957 7425] contains the following section:

All the beds belong to the hemisphaerica–dichotoma Zone. Bedding and cleavage dip at 30°/135°. Scree above the quarry face has yielded rare small lingulid brachiopods, juvenile bivalves, M. dichotoma, M. hemisphaerica, R. (R.) costata, R. (R.) striatula, R. (R.) sp. nov., and R. (F.) semen; track exposures 18 m above the quarry contain M. dichotoma, R. (R.) costata and R. (R.) striatula, and debris alongside has yielded in addition Neochilina cf. parvula, rare flattened ammonoids and fragmentary conodonts. The ostracod fauna again belongs to the hemisphaerica–dichotoma Zone.

Slates in a stream section [SX 7942 7436] to [SX 7949 7472] commonly show cleavage dipping at 19°–28°/139°. At [SX 7944 7455] fragmentary trilobites, a single specimen of M. dichotoma, R. (R.) costata and R. (R.) cf. striatula (hemisphaerica–dichotoma Zone) were obtained; at [SX 7949 7469] slightly bioturbated laminated calcareous siltstones exhibit a vertical fracture cleavage trending 258°; at [SX 7949 7472] slates show cleavage inclined at 70°/145°. In a track [SX 7952 7472] on the western slopes of Mount Ararat slates contain a 65-mm band of rusty-weathered fine-grained limestone dipping at 60°/070°. Bristow (1962) collected Kosmoclymenia sp.and Phacops granulatus of toVβ–VI from this limestone; the slates yield M. dichotoma, R. (R.) costata and R. (R.) striatula. Crags nearby [SX 7953 7470] on the western slopes of Mount Ararat show slates with decalcified nodules up to 75 mm across and cleavage dipping at 60° south-east; Bristow (1962) collected Kosmoclymenia undulata of toVβ–VI. In a track [SX 7961 7461] to [SX 7979 7464] to the south-east green slates with purple bands and cleavage dipping at 40°/162° contain crinoid debris, bivalves, occasional flattened clymenids, including Kosmoclymenia sp.; at [SX 7961 7461] a 35-mm rusty-weathered fine-grained limestone dipping at 45°/158° yielded D. anophthalmus of toV–VI. Below Mount Ararat slates and siltstones, with thin limestones and locally decalcified nodules which yielded Cymaclymenia cordata and Imitoceras sp.(toVβ–VI), occur in faulted sequences with Ararat Slate and Mount Ararat Chert; beds dip at 80°/336°–020°.

Surface slate debris [SX 7905 7513] east of Lounston yielded Richterina (R.) striatula and Sanguinolites? ellipticus. A track section [SX 7908 7515] in 0.6 m of weathered slates, bedding and cleavage dip 22°/168°, yielded a few crinoid fragments, Petraia sp., juvenile chonetid brachiopods, juvenile bivalves (including Guerichia venusta and S.? ellipticus), A. (W.) cf. warsteinensis, D. anophthalmus, D. aff. anophthalmus, Phacops granulatus, P. aff. wedekindi, P. cf. wedekindi, P. cf. wocklumeriae, Pseudowaribole (P.) cf. conifera, P. (P.) aff. octofera, Typhloproetus subcarintiacus, Maternella dichotoma, M. hemisphaerica, and abundant Richterina (R.) costata, R. (R.) striatula, rare leperditellids and indeterminate smooth ostracods. The trilobite fauna is referable to the Wocklumeria Stufe (toVI) and the ostracods to the hemisphaerica–dichotoma Zone. On the west side of the track at the bridge 0.6 m of green slates with buff delcalcified lenses a few millimetres in diameter yielded crinoid debris, juvenile chonetids, Cymaclymenia cf. striata, Cyrtoclymenia sp., Kalloclymenia sp., Kosmoclymenia sp., Parawocklumeria?, Dianops aff. griffithides, Phacops cf. granulatus, P. cf. .wocklumeriae, Chilobolbina? aff. rhenana, M. dichotoma, R. (R.) costata, R. (R.) striatula and indeterminate smooth ostracods. The trilobite and ammonoid fauna belongs to toVI, perhaps to the upper part; the ostracods indicate the hemisphaerica–dichotoma Zone. These two localities are the "Lounston fossil beds" of Bristow (1962) and were collected by Ussher (1913).

Siliceous slates [SX 7937 7517] yielded crinoid debris, bivalve fragments, M. hemisphaerica, R. (R.) costata, R. (R.) striatula and R. (F.) semen. The fauna is referable to the hemisphaerica–dichotoma Zone. A nearby section [SX 7940 7516] provided crinoid debris, Paracyclas? and R. (R.) costata.

A lane near Lenda shows [SX 7923 7574] slates with small decalcified lenses and mottled green mudstones yielding radiolarian casts, crinoid debris, indeterminate juvenile brachiopods and bivalves, T. cf. subcarintiacus, M. dichotoma, R. (R.) striatula and a conodont fragment; the ostracod and trilobite fauna is referable to toVI. The nearby road section [SX 7952 7582] to [SX 7963 7572] shows slates with a gently undulating slaty cleavage cut by a crenulation cleavage dipping at 60°/180°, and [SX 7986 7541] to [SX 7964 7571] slates with slaty cleavage dipping at 32°/351° and crenulation cleavage at 60°/350°. Exposures in a bank [SX 7974 7562] show slates with cleavage dipping at 58°/171° and with a few decalcified nodules that yielded Imitoceras sp.and Sporadoceras sp.Temporary excavation around a new house at Pool (Pete) Farm [SX 7987 7540] exposed slates with decalcified laminae, and nodules that yielded crinoid ossicles, brachiopods and Kosmoclymenia?. The road section [SX 8012 7523] to [SX 8024 7516] to the south-east exposes 9 m of green and purple slates with fragmentary flattened clymenids [SX 8023 7516] and a single specimen of Kosmoclymenia? [SX 8021 7517], overlain by 55 m of green slates with some siliceous and decalcified nodules up to 12 mm across aligned parallel to the slaty cleavage, rare silty beds, and at [SX 8020 7518] a band of grey slates with Cladochonus?, solitary corals, crinoid debris and rare bivalves; the prominent slaty cleavage dips at 25° to 40° south-east and is cut by a later crenulation cleavage inclined at 80°/150°, producing 'pencilled' zones.

On the east side of the Liverton valley a quarry [SX 8034 7519] exposes 4.6 m of slates with radiolarian casts, R. (F.) semen and R. (F.) aff. semen, and with scattered thin beds of silty sandstone that commonly exhibit small-scale cross bedding, faulted against overlying rather massive slates 4.6 m thick and containing a bed rich in solitary corals, crinoid debris, juvenile brachiopods and bivalves. Cleavage and bedding dip at 30°/129°. Ussher's (1913) fauna includes Cladochonus? and indeterminate ammonoids.

A quarry [SX 8001 7552], in 1.83 m of slates with scattered corals, bivalves, flattened ammonoids, M. hemisphaerica, R. (R.) costata, R. (R.) striatula and R. (F.) semen (hemisphaerica–dichotoma Zone), shows slaty cleavage dipping at 22°/070° cut by crenulation cleavage dipping at 72°/174°.

A track section [SX 7998 7560] to [SX 7994 7567] shows micaceous mudstones with juvenile bivalves of the uppermost Rora Slate, associated with micaceous slates and limestones with radiolarian casts and juvenile ammonoids, probably belonging to the Mount Ararat Chert.

Kate Brook Slate

Green slates in the field south-east of the Welcome Stranger Inn, near Blackpool, contain spiriferoid brachiopods [SX 8102 7365]. Brachiopods also occur in slates in the roadside opposite Higher Staplehill Farm [SX 8253 7369].

In the Kate Brook a calcareous sandy lens yielded Cyrtospirifer sp. [SX 8664 7867]; elsewhere cleavage attitude ranges from 25°/072° [SX 8783 8018] to 40°/160° [SX 8792 8027], and slates [SX 8789 8023] yield rare crinoid debris and fragmentary brachiopods. Immediately north of Crammers Bridge slates with cleavage inclined at 35°/127° yield fragmentary brachiopods [SX 8811 8102].

Cuttings for the A38 road [SX 8823 8173] to [SX 8823 8193] exposed slates with micaceous partings and brown decalcified sandy limestones up to 15 mm thick. The thin limestones yielded a few brachiopods, and the slates contained a poorly preserved fauna of crinoid debris, brachiopods, juvenile bivalves and ostracods. Weathered slates in a stream [SX 8813 8179] south of Rock Lane yielded crinoid debris and fragmentary brachiopods.

Gurrington Slate

Woodland to Morley

Grey slates with ferruginous lenticles in the lane [SX 7972 6892] east-north-east of Woodland church yielded a specimen of Phacops [SX 7967 6881]. Slates with dark grey blebs show cleavage dipping at up to 40° north-east [SX 7985 6910]; quartz veins parallel to cleavage are present locally. Around Wotton [SX 8010 6955] it is commonly difficult to differentiate between dolerite and spilite; the latter is dark grey with numerous amygdales. Near Lower Wotton [SX 8013 6944] dark grey slates show cleavage dipping at 22°/126° and laminations dipping at 18°/306°. At Wotton Cross grey slates with weakly defined laminations [SX 8023 6982] to [SX 8030 6981] dip at 30° eastward.

Ussher (1913) recorded indeterminate styliolinids and ammonoids in a lane north of Metley [SX 8105 7034]; the styliolinids suggest an age not younger than mid Famennian. Slates along the Kester Brook have cleavage dipping southward except at [SX 8193 7168], where green slates are inclined at 40°/143°. Grey slates [SX 8257 7132] near Morley yielded Probeloceras sp., referred by Professor M. R. House to the cordatum Zone.

Forder Green–Mallands–Knowles Hill

West of Halswell cleavage in light green slates dips at 14°/100° and shows lineation plunging at 10°/066° [SX 7807 6784]. Exposures in Bovey Lane [SX 7957 6855] to [SX 7968 6868] yielded a probable splendens Zone fauna including Entomoprimitia (Entomoprimitia) cf. splendens, E. (E.) cf. triangulata, Entomozoe (Nehdentomis) sp. nov., Entomozoe? aff. E. (N.) pseudorichterina and rare tentaculitoideans.

Purple slates in the roadside [SX 7916 6885] north of Woodland church have yielded a probable late Frasnian fauna including Entomoprimitia (E.) cf. triangulata, Entomozoe (N.) aff. pseudophthalmus, E. (N.) sp. nov.and rare tentaculitoideans. To the south-east of Orlycombe Bridge [SX 7875 6872] greenish grey slates with interbedded spilite dip at 20°/040°. Westwards to beyond Gurrington House rust-coloured lenticles are common and cleavage dips at 10° to 20° southward.

Gurrington Quarry [SX 7842 6958] is excavated in slates with cleavage dipping at 50°–65°/175°. Purple and green colour banding, sporadic silty laminations and cleavage are commonly parallel. A few kink folds up to 50 mm wide are present; narrow vertical quartz veins trend with the cleavage. A nearby roadside section [SX 7831 6975] yielded Richterina (R.) aff. striatula and Richterina sp. nov. aff. R. (F.) intercostata, probably indicating the intercostata Zone.

North-west of Lower Wickeridge Farm [SX 7840 6992] purple slates with spilite are inclined at 40°/160° and contain Guerichia sp., M. dichotoma, R. (R.) costata and R. (R.) striatula, a hemisphaerica-dichotoma Zone fauna. Just north of the Rising Sun Inn [SX 7901 6995] cleavage in purple and green slates dips at 40°/140° and the slates show weakly defined kink folds trending 144°.

Near West Down [SX 7905 7018] purple and green slates have cleavage dipping at 30°/172° and show bedding/cleavage lineation plunging at 23°/110°; at [SX 7935 7033] this lineation is parallel to the strike of the cleavage, which dips at 50°/165°. A disused quarry [SX 7920 7059] to the north-north-west contains purple slates, cleavage dip 40°/163°, which yielded Entomoprimitia (Reptiprimitia) rabieni, Maternella dichotoma, M. hemisphaerica, Richterina (Fossirichterina) semen and R. (Richterina) striatula, a hemisphaerica-dichotoma Zone fauna.

Another quarry [SX 7936 7051] contains purple and green slates with cleavage dipping at 35°/152° and bedding/cleavage lineation plunging at 10°/082°. Quartz veins are common hereabouts, up to 75 mm thick and generally aligned with the cleavage. A small quarry at [SX 7975 7104] shows purple and green slates intruded by dolerite, the junction dipping at 40°/1710; the slates yielded Guerichia sp., Richterina (R.) cf. costata, R. (R.) cf. striatula and R. (F.) semen.

To the north-east of Combe Farm, spotted purple slates [SX 8037 7153] dip southward in contact with dolerite. Near Higher Herebere [SX 8073 7150] purple slates with cleavage dipping at 20°/174° yielded M. dichotoma, M. cf. exornata, M. cf. hemisphaerica, R. (R.) costata, R. (R.) striatula and R. (F.) semen, a lower hemisphaericadichotoma Zone fauna. Poorly preserved ostracods in green slate at [SX 8075 7147] indicate the hemisphaerica-dichotoma Zone and include M. dichotoma and R. (F.) semen. A road section [SX 8102 7157] provided abundant ostracods including Entomoprimitia sp.aff. E. (E.) nitida, E. (R.) rabieni, Richterina sp. nov. aff. R. (F.) intercostata and R. (Volkina?) latecostata; the same cutting [SX 8107 7156] yielded R. (V.?) aff. latecostata, Richterina sp. nov. aff. R. (F.) intercostata and Bertillonella sp.; these faunas indicate the intercostata Zone.

Ussher (1890, 1913) recorded ammonoids, a styliolinid and a bactritoid from a road cutting near Wrigwell [SX 8118 7145]; these include specimens identified by Professor House as Aulatornoceras sp.and Manticoceras sp., indicating mid to late Frasnian.

Slates in another lane section [SX 8117 7179], near Wrigwell Cottages, yielded bivalves, M. cf. dichotoma, M. cf. exornata, R. (R.) cf. costata and R. (R.) striatula; those in a road section by Chipley Bridge [SX 8109 7182] contained bivalves, M. hemisphaerica and R. (R.) striatula; both faunas indicate the hemisphaerica-dichotoma Zone. Near Mill Cross, greenish grey slates with cleavage dipping at 44°/150° show bedding/cleavage lineation plunging at 38°/130° [SX 8168 7183], and green and purple slates yielded Entomoprimitia? sandbergeri and Entomozoe (Nehdentomis) nehdensis [SX 8177 7202], a serratostriata-nehdensis Zone fauna.

At 550 m north of Mallands purple and green slates exposed in a building site [SX 8230 7270] have cleavage inclined at 36°/112°. Ussher (1913) recorded Dianops anophthalmus, referable to to?V–VI, from slates in the roadside at Western House [SX 8408 7238], which have yielded a hemisphaerica-dichotoma Zone fauna. Purple slates by the road [SX 8384 7236] near Western House have been intruded by dolerite and contain bivalves, R. (R.) costata and R. (R.) striatula. Similar rocks in Foxwell Lane [SX 8517 7184] yielded bivalves, R. (R.) cf. striatula and R. (F.) semen.

Laminated grey to buff slates with sporadic rust-coloured nodules in Whitehill Road [SX 8498 7216] have yielded M. dichotoma and R. (R.) costata, indicating the hemisphaerica-dichotoma Zone. Knowles Hill Quarry, now obscured, provided Ussher (1913) with ostracods and abundant Trimerocephalus mastophthalmus, a Cheiloceras Stufe species.

Bremridge to Chipley

About 450 m west of Bremridge Farm grey slates have cleavage dipping at 50°/162° [SX 7812 7006] and contain boulders of vesicular spilite [SX 7824 7007]. Exposures [SX 7832 7056] near Combe Cross show grey silty slates with laminations sub-parallel to cleavage, dip 34°/165°; rust-coloured lenticles are common.

Grey slates cropping out sporadically along the road north-west from Combe Cross give way [SX 7814 7082] to hard calcareous laminated siltstones, weathering buff; cleavage dips at 40°/155°. Spilite boulders are present in places, one [SX 7808 7083] being 0.6 m across. Beyond this last locality grey slates with rust-coloured lenticles pass into lapilli tuff [SX 7803 7088]. The tuff has been sheared parallel to bedding, which dips at 35°/170° [SX 7798 7096]. Thin quartz veins occur locally.

South-east of Goodstone spilite occurs in tuffs and slates overlying Chercombe Bridge Limestone [SX 7871 7143]; it is strongly amygdaloidal with a few plagioclase laths forming phenocrysts. Weakly folded laminated siltstones crop out near Hole [SX 7873 7090], and slates in a small pit [SX 7882 7102] contain a spilite boulder 300 mm long.

Grey silty slates adjoin tuffs and interbedded spilites 630 m west of Gale; their cleavage dips at 20°/130° [SX 7879 7147]. The lane [SX 7946 7160] north from Gale exposes dark grey slates with rust-coloured lenses dipping at 35°/180° and slates with calcareous lenses up to 10 mm thick dipping at 18°/200° [SX 7948 7180]. On the south side of the A383 road [SX 7960 7195] siltstones and mudstones lie in small tight folds overturned to 246° and plunging at 45° eastward; a weakly defined cleavage dips at 20°/166°.

Greenish grey slates at the top of a 15-m face in a quarry [SX 8070 7218] opened for spilite at Chipley (Plate 2) yield a probable hemisphaerica-dichotoma Zone fauna, including E. (R.) cf. rabieni, M. cf. dichotoma, R. (R.) costata and R. (R.) striatula. Similar slates at South Knighton [SX 8116 7245] have cleavage inclined at 10°/041°. The rocks are generally poorly cleaved and tend to split along silty laminae; they have yielded a few brachiopods, bivalves, bactritoids, fragments of goniatites and abundant R. (R.) striatula [SX 8106 7298]. The ostracods suggest that these slates are not older than toIII.

Luxton Nodular Limestone

Lindridge Hill–Whiteway Barton

Ussher (1913) recorded shales with tentaculitoideans and ostracods at Lindridge Hill and collected Bertillonella (Rabienella) cf. cicatricosa, suggesting the cicatricosa Zone. An exposure of Kiln Wood beds [SX 8767 7468] north of Lyndridge Hill Cottages shows shales yielding brachiopods and a few small bivalves interbedded with limestones containing tentaculitoideans and a conodont fauna (loc. 52) referable to the upper gigas Zone. In a steep field nearby [SX 8766 7470] nodular limestones and shales with nodules dip steeply south-eastward; the limestones yielded rare brachiopods and conodonts (loc. 51) of the upper crepida Zone.

Whiteway Barton Quarry [SX 8820 7500] is largely overgrown but contains three small exposures of lenticular and nodular limestones with shale partings. In the western part of the quarry lenticular limestones, commonly heavily veined by calcite, occur in a jumbled mass faulted against East Ogwell Limestone. In the middle part of the face 1.2 m of similar limestones are horizontal; conodonts (loc. 60) indicate the lower marginifera Zone. Near the eastern end of the quarry thinly bedded limestone faulted against East Ogwell Limestone yielded conodonts (loc. 61) referable to the middle velifer Zone. This is the youngest fauna recognised in the area, and includes a relatively early appearance of Polygnathus hassi. Ussher (1913, p. 38) noted a slab of red shaly limestone, probably derived from the quarry, containing goniatites which include Beloceras sagittarium of the cordatum Zone (House, 1963); this suggests that cordatum Zone nodular limestone was formerly exposed in the quarry. In the field east of the quarry small crags [SX 8821 7502] of thinly bedded nodular limestones yielded conodonts from the most northerly outcrop (loc. 62) referable to the crepida Zone and from the most southerly crag (loc. 63) to the upper crepida Zone.

In Strongs Cover [SX 8880 7520] 2.4 m of red thinly bedded and rubbly fine-grained limestones, reminiscent of the Lower Dunscombe Goniatite bed, are faulted against massive East Ogwell Limestone and contain conodonts (loc. 64) referable to the crepida Zone. This locality is noteworthy in also yielding forms from at least the two earlier conodont zones, which could indicate reworking or, more likely, condensation of faunas.

Well area

Near Well [SX 8776 7704], Kiln Wood beds of comparable age and thickness to those of the Chudleigh escarpment are exposed and a red goniatite-bearing limestone similar to the Lower Dunscombe Goniatite bed was recorded by Vssher (1913) and Anniss (1933) above the massive limestones in the eastern part of the Well inlier. The nodular limestones above these basal members are about 10.5 m thick and occupy a triangular outcrop. They are bounded by a north–south fault to the west, overthrust by Carboniferous rocks to the south, and faulted against East Ogwell Limestone to the north-east.

Ussher (1913) mentioned red shaly limestones with clymenids on the west side of the road west of Well Farm. The distribution of debris of Winstow Chert and Posidonia beds suggests a reduced inverted succession younging north-west. A trench [SX 8763 7712] to [SX 8767 7708] in shales and nodular limestone (below) ended not far from Winstow Chert brash, and the Whiteway Slate is probably either faulted out or condensed.

A small temporary section [SX 8798 7700] at the eastern side of Well farmhouse showed micaceous shales faulted against East Ogwell Limestone. A track section [SX 8801 7699] south of the farmyard exposes, from north to south, 0.6 m of thinly bedded, reddish pink nodular limestone with red shale partings yielding rhomboidea or low marginifera Zone conodonts (loc. 55), faulted against 1.3 m of red shales with abundant nodules and occasional bands of nodular limestone, followed by 2.9 m of red shales with fewer limestone nodules and rare bands of nodular limestone. The top metre of this limestone yielded a rhomboidea Zone conodont fauna (loc. 56) with some evidence of reworking, and the lowest 1.9 m conodonts (loc. 57) of the rhomboidea Zone. Another track [SX 8803 7697] shows 2.4 m of limestones dipping at 60°/010°, apparently in inverted sequence; the upper 1.2 m yield (loc. 58) lower marginifera Zone and the lower 1.2 m (loc. 59) upper marginifera Zone conodonts. Bivalves, brachiopods and indeterminate ammonoids also occur.

A bank [SX 8776 7704] at the entrance to a field west of Well contains silty micaceous shales with locally decalcified partings, nodules and horizons of nodular limestone which are referred to the Kiln Wood beds. The shales yielded Bertillonella (B.) erecta, B. (Rabienella) cicatricosa, B.? aff. fabaeformis, Entomozoe (Nehdentomis) pseudorichterina, E. (N.) tenera, Neochilina cf. parvula and Ungerella calcarata of the cicatricosa Zone. Tentaculitoideans of the Homoctenus tenuicinctus group and the trilobite Cgphops cf. acuticeps were also present. The nodular limestones yielded foraminifera, brachiopods and conodonts (loc. 53) belonging to the A. triangularis Zone. A borehole drilled adjacent to this exposure proved about 8 m of grey micaceous shales with sporadic horizons of nodular limestone up to 50 mm thick; the limestone contained a probable gigas Zone fauna (loc. 54). A pit for a telegraph pole [SX 8774 7706], also within the Kiln Wood beds, exposed shales with thin decalcified partings yielding B. (R.) cicatricosa, E. (N.) pseudorichterina, E. (N.) tenera and U. calcarata of the cicatricosa Zone. Tentaculitoideans are also present.

An oblique dip section in a trench [SX 8763 7712] to [SX 8767 7708] showed the following sequence from south to north (Figure 10):

Fossils from nodular limestone [SX 8767 7708] (Ussher, 1913; Anniss, 1933) have been identified by Professor House as very early platyclymenids.

Gappah to Lower Dunscombe

The Kiln Wood beds of Kiln Wood Quarry [SX 8612 7792] have yielded only mid-Frasnian fossils. In the quarry the following sequence dips at 8°/098° and is separated by a possible slip zone from the underlying massive Chercombe Bridge Limestone:

The sequence comprises three cycles, each consisting of grey micaceous shales, overlain by blocky bedded siliceous mudstones, less siliceous towards the base and commonly with a tuff horizon, followed by grey or pinkish grey nodular or streaked limestone. The tuffs consist of angular fragments of quartz, tabular feldspar crystals, commonly sericitised, and a little biotite, in a matrix of chlorite and argillaceous material. Lack of rounding and the isolation of many grains in the matrix suggest little reworking and absence of current activity. Crinoid debris, Styliolina sp., and bivalves with foraminifera attached to some of the valves, occur in the limestones. Tucker and van Straaten (1970) showed that conodonts of the lowermost limestone ranged from upper asymmetricus to lowermost gigas Zone, and those of the topmost limestone, lying below the goniatite horizon, from the A. triangularis Zone to low in the upper gigas Zone.

The road running past Kiln Wood [SX 8612 7803] to [SX 8614 7793] exposes nodular limestones, shales with nodules and shales dipping gently eastward. Immature bivalves and crinoid debris have been obtained from micaceous shales, and a single rugose coral from a nodular limestone. Conodonts from nodular limestones in the lower part of the section (Tucker and van Straatan, 1970) proved all the conodont zones from lower or middle crepida Zone to lower marginifera.

At Mount Pleasant, House and Butcher (1973) trenched through grey micaceous shales with a limestone horizon, referred to the upper part of the Kiln Wood beds, which were faulted against massive Chercombe Bridge Limestone [SX 8696 7874] and overlain by massive nodular limestones of the main part of the formation. From nodular limestones and shales on the southern side of a hedge [SX 8699 7873] House (1963) listed the following fossils of the annulata Zone: Imitoceras sp., Platyclymenia (Platyclymenia) aff. walcotti and P. (Trigonoclymenia) protacta. He recorded Cheiloceras aff. amblylobum of the curvispina Zone from the northern side of same hedge. House and Butcher (1973) recorded goniatites of the Imitoceras sulcatum group from a track [SX 8696 7871].

North-east of Mount Pleasant the Luxton Nodular Limestone is largely faulted out. However, in Lower Dunscombe Quarry [SX 8855 7907] at the top of the western side of the north face, greyish pink to cream-coloured thin irregularly bedded fine-grained limestones with shaly partings, belonging to the Lower Dunscombe Goniatite bed, dip at 30°/217°. The contact with underlying East Ogwell Limestone is obscured by rubbish. House (1963) listed Manticoceras cordatum, M. cf. intumescens, Beloceras sagittarium and Tornoceras (T.) sp.of the cordatum Zone. Tucker and van Straaten (1970) recorded Syringaxon sp., bivalves including Buchiola sp., and a rich late Frasnian conodont fauna ranging throughout the upper gigas Zone. Trilobites include Cryphops acuticeps (in the Sedgwick Museum and located as "Lower Dunscombe") and Roemer (1880) noted a median dorsal plate of Coccosteus sp.Brash of red nodular limestone and slate occurs on the dip slope of the escarpment in the field above Lower Dunscombe Quarry, stratigraphically above the Lower Dunscombe Goniatite bed. House (1963) recorded Sporadoceras cf. contiguum of the sandbergeri Zone from the ploughed field above the quarry.

Teign Estuary

A discontinuous cliff section on the northern side of the Teign Estuary [SX 8962 7302] to [SX 8969 7300] exposes slates with thin nodular limestones. Near the eastern end decalcified nodules yielded crinoid debris and bivalves and also the trilobite Trimerocephalus mastophthalmus of Cheiloceras Stufe age.

The type section of the Luxton Nodular Limestone is on the foreshore [SX 9000 7288] west of Luxton's Steps; slates with scattered calcareous nodules show cleavage dipping at 35°/315°. In a small cliff immediately to the north micaceous slates contain both bands and nodules of limestone. Bedding dips at 35°/315° parallel to the slaty cleavage. The nodules have yielded conodonts (loc. 65) ranging from the crepida Zone to the marginifera Zone. Ussher (1913) recorded an ammonoid from "opposite Coombe Cellars" which has been redetermined as Rectoclymenia aff. subflexuosa, probably derived from the delphinus Zone of Platyclymenia Stufe age.

The junction with the overlying Whiteway Slate is taken at the top of a 75-mm nodular limestone which occurs in green slates in a stream section [SX 9004 7313] to [SX 9005 7308]; the limestone yields an upper marginifera to lower velifer Zone conodont fauna (loc. 66). A small cliff [SX 9015 7306] north-west of Luxton's Steps exposes 3 m of massive nodular limestones and slates with nodules dipping at 10°/153°. Conodonts (loc. 67) from this section are referable to the upper marginifera Zone.

On the west side of Bishopsteignton purple and green slates with scattered limestone nodules, which cannot be separated from Whiteway Slate, have yielded bivalves and ostracods at [SX 9005 7358], including Entomozoe (Nehdentomis) nehdensis and E. (N.) asymmetrica, of the serratostriata-nehdensis Zone. E. (N.) nehdensis has also been obtained from slate debris in a field at [SX 9049 7420]. Slates in a road section [SX 9040 7402] east of Clanage Cross contain a 50-mm band of fine tuff A borehole in the southern part of Bishopsteignton passed through terrace gravel and East Ogwell Limestone into Luxton Nodular Limestone whose conodonts (Rhodes and Dinley, 1957) suggest an anomalous late Givetian age.

Emblett Hill–Abbotskerswell–Barton

In the Denbury Unit, the only record of the Luxton Nodular Limestone is in the Emblett Hill Borehole (p. 184), where a faulted mass of red slates with limestone nodules occurs to a depth of 11.68 m. It lies above East Ogwell Limestone and yields a marginifera fauna (loc. 37). Luxton Nodular Limestone was recovered from three levels in the Rydon Ball Farm Borehole (p. 183).

Lenticular micritic limestone 2.4 m thick, probably Luxton Nodular Limestone (p. 49), overlies massive East Ogwell Limestone in the southern face of Ransley Quarry, but precise relationships are obscured by faulting. This limestone has yielded goniatites of probable cordatum Zone age (House, 1963) and conodonts (loc. 25) [SX 8443 7018]. South of Newton Abbot Cemetery [SX 8526 6985] nodular limestone in obscure relationship with the East Ogwell Limestone has yielded conodonts (loc. 44) ranging from the base of the A. triangularis Zone to low gigas Zone. Structurally isolated Kiln Wood beds exposed near Conitor Quarry [SX 8508 6942], 0.8 km south of Newton Abbot Hospital, comprising laminated grey slates with limestone lenses, have yielded a mid-Frasnian cordatum Zone fauna. Pyritised goniatites from the slates include Manticoceras cf. cinctum (?= retrorsum), M. cf. cordatum, M. cf. tuberculatum and Tornoceras (Linguatornoceras) cf. linguum. Bactritoids and Buchiola? also occur. The limestones yield poorly preserved ostracods, including Entomozoe (Nehdentomis) pseudorichtertha, and a rich but mixed conodont assemblage (loc. 43) in which the youngest forms indicate the middle and upper Palmatolepis triangularis Zone; other forms are derived from the Ancyrognathus triangularis or lower gigas Zone.

Conodonts (loc. 42) of the lower marginifera Zone of Sandberg and Ziegler (1973) were obtained from samples taken from a trench [SX 8483 6895] through slates with abundant calcareous lenticles which crop out [SX 8474 6892] west of Abbotskerswell, in an area of Whiteway Slate. Similarly, red nodular limestones in Fore Street, Barton [SX 9095 6714], are referred to the Luxton Nodular Limestone but have not been mapped separately from the Whiteway Slate; they yielded a conodont fauna of velifer Zone age (loc. 68).

Whiteway Slate

Chudleigh

In the Ugbrooke Unit the Whiteway Slate lies conformably between the Luxton Nodular Limestone and the Winstow Chert.

Its narrow outcrop is repeated by folding, and facies changes occur from north to south. On the Chudleigh escarpment trenching right across the formation proved (House and Butcher, 1962, 1973) 17 m of red, purple, grey and green slates with fossiliferous limestone nodules, overlain by 8.5 m of siliceous green shales with rusty-weathering limestone nodules bearing a rich fauna of ammonoids, ostracods and trilobites mostly of Wocklumeria Stufe age. A strip of Whiteway Slate extends north from Gappah [SX 8622 7738] past Kiln Wood, and is almost completely faulted out in the road section east of the wood.

In pits near Winstow Cottages House and Butcher (1962, 1973) showed the Whiteway Slate to comprise shales and slates with ferruginous-weathered calcareous nodules bearing faunas from post lower Clymenia Stufe to Gattendo0a Stufe age. The same workers obtained a Wocklumeria Stufe fauna from a nearby field [SX 8652 7813].

A track section [SX 8621 7804] to [SX 8628 7803] adjacent to Winstow Cottages shows slates with rusty-weathered calcareous nodules overlain in the east by Winstow Slate, and the ammonoid, ostracod, trilobite and conodont faunas from this section have been described by House and Butcher (1973) and Tucker and van Straaten (1970).

Pits at Mount Pleasant, above the outcrops of Luxton Nodular Limestone, revealed shales with ferruginous nodules and nodular limestones ranging in age from lower Clymenia to Wocklumeria Stufen (House and Butcher, 1973); the top of the Whiteway Slate was taken below shales thought to resemble Tournaisian strata at Winstow Cottages. Pitting farther north-east [SX 8727 7884] showed Winstow Chert faulted against Chercombe Bridge Limestone; and eastwards from here both Luxton Nodular Limestone and Whiteway Slate appeared to be faulted out, perhaps by a bedding-plane thrust. However, House (1963) recorded Cymaclymenia cordata from "Lower Dunscombe", which suggests the presence there of Whiteway Slate.

The lower part of the Whiteway Slate yielded the following fossils from trenches at Mount Pleasant and Winstow Cottages: Clymenia?, Cymaclymenia sp., Cyrtoclymenia sp., Kosmoclymenia cf. undulata, Imitoceras sp., Dianops aff. griffithides, Maternella sp.and Richterina (Richterina) sp.Adjoining exposures have yielded (Tucker and van Straaten, 1970) a styriacus Zone fauna, together with Palmatolepis gracilis sigmoidalis of the lower costatus Zone; the ammonoid evidence suggests that the styriacus Zone fauna may be derived.

The upper part of the formation is mainly of Wocklumeria Stufe age. Fossils from Mount Pleasant and Winstow Cottages include: Cymaclymenia constricta, C. cordata, C. aff. cordata, Gonioclymenia? (Kallocylmenia) sp., Kosmoclymenia bisulcata, K. aff. undulata, Parawocklumeria distorta, P. laevigata, Imitoceras spp., Kenseyoceras sp., Mayneoceras sp., Sporadoceras cf. posthumum, Dianops anophthalmus, Phacops (Omegops) accipitrinus, P. cf. ensae, P. granulatus, Maternella sp.and Richterina (R.) striatula. The presence of Archegonus (Phillibole) drewerensis at Winstow Cottages suggests an upward passage into beds of Gattendorfia Stufe age.

Whiteway Barton

At Well the Whiteway Slate is presumed to be condensed or faulted out, Luxton Nodular Limestone being thrown against Winstow Chert. From Lindridge Hill to west of Whiteway Barton the Whiteway Slate is similarly faulted out. At Whiteway Barton the formation crops out to the west of a north–south fault, where it is inverted and faulted against Luxton Nodular Limestone and East Ogwell Limestone to the south. The formation here is probably about 45 m thick, though there is no evidence for slates older than upper Clymenia Stufe and the lowest beds may be faulted out.

Material collected by Ussher (1913) and I. Thomas (1909) from "Whiteway Farm" included the type specimen of Cyrtosymbole anglica, probably the senior synonym of A. (P.) drewerensis. The type section of the Whiteway Slate [SX 8844 7525], a cutting alongside a barn, exposes buff to pale green slates yielding small brachiopods, Cyrtosymbole (Calybole) ussheri, Pseudowaribole (P.) conifera, Typhloproetus subcarintiacus, Maternella dichotoma, M. hemisphaerica, Richterina (R.) latior, R. (R.) striatula, indeterminate smooth ostracods and rare conodont fragments. The ostracods indicate the hemisphaericadichotoma and latior zones and the trilobite fauna is referable to toVI. Both here and in a trench north of Whiteway Barton ((Figure 11)a) it has been shown that the top of the formation lies within the Gattendorfia Stufe; in the trench pale green slates, with earthy siliceous limestone nodules containing bivalves and trilobites, and also greyish black siliceous slates with the same fauna, probably lie close beneath the black sooty slates of the Winstow Slate.

A temporary pit [SX 8851 7518] in Whiteway Barton farmyard exposed slates yielding rare solitary corals, crinoid debris, lingulid brachiopods, Archegonus ( Waribole) sp., Cyrtosymbole (Calybole) ussheri, Typhloproetus cf. subcarintiacus, Maternella dichotoma, M. hemisphaerica, Richterina (R.) striatula and conodont fragments. The ostracods indicate the hemisphaerica-dichotoma Zone and the trilobite fauna is referable to toVI. Slate brash in a ploughed field [SX 8876 7530] yielded a rich ostracod fauna of Maternella exornata, M. dichotoma and R. (R.) striatula, referable to the hemisphaerica-dichotoma Zone. This could indicate that the beds range down into the Clymenia Stufe. Other outcrops hereabouts yielded a specifically Wocklumeria Stufe fauna, including Archegonus (Waribole) cf. warsteinensis, Cyrtosymbole (Calybole) ussheri, C. (C.) aff. ussheri, D. anophthalmus, P. ensae, P. wocklumeriae, Pseudowaribole (P.) conifera, T. subcarintiacus, E. (R.) rabieni, M. dichotoma, M. exornata, M. hemisphaerica, R. (R.) costata and R. (R.) striatula.

Rubble in a field [SX 8847 7533] to [SX 8860 7545] yielded solitary corals, crinoid debris, brachiopods, small bivalves, Kosmoclymenia?, Archegonus (Waribole) cf. warsteinensis, C. (C.) aff. ussheri, Dianops anophthalmus, Phacops ensae, T. subcarintiacus, E. (Reptiprimitia) rabieni, M. hemisphaerica, Richterina (R.) costata, R. (R.) striatula and rare conodonts; the trilobites are referable to toVI, the ostracods to the hemisphaerica-dichotoma Zone; at [SX 8851 7534] crinoid debris and fragmentary bivalves, Entomoprimitia (Reptiprimitia) rabieni, M. hemisphaerica, Richterina (R.) costata, R. (R.) striatula, the ostracod fauna referable to the hemisphaerica-dichotoma Zone; from [SX 8855 7539] to [SX 8856 7544] E. (Reptiprimitia) rabieni, Richterina (R.) costata, and R. (R.) striatula, (late Famennian); between [SX 8860 7545] and [SX 8857 7554] debris yielding rare solitary corals, crinoid fragments, bivalves, Archegonus (Phillibole) drewerensis, Maternella sp. nov. aff. M. circumcostata, M. clathrata and Richterina (R.) latior, the trilobites and ostracod faunas referable to the Gattendorfia Stufe.

Similar Carboniferous faunas were obtained from 10.9 m of pale green slate including occasional horizons of dark siliceous slates, bearing rare decalcified limestone nodules, which were exposed in the south-east part of a trench [SX 8857 7533] to [SX 8856 7555] (trench 3, (Figure 11)b). These beds yielded Archegonus (Latilobe) cf. laticampus, A. (P.) drewerensis, A. (P.) duodecimae, Diacoryphe strenuispina and Liobolina sp.of Gattendorfia Stufe age. Entomozoid ostracods included Maternella aff. arcuata, M.? aff. empleura, M. pfaffenbergensis, M. schindewolfi, M. aff. seilerensis, R. (R.) latior, Ungerella postmulticostata and U. stockumensis. Other ostracods included Bairdiocypris sp., Hypotetragona sp., jonesina? and Paraparchites cf. okeni.

Teign Estuary

In the Kingsteignton Unit, Whiteway Slate occupies all the ground along the Teign Estuary south of the Ugbrooke Unit, except for a probably conformable inlier of Luxton Nodular Limestone. It consists of red, green and grey laminated slates with rare thin bands and nodules of siliceous or sideritic material, and with some calcareous nodules and laminae mainly in the lower beds. Ostracod faunas indicate the hemisphaerica-dichotoma Zone and may range down into the intercostata Zone. Grey laminated slates with three flows of spilite lava, showing pillow structures that dip at about 60° south-westward in normal succession, are exposed on the northern and southern shores of the estuary. In thin section the spilites are indistinguishable from those occurring in the lower part of the Kingsteignton Volcanic Group (p. 26). The grey colour of the slates adjacent to the spilites, in a succession of purple slates, suggests that volcanicity affected the diagenesis of the mud.

A railway cutting [SX 8752 7230] to [SX 8823 7251] shows purple slates with cleavage dipping at 28° to 54° south-westward. Small crags [SX 8785 7241] have yielded bivalves and ostracods including Maternella sp., Neochilina cf. parvula and Richterina sp. nov. aff. R. (Fossirichterina) intercostata, probably referable to the intercostata Zone. At [SX 8805 7246] a band of siliceous nodules up to 13 mm thick runs parallel to the slaty cleavage. Between [SX 8809 7247] and [SX 8813 7248] the slates contain four spilite lava flows from 0.6 to 4.6 m thick; contacts dip at 60°/216°, parallel to the slaty cleavage. One of these flows crops out with slates on the shore [SX 8816 7246], the contact dipping at 44°/193°.

In Hackney Channel [SX 8760 7226] to [SX 8775 7226] purple slates have a cleavage inclined at 52°/190°. Behind the Passage House Inn [SX 8793 7234] purple slates with thin decalcified seams yielded bivalves and Richterina sp. nov. aff. R. (F.) intercostata. Purple slates at Kingsteignton sewage works [SX 8766 7251] yielded bivalves and poorly preserved ostracods including Maternella dichotoma, Richterina (R.) costata, R. (R.) striatula and R. (Fossirichterina) cf. semen, referable to the hemisphaerica-dichotoma Zone. In a wood north of the railway pale green slates with thin calcareous seams and tiny decalcified nodules show cleavage dipping at 25°/185° [SX 8833 7264] and 26°/195° [SX 8838 7266]. Small reefs on the foreshore [SX 8887 7263] shows black slates with a 100 mm band of black pyritiferous sideritic mudstone dipping at 28°/045° parallel to the slaty cleavage.

Pale lilac slates not far above the Luxton Nodular Limestone were exposed [SX 8976 7319] by ploughing in a gateway south of Forder Cross; they yielded a few brachiopods, bivalves, Archegonus (Waribole) sp., Phacops granulatus, M. dichotoma, R. (R.) costata, R. (R.) striatula, R. (F.) semen, smooth ostracods and palaeocopids. The ostracods are referred to the hemisphaerica-dichotoma Zone.

Grey slates at Bishopsteignton, referred by Ussher (1913) to the Culm Measures on the evidence of an outcrop of radiolarian chert, are here considered to resemble Whiteway Slate. In a building site [SX 9044 7348] south-east of Bishopsteignton House grey laminated slates showed cleavage dipping at 28°/164°. In the lane [SX 9075 7387] to [SX 9081 7390] joining Smith's Hill and Teignview Road similar slates exhibit a flat-lying cleavage. A nearby track section [SX 9083 7393] to [SX 9084 7406] exposes grey slates with scattered siliceous bands and nodules and with cleavage dipping at 18°/037°. Alongside a road [SX 9132 7398] near Ashill Farm laminated grey slates with cherty bands exhibit gently inclined cleavage. A cutting [SX 9128 7403] behind Ashill Farm showed grey slates with thin lenses of dark grey bituminous pyrite-rich limestone, cut by a horizontal cleavage.

Near Buckland Barton purple slates with faint spotting and ostracod impressions adjoin dolerite at the sewage works [SX 8812 7202]; [SX 8825 7203]. Foreshore exposures [SX 8867 7215] to [SX 8875 7216] to the east show three spilite pillow lavas up to 4.6 m thick interbedded with grey laminated slates. Cherty mudstones are locally present above the pillows. Farther east [SX 8875 7216] to [SX 8903 7216] grey, purple and green slates contain scattered cherty nodules and bands dipping at about 30° south or south-west parallel to the slaty cleavage. Ussher (1913) recorded clymenids, trilobites, bivalves and ostracods from hereabouts. The material is indeterminable except for a single specimen of M. dichotoma referable to the hemisphaerica-dichotoma Zone.

Between Netherton Point [SX 8915 7224] and Arch Brook Bridge grey and purple slates are exposed in cliff sections up to 6 m high. Immediately west of Coombe Cellars the slates are red and purple, micaceous and faintly calcareous; they contain calcareous laminae up to 5 mm in thickness. These laminae exhibit micro-crenulations whose fold axes strike 246° and whose axial planes dip at 5° southward, parallel to the slaty cleavage. A large open recumbent fold with axis trending 244° and micro-crenulated laminae on the limbs is seen at [SX 8989 7235]. Farther west the slates are laminated grey and purple in colour and small siliceous lenses, nodules and bands [SX 8926 7226]; [SX 8934 7225] pick out recumbent folds whose axes plunge gently southward. On the foreshore [SX 9060 7227] to [SX 9067 7224] immediately east of Coombe Cellars purple slates contain calcareous bands up to 5 mm thick.

Ipplepen–Wolborough–Barton

In the Abbotskerswell Unit the Whiteway Slate is extensive south of Newton Abbot, between the Ipplepen Fault and the New Red Sandstone, and crops out from beneath Middle Devonian and Carboniferous klippen of the East Ogwell Unit. It comprises mainly grey and green slates, with some purple slates and with sporadic developments of siliceous and calcareous nodules and yields ostracods of the hemisphaerica-dichotoma Zone.

Specimens from a building site [SX 8358 6677] near Ipplepen included M. dichotoma, M. hemisphaerica, R. (R.) costata and R. (R.) striatula. A similar fauna was obtained from slate debris north-east of Rydon Hill [SX 8420 6891] and from near East Ogwell, where tightly folded laminated siltstones were exposed in a building site [SX 8397 7010] in two closely adjacent anticlines. Only one of the folds was well displayed, plunging at 15° to the south-south-west; the upper limb dipped at 15° south-east and the lower at 40° south-east. Beneath the lower limb the siltstones lay in small tight folds underlain by less deformed beds dipping at 20°/150°. Slates in an adjacent exposure [SX 8398 7014] contained M. hemisphaerica, R. (R.) costata and R. (R.) striatula, and those from a temporary trench [SX 8387 7016] yielded poorly preserved trilobite pygidia which appear to show Lower Carboniferous affinities.

A cutting in Ogwell Road [SX 8443 7016] shows soft red shales that may belong to the Whiteway Slate. They appear to overlie the East Ogwell Limestone and Luxton Nodular Limestone of the nearby Ransley Quarry, but relations are confused by faulting and the outcrop has not been delineated. The Rydon Ball Borehole was drilled through a klippe of Ugbrooke Sandstone of the East Ogwell Unit into the underlying strata, which included Whiteway Slate (p. 183).

Near Higher Langford [SX 8528 6940] slates yielded Entomoprimitia (Reptiprimitia) rabieni, M. dichotoma, M. hemisphaerica, Richterina (R.) costata and R. (R.) striatula of the hemisphaerica-dichotoma Zone, and north-north-east of Abbotskerswell [SX 8579 6909] laminated slates lie in small tight horizontal folds with cleavage dipping at 28°/078°. In the Barton area, slates referred to the Whiteway Slate are indifferently exposed, as at [SX 9168 6660].

Chapter 5 Carboniferous

General account

Following the pioneer work of De la Beebe (1839) and Sedgwick and Murchison (1840), Godwin-Austen (1842) carried out a detailed study of the local successions of the Devonian and Carboniferous rocks around Chudleigh, Ugbrooke, East Ogwell and Newton Abbot. A unique contribution to the understanding of the Carboniferous, or Culm Measures, was made by Ussher in a series of memoirs (1902, 1908, 1912, 1913) and other publications (1887, 1891, 1892b, 1900, 1901, 1906). However his published maps do not show all the refinements of the successions he had established. Elements of his Lower Culm stratigraphy in the middle Teign Valley are reflected in this Memoir and on the current revision of the map.

Ussher and others following him have established that the Carboniferous represented on the Newton Abbot Sheet exhibits lateral and vertical changes of facies. That these facies changes are significantly less marked in the Carboniferous than in the Devonian reflects the progressive establishment of more uniform conditions over the area, though the removal of the Carboniferous from the southern part of the district may have simplified the picture.

Lower Carboniferous

The boundary between Devonian and Carboniferous strata, as defined palaeontologically, does not coincide with a lithological change (Figure 6). Trenching carried out on the Chudleigh escarpment by House and Butcher (1973) showed complete conformity between the two systems within the Whiteway Slate in the Ugbrooke–East Ogwell Succession of this Memoir. During the resurvey additional evidence has shown that a conformable passage is also present in the Hyner Shale–Trusham Shale sequence of the Teign Valley Succession, and probably also in the Gurrington Slate of the Bickington–Beacon Hill Succession. No sedimentary break has been observed in the Rora Slate of the Liverton Succession, where lowest Carboniferous faunas have not been found. The Devonian and lowest Carboniferous ostracod slates are described in Chapter 4, except for the Trusham Shale, which is wholly Carboniferous.

The thickest Lower Carboniferous sequence occurs in the Teign Valley Succession. Here the Trusham Shale, comprising olive green to pale grey micaceous shales, succeeds the dark grey Hyner Shale. The Trusham Shale is overlain by the black Combe Shale. The ostracod slates are similarly followed in the Liverton Succession by the black Ararat Slate and in the Ugbrooke–East Ogwell Succession by the black Winstow Slate, neither of which has been mapped separately from the overlying cherts. Only bivalves, immature goniatites, and conodonts occur in the Ararat Slate, and conodonts and Chondrites in the Winstow Slate; no fauna has been obtained from the Combe Shale. Probably euxinic conditions prevailed.

Bedded radiolarian cherts above the shales are associated with siliceous shales, mudstones, tuffs, lavas and manganiferous beds. The chert formations, Teign Chert, Ararat Chert and Winstow Chert, thin from north to south, as do the associated volcanic rocks, and possibly the amount of sedimentation was controlled by the submarine topography. The cherts are generally thought to have originated by direct precipitation of colloidal silica on the sea floor, possibly in response to an increase in silica content of the water because of nearby submarine volcanic eruptions. Some of them may, however, represent mudstones in various stages of silicification, and the highest may have replaced limestones.

Apart from radiolaria, fossils are rare in the cherts. Because the several successions are thought to be conformable, both below with lowest Dinantian strata and above with high Dinantian rocks, it seems that the cherts are a condensed sequence that represents much of Dinantian times and possibly accumulated within a starved basin that extended northwards to end at the foot of the shelf on which the thick Carboniferous Limestone succession of Somerset was contemporaneously forming.

Several bands of keratophyric crystal-tuffs, individually up to 12 m thick, lie within the Teign Chert and the northernmost outcrops of the Mount Ararat Chert, and probably represent fall-out from submarine eruptions. They die out southwards and the volcanic centres thus probably lay to the north and may now be hidden by Upper Carboniferous strata. Graded tuffs could have been brought in by turbidity currents. Vesicular spilite lavas and vitric-flow tuffs in the Teign Chert and Mount Ararat Chert represent local submarine eruptions. The spilites are chemically similar to the dolerite sills in the area and may have formed where shallow intrusions broke through to the sea floor. Felsitic lavas in the Teign Chert contain shard-like structures and may represent uncontaminated near-shore submarine eruptions of glass shards that flowed along the sea floor for a short distance.

The highest Dinantian strata in the district are the Posidonia beds. This informal term is applied to widespread shales with some thin cherts and limestones, in which species of Posidonia are common. These shales have been included within the Teign, Mount Ararat and Winstow chert formations. They are 30 to 45 m thick in the Teign Valley, but become thinner towards the south. A striking feature is the sudden appearance of abundant pelagic fauna. Benthos is variably present and Ramsbottom (1970) has commented that the goniatite-pectinoid fauna of Exeter and the northern part of the Newton Abbot district changes southwards in the Teign Valley to include first phillibolid trilobites and then brachiopods. The goniatite-pectinoid fauna with a shelly benthos continues into the Liverton Succession and is joined by ostracods in the southern successions. Ramsbottom saw this faunal distribution as indicating an approach to the southern margin of the basin. Goniatites indicate ages for the Posidonia beds ranging from B₂ to P₂ (pp. 176–177).

Trusham Shale

The Trusham Shale consists of about 60 m of olive-green shales and pale grey micaceous shales. The former show a fine even texture which has allowed a good conchoidal fracture to develop and which imparts a characteristic soapy feel. The pale grey micaceous shales are rather flaggy and commonly banded, the micas being concentrated in thin somewhat calcareous laminae. The carbonate has almost invariably been leached, leaving thin limonitic horizons clearly picking out the bedding. The lower boundary has been mapped at the change from the dark blue or bluish green mottled shales and calcareous beds at the top of the Hyner Shale to grey micaceous and green shales, and may be observed at Teign Lane, Trusham [SX 8475 8225], and near Christow [SX 8405 8482] to [SX 8403 8477]; the upper limit is at the change to bluish black shales of the Combe Shale. A track section north of Leigh [SX 8426 8113] is the type locality.

Readily recognised outside the metamorphic aureole, the Trusham Shale becomes less easy to distinguish from the Hyner Shale as the granite is approached, particularly where the micaceous shales are absent. For this reason the Hyner and Trusham shales have not been separated in the area between the Lustleigh and Bovey Tracey faults and also farther west. West of the fault zone the slates are lightly spotted, dull green to pale greyish green in colour and commonly micaceous and mottled. Bands of black siliceous mudstone up to 0.25 m thick and small siliceous nodules in the slates in Ilsington are reminiscent of those in the Hyner Shale; they lie not far below the base of the Combe Shale, and a slight facies change is indicated.

Ostracods from greenish grey slates at localities [SX 8430 8130]; [SX 8426 8113] near Leigh Farm include Maternella circumcostata, M. aff. seilerensis and Richterina (R.) latior of the latior Zone. The bivalve Sanguinolites? ellipticus is abundant locally at somewhat higher horizons, with bryozoans and indeterminate juvenile brachiopods. In the Ilsington area no fauna has yet been found.

Combe Shale

The Combe Shale consists of highly fissile bluish black or black shales and slates, locally characterised by white silty laminations and closely spaced vertical joints particularly in higher horizons. Siliceous nodules become increasingly numerous in the upper parts of the formation and appear to be the precursors of the bedded cherts of the the succeeding formation. The road cutting [SX 8402 8476] to [SX 8411 8446] immediately north of Lower Ashton has been selected as the type locality. The lower boundary of the formation, mapped at the change from grey micaceous and green shale of the Trusham Shale to the typical bluish black shales of the Combe Shale, is exposed in the road cutting from Ilsington to Woodhouse [SX 7860 7637]. The slates south of the Narracombe Thrust do not show the white laminations common in the Teign Valley. The Combe Shale has yielded no fossils except small fragments of wood in the Teign Valley.

Dolerite sills locally increase the overall width of outcrop of the formation. In the Teign Valley, and between the Lustleigh and Bovey Tracey faults, the Combe Shale ranges from 90 to 150 m thick. South of the Yarner–Brimley ridge the formation is probably less than 90 m thick, and in Ilsington it is no more than 45 m. The contrast in hardness between shales and dolerites is strongly reflected in the topography. The shales adjacent to the intrusions are indurated and bleached, and in places the contact-altered shale is more resistant to erosion than the dolerite.

Teign Chert

The Teign Chert forms a prominent topographic feature in the Teign Valley and picks out the broad structure. It comprises mainly well-bedded cherts which vary from white, grey and green to bluish black and black; rarely they are jasper red. Individual chert beds range in thickness from a few millimetres to over 0.3 m and are generally separated by partings of siliceous shale 3 to 6 mm thick. Radiolaria are locally abundant within the cherts, their presence being indicated by white casts and hollows.

Interbedded bluish black shales, limestones, quartzkeratophyre-tuffs, vesicular lavas and manganiferous beds and nodules diversify the relief of the broadly rounded chert hills. The tuffs commonly contain radiolaria and range in thickness from a few millimetres to 12 m. They are present throughout the chert sequence but occur mainly in the middle of the formation, where an impersistent tuff horizon can be traced along the length of the middle Teign Valley. Farther south volcanic rocks are less common. Towards the northern limit of outcrop, spilites are developed near the base of the formation, as in Scatter Rock Quarry [SX 8228 8561] and Scanniclift Copse [SX 8448 8628]. Within the granite's metamorphic aureole, calcareous beds altered to calc-flintas alternate with chert.

In the Posidonia beds at the top of the Teign Chert basal shales containing Posidonia becheri are overlain by interbedded shales, cherts and thin limestones. A band with spirally striate goniatites above the beds containing Posidonia acts as a useful marker horizon. The Posidonia beds have yet to be identified in the Teign Valley Succession west of the Lustleigh Fault.

Almost continuous sequences through the Teign Chert may be seen in the type locality, a road section west of Lower Ashton (p. 69), and in a supplementary section along Teign Lane east of Crockham Quarry [SX 8528 8104] to [SX 8545 8070] The base of the formation is taken at the bottom of the lowest bedded chert above the Combe Shale and the top is placed to include the Posidonia beds. The apparent thickness of the formation is between 150 and 230 m, depending largely on the presence of volcanic rocks and igneous intrusions. The Posidonia beds are fairly constant at 30 to 45 m.

With the exception of radiolaria (Hinde and Fox, 1895), no fossils have been reported from the Teign Chert below the Posidonia beds. P. becheri is restricted to the lower part of the Posidonia beds and is regarded as occurring within the P₁ Zone in Great Britain. However, the presence of small forms of P. becheri in shales at the base of these beds, together with Entogonites grimmeri, indicates a B₂ age. This is confirmed by conodont evidence and by the presence of Bollandoceras micronotum in limestones within these basal shales. In slightly higher beds extremely large specimens of P. becheri occur together with trilobite fragments and goniatites including Goniatites aff. falcatus, which are indicative of the P_1a and P_1b, subzones. Neoglyphioceras spirale, present above the beds containing P. becheri, occurs in the Pld Subzone in northern England. However, its occurrence with Girtyoceras sp.at [SX 8510 8625], Hibernicoceras cf. carraunense at [SX 8546 8651], and Sudeticeras aff. ordinatum at [SX 8570 8630], all within the Posidonia beds, indicates the presence of subzones P_1c, P_1d, and P_2a respectively at these localities within the middle Teign Valley. Thus the Posidonia beds range from B₂ to P_2a, all the intermediate P, subzones being present.

Volcanic rocks within the Teign Chert

The volcanic rocks interbedded in the Teign Chert in the middle Teign Valley are mainly spilites and various types of quartz-keratophyre-tuffs. The spilites occur near the base of the formation in the northern part of the Teign Valley. Tuffs occur throughout the formation and range in thickness from a few millimetres to a mappable unit 12 m thick in the middle Teign Valley. Similar volcanic rocks occur west of Bovey Tracey but become less common southwards and are rare west of the Lustleigh Fault.

Quartz-keratophyre-tuffs

The quartz-keratophyre-tuffs have been divided into crystal-tuffs, lithic tuffs and vitric-flow tuffs, following Pirsson (1915). The mineralogy and chemical composition of the pure crystal-tuffs and vitric-flow tuffs (Table 5) show them to be soda rhyolitic, and the presence of albite or albite-rich feldspar and quartz makes them equivalent to quartzkeratophyre. The presence of (3-quartz shows that the rhyolitic magma crystallised at a fairly high temperature, between 573°C and 870°C, and the presence of large euhedral crystals suggests that the magma was in an advanced state of crystallisation prior to eruption.

Crystal-tuffs are by far the most common in the area. They consist largely of crystal fragments with subsidiary amounts of vitric and lithic material; radiolarian tests are a minor but not uncommon component. Coarse crystal-tuffs occur throughout the mapped tuff horizon and in other thick units; fine crystal-tuffs generally occur as thin layers interbedded with the cherts. Both the coarse and fine tuffs are generally well bedded. Lenticles of primary chert up to 0.2 m long are common in the coarser tuffs; they are parallel to bedding, contain radiolaria and a few crystal fragments in places, and appear to have formed as the tuffs were deposited. Graded bedding is seen only in the coarse tuffs, which in places grade upwards into finer tuffs. Their extent, bedding, grading, interstratification with chert, and contained radiolaria, indicate that the crystal-tuffs were deposited in submarine conditions from airborne pyroclastic material. The abundance of coarse crystal-tuffs suggests that a subaerial source was relatively near.

Lithic fragments of chert and lava make up less than 10 per cent of the crystal tuffs. Since there is no evidence of penecontemporaneous erosion within the bedded cherts, the chert fragments in the tuffs probably came from volcanic centres outside the area. The lava fragments generally occur as rounded brown masses up to 2 mm across, consisting of small (0.1-mm) albite microliths, commonly showing trachytic or subparallel alignment, in an amorphous fine-grained groundmass. The texture of the fragments is comparable to that of the groundmass of the spilites (p. 63). These lava fragments were derived either from pre-existing lavas or from spilitic lava brought up from depth and erupted at the same time as the acid residuum of quartz-keratophyre. Scattered devitrified small shards and fragments of vesicular pumice provide evidence of an original vitric component.

In hand specimen the coarse crystal-tuffs are seen to consist of abundant large angular fragments of white feldspar and colourless quartz, with scattered larger blebs and fragments of greenish grey lava and fragments and lenticles of grey chert, set in a fine greenish grey groundmass. Thin sections show that crystal fragments, of which feldspar is the predominant mineral, constitute between 50 and 70 per cent of the rock. The feldspar occurs as broken anhedral to subhedral laths and tablets of albite up to 3 mm in length, which generally appear cloudy owing to alteration to sericite and chlorite. The feldspars are characterised by chessboard twinning, although other types of twinning may also be present. Scattered crystals of untwinned orthoclase also occur. Colourless anhedral and euhedral quartz crystals up to 1.5 mm in size commonly show faint cracks; many are rounded and small bottle-like apophyses provide evidence of previous magmatic corrosion; many crystals also show a (3 habit. Small rounded crystals and broken euhedral prisms of colourless zircon (0.5 mm) occur sporadically.

The groundmass is mostly of pale green or yellowish green chlorite and colourless sericite forming streaky aggregates parallel to the bedding, together with varying amounts of chalcedonic silica commonly showing a slightly fibrous texture, and a fine-grained mosaic of quartz. A little calcite may also be present. In places the groundmass is composed entirely of quartz and chalcedony to give a silicified tuff, and it is possible that these tuffs originally had a considerable vitric component. The groundmass generally has a dirty appearance owing to the inclusion of fine volcanic dust. Radiolaria are locally abundant as small spherical masses, either with dark centres and clear outer zones of microcrystalline quartz or chalcedony, or composed entirely of a granular mosaic of quartz. The spire bosses, where preserved, occur as small tooth-like projections around the perimeter of the fossil.

Fine crystal-tuffs are present throughout the chert sequence of the middle Teign Valley. In hand specimen they are grey or green and very fine grained, with small crystal fragments and quartz, and a little vitric and lithic material. Thin sections show that crystal fragments constitute between 20 and 30 per cent of the rock. Feldspar occurs as small angular fragments up to 0.4 mm across, and is much altered to sericite and chlorite. A few of the crystals still show faint chessboard twinning. Quartz forms colourless rounded masses, commonly showing signs of magmatic corrosion, and small angular fragments 0.3 mm in size. Small zircons (0.05 mm) and cubes of pyrite (0.1 mm) are also present. Lithic fragments, comprising blebs of lava and rare pieces and lenticles of chert, are small, as are vitric fragments of devitrified pumice and glass shards.

The groundmass of the fine tuffs, up to 80 per cent of the rock, is composed of fine-grained quartz and chalcedony with green chlorite and colourless sericite. The chlorite and sericite occur in small shreds that are often aggregated into streaky masses parallel to the bedding. A little pleochroic biotite is also present, and the whole groundmass has a dirty appearance. Small spherical radiolaria (0.2 mm) are abundant.

The vitric tuffs vary in thickness from 0.5 to 3 m. They were first noted by Bristow (1963), who described them as deformed shard rocks of submarine origin. Thin sections show the vitric material to constitute up to 70 per cent of the rock. Most of the shards are Y-shaped, but crescent and rod shapes also occur. Both shards and pumice fragments are commonly flattened and elongated so that vesicles appear as thin streaks. Deformed shards and pumice fragments are wrapped around each other, and around the lithic and crystal fragments, and give an impression of 'flow' and welding. They are dark brown and almost isotropic under crossed polars, indicating that only slight devitrification has taken place. The amygdales are filled by pale green chlorite, chalcedonic silica, quartz and calcite. Quartz and feldspar fragments show signs of corrosion. Lithic fragments of chert and lava also occur. The matrix is a finely crystalline aggregate of chlorite and sericite, with a little calcite. In the silicified vitric tuff at Mistleigh Copse [SX 8540 8587] radiolaria occur in the groundmass.

Probably the vitric tuffs were deposited when a volatile-rich quartz-keratophyre magma erupted explosively on the sea floor and the resultant hot pyroclastic material settled nearby, possibly after flowing down a submarine slope. Heat, volatiles and physico-chemical reactions facilitated deformation of the shards under a load of only a few metres of ash. Living and dead radiolaria became incorporated in the flow of hot ash (Fiske, 1963; Mutti, 1965; Howells and others, 1973).

Spilites

Spilites occur in the northern part of the middle Teign Valley at Scatter Rock [SX 8220 8561] and also just outside the present district at Scanniclift Copse [SX 8448 8628] to [SX 8439 8606], and South Wood [SX 8261 8638].

At Scatter Rock the spilite occurs in the core of the Christow Common Syncline. The alignment and elongation of the amygdales gives the impression of 'swirl' within the flow. Dark veins cutting the spilite may represent mobilised chert caught up within the flow, or later basic injections. Similar rocks have been noted west of the Lustleigh Fault. Crags of black hornfels with remnant amygdales at Pullabrook [SX 7924 7955] have been recognised in thin section as thermally metamorphosed spilite lavas. Similar hornfelses crop out in the railway cuttings to the north [SX 7922 8006].

Chemical analyses of the spilites show them to be of similar composition to the chilled margins of the dolerite sills (Table 5). The texture of the amygdaloidal spilites is also like that of the amygdaloidal margins of the dolerite sills. Probably many albite-dolerites were intruded as near-surface sills and occasionally broke through the sea floor to be extruded as submarine spilites, commonly forming pillow structures.

The Scanniclift Copse spilite lies outside the metamorphic aureole of the granite and is 9 to 12 m thick, with pillows averaging 0.5 to 1 m in diameter. Thin sections show a groundmass composed of numerous microliths of albite, approximately 0.2 mm in length, set in an amorphous cryptocrystalline mass of black ore minerals, yellowish green and green chlorite, and a little calcite. The feldspar microliths show both multiple and simple twinning, and are generally full of inclusions of green chlorite and sericite with a little calcite. They occur in stellate clusters in some places and in sub-parallel 'flow' alignments in others. A few larger feldspars (0.5 to 1.5 mm) exhibit simple, multiple and chessboard twinning; they are much altered to chlorite, sericite and a little calcite, especially along the twin planes. Some of these feldspars show a central core of chlorite with a relatively clear outer rim of feldspar. Amygdales constitute about 50 per cent of the rock. They range from 0.5 to 5 mm, spherical to ellipsoidal or streaked, and are filled with chlorite, quartz, calcite and a little albite.

The amygdaloidal lava at Scatter Rock is identical in thin section, except that the original material filling the vesicles has been largely replaced by highly pleochroic green hornblende and a little biotite, and the majority of the feldspar altered to labradorite. These changes reflect metamorphism by the granite.

Felsitic lavas

The term 'felsitic lavas' is applied to finely banded, pale grey, very fine-grained rocks that exhibit a felsitic texture and occur interbedded amongst the cherts. They are exposed only in Brimley Copse [SX 7977 7702], where they form beds or flow up to 0.6 m thick. Fine laminae, commonly picked out by hematite stringers, exhibit small 'flow' folds of a recumbent nature. In thin section the rock appears largely recrystallised, consisting of a mosaic of quartz and a little chalcedony; feldspar has not been identified positively. Pyrite is typically abundant and hematite veinlets are common. In plane-polarised light a vague remnant texture, in which elongate lenses of the quartz mosaic are picked out by stringers of hematite and other oxides, gives the impression of welded glass shards that have been completely devitrified and silicified.

Mount Ararat Chert

The Mount Ararat Chert occurs only in the Liverton Unit. It is significantly thinner than the equivalent beds in the Teign Valley and contains only subordinate volcanic rocks.

The bedded cherts of Ramshorn Down and their associated radiolarian fauna were first described by Hinde and Fox (1895). Ussher (1913) noted that these beds rested directly on Upper Devonian slates, with no dark grey shales and mudstones [Combe Shale] between. Bristow (1962, 1963) used the term "Ramshorn Down Cherts" to describe all the cherts south of Ramshorn Down and Rora Down, and included siliceous rocks west of the Ilsington Fault in his "Sigford Beds". The cherts on both sides of the Ilsington Fault and south of the Silverbrook Thrust are here considered to belong to the Mount Ararat Chert.

The formation consists mainly of bedded siliceous rocks and slates, with local interbedded pyroclastic rocks. The type section is a Forestry Commission track immediately west of Mount Ararat [SX 7973 7477] to [SX 7972 7482]. Here the junction with the Rora Slate is taken at the top of the uppermost siltstone of the latter formation (p. 52). The Ararat Slate at the base, less than 10 m thick and not separately mapped, comprises black slates weathering to greyish white, their lower part siliceous and containing scattered orange decalcified nodules up to 100 mm across. The Ararat Slate passes upwards within 0.6 m of strata into regularly alternating black bedded chert and black siliceous shale. The type section has yielded a single oral imprint of Pseudopolygnathus triangulus subsp. indet., and Dr S.C. Matthews considered this conodont to range from late Siphonodella triangula inaequalis Zone to top Scaliognathus anchoralis Zone. Probably the base of the Ararat Slate lies within the Gattendorfia Stufe.

The cherty division of the Mount Ararat Chert is unlikely to exceed 60 m in thickness. Its basal part consists of black faintly laminated cherts up to 0.25 m thick interbedded with black shales. Above this occur alternating bedded chert and more or less siliceous mudstone with thin shale beds. The siliceous beds are usually 25 to 100 mm thick though some range up to 0.6 m. They are well defined, blocky, and well jointed, and show every gradation from very fine-grained flinty chert to fine-grained siliceous mudstone. Coarser grained varieties range from siliceous mudstones to mud-stones, and in thin section contain some detrital quartz in the silt grade. Black and grey colours predominate over green and purple. Radiolaria are locally abundant. The bedded cherts of the formation have yielded only a single bar-type conodont mould from loose debris west of Bethelcombe Cross [SX 7857 7427].

The fauna of the Posidonia beds in the Mount Ararat Chert (p. 176) consists predominantly of a pelagic goniatite-pectinoid association, but there is also some evidence of shelly benthos in the form of crinoid debris, bivalves and costate brachiopods. The lowest Posidonia beds recorded are grey buff-weathering shales of Viséan P_1b age, locally siliceous and containing abundant P. becheri. Either the basal strata of the Posidonia beds as seen elsewhere are not exposed or there is a facies change, the uppermost cherts being of B₂ to P_1a age. In the Liverton Unit beds of P_1d, to P_2a age are in some places black, grey and green siliceous slates, mudstones and cherts, and in others black shales, commonly slightly siliceous.

Volcanic rocks within the Mount Ararat Chert

Volcanic rocks in the Mount Ararat Chert are virtually confined to the northern edge of the outcrop. They consist of vesicular spilite lavas and various types of quartzkeratophyre-tuff, as in the Teign Chert. Fine to coarse, dark greenish grey and black, quartz-keratophyre crystal-tuffs are abundant. In addition to shards of pumice and glass, they contain albite crystals and fragments of lava, the latter commonly flattened. Bristow (1963) noted the presence of flattened and deformed shards. The only evidence of pyroclastic rocks in the southern part of the outcrop is a 20-mm band of red clay interbedded with cherts on Mount Ararat [SX 7973 7477]; it is interpreted as a decomposed fine-grained tuff.

Winstow Chert

The Winstow Chert near Chudleigh was described by House and Butcher (1973) as about 23 m of cherts with radiolaria, sponge spicules and conodonts. They defined the base as the lowest black siliceous and sooty slates above the Whiteway Slate, the top as a conodont horizon within a lithological passage into Posidonia beds, and they cited a quarry [SX 8630 7801] by Winstow Cottages (Ussher, 1913; Hinde and Fox, 1895) as the type locality. The Posidonia beds are here included within the Winstow Chert and are taken to include all the siliceous shales with Posidonia below the Ugbrooke Sandstone. No volcanic rock has been found within the formation. The Winstow Slate, a basal member consisting of black siliceous and sooty slates, is not separately mappable. Its junction with the underlying Whiteway Slate seems to be sharp [SX 8630 7801]. Trenching at Whiteway Barton (Figure 11) suggested a thickness of about 14 m, but the slates are much thinner at Winstow Cottages. Probably the latter sequence is condensed, since the thicknesses of the Gattendorfia Stufe at the two localities are 9 to 10 m and 1 to 2 m respectively, but it is possible that the thickness has been reduced by faulting (House and Butcher, 1973).

The main part of the formation above the Winstow Slate consists of green or greyish green, black and purple cherts, blocky siliceous mudstones and shales. The more siliceous beds are well jointed and rarely exceed 76 mm in thickness. There is no lamination or any evidence of current action. Radiolaria are common. Some of the siliceous mudstones are indistinguishable from unfossiliferous Posidonia beds. The latter consist of about 10.5 m of shales, cherts and limestones. The shales are black, greyish green, green, buff, red and purple, and are interbedded with blocky siliceous mudstones and thinly bedded cherts which locally bear abundant radiolaria. Limestones occur as thin beds in dark green siliceous shales and as partings and lenses in black shales.

The Winstow Slate is generally barren but House and Butcher (1973) obtained Chondrites and other ichnofossils, as well as external moulds of conodonts, from the pit [SX 8630 7801] near Winstow Cottages. Matthews (1969) referred the conodonts to the Siphonodella crenulata Zone and reported (1970) that this horizon post-dated the Gattendorfia Stufe and predated any known Western European Ammonellipsites Stufe cephalopod fauna. The cherts and siliceous mudstones above the Winstow Slate yielded only sponge spicules and radiolaria during the recent survey. However, House and Butcher (1973) recorded conodonts from a trench [SX 8628 7801] in the upper part of the formation near Winstow Cottages which Matthews (1969) placed in the anchoralis Zone. Other pelagic elements included chitinophosphatic brachiopods and ostracods. At Mount Pleasant, siliceous black shales and cherts just below the Posidonia beds yielded (House and Butcher, 1973) Phillipsia leei, Archegonus (Phillibole) aprathensis and Archegonus (Waribole?) laevicauda.

The Posidonia beds bear a rich fauna of solitary corals, brachiopods, posidoniid bivalves, goniatites, some orthocone nautiloids, trilobites and conodonts (Appendix 2). Ostracods are locally abundant and radiolaria are usually present in the more siliceous rocks. During the recent survey the Visean subzones PI, and Pea have been recorded, but House and Butcher (1973) suggested a range from B₂ to B_2c. They noted a fauna from a trench north-west of Castle Dyke, including P. becheri, P. membranacea, Goniatites granosus, Hibernicoceras sp., N. spirale, Pronorites ludfordi, Sudeticeras crenistriatum, Archegonus (Phillibole) sp., A. (Waribole?) sp.and Phillipsia leei. A goniatite collected by W.A.E. Ussher from "¾ mile south of Ugbrooke House, near Chudleigh", was referred to the Paragoniatites newsomi Group of P₂, age by Butcher and Hodson (1960), but has been redetermined as Lyrogoniatites sp.of P₂ age in George and others (1976). It is therefore most likely that, as elsewhere in the area, the Posidonia beds are no younger than Pea.

Upper Carboniferous

Ussher's (1892b) Exeter-type Culm and Ugbrooke Park Beds are now identified as Crackington Formation and Ugbrooke Sandstone. Both appear to be youngest Visean to Namurian in age; the Crackington Formation contains E₂ fossils in its lower part and the Ugbrooke Sandstone a tentatively dated E or H fauna at its base.

The Crackington Formation consists of shales with subordinate turbiditic sandstones equivalent to similar strata north of Dartmoor (Edmonds and others, 1968). The Ugbrooke Sandstone comprises bluish grey, locally pebbly sandstones with subordinate conglomerates, mudstones and slates and is interpreted as a marginal flysch facies. Several authors have suggested that its base is an unconformity (Godwin-Austen, 1842; Holl, 1868; Ussher, 1892b). Somervail (1898) regarded the Ugbrooke Sandstone as post orogenic as well as unconformable, a view endorsed by Simpson (1959) and Prentice (1962), but House and Butcher (1973) showed by trenching that on the Chudleigh escarpment Posidonia beds were overlain conformably by Ugbrooke Sandstone. More recent trenching at Whiteway Barton has confirmed this relationship.

Crackington Formation

The Crackington Formation conformably succeeds the Posidonia beds of both the Teign Valley and Liverton successions and is represented in the Teign Valley, Ilsington and Liverton structural units. The Ashton Shale has been mapped at the base of the formation. Its type locality is 229 m south of Spara Bridge (p. 73) and its base is exposed in Ruggadon Lane [SX 8560 8166]. It consists of micaceous shales with sporadic siltstones and sandstone bands which rarely exceed 25 mm in thickness. The shales are pale greyish blue when fresh, but commonly weather to rusty red. Pencil cleavage is locally characteristic. Dark sooty bands, usually less than 5 mm in thickness, in the upper part of these shales have yielded goniatites of Namurian Etc age. The change from Teign Chert to Ashton Shale is reflected in the topography, but the top of the Ashton Shale, mapped at the base of the lowest prominent sandstone of the Crackington Formation, is less clear. The thickness of the Ashton Shale in the Teign Valley ranges from 90 to 210 m; west of the Lustleigh Fault it is either attenuated or absent.

Above the Ashton Shale individual sandstones range in thickness from a few millimetres to a metre. Elongate flute casts occur on the soles of thick sandstones. The soles of thinner-bedded sandstones commonly have small groove casts a few millimetres across. Graded bedding is widespread. However, medium to fine-grained sandstones commonly pass up into laminated silty sandstones rich in mica and plant fragments, which are usually current orientated, and these in turn fine upwards through siltstones into black shales. Ripple-drift bedding is common near the top of such turbidite units, usually in fine siltstone. Parallel lamination may occur in the upper part of the thicker units and throughout the thinner sandstones.

Palaeocurrent evidence is insufficient to define sources of sediment, but the unusually coarse sandstones and conglomerates in the Crackington Formation of the southern part of the outcrop suggest a passage into the Ugbrooke Sandstone, a more proximal group of turbidites lying farther to the south.

Fossils are rare in the Crackington Formation. Goniatites have, however, been found in thin sooty bands within dark grey slates in the middle Teign Valley. Cravenoceratoides sp.from the Ashton Shale at [SX 8432 8385] indicated an E_2b age, and fossils from the lower parts of the overlying strata, including Nuculoceras nuculum from [SX 8642 8485] were of E_2c age. Some sandstones and silty mudstones are characterised by comminuted plant fragments, presumably carried by the turbidity currents.

Ugbrooke Sandstone

The Ugbrooke Sandstone occurs only in the region of recumbent folds south of the Bickington Thrust; it is represented at the surface in the Ugbrooke and East Ogwell units and was proved in the Abbotskerswell Unit in the Rydon Ball Farm Borehole [SX 8436 6929] (Appendix 5). The formation rests conformably on Posidonia beds and a trench at Mount Pleasant [SX 8703 7865] has yielded immature goniatites, possibly belonging to the Homoceras beyrichianum group or Cravenoceratoides sp.; Phillipsia sp., orthocone nautiloids and poorly-preserved ostracods were also present. This fauna is believed to represent Namurian E or H (House and Butcher, 1973). The type section of the formation is a quarry [SX 8690 7802] in Ugbrooke Park. It has been possible to map members consisting mainly of sandstone and conglomerate or of mudstone and slate. The basal facies consists of about 9 m of black and greenish grey laminated mudstones and siltstones with sporadic thin sandstones. It is overlain by the massive sandstone facies, which was penetrated by the Rydon Ball Farm Borehole (p. 183), and consists dominantly of coarse- and medium-grained sandstones which are locally pebbly and in places pass into fine conglomerates. Subordinate lithologies within this sandstone facies include fine sandstones and siltstones with rare black shales and mudstones. The sandstones occur in massive units from 0.3 to 5.2 m thick, commonly separated by thin black shales. Many of the sandstones seem structureless, but a crude internal grading is seen in some larger exposures. Characteristically a basal erosion surface is overlain by massive sandstones, coarse or even conglomeratic at their bases and fining upwards. Some thin beds show good grading, and repeated and reverse grading occur; thus pebbly sandstones are locally present high in the sequence. Pebbles are scattered and generally sub-rounded or rounded and less than 10 mm in diameter. Angular fragments of black shale, mainly in the basal part of each unit, were probably derived by penecontemporaneous erosion. Chips of black chert and green siliceous mudstones are less common. Larger fragments (up to 0.45 m long) and pebbles (up to 0.15 m diameter) of shale, mudstone and siltstone show local flame structures indicating that they were still plastic when incorporated in the sediment. Plant debris is common, and fossil logs up to 0.2 m long have been noted. Above the massive sandstones are finer sandstones and siltstones showing parallel lamination and ripple-drift bedding and containing many carbonised plant fragments. The massive sandstone facies is believed to represent proximal turbidites deposited from dense turbulent suspensions.

The pebbles and cobbles of the coarser conglomerates are sub-rounded to rounded and commonly of chert, sandstone and vein-quartz. The larger cobbles and boulders of sandstone, and angular pieces of fine-grained laminated sandstone, siltstone and shales, are characteristic Ugbrooke Sandstone lithologies. Some show wet sediment distortion within their matrices, and the black shale slabs are commonly slickensided. A sandstone boulder [SX 8770 7577] near New Linhay measured 0.5 X 0.4 m. The massive cobble-conglomerates and pebble-conglomerates lack grading, and their pebble imbrication suggests traction flow; possibly very immature turbidites initiated on steep slopes could carry this coarse material short distances. The coarse sandstones and conglomerates contain a variety of rocks.

Granite abounds, generally a fine-grained granodiorite with feldspar laths up to 5 mm long, brown biotite and anhedral quartz; plagioclase exceeds two-thirds of the total feldspar, but partly kaolinised orthoclase and microperthite occur.

Alkali microgranite, with dominant orthoclase and abundant biotite, is less common. Aplite and microgranophyre are rare. Quartz- and feldspar-porphyries are common.

Spilite fragments are plentiful, and commonly deeply weathered and iron stained. They show albite in a chlorite matrix, and in some places contain amygdales filled with chlorite and chalcedony. Similar rocks with albite, a little quartz and irregular patches of chlorite may be spilites or fine-grained dolerites.

A single fragment of gneissic rock was found, and a few of quartz-mica-schist. Quartzite is abundant, some wholly recrystallised, some with a relict detrital texture. Other fragments include much sheared greywacke, slate, chert, sandstone and vein quartz, and rare limestone. Pebbles, angular fragments and constituent mineral grains of quartz, feldspar and mica of the sandstones and conglomerates are embedded in chlorite and sericite.

The dominant rock fragments suggest a southern provenance; the metasediments could have come from a metamorphic province which included the Start schists, and the quartzites, sandstones and slates could have been derived from Devonian rocks similar to those now found north of the Start schists. A metagreywacke, similar to the Ugbrooke sandstones, may indicate the presence of beds of deformed Ugbrooke facies in the source area. Deformed and undeformed Lower Carboniferous chert pebbles occur, but Middle Devonian limestone pebbles are absent. The source of the granite fragments is unknown, but the spilite and fine-grained basic igneous rocks resemble those of the Middle and Upper Devonian of the southern successions.

Pebbly mudstone, consisting of pebbles of quartz and feldspar and rock fragments up to 10 mm in diameter scattered through a silty mudstone matrix, occurs locally [SX 8440 6990]; [SX 8440 6810]. Black sandstones with a muddy matrix occur as distinctive horizons in the massive sandstones of the Rydon Ball Farm Borehole (p. 183). They contain many fragments of mudstone, silty mudstone and fine-grained laminated sandstone, randomly orientated and commonly folded into 'slump overfolds' (Crowell, 1957), lumps of coarse cleaner sandstone of massive sandstone facies up to 50 mm across, and siderite nodules; all of which point to considerable penecontemporaneous erosion. An incipient cleavage, commonly picked out by carbonaceous streaks, is characteristic of these muddy sandstones, but has not been noted in other sandstones of the Ugbrooke Sandstone.

A facies of black slates and mudstones with laminae of siltstone and fine sandstone crops out extensively. Thin laminae of coarse sandstone were noted south-west of Combe Holdridge, and bands of small siderite nodules in a lane [SX 8720 7582] south of Combe. The mudstones commonly exhibit incipient bullions up to 0.15 m diameter that give many outcrops a nodular appearance. Thin silty and sandy bands within the mudstones occasionally show parallel or small-scale cross lamination, and are turbidites. Thicker turbidite sandstones are rare. They are commonly 0.15 to 0.45 m thick, have flat bases, locally display small groove casts and scours and contain chips of black mudstone in their basal parts, and grade upwards into silty sandstone rich in plant fragments and through siltstone into black mudstone.

The Ugbrooke Sandstone probably formed at or near the base of a continental slope. The proximal turbidites of the massive sandstone facies are thought to be near-source, submarine-fan deposits, laid down at the slope base, and the muddy sandstones in the massive sandstones may represent rapidly deposited immature dense turbidity currents. The lenticular conglomerates are interpreted as infilled braided or meandering channels on the fans. Fluid or semi-fluid mudflows probably gave rise to the pebbly mudstones, which are typical lower-slope or base-of-slope deposits (Crowell, 1957). The background sediments of the lower slope or base of slope are probably represented by the black mudstone facies, though the variable amount of thin turbidites within the facies demonstrates that turbidity currents invaded most areas from time to time.

Details

Lower Carboniferous

Trusham Shale

Ilsington–Narracombe

A road section [SX 7853 7628] north of Ilsington School exposes 1.2 m of dull green micaceous slates. Near the base of the cutting a band of black siliceous mudstone 0.25 m thick runs parallel to the slaty cleavage. Rare small siliceous nodules occur in the slates. The slaty cleavage dips at 16°/240° and a second cleavage at 52°/135°. Greyish green mottled slates exposed in the road sections [SX 7850 7610] near Court Barton Farm, show a slaty cleavage dipping at 20°/174° and later cleavage 50°/128°. Dull green slates with very faint sporadic rusty spots crop out behind a barn at Narracombe [SX 7867 7678] and dip at 35°/048°.

Area south of Lustleigh

Within the Lustleigh–Bovey Tracey fault zone metamorphosed rocks mapped as undifferentiated Hyner and Trusham shales are recognised solely by their stratigraphical position. They occur in an anticlinal core. Exposure is extremely poor but the dominant lithology is light to medium grey spotted fine-grained quartzbiotite-cordierite-hornfels, though darker, rather more siliceous, hornfelses do occur. The presence of dolerite sills within the hornfelses suggests that the beds are older than the Crackington Formation, for in south-east Dartmoor no intrusive sills are known in the latter.

At the side of a forestry track [SX 7872 8002] 0.6 m of grey fine-grained siliceous hornfels occur below a crag of microgranite. West of Packsaddle Bridge, a small road section [SX 7875 8025] exposes grey weathered spotted slaty hornfels and black quartzschorlite-hornfels in contact with rotten microgranite. Massive grey hornfels 3 m thick is in irregular contact with a microgranite sill in the northern part of Packsaddle Bridge Quarry [SX 7889 8024]. In the railway cutting at [SX 7890 8032] crags show grey very fine-grained quartz-biotite-cordierite-hornfels dipping at 30°/120°, invaded by two large microgranite sills. A specimen from 0.6 m below one microgranite contains subrounded biotite crystals, irregular and rounded areas of sericitised cordierite, recrystallised quartz mosaics and muscovite.

Teign Village–Trusham–Christow

A roadside exposure [SX 8403 8097] east of Teign Village shows 3.5 m of pale grey and greenish grey micaceous shales which yielded indeterminate bryozoa and brachiopods. A track north of Leigh [SX 8426 8113] exposes 64 m of green shales with Bertillonella (B.) sp. nov; Maternella sp. nov. aff. M. circumcostata and Richterina (R.) latior. A similar fauna was obtained from shales exposed by ploughing [SX 8430 8130].

Along Teign Lane, in the core of the Ranscombe Anticline, mottled green shales at the top of the Hyner Shale pass into more micaceous pale grey shales of the Trusham Shale [SX 8475 8225]. The latter at Manor Farm [SX 8488 8224] and to the east [SX 8513 8221]; [SX 8533 8220] contain brown decalcified bands; dips are steep to north and south and a pronounced east-west vertical cleavage is also present. Juvenile brachiopods and other indeterminate fossils have been found at the last two outcrops.

Greyish green micaceous shales overlying calcareous horizons at the top of the Hyner Shale near Christow [SX 8405 8482] to [SX 8403 8477] dip at 70°/174° and have yielded Sanguinolites? ellipticus. On the north side of the road south-east of Southleigh Cottages [SX 8505 8540] to [SX 8515 8544] S.? ellipticus and juvenile brachiopods are present in pale grey micaceous shales dipping at 70°/176° with joints dipping at 88°/250° and 70°/012°.

Combe Shale

Ilsington area

At the western end [SX 7860 7637] of the road section north-east of Ilsington the transition from green slates of the Trusham Shale is seen. Discontinuous exposures of greyish black slates with small siliceous nodules occur along the road, and at the eastern end the slates are more siliceous and lie close beneath the metamorphosed cherts and limestones of the Teign Chert seen in the adjacent Slipping Stones Quarry [SX 7897 7647]. Both Ussher (1913) and Bristow (1963) referred the black slates to the Upper Devonian.

Gradner Rocks [SX 7812 8022] to Knowle

Immediately to the south of the granite that forms Gradner Rocks occur large blocks of pale grey granular hornfels and black quartz-schorlite-hornfels. In thin section the latter is seen to consist of quartz and zoned bluish grey tourmaline (schorlite) with a little biotite, and to show traces of possible relict bedding.

In Drakeford Bridge Quarry [SX 7895 8009] pale grey fine-grained biotite-rich hornfels, 4.5 m thick, contains horizons of tourmalinised dark hornfels and appears to dip at 48°/080°. Zones of deeply weathered hornfels, probably hydrothermally altered by an adjacent microgranite, run parallel to the bedding. On the western side of the quarry a track exposure shows 1.5 m of greyish black fine-grained siliceous hornfels, weathering pale grey, in contact with a microgranite, the contact dipping at 5°/075°. Small pits adjacent to the track show rotten chloritised microgranite containing blocks of tourmalinised hornfels up to 0.3 m diameter.

Crockham Quarry–Trusham–Lower Ashton

Black shales in Crockham Quarry crop out above and below the main dolerite intrusion and dip at 50°–60°/150°. Shales adjacent to the dolerite are baked white. Kink bands occur in the south-east corner of the quarry [SX 8476 8068]. A roadside exposure [SX 8500 8081] to [SX 8505 8078] shows similar black shales with fine white laminations. They become increasingly siliceous upwards and pass conformably into thinly bedded cherts of the Teign Chert.

Black siliceous shales in the roadside above Tinkley Quarry [SX 8480 8160] show darker silty laminations. In the track above Whetcombe Quarry bluish black shales dip at 79°/178° and contain small siliceous nodules [SX 8452 8168] to [SX 8445 8186] and fossil wood [SX 8452 8169].

At Hill Crest [SX 8537 8203] pale grey vertical metamorphosed shales darken away from the dolerite and pass into black siliceous shales dipping at 80°/350°. Jet-black fissile shales along the Bramble Brook [SX 8590 8200] to [SX 8650 8182] are in the core of the Ranscombe Anticline. Along the railway west of Rydon hard bluish black shales [SX 8428 8256] with minor easterly plunging open flexures dip at 70°/006°. Dendritic manganese is common on the weathered surfaces. In the stream section east of Broom Hill [SX 8451 8314] to [SX 8454 8316] black shales at the top of the Combe Shale are overlain by bedded cherts of the Teign Chert and dip at 80°/135°.

In a field north-east of Warren Clumps [SX 8494 8408] small outcrops of black shales dip at 86°/168° and are characterised by fine vertical laminations, commonly filled with white material, dipping at 81°/070°. The type locality for the Combe Shale is the road section east of Ashton Mill Weir [SX 8402 8470] to [SX 8411 8440]; it shows bluish black shales which dip at 84°/160°. At their base the shales contain small siliceous lenticles and are greenish grey and micaceous. In a quarry at Hill Copse [SX 8407 8445] black shales underlie a dolerite sill. Within the sill a 1.2-m band of adinole represents a raft of shale caught up in the intrusion. Above the sill 0.3 m of adinolised sediments are overlain by 1.2 m of white spilosites capped by black shales.

In the railway cutting north-west of Lower Ashton [SX 8417 8459] to [SX 8421 8447] white spotted spilosites, which overlie the dolerite of Ryecroft Quarry, pass up into 9 m of soft black shales with fine white laminations parallel to the bedding, dip 60°/132°. Black shales, becoming increasingly siliceous upwards, pass into cherty mudstones of the overlying Teign Chert at [SX 8421 8447]. Near Great Cleave Barn [SX 8588 8568] black shales dip at 50°/150°. Siliceous black shales in a track to the north of Mistleigh [SX 8546 8567] dip at 75°/170° and pass conformably into overlying cherts of the Teign Chert.

Teign Chert

Ilsington–Brimley

Within the Ilsington Unit, south of the Narracombe Thrust, the Teign Chert dips gently north-east, east or south-east within recumbent folds. A well-marked calcareous horizon at the base has suffered selective thermal metamorphism during the emplacement of the Dartmoor Granite. It contains calc-silicate assemblages in contrast to the Combe Shale and Trusham Shale immediately to the west, which are only faintly and sporadically spotted.

In the north bank of the track [SX 7895 7584] running past Silverbrook Mine, thickly bedded black cherts and brownish pink, faintly calcareous, cherts dip at 35°/140° and are overlain by grey siliceous slates with two 75-mm bands of greyish green and white banded chert. In Lenda Wood Quarry [SX 7920 7584] 12 m of pale green, mauve, grey and cream-coloured cherts contain scattered bluish black cherts and greyish pink sugary-textured calc-flintas. In some parts of the quarry the cherts are cut by fine veinlets of dark green epidote a few millimetres thick. Hinde and Fox (1895) reported ill-defined radiolarian casts in thin sections of these cherts. The eastern face of the quarry shows part of an open recumbent fold whose axis trends 140°; a horizontal fracture cleavage is present in the fold core.

In the southern side of Old Town Hill [SX 7937 7595] buff-weathering siliceous blocky mudstones with a few horizons of bluish black chert dip at 10°/134°. A 75-mm band of fine- to medium-grained crystal-tuff is interbedded with the mudstones at the western end of the section.

Slipping Stones Quarry [SX 7897 7647] exposes a strike section in 7.5 m of cherts, calcareous cherts and limestones dipping at 26°/026°. The bottom of the quarry is situated about 9 m from an exposure of Combe Shale in the Ilsington road section (p. 67) and 3 m stratigraphically above it. In the quarry, black and dark grey cherts and pale green, mauve and cream calc-flintas pass up into interbedded calc-flintas and decalcified calcareous cherts, limestones and shales. The regular interbedding of cherts and calc-flintas with decalcified siliceous limestones, now represented as brown rottenstones, is seen only in the upper half of the quarry. Locally they show poorly developed load casts and small scale slumping of laminae.

Bluish black cherts with black slate partings are common in Narracombe Wood. East of the easternmost tributary of Liverton Brook the cherts are more thinly bedded, include horizons of blocky buff-weathering siliceous mudstone and appear to contain more radiolarian casts. Such cherts at Woodhouse [SX 7930 7638] are bluish black and contain abundant radiolarian casts. Ussher (1913) described buff shales with Posidonia and goniatites at Woodhouse Cross [SX 7944 7641] and at Woodhouse; probably a narrow outcrop of Posidonia beds in this area is truncated by the Narracombe Thrust.

A quarry [SX 7804 7779] west of Ullacombe shows about 12 m of bluish black cherts with sporadic spotting in the less siliceous bands, in beds 51 mm to 0.25 m thick, dipping at 38°/041°. At the eastern edge of Brimley Copse [SX 7982 7706] blocks of chert breccia consist of angular fragments of grey or black chert up to 25 mm across set in a siliceous cement. At the north-western edge of the copse [SX 7977 7702] small crags show similar breccia overlain by black cherts which are capped by 0.6 m of grey finely banded felsitic 'lava'. Crags of such 'lava' occur above, with flowage folds developed in the banding.

Pullabrook–Parke

Between the Lustleigh and Bovey Tracey faults the Teign Chert and its dolerite sills lie in the south limb of the southerly overturned Trusham Regional Anticline and form prominent ridges. Crags by the bank of the River Bovey [SX 7936 7978] to [SX 7949 7977] north-east of Pullabrook expose about 6 m of black bedded cherts that grade into spotted black siliceous slates. Scattered thin bands of pale-coloured blebs a few millimetres in diameter occur in the cherts. At [SX 7954 7968] slates form about 25 per cent of the exposed rock, the cherts are flinty and contain thin pale streaks of fine-grained ?calc-silicate, and the dip is 40°/009°.

Immediately south of Pullabrook Halt, on the eastern side of the disused railway [SX 7929 8002], occur small crags of dark green sugary-textured pyroxene-plagioclase-hornfels. Banding, by zones composed of sheaves of pale green to colourless diopside, probably represents the original bedding. The rest of the rock is composed of scattered large plagioclase crystals set in granular aggregates of pyroxene and plagioclase with a little quartz and sphene. Diopside is locally rimmed by green amphibole. At the southern end of the railway cutting to the south-east [SX 7990 7976] 9 m of thickly bedded black cherts dip at 35°/100°.

At the southern end of a road section [SX 7988 7973] to [SX 7959 7991] 6 m of rusty-weathering and brecciated black cherts with faint white laminations and occasional partings of black-spotted slate dip at 32°/120°. In the northern part of the section slates up to 1 m thick are interbedded with the cherts; bedding dips at 75°/076°. The southern edge of Ledge Wood is just outside the aureole. Black siliceous slates with thin chert bands dip at 65°/027° [SX 8009 7887]. At the eastern end of the wood [SX 8052 7874] a cutting exposes greenish grey siliceous mudstones dipping at 50°/102°. Immediately to the north a crag exposes massive white-weathering black pebbly slate that contains numerous clasts of black mudstone, siliceous mudstone and bluish black chert, and is presumably interbedded with the cherts and siliceous mudstones. The less siliceous clasts are sub-rounded or disc shaped, but the chert fragments have remained rhomboidal. Clasts form up to 45 per cent of the rock and occur in the slate matrix with their long axes parallel to the bedding or at an angle of up to 20° to it; they vary in size from 10 to 15 mm. Large blocks of similar rocks, though extensively silicified, have been noted in Brimley Copse [SX 7983 7706].

Bovey Tracey–Ranscombe–Beardon Hill

At the roadside north of Frost Cross [SX 8330 7972] bluish grey bedded cherts, full of radiolaria, 3 m, lie south of a quarry in coarse crystal-tuffs. In a road section at Lower Crockham Copse [SX 8505 8078] to [SX 8515 8062] cherts dip at 50° to 55° south-eastward. A unit of coarse crystal-tuff 18 m thick crops out at [SX 8513 8070] in the middle of the section. At [SX 8522 8057] soft pale grey white-weathering mudstones are 9.1 m thick, dip at 55°/146° and contain P. becheri; Goniatites sp.in the higher beds indicates a high P, age.

Beside Teign Lane, in Farley Copse, cherts with interbedded coarse crystal-tuffs dip at 50° to 70° south-eastward [SX 8540 8094]. At [SX 8544 8085], 4.6 m of well-bedded white siliceous cherts and mudstones, overlying well-bedded pale grey to bluish grey cherts, include a thin silicified limestone band at their base and dip at 70°/160°. The following fossils were collected: crinoid debris, zaphrentoid coral, Crurithyris? and P. cf. becheri, probably belonging to the P, Zone. Hinde and Fox (1895) recorded unidentifiable radiolaria from this locality. Bedded bluish black cherts 18 m thick, seen in a pylon excavation [SX 8505 8145] north-east of Crocombe Bridge, are folded into a 6-m-high anticline overturned to the south-east and pass down into black siliceous shales of the underlying Combe Shale in the south-western part of the exposure. At Ruggadon Farm [SX 8596 8163] soft grey mudstones and shales, 0.9 m, dip at 76°/182° in a bank just outside the orchard and contain P. aff. becheri of the B₂ Zone. The cherts in the core of the Ranscombe Anticline are cut by the Bramble Brook at Higher Ranscombe, and stream sections reveal that minor faults have developed as accommodation structures near the crest of the fold. This faulting, which is not shown on the map, disrupts the stratigraphy locally; thus the Posidonia beds are infaulted at [SX 8645 8170].

A pylon excavation [SX 8629 8229] north-west of Shortridge revealed pale grey shales and mudstones and thinly bedded cherts, 4.6 m, dip at 70°/030°. Debris yielded Euchondria dentistria, Coleolus sp., P. aff. becheri, Entogonites grimmeri, Goniatites cf. moorei and indeterminate orthoceratids and prolecanitids, of B₂ age.

Black sooty shales in the bank of the stream south-east of Trusham Cross [SX 8588 8257], 0.9 m, are overlain successively by thinly bedded cherts and siliceous shales dipping at 66°/030° in the stream bed, 0.6 m, thin hard black mudstone containing H. cf. carraunense and N. spirale of the Pk, Subzone, 0.05 m, and thinly bedded shales and siliceous black mudstones, 1.1 m. The track to Rydon Farm [SX 8570 8268] shows black siltstones and shales, 6.1 m, which yielded Girtyoceras? and N. cf. spirale of P_1d, age. In the stream east of Broom Hill cherts overlie black shales of the Combe Shale [SX 8456 8317] and dip at 80°/120°. Farther upstream a porphyritic dolerite is overlain by cherts dipping at 70°/122° [SX 8475 8314] and including 9 m of green fine-grained tuffs [SX 8472 8314].

On Beardon Hill shallow trenches [SX 8557 8405], presumably old excavations for manganese, expose white-veined cherts dipping at 45°/090° and containing radiolaria. To the north-west, vertically bedded red jasper chert strikes east-west in the old manganese workings of Ashton Mine [SX 8535 8418]. These beds are impregnated with black manganese ore. Thin sections show the jasper to be composed mainly of chalcedonic silica, with scattered patches of granular quartz. Red flecks of iron oxide, probably hematite, give the rock its red or pink colour. This hematite probably originated as volcanic dust. Radiolaria are abundant and joints dip at 88°/270°.

West of Ashton Mine the chert outcrop is disrupted by the Lower Ashton Stretch Thrust, which cuts through the southern limb of the Ashton Syncline and carries the formation southward to Canonteign Barton.

Canonteign House–Higher Barton–Christow Common

At the roadside north of Canonteign Barton [SX 8382 8334] grey shales and siliceous mudstones, 6.1 m, dipping at 64°/134°, yielded fragments of P. becheri of P₁ age. A quarry [SX 8412 8375] in Byteign Plantation reveals black thinly bedded cherts and white mudstones, 1.8 m, dip 60°/133°. P. becheri within the mudstones indicates a B₂ age.

The road section west of Lower Ashton serves as the type locality for the Teign Chert. The following upward succession is exposed from north to south: Combe Shale overlain by 45 m cherts and siliceous shales containing 6 m coarse crystal-tuff [SX 8413 8439], dip 40°/160°; 21 m cherts and shales with a few thin bands ofjasper; 6 m brecciated mineralised bluish grey cherts with small veins and vughs of quartz and pyrite and irregular pockets of botryoidal pyrolusite and psilomelane, a manganese deposit worked in the past in the adjoining wood [SX 8414 8432]; 94 m radiolarian cherts and siliceous shales. The succeeding 9 m of radiolarian cherts and siliceous shales, which are poorly exposed, are followed by 6 m black thinly bedded cherts, shales and fine tuffs folded into a tight northerly overturned anticline plunging at 58°/080° [SX 8415 8432]. Following a gap with black shale debris [SX 8415 8430] to [SX 8416 8426] the following upward succession is exposed at road level; a coarse tuff band full of angular fragments of black chert, 0.3 m, overlain by bluish grey bedded cherts, 1.8 m, and black siliceous shales with thin chert bands dipping at 50°/150°, 5.5 m. About 9 m above road level, green fine-grained well-bedded tuffs, dip 60°/140° 4.6 m, are overlain by thinly bedded cherts and buff-weathering shales folded into northerly overturned flexures, 0.9 m, black chert bands with interbedded fine green tuffs, 10.7 m, coarse crystal-tuffs dipping at 53°/157° and full of lenticles of chert aligned parallel to bedding, 6.1 m. At road level 1.8 m of bluish black cherts form open folds plunging at 22°/092° [SX 8423 8417]. In a recess 4.6 m above road level 2.4 m of cherts and vitric tuffs appear in tight sheared folds overturned to the north, with axial planes dipping at 60° southward and plunging at 35°/098° [SX 8424 8415]; succeeded by 9 m of thinly bedded cherts, shales and fine green tuffs with fine quartz veining in cherts and with minor flexures plunging at 30°/090°. At road level from [SX 8422 8419] to [SX 8427 8415] 55 m of well-bedded bluish black and greenish grey radiolarian cherts and siliceous mudstones with thin seams of siliceous shale and fine green tuff bands dip at 60°/140°; minor folds at [SX 8425 8417] and [SX 8427 8416] are overturned to the north, with axial planes dipping at 60° to 70° southward and axes plunging at 36°/096° and 44°/096°; 13 m of pale grey calcareous cherts and mudstones with radiolaria including Porodiscus sp., Cenosphaera sp., Cenellipsis sp.and Lithocyclia devoniensis (Hinde and Fox, 1895; Ussher, 1913) are overlain at [SX 8427 8415] by Posidonia beds as follows: black siliceous shales, 0.23 m, overlain by grey limestone with fine calcite veining, 0.08 m, black shales dipping at 56°/130° and containing P. becheri and also small goniatites in thin sooty bands in the upper shales indicating a B₂ age, 1.8 m, grey rusty-weathering fissile slightly calcareous shales crowded with large specimens of P. becheri of P_1b age, 0.9 m, grey limestone, 0.025 m, and bluish grey limestone with shales dipping at 55°/131°, 0.6 m.

Another road section [SX 8432 8405] 27 m south-west of Spara Bridge shows: siliceous bluish black mudstones and shales with quartz veining, 0.9 m, dip 60°/136°, which have yielded Neoglyphioceras sp.and Sudeticeras aff. ordinatum of P_2a age, overlain by bluish black shales and siliceous mudstones, 3.1 m dip 60°/136°, and thinly bedded cherts with thin seams of siliceous shale, 5.5 m, showing small ripples on the bottom of some beds which may be sedimentary or tectonic in origin; these beds dip at 50°/188°.

The railway cutting near Lower Ashton [SX 8422 8448] to [SX 8424 8441] shows Combe Shale overlain successively by pale grey cherts, dip 50°/140°, 12.2 m; thickly bedded fine-grained tuff dipping at 61°/129°, 3.7 m; massive black chert, 0.6 m; thinly bedded fine-grained tuffs with thin chert bands and lenticles, 3.7 m; coarse-grained crystal-tuffs with chert bands and lenticles dipping at 67°/128°, 27.4 m; and cherts with thin seams of siliceous shale and some radiolaria, 3.1 m.

A section behind a cottage [SX 8442 8426] north of Lower Ashton shows the following beds dipping at 60°/150°:

Debris along the adjacent railway track [SX 8440 8415] yielded P. becheri and G. aff. falcatus, referable to P_1b.

At the roadside north-north-east of Lower Ashton [SX 8448 8424] grey shales with thin dark siliceous bands, yielding Posidonia?, dip 50°/190°, 2.4 m, are overlain by bluish black siliceous shales, 3.1 m, bluish black chert, 0.3 m, and bluish black thinly bedded cherts and siliceous mudstones in which Neoglyphioceras sp.in the upper horizons indicates P₁ age, 4.6 m.

Scattered outcrops north-east of Lower Ashton, along the southern limb of the Ashton Anticline, dip consistently south-east. Chert beds impregnated with black manganese can still be seen in old workings above Higher Barton [SX 8555 8510].

The Teign Chert of the Court Barton Anticline and Syncline is bounded by the Scanniclift Stretch Thrust and a minor north-south fault. In Church Lane, Christow [SX 8354 8471] to [SX 8360 8487], well-bedded cherts and siliceous shales dip at 70°/146° and are inverted. Ussher (1913) recorded Xiphosphaera sp.and Sethocapsa sp.from this locality. On the hillside near Christow Wood Cottages a temporary trench [SX 8375 8574] exposed thinly bedded bluish black siliceous mudstones and shales dipping at 50°/115°. Specimens of N . cf. spirale were collected, probably of P₂ Zone. Ussher (1913) recorded P. becheri and Neoglyphioceras from a field at the end of Priory Lane [SX 8345 8595], the lane then being called Friar's Lane. The road section north of Christow [SX 8360 8582] to [SX 8359 8598] shows grey cherts and green siliceous mudstones with fine tuff bands, 91 m thick, dipping at 40° to 50° south-east. An overlying intrusive dolerite is capped by pale grey thinly bedded cherts, shales and mudstones dipping at 50°/153°.

Vertically bedded cherts' striking east-west in the northern and southern faces of Scatter Rock Quarry [SX 8220 8565] show colour banding in green, grey, white and black. They are associated with spilitic lavas near the base of the Teign Chert (p. 63), and lie in the core of the Christow Common Syncline within the metamorphic aureole of the Dartmoor Granite.

Mount Ararat Chert

Higher Sigford–Yeo–Penpark

An outlier of Mount Ararat Chert caps a hill at Higher Sigford [SX 7815 7440] and is bounded to the east by a north-south fault. In the east bank of a stream [SX 7802 7462] massive black cherts with bands of pale green, pink and grey calc-flinta containing small blebs and veinlets of green epidote, are underlain by black siliceous slates with rare P. becheri of P₁ age. Bedding dips at 15°/150°, inverted. Thus Posidonia occurs between the cherts above and the Cracking-ton Formation below, and the relationship may be conformable. Farther east, but west of the Ilsington Fault, the Mount Ararat Chert caps two hills; Posidonia beds have not been found and the junction is probably a structural break.

Crackington Formation slates with thin sandstones at Yeo [SX 7916 7327] contain a slice of black siliceous slates and thin cherts dipping at 34°/126° between faults dipping at a. low angle southwards. The black slates yielded crinoid debris, a costate brachiopod (spiriferoid or rhynchonelloid), Posidonia corrugata and Neoglyphioceras sp., referable to the P_2a Subzone.

At Penpark [SX 783 725], a slice of Mount Ararat Chert has been faulted into the Crackington Formation. Banded cherts in crags [SX 7801 7240] in a copse south-west of Penpark are disposed in an open recumbent fold, axial trend 079°, axial-plane dip 28° south, and gentle westerly plunge.

Bickington to Ilsington

East of the Ilsington Fault the Mount Ararat Chert forms high ground such as the ridges of Ramshorn Down, Rora Down and the conical hill at Penn Wood. The formation is disposed in open to close recumbent folds in a sheet above the Rora Slate; the junction is nearly horizontal, though with a slight easterly or south-easterly dip. The Posidonia beds occur at scattered localities that may represent fold cores; only north of Bickington and south of Lounston is the succession from cherts through Posidonia beds into Cracking-ton Formation present.

A road section [SX 7983 7289] near Lee Farm exposes pale grey cherts with greenish grey siliceous shales, steeply dipping to vertical, strike 227°; a flat-lying fracture cleavage is well developed; 50 m to the north, black slates yielded abundant P. becheri of P₁ age. A similar section [SX 7994 7295] gave Hibernicoceras? of P_1d age. Black shales in a gully [SX 8003 7334] south-east of Lurcombe yielded Hibernicoceras cf. kajlovecense referable to P_1d.

A stream [SX 7996 7346] north-east of Lurcombe has exposed black, blocky, siliceous mudstones and black pyritous, laminated, silty mudstone exhibiting spheroidal weathering. P. becheri, Hibernicoceras? and Neoglyphioceras sp.indicate a P_1d age. Near an old shaft [SX 8013 7394] of a manganese mine, red jasperised chert, extensively quartz veined, dips at 43°/263°.

A quarry [SX 8036 7436] shows 2.45 m ofweathered greyish buffand brown siliceous mudstones folded into a tight slightly overturned anticline and syncline. Fold axes strike 065° and axial planes dip at 70° northward. A well-developed fracture cleavage occurs in the fold cores. Another quarry [SX 8035 7438] shows 6 m of bluish black cherts folded into two large open upright folds with axes trending 240°, plunge 24° westward.

At Ramshorn Down Quarry [SX 7902 7402] about 18 m of black laminated cherts, siliceous mudstones, blocky mudstones and slates lie in recumbent zigzag folds whose axes trend 102°. Radiolaria are common and Hinde and Fox (1895) published a faunal list. In the eastern side of the road [SX 7898 7417] south-east of Bethelcombe Cross bluish black and brownish grey cherts are interbedded with tuffaceous slates and overlain by buff-weathering green and black siliceous slates.

Mount Ararat Quarry [SX 7976 7474] is largely filled in, but small sections at the top of the quarry show grey, rather coarse siliceous mudstones with white laminae and banded cherts. The axis of a recumbent fold trends 103° at the top of the quarry. A track [SX 7973 7477] to [SX 7972 7482] crosses a faulted syncline, showing the following succession:

Zigzag folds in the bedded cherts (Plate 8) show axes trending 085° and subhorizontal axial planes. Another track section [SX 7975 7485] to [SX 7975 7488] shows similar zigzag-folded cherts, mudstones and slates faulted against siltstones of the upper Rora Slate to the south and against red shales of that formation to the north.

Field brash [SX 7884 7469] to [SX 7895 7469] of pale green and greyish black siliceous platy mudstones and cherts yielded Posidonia corrugata, Catastroboceras sp Goniatites granosus, Neoglyphioceras sp.and Sudeticeras sp.of P_2a age, and [SX 7903 7453] Hibernicoceras sp., Neoglyphioceras sp.and an orthocone nautiloid of P_1d age. Fossils collected by Bristow (1962) from a temporary exposure in thinly bedded siliceous slates and cherts [SX 7890 7463] have been redetermined as P. cf. corrugata, Hibernicoceras? and N. cf. spirale of P_1d age. Nearby [SX 7891 7464] surface debris of greyish black siliceous slates and mudstones contains Coleolus sp., Chaenocardiola sp. nov., G. cf. granosus, Neoglyphioceras sp.and Sudeticeras? referable to P_2a. Grey siliceous slates, completely buff weathered, in a lane [SX 7903 7472] north-east of Bethelcombe Cross, are vertical, strike 200°. Bedding planes are crowded with P. becheri of P_1b age. Farther south in the same lane [SX 7905 7470] green siliceous slates yielded crinoid debris and fragments of Posidonia sp., referable to P₁.

A temporary cutting behind a barn at Lenda Farm [SX 7931 7560] exposed 2 m of bluish black cherts with paler laminae and some thin black shale partings; the axis of a zigzag fold trends 284°. Black slates and slaty mudstones adjacent to Rora Slate in a track [SX 7994 7567] to [SX 7998 7560] south of Higher Brimley contain radiolarian casts, large bivalves and rare juvenile ammonoids, and may belong to the Ararat Slate.

Winstow Chert

Whiteway Barton–Ideford–Well

Red and purple shales in a bank [SX 8791 7495] yielded P. corrugata. A similar exposure [SX 8818 7506] alongside the road leading to

Whiteway Barton reveals purple cherts and siliceous mudstones overlying siliceous black shales with horizons of lilac chert of the Posidonia beds in inverted succession, dip 48°/138°; the Posidonia beds contain P. becheri and trilobites. In a trench [SX 8842 7533] ((Figure 11), trench 1) north of Whiteway Barton about 5.2 m of Winstow Slate were exposed dipping at about 20° south or south-east. The slates were uniform, black and slightly siliceous. Adjoining exploration pits showed that the Winstow Slate exceeds 10.7 m in thickness hereabouts. Decalcified limestone with occasional chert fragments yielding P. becheri and trilobite fragments were exposed in trench 2 [SX 8842 7538]; 2.4 m from the northern limit of the trench the Posidonia beds passed invertedly down into greyish green micaceous siltstones with mudstone laminae and shales of the Ugbrooke Sandstone. The first sandstone in the latter appeared 1.22 m below the junction. A major fault with a considerable smash zone was recorded 26.2 m from the southern end of trench 3 [SX 8855 7555]; it cut out most of the Winstow Chert and brought Winstow Slate against Posidonia beds. At the northern end of the trench the Posidonia beds consisted of thinly bedded cherts and siliceous dark green shales with a rich fauna of P. corrugata, P. cf. membranacea, Hibernicoceras sp., N. cf. spirale, a juvenile prolecanitid and orthocone nautiloids referable to P, d. Black shales and cherts exposed in a field north of Whiteway Barton [SX 8848 7554] yielded a P_2d fauna including Tornquistia sp., Posidonia sp.and N. spirale.

Between Whiteway Barton and Larcombe Bridge a north-east-south-west fault throws Winstow Chert against Whiteway Slate and New Red Sandstone, and probably most of the lower part of the formation is missing. Debris of chert and shale [SX 8888 7619] to [SX 8893 7631] north-east of Hestow Farm yielded P. corrugata and Neoglyphioceras sp.of P_1d Subzone age. A small exposure [SX 8908 7665] adjacent to Larcombe Bridge shows thinly bedded cherts and siliceous mudstones of P_1d age which yielded P. cf. becheri, Neoglyphioceras sp.and a trilobite pygidium. Fragments of chert with limestone laminae in a bank to the west [SX 8885 7677] yielded a fauna including P. corrugata, N. cf. spirale and a trilobite pygidium, and shaly debris adjoining Hamblecombe Lane contains P. corrugata, Neoglyphioceras sp.and a small trilobite [SX 8883 7691] to [SX 8874 7700], and P. becheri, P. corrugata, G. aff. granosus, Neoglyphioceras sp.and Hibernicoceras sp., as well as ostracods and trilobites [SX 8858 7733] to [SX 8864 7737], all referable to P1d.

In the vicinity of Well, a north–south band of purple cherts [SX 8755 7735] is conformable with the overlying Posidonia beds, but much of the lower part of the formation is probably faulted out against Ugbrooke Sandstone and Luxton Nodular Limestone. An east–west band immediately south of Well is thrust over the Luxton Nodular Limestone and probably contains only the upper part of the Winstow Chert. Similarly, a small faulted inlier in the deep valley south of Well also probably exposes only the upper part of the formation; the Posidonia beds form a large outcrop round the edge of the inlier.

A trackside exposure [SX 8821 7677] consists of fine-grained limestones and shales dipping at 35°/150°; the limestones yield P. becheri of P, age. A stream section [SX 8811 7668] on the northern side of Well Covert shows dark grey, thinly bedded siliceous limestones weathering to brown rottenstones and with a P₂ fauna including P. corrugata. Cherts in a bank [SX 8784 7695] west-south-west of Well Farm dip at 35°/168° and yielded N. cf. spirale and Hibernicoceras sp.of P_1d age.

North of Sedgewell House, debris [SX 8742 7665] of greyish green siliceous mudstones and black shales yielded P. corrugata and N. cf. spirale of P_1d age. A nearby section [SX 8763 7668] showed thinly bedded greyish blue siliceous limestones with black calcareous shale partings, 0.3 m, overlain by black siliceous shales with calcareous laminae and streaks, 0.6 m, dip 68°/335°; the fauna here includes P. becheri, P. corrugata, Dimorphoceras?, Hibernicoceras sp., N. cf. spirale and an orthocone nautiloid of P_1d Debris [SX 8754 7706] to [SX 8764 7694] north of Olchard Lane consists largely of buff-weathering greyish green silty shales and micaceous and black siliceous shales and yields crinoid ossicles, brachiopods, bivalves, goniatites, ostracods and rare trilobites. Conodont moulds include Gnathodus cf. commutatus and G. bilineatus; the latter species indicates a maximum age of late Ammonellipsites Stufe. Fragments of greyish green cherty mudstone at [SX 8760 7715] include radiolarian casts and decalcified partings and contain brachiopods, bivalves, goniatites and trilobites. Rubble of red mudstone and siliceous mudstone [SX 8754 7728] to [SX 8760 7734] with radiolarian casts yielded P. becheri, P. corrugata, Hibernicoceras sp., N. spirale, an indeterminate prolecanitid and an orthocone nautiloid, referable to P_1d.

Gappah to Waddon

Near Gappah the Winstow Chert is faulted against Ugbrooke Sandstone to the north and forms a prominent feature. From north of Gappah to Winstow Cottages, the Whiteway Slate, cherts and Posidonia beds are conformable. The quarry [SX 8630 7801] near the cottages exposes 1.5 m of greyish green siliceous blocky mudstones and cherts in courses up to 51 mm thick, with rare paper-thin shale partings. In the southern bank of the track immediately west of the quarry, black siliceous shales of the Winstow Slate overlie Whiteway Slate. House and Butcher (1973) recorded that a pit dug beside this exposure yielded conodonts, and a nearby trench [SX 8628 7801] exposed siliceous and blocky shales containing conodonts and grading up into the Posidonia beds.

A roadside exposure [SX 8632 7778] shows black shales with laminae of yellow micaceous silt and decalcified siltstones; the shales contain Posidonia sp., ostracods, trilobite fragments and bar-type conodont moulds.

A temporary pit [SX 8625 7815] in Lawelldown Wood showed black shales with a few decalcified partings and lenses and rare decalcified limestones; crinoid debris, P. becheri, an orthocone nautiloid fragment, Hibernicoceras sp.and N. cf. spirale of Pic' age were obtained. Debris of black shale in a ploughed field [SX 8653 7823] yielded P. becheri, P. corrugata, Dimorphoceras?, Hibernicoceras sp., N. spirale and orthocone nautiloids referable to P_1d.

At Mount Pleasant, House and Butcher (1962, 1973) recorded that one pit [SX 8701 7867] showed a conformable junction with Whiteway Slate, another pit [SX 8702 7867] showed pale green cherts with poorly preserved conodonts on some bedding planes, and a trench [SX 8701 7865] exposed the following sequence: greyish green tough blocky siliceous shales with radiolarian chert bands at base, 1.68 m, overlain by greyish green and yellow-weathering shales and blocky siliceous green shales, 0.62 m, blocky cherts with trilobites, 0.13 m, and yellow-weathering fossiliferous shales (Posidonia beds), 4.74 m; the Posidonia beds were faulted against Ugbrooke Sandstone.

North-west of Castle Dyke, near Mount Pleasant, more pits [SX 8729 7882] exposed greyish green siliceous shales in the middle and upper parts of the formation. In this area the Winstow Chert is faulted against Chercombe Bridge Limestone, and this fracture can be traced along the strike to near Waddon [SX 892 802], where the cherts pass beneath New Red Sandstone. It is likely that the Winstow Chert is tectonically thinned adjacent to the fault, and much of the lower part may be missing. Debris near Waddon Barton includes brick-red shales, maroon and lilac siliceous mudstones, pale grey cherts and decalcified limestones yielding P. becheri and N. cf. spirale of P_1d [SX 8847 7943], and pale to dark lilac cherts, siliceous mudstones and brick-red shales yielding P. corrugata and Sudeticeras? of P₂ age [SX 8905 7980].

Immediately south of the Chudleigh escarpment, in the vicinity of Dunscombe Covert, thinly bedded black cherts [SX 8845 7870] are overlain by Posidonia beds and sandstones of the Ugbrooke Sandstone. Locally the cherts have been thrust over mudstones of the Ugbrooke Sandstone.

Undifferentiated cherts

Three fault-bounded outcrops of cherts of Lower Carboniferous aspect occur along the Holne Thrust near Chudleigh. At Heightley, a faulted slice between the Crackington Formation of the Teign Valley Unit and that of the Kate Brook Unit is exposed in a field [SX 8618 7858] where 1.25 m of thinly bedded black cherts pass up into fine-grained volcanic rocks that are completely rotted to a brown earth. Bedding dips at 20°/355°. Similar cherts with black shale partings dip at 34°/150° in a small waterfall [SX 8748 7977] north of Old Kate Bridge and in a field [SX 8771 8042] west of Hams Barton black laminated cherts dip at 45°/136°. The stratigraphical relations of these cherts are unknown and they are arbitrarily assigned to the Kate Brook Succession.

Ussher (1913, p. 48) found "signs of Posidonomya-beds and of chert" north-west of Abbotskerswell near Conitor, adjacent to a klippe of Ugbrooke Sandstone, but none was reported in the resurvey.

Upper Carboniferous

Crackington Formation: Ashton Shale

The Ashton Shale has been traced east of the fault zone at Bovey Tracey, on the edge of, or outside, the metamorphic aureole of the Dartmoor Granite, north-eastwards in the southern limb of the Trusham Regional Anticline, the outcrop narrowing towards the nose of the fold. In this area the Crackington Formation, bounded to the east by the Holne Thrust, probably reaches its greatest thickness within the district. A short distance north of Chudleigh, between the Holne Thrust and the Lower Carboniferous rocks in the nose of the Trusham Anticline, only the Ashton Shale and a few metres of overlying shales and sandstones are present.

Rusty-weathering grey and black shales and sandstones are exposed [SX 8317 7946] to [SX 8325 7927] in the track leading to Higher Combe. In Ruggadon Lane [SX 8560 8166] to [SX 8567 8161] shales with thin lenses of fine-grained sandstone, overlying Posidonia beds, dip at 70°/188°. In the lane [SX 8666 8148] north of Coombeshead, rusty-weathering grey shales with a pronounced pencil cleavage are full of plant fragments.

At 229 m south of Spara Bridge [SX 8432 8385], the type locality for the Ashton Shale, grey shales with plentiful plant fragments are 1.8 m thick and dip at 80°/150°; thin (12-mm) black soft sooty bands yielded well-ornamented goniatite fragments (Cravenoceratoides sp.) of E_2b age; overlying shales, 3 m thick, show pencil cleavage. A stream section [SX 8562 8475] to [SX 8576 8484] shows shales and sandstone bands in easterly plunging open folds. At Higher Ashton [SX 8560 8467], 73 m of grey shales, dipping at 70°/184°, are full of plant debris and contain slightly harder dark silty bands. At [SX 8660 8554] shales with a 90-mm sandstone band lie in an open fold plunging at 25°/090°.

Grey micaceous shale fragments containing plant debris are common. A short distance outside the district grey silty shales dipping at 68°/270° [SX 8359 8605] conformably overlie bedded cherts of the Teign Chert.

Crackington Formation undifferentiated

Bickington–Ilsington–Parke

Near Bickington, Crackington Formation is overthrust by Devonian rocks in the Bickington thrust complex (Plate 9). In the Liverton Unit, black slates with thin siltstones and sandstones contain groups of sandstones in beds up to 0.6 m thick. Only around Lemonford Cottage [SX 7925 7270] are coarse sandstones present, in beds up to 0.25 m thick. The slates locally exhibit spheroidal weathering and contain small siderite nodules.

In a field [SX 7895 7241] north-east of Goodstone Cross, thin fine-grained sandstones dipping at 43°/142° are shown by their groove casts to be inverted. North-north-west of Lemonford [SX 7906 7289] black slates show spheroidal weathering and contain silt laminae dipping at 30°/130°. By the River Lemon at [SX 7899 7338] black mudstones with fine-grained sandstones show ripple-drift bedding and groove casts, section inverted, and 45 m upstream black slates with thin silty horizons exhibit bedding/cleavage lineation that trends 285°, cleavage dip 40° south, bedding dip variable but usually steep.

At Yeo Farm [SX 7912 7332] 9 m of black slates contain three thin sandstones whose irregular bases and graded bedding show that the sequence is inverted. At the south-east end of the section black slates dipping at 38°/122° are cut by low-angle faults parallel to, or slightly steeper than, the bedding and are structurally overlain by about 6 m of Posidonia beds (p. 70).

In a wood [SX 7853 7357] adjacent to the River Lemon, black slates with fine-grained sandstones lie in a recumbent fold whose axis trends 265°. Locally there is a gradation from fine-grained sandstone through silty sandstone into slates. To the north, slates in a lane [SX 7805 7398] show spheroidal weathering and scattered siderite nodules up to 10 mm diameter. At Higher Sigford [SX 7813 7448] black, buff-weathering slates, with thin, medium-grained sandstones, dip at 35°/153°.

Between Bickington and Blackpool exposures include: slates with coarse well-graded sandstones that fine upwards into silty sandstones rich in plant fragments, dip 30° eastward [SX 8036 7374]; slates with thin sandstones, commonly with coarse bases, dip 45° between east and south-east [SX 8048 7370].

In a small copse [SX 8006 7658] north of Wilsworthy blocky buff-weathering black mudstones contain small nodules rich in pyrite. A 0.5-m fine-grained sandstone near the base of the section dips at 45°/049°. A pit to the south [SX 8007 7655] shows 3 m of fine-grained laminated sandstones and mudstones dipping at 60°/221°. In an adjacent quarry [SX 8008 7655] 7.5 m of blocky black mudstones contain scattered thin cross-laminated siltstones and bluish grey very fine-grained sandstones showing ripple-drift bedding. A thicker sandstone exhibits elongate flute casts trending 120°. Bedding dips at 48°/213°, steepening to vertical. Behind Wilsworthy Farm [SX 8013 7647] weathered slates and sandstones lie in a recumbent fold whose axis trends 165°.

In the Ilsington Unit, south of the Narracombe Thrust and west of the Ilsington Fault, occur spotted bluish black shales, locally faintly laminated. In a field exposure [SX 7820 7572] a thin sandstone bed within black spotted slates dips at 17/107° and provides the only in situ evidence of sandstones interbedded with the slates. However, surface debris includes a little thinly bedded silty sandstone. A quarry in Middlecott Wood [SX 7825 7644] exposes 6 m of bluish black spotted slates with a 20-mm siliceous band at their base; the slaty cleavage dips at 34°/062° and a later crenulation cleavage at 50°/135°.

In the Teign Valley Unit west of the Lustleigh Fault the Crackington Formation dips north-eastwards and presumably youngs in that direction. Between Ullacombe [SX 7838 7770] and Whisselwell [SX 7940 7757] the abundance of sandstone debris suggests that the Ashton Shale is absent or very thin. Forestry tracks in Yarner Wood [SX 784 787] expose black spotted slates and thin sandstones with local packets of bluish grey and pale grey quartzitic sandstones. Beds 0.2 m thick are typical, though some range up to a metre. The quartzitic sandstones are partly recrystallised.

Within the Lustleigh–Bovey Tracey fault zone south-east of Pullabrook [SX 7930 7952] the Crackington Formation lies in the north-eastern limb of a southerly overturned anticline. Black slates, locally faintly spotted, predominate over thin greyish blue sandstones up to 0.15 m thick and much of the sequence could belong to the Ashton Shale. It is best exposed in the railway cutting [SX 8076 7895] north of Parke.

Bovey Tracey – Ranscombe – Christow

Near Bovey Tracey the Crackington Formation comprises a sequence of distal turbidites, the beds being generally thin, fine grained and regularly stratified, with sharp bases and with tops grading into finer sediment. Laminations and ripple-drift bedding are common. The sandstones are bluish grey, weathering to pale grey, and consist of quartz, a little feldspar and rock fragments in a quartz-sericite-chlorite matrix. Most of the rock fragments are of slate, phyllite or fine-grained sandstone. A section on Hatherdown Hill [SX 8320 7899] to [SX 8330 7903] in shale and sandstone shows dips of 35°/145°; some of the shales are full of plant fragments, and gentle open flexures are present. East of Dunley House [SX 8407 7834] a flat-lying easterly plunging minor fold in shales with thin sandstones is exposed in a stream bed.

Beside the road [SX 8536 8037] north of Lyneham sandstones grade upwards into laminated micaceous shales full of plant debris and show bottom structures. Thinly bedded sandstones in thick shales, locally with minor faults, are exposed in a stream east of Ranscombe [SX 8700 8197] to [SX 8768 8193]; at [SX 8783 8194] a fold plunges at 35°/094°.

North of Whiteway House the Bickington Thrust runs beneath New Red Sandstone. In this vicinity the Crackington Formation occupies the core of the George Teign Barton Regional Syncline. A forest track [SX 8686 8229] to [SX 8740 8321] shows thin regularly bedded sandstones and shales in folds plunging at around 50°/092°. Load casts and flute casts on the sandstones show them to be right way up. Thin sections of the sandstones show up to 80 per cent quartz. The quartz occurs as subangular to subrounded grains up to 0.3 mm across and locally strained. The matrix consists of scattered feldspar, sericite, chlorite and a few altered rock fragments.

A roadside section north-west of Holden Cross [SX 8620 8382] to [SX 8607 8392] exposes shales dipping at 50° to 60° southwards. They have a pronounced pencil cleavage and contain abundant plant debris and scattered thin lenses of argillaceous sandstone. Shales and sandstones in a nearby stream [SX 8585 8455] to [SX 8670 8545] dip generally at 50° to 80° southwards, but locally northwards; easterly plunging folds are seen in places to be overturned to the north. At [SX 8642 8485] thin sooty bands within grey shales yielded goniatite fragments of E₂ age (p. 177 ).

In Kiddens Plantation [SX 8748 8433] micaceous shales and sandstones dip at 75°/045°. Minor open flexures plunge at 45°/077°. An incipient slaty or fracture cleavage in the shales in the cores of the folds is refracted across the sandstone beds. At [SX 8723 8487] inverted sandstones dipping at 65°/160° carry flute casts striking 160°.

In a copse north-west of Great Leigh sandstones and shales brought against Hyner Shale by the Scanniclift Stretch Thrust dip at 58°/162' [SX 8420 8570]. The minor road running north-eastward from Christow church cuts through shales and sandstones, locally inverted. A section alongside the B3193 road [SX 8395 8528] to [SX 8396 8580] shows dips of 60°/139°, abundant plant fragments and [SX 8396 8580] upright folds plunging at 40°/080°.

Bellamarsh Barton to Lawell House

The Crackington Formation of the Kate Brook Unit has been carried north-westwards by the Holne Thrust. The percentage of sandstone is variable, but generally exceeds that of mudstone. The sandstones are only rarely coarser than medium sand grade. Most beds are graded, and range from 75 mm to 0.3 m in thickness, the maximum recorded being 1.37 m.

Site investigations for the Chudleigh by-pass road on the steep slopes of Bellamarsh Copse [SX 859 782] to [SX 853 775] adjacent to the River Teign proved abundant turbiditic sandstones and subordinate greyish black shales and mudstones in a north-facing recumbent fold. During road construction Dr R. T. Taylor (personal communication) noted a poorly sorted conglomerate bed 3 m thick, which could be traced laterally for 20 to 30 m. It varied from grey muddy, sandy, pebble-conglomerate to muddy, pebbly, coarse sandstone; most pebbles were about 2 to 4 mm long but some ranged up to 20 mm. Subrounded to rounded pebbles of black chert and grey siliceous mudstone predominated, with subordinate vein quartz and weathered feldspar. Angular black mudstone clasts were also noted. The siliceous lithologies were of Lower Carboniferous aspect.

Ugbrooke Sandstone

Two Mile Oak Cross to Bradley Manor

West of Two Mile Oak Cross, dark grey shales with abundant quartz pebbles are exposed [SX 8440 6810], and at the cross-roads [SX 8480 6798] black slates were seen in a temporary exposure.

Boulders of sandstone and conglomerate occur locally between Rydon Hill and Bradley Manor [SX 848 708], and there are scattered exposures of sandstone, conglomerate, siltstone and grey slates. For example: near the entrance to Conitor Quarry coarse-grained sandstone [SX 8513 6930]; grey slates containing large rounded pebbles in a hedge [SX 8440 6990] and along Ogwell Road east from [SX 8455 7011] dip at 45°/125°; coarse-grained sandstone at Under-cleave Farm [SX 8460 7044]; conglomerate with quartz pebbles up to 50 mm diameter [SX 8460 7058]. The formation was penetrated by the Rydon Ball Farm Borehole (p. 183).

Whiteway Barton–Ugbrooke Park–Dunscombe Farm

Between Ponswine [SX 8750 7511] and Combe, black laminated slates with some silty beds underlie the massive sandstones of the Ugbrooke Sandstone. A track [SX 8726 7527] south-south-west of Combe Holdridge shows steeply dipping mudstones with silt laminae cut by cleavage dipping at 30°/090°. Sooty black slates in the lane [SX 8720 7582] from Combe Hill Cross to Combe contain bands of siderite nodules and show cleavage dipping at 10°/101°.

A small disused quarry [SX 8740 7543] south of Combe Holdridge contains 0.6 m of fine-grained sandstones and black mudstones, the sandstone beds around 25 mm thick and cross-laminated in their upper parts, overlain by 0.6 m of massive medium-grained sandstone, coarse in the middle; the strata dip at 40°/l49°. Between Combe Holdridge and Whiteway Barton outcrops of conglomerate have been mapped within the massive sandstones of the Ugbrooke Sandstone in areas where conglomerate makes up more than half of the rock. Massive conglomerate between Combe Cottages and Tor's Hill is mostly of rounded to subrounded pebbles up to 25 mm diameter. However, many angular pieces of sandstone, siltstone, shale and chert are present, one large sandstone fragment measuring 0.5 X 0.4m. Above the conglomerate are 0.61 m of medium- and coarse-grained sandstones with an erosional base; they contain numerous plant fragments and dip at 35°/079° [SX 8770 7577].

Beside a path [SX 8803 7640] north-east of Combe Holdridge loose debris of Ugbrooke Sandstone yielded a specimen which Dr C. Patterson has reported is not unlike a scale of the fish Strepsodus, but the typical markings are not preserved. A disused quarry [SX 8792 7599] east of Combe Holdridge exposes the following sequence, with beds dipping at 28°/075°:

At Whiteway Barton a trench [SX 8842 7538] exposed decalcified cherts and limestones of the Posidonia beds, overlain by greyish green micaceous siltstones with mudstone laminae and shales and with a thin sandstone 1.2m above the junction. From Whiteway Barton to Hestow Farm [SX 8883 7608] black shales and mudstones occur in the surface debris immediately to the west of the Posidonia beds outcrop. Pebble conglomerates are overlain by well-bedded, laminated medium- and fine-grained sandstones and siltstones along a track [SX 8829 7552] to [SX 8833 7556]; the beds dip at 16°/120° and are inverted. In an old quarry [SX 8838 7567] immediately north of New Linhay 3 m of thickly bedded, well-jointed, medium-grained sandstones with a few pebbly layers dip at 65°/335°

At Sedgewell Farm [SX 8752 7679] 1.2 m of massive pebbly, coarse-and medium-grained sandstones with abundant shale pellets, passing upwards into 1.8 m of coarse-grained sandstone, dip at 20°/010°; at [SX 8751 7677] 2.4m of massive, coarse-, medium- and fine-grained sandstones dip at 30°/3l6°. Outcrops [SX 8767 7658] south-south-east of Sedgewell are of conglomerates and coarse-grained sandstones dipping at 42°/128°; contained angular fragments of shale, chert and siliceous mudstone are up to 51 mm across. Old overgrown quarries and crags in Cleeve Copse [SX 8868 7672] to [SX 8874 7675] expose medium- and coarse-grained sandstones, with shale fragments up to 0.17 m long, which locally pass up into laminated, medium-grained sandstones rich in plant fragments.

Black, faintly laminated slates whose cleavage dips at 28°/030° crop out at Gappah [SX 8635 7739]; [SX 8640 7743] and to the south of it [SX 8659 7657]; [SX 8650 7617].

An old quarry [SX 8690 7802] in Ugbrooke Park exposes the following sequence of beds which dip at 49°/069°;

A track [SX 8698 7794] to [SX 8713 7808] along the north side of Lower Water Lake, Ugbrooke Park, shows conglomerates and coarse-grained sandstones dipping at 50°/060°. The southern part of the section is mostly pebble conglomerates and coarse sandstones. The northern part shows pebbly conglomerates with coarse- and medium-grained sandstones, locally roughly bedded. On the south side of Lower Water Lake [SX 8698 7783] occur crags of massive, sandy pebble-conglomerate, with cobbles up to 102 mm, and pebbly, coarse-grained sandstones.

House and Butcher (1973) exposed the following section in a trench [SX 8703 7865] at Mount Pleasant: green slates with some siliceous bands, blocky-bedded cherts and shales (Posidonia beds), separated by a small fault from overlying weathered rubbly micaceous siltstones with arkosic sandstones containing green and buff laminated silts and siliceous beds, the lowest beds yielding immature goniatites, trilobites and poorly preserved ostracods, 2.4 m, which are capped by micaceous laminated silts with arkosic sandstones and grey shales. An adjacent trench confirmed that the junction with Posidonia beds is conformable. A nearby pit [SX 8732 7879] exposed coarse arkosic sandstones with dark grey micaceous shales in the basal part of the Ugbrooke Sandstone.

Between Ugbrooke House and Lower Dunscombe Farm [SX 8857 7895] black mudstones predominate over shales, and thin siltstones and graded sandstones are characteristic. The shale–mudstone facies seems to replace the massive sandstone facies near Biddlecombe Cross [SX 8810 7906], leaving only a thin development of the latter above the basal beds.

The track [SX 8714 7812] to [SX 8719 7818] along the northern side of Higher Water Lake exposes black, faintly laminated mudstones dipping at 44°/070°. Scattered ribs of micaceous siltstone or very fine-grained sandstone rarely exceed 6 mm in thickness. In the southern part of the section a 150-mm fine-grained, laminated and graded sandstone with ripple-drift bedding passes up into black mudstones and has groove casts on its base. Farther along the track [SX 8723 7820] to [SX 8728 7820] black laminated mudstones contain incipient bullions up to 0.15 m diameter. Towards the eastern end of the section a 0.42-m sandstone has a sharp base and grades upwards from medium-grained sandstone into silty sandstone, siltstone and mudstone; bedding dips at 38°/065°.

A section [SX 8792 7837] along the track leading north-west from Ashwell Lodge shows black mudstones with fine silt laminae and with ribs of laminated siltstone or fine-grained sandstone that commonly display ripple-drift bedding and asymmetrical ripple marks. A 0.12-m laminated sandstone, with sharp erosional base bearing groove casts and rare flute casts, contains shale chips in its lower parts and numerous plant fragments throughout. It grades up through 50 mm of laminated micaceous siltstone into black mudstone. Bedding dips at 20°/094°. An old quarry [SX 8777 7897] north-east of Castle Dyke exposes 3 m of fine conglomerates and coarse sandstone dipping at 18°/080°.

At Biddlecombe Cross [SX 8802 7908] black shales contain graded sandstones, up to 0.15 m thick with ripple-drift bedding in their upper parts, and dip at 35°/085°. A stream [SX 8901 7892] to [SX 8904 7898] south-west of Higher Dunscombe has cut through massive and poorly bedded conglomerates, coarse- and medium-grained sandstones dipping at 38°/018°; at the southern end of the section they are faulted against New Red Sandstone breccias. An old quarry nearby [SX 8917 7894] contains 1.22 m of massive conglomerate.

Chapter 6 Igneous rocks and associated metamorphism

Introduction

The Dartmoor Granite forms high moorland in the north-western corner of the district. Elsewhere, a spilitic igneous suite includes picrite, teschenitic dolerite, albitedolerite, spilite and quartz-keratophyre-tuff. The extrusive rocks occur in the Teign Valley, Bickington–Beacon Hill and Ugbrooke–East Ogwell successions. They range in age from Middle Devonian to Lower Carboniferous, and are described in the appropriate stratigraphical accounts. Two small masses of quartz-porphyry south-east of Christow are probably part of one east–west-trending intrusion. Lavas at the base of the Permian crop out at two localities in the north of the district.

Basic igneous rocks

Picrite

Picrite has been reported (Busz, 1896; Shannon, 1922) from a single locality, now obscured, at Knowles Hill [SX 8599 7180]. Shannon described a range of rocks, produced by differentiation, in which picrite formed the base of a mainly doleritic intrusion which was veined by bostonite. Contacts were sharp, suggesting that differentiation did not take place in situ. Some quartz was thought to have been introduced by contamination with country rock.

Teschenitic dolerite

Fairly coarse-grained bluish grey dolerite west of Stancombe Farm [SX 8038 7357], near Bickington, trends north–south through the Mount Ararat Chert. Lowe (1908) described it as a small differentiated intrusion and identified four rock types in thin section:

1. Camptonite or hornblende-lamprophyre, consisting mainly of euhedral brown hornblende and plagioclase.
2. Teschenitic dolerite, consisting of euhedral brown hornblende and euhedral augite with plagioclase and interstitial analcime.
3. Teschenite, consisting of plagioclase and euhedral augite, the latter exhibiting peripheral brown hornblende crystals, with hornblende and interstitial analcime.
4. Analcime-diabase or augite-teschenite, consisting of plagioclase with non-euhedral subophitic augite, interstitial analcime and no hornblende.

Lowe noted that camptonite occurred at the top and sides of the intrusion and that hornblende petered out downwards, and he proposed that gravity differentiation had concentrated augite towards the base. The single remaining exposure, a small disused quarry [SX 8015 7344], shows types 3 and 4.

Albite-dolerite

The albite-dolerites have mainly intruded argillaceous sequences and are particularly abundant in the Teign Valley and Bickington successions. They are pre-orogenic or synorogenic. Porphyritic and non-porphyritic types have the same mineral and chemical compositions. Contact metamorphism of country rocks has produced spilosites, locally altered to albite-rich spilosites and adinoles by metasomatic introduction of sodium. Within the Teign Valley Succession the dolerites form sills in the Hyner Shale, Trusham Shale and Combe Shale, rarely in the Teign Chert. The sills range in thickness from a few metres to over 60 m; some are laterally extensive and others comprise several lenticular masses.

Small plug-like masses of albite-dolerite which occur within the recumbently folded slates of the Gurrington Slate and Nordon Slate may represent original sills tectonically disrupted. Dolerites are particularly common at the contacts between the belts of purple and grey slates within the Gurrington Slate; they have caused spotting and loss of colour. Contacts with country rocks are exposed in old quarries near Burne Cross [SX 7973 7103 and 7980 7130] and Lower Wickeridge Farm [SX 7840 6993].

A few albite-dolerites occur in the Whiteway Slate of the Kingsteignton and Abbotskerswell units. A large dolerite mass is exposed on the south side of the Teign Estuary between Wildwoods Point [SX 8758 7190] and Buckland Point [SX 8838 7214]. The rock is fine to medium grained and bluish green where fresh, though most of the crags exhibit thick, spheroidally weathered skins. On the north side of the estuary a fairly coarse-grained greenish grey dolerite is exposed [SX 8771 7243] north of Hackney.

A small exposure [SX 8945 7343] of black, fine-grained basic rock west of Forder Lane is the only dolerite noted in the Kingsteignton Volcanic Group.

Petrography

Porphyritic albite-dolerites with feldspar phenocrysts up to 5 mm across occur at Crockham Quarry [SX 8490 8085] (Figure 12), Trusham Quarry [SX 850 812], Tinkley Quarry [SX 8470 8167], Hill Copse Quarry [SX 8410 8444], Higher Herebere Quarry [SX 8036 7155], Beacon Hill Quarry [SX 8010 6737] and near Lower Wickeridge Farm. The feldspar tablets are commonly pale green, owing to the presence of small crystals of chlorite, and lie in a dark green, grey or mauve groundmass of augite, chlorite and opaque minerals. Locally the feldspars are pink and alkali-rich. Non-porphyritic albite-dolerites show white, green or pink feldspar laths. Trachytic textures with subparallel feldspars were noted in rocks from Ryecroft Quarry [SX 8432 8475] (Figure 12), Broom Hill [SX 8448 8330], Great Cleave Barn [SX 8582 8565], near Holbeam [SX 8257 7145], near Burne Cross [SX 7973 7103]; [SX 7980 7130], Knowles Hill Quarry [SX 8601 7180] and alongside the Teign Estuary [SX 8771 7243].

All the dolerites have fine-grained leucocratic margins which are commonly vesicular and form sharp contacts with the adjacent shales. The amygdales are usually spherical, about 2.5 mm across and filled by quartz or, more rarely, calcite; locally they are up to 25 mm in diameter, as at the upper contact of the Ryecroft sill. The chilled edges consist of microliths of albite and a few larger feldspar phenocrysts in a cryptocrystal-line groundmass of chlorite, opaque ore minerals, apatite and a little biotite.

The feldspar is mostly albite with traces of oligoclase or andesine. It shows albite twinning with some pericline and Carlsbad twinning and chessboard twinning. A little untwinned feldspar is probably orthoclase. The albite occurs as euhedral to subhedral laths and tablets, and any orthoclase as small aggregates of anhedral crystals. The smaller laths of plagioclase feldspar in both porphyritic and non-porphyritic dolerites are strongly saussuritised and show ophitic and subophitic textures with the augite. Small clear crystals of albite within areas of chlorite may be of secondary origin. Most of the feldspars show alteration to chlorite and some to sericite, calcite, epidote, zoisite and prehnite. Many have a core of chlorite surrounded by relatively clear feldspar.

Augite occurs as euhedral crystals and anhedral plates 1 to 4 mm across. It is usually colourless to pale brown, or rarely purple and pleochroic (titanaugite). Much is altered to chlorite and some to calcite. A few of the intrusions contain no pyroxene and hornblende is very rare, except in one body near Newton Abbot (Middleton, 1960, pp. 204–205). Biotite is present as rare small brown pleochroic flakes largely altered to chlorite.

Apatite occurs as abundant colourless euhedral prisms and acicular crystals up to 2 mm long. Many have a central six-sided cavity filled with green chlorite, and some are rendered semi-opaque by small dark fluid inclusions. The iron oxides are titaniferous magnetite and ilmenite in euhedral latticed growths generally encrusted with white leucoxene. A little pyrite occurs as small cubes or anhedral masses. Calcite forms veins and fills interstices. As an alteration product it locally occupies the cores of feldspar laths. Quartz occurs as rare clear anhedral interstitial patches, generally in the upper parts of the sills. Chlorite may form up to 5 per cent of the rock, as a pale green fibrous alteration product and as a primary interstitial mesostasis. The interstitial chlorite is generally darker green and its fibres are commonly arranged in radiating spherulites. Tiny granular crystals of sphene are present locally in the chlorite.

Order of crystallisation

The first minerals to crystallise were apatite and ilmenite, probably in that order since enclosures of euhedral apatite occur within ilmenite. The feldspars also began to form early, as shown by large euhedral tabular phenocrysts and laths enclosing apatite and ilmenite. Probably the first feldspars crystallised before final emplacement, since scattered phenocrysts of feldspar are present in the chilled margins. Augite followed, and continued to grow around feldspar laths to produce ophitic and subophitic textures. A little biotite crystallised towards the end of augite crystallisation. Chlorite crystallised last, and Lehmann (1965) considered it a primary mineral of the Trusham and other dolerites and argued for a deep-seated magma source.

Metamorphism and metasomatism associated with the dolerites

Most dolerites have effected contact alteration, but rarely for more than a metre. Spilosites, and locally adinoles, were noted at the margins of sills at Ryecroft, Hill Copse, Tinkley and Crockham quarries.

At the outer limit of the metamorphism the spilosites are pale grey with tiny dark green or black spots preferentially developed along the bedding planes. Towards the dolerite the spilosites become more compact, less fissile and paler, almost white, and the spotting is more pronounced. At the contact the spilosites are locally replaced by adinoles.

The first stage of alteration, as observed in the Combe Shale, is the development on the bedding planes of dark spots about 0.5 mm in diameter. These spots are aggregates of pale green chlorite with some xenoblastic quartz, and a little idioblastic magnetite which commonly shows a square cross-section. In the next stage the rock has lost its carbonaceous material and appears paler in colour. The spots have grown in a porphyroblastic manner, pushing the micas of the matrix into apparent flow structures. Nearer still to the dolerite, metasomatism has occurred. The spots are angular and consist of a fine granoblastic mosaic of xenoblastic quartz and albite, with only a little chlorite. Square cross-sections and patterns of inclusions are reminiscent of chiastolite, which may have been present and been replaced. The mica of the matrix shows replacement by a fine granular aggregate containing much albite. Adinoles near the contact contain no mica and few sedimentary features and consist of a microcrystalline mosaic of quartz and albite with some small tabular spots. At the contact they show no trace of sedimentary origin and the mosaic includes rutile needles, pyrite and a little magnetite.

(Table 4) presents analytical data from contact-altered Combe Shale. Major and minor elements, except ferrous iron, were determined by colorimetric and atomic absorption procedures; FeO was determined by titration with standard dichromate solution, and all trace elements were analysed by X-ray fluorescence spectrometry. It appears that dolerites with many vesicles, and hence probably abundant volatiles (e.g. Ryecroft and Hill Copse sills), have given rise to pronounced sodium metasomatism of the shales, whereas those with less vesicular margins and relatively free from volatiles produced negligible metasomatic effects (e.g. Crockham sill).

Chemistry of the basic igneous rocks

(Table 5) and (Figure 13) give analyses of the Ryecroft and Crockham sills, a 1.5-m chilled sill lying 6 m above the main Crockham sill, the spilitic pillow lavas of Scanniclift Copse [SX 8448 8628] to [SX 8439 8606], the quartz-keratophyre vitric-flow tuffs of Christendown Clump [SX 8508 8635], and the pure crystal-tuffs of the Lower Ashton railway section [SX 8422 8445].

The mineralogical and chemical compositions of the Ryecroft and Crockham sills reflect the differentiation trends of the dolerite magma. The rock at the base of these sills shows a concentration of titanaugite, ilmenite, magnetite, chlorite and apatite. Albite content increases upwards and the Ryecroft non-porphyritic dolerite shows an accompanying increase in feldspar size. Both sills contain a little interstitial quartz in their upper parts and have a zone of coarse-lath dolerite at the top. Upward increases of SiO₂, Al₂O₃ and alkalies (Figure 13) reflect an increase in the amount of albite. A slight decrease in K₂O at the tops of the sills may be due to displacement by Na₂O. Percentages of FeO, CaO, MgO, TiO₂, P₂O₅ and Fe₂O₃ show a general decrease towards the tops of the sills. The trends in iron and MgO may be related to the spatial distribution of titanaugite, chlorite and biotite, that of CaO to titanaugite and apatite, that of P₂O₅ to apatite, and that of TiO₂ to titanaugite, titaniferous magnetite, ilmenite and leucoxene. Of the trace elements, Rb, Ba, Ga, Li and Y show slight increases towards the tops of the sills and Cr, Ni, Co and Cu show slight increases at the bottoms.

(Figure 14) illustrates the course of differentiation of the Ryecroft and Crockham sills and relates the trend to the analyses of spilitic pillow lavas and quartz-keratophyre tuffs. Plotting total alkalies against silica (Figure 15) shows all rocks of the suite to lie in the alkaline basalt field as defined by Macdonald and Katsura (1964). The presence of normative nepheline in some of the rocks further confirms the alkaline affinities of the magma (Yoder and Tilley, 1962), as does the trace element concentration (Gast, 1968).

The probable sequence of events was as follows:

1. An alkaline basalt magma was introduced into deep-seated reservoirs.
2. This magma was intruded as near-surface albite-dolerite sills, which occasionally broke through the sea floor as spilitic lavas.
3. Differentiation of the magma at depth followed the same course as seen in the dolerite sills. The early euhedral crystals of apatite, ilmenite, pyroxene and biotite sank to the bottom and the larger feldspar laths were displaced upwards by flotation or gaseous action.
4. Continuation of these processes led to an enrichment in silica and alkalies, and the upper parts of the magma became a raft-like mush of albite laths with subordinate acid material rich in volatiles. At the bottom of the magma there was a concentration of pyroxene, biotite, ilmenite and smaller feldspar laths in ophitic and subophitic intergrowth with pyroxene.
5. Further differentiation and build-up of volatiles led to increased pressure which ruptured the overlying strata, the volatiles and acid residuum being erupted as quartz-keratophyre-tuffs. The quartz-keratophyre flow tuffs were probably formed during quieter periods around this time.
6. The igneous activity which produced the spilitic suite ceased at the end of the Lower Carboniferous.

Acid igneous rocks

Dartmoor Granite

The most easterly extension of the Dartmoor Granite crops out in the north-west region of the Newton Abbot district. No marked deflections of the boundary are apparent where streams like the deeply cut Shuttamoor and Beadon brooks cross the contact, suggesting that the eastern boundary of the granite is, in general, steep or nearly vertical, and this is confirmed in Great Rock Mine [SX 8278 8163], where an almost-vertical contact of granite with hornfelsed sediments is exposed in several of the drifts. At Whitstone Quarry [SX 8145 7923] the interface undulates along both strike and dip directions, although in the main quarry face it appears to dip at about 60° to 65° south-westward. A hump of granite penetrating shales in the floor of the quarry adds to an impression of large-scale down-dip irregularity. Blyth (1962, p. 441) considered the granite here, and at Whitstone Rock [SX 8134 7926], 110 m to the north-west, to be part of a major steeply-dipping east–west vein rather than the main intrusion. Fragments of sediment certainly occur in fields north of the tor. However, as the Whitstone granite is texturally indistinguishable from normal and contaminated big-feldspar material, it seems more likely that the mass is isolated from outcrops farther north either by a shallow saddle of country rocks or by faulting. The granite shows no evidence of chilling in the quarry. The characteristic megacrystic texture is maintained right to the contact. Several small fine-grained granitic veins penetrate the hornfels, and Blyth recorded the presence of steep, southerly-dipping, east-west "porphyritic granite" veins in sediments to the south.

Geological relationships confirm that the granite postdates the folding of the Carboniferous strata. The easterly plunge of folds in the Teign Valley may or may not be related to the emplacement of the granite. Irving (1888) and Worth (1890) identified Dartmoor granite detritus in the Permo-Triassic rocks, and Dangerfield and Hawkes (1969) argued that constituents of the Permo-Triassic, such as rock fragments which match xenoliths and aplogranitic material in the Dartmoor pluton, point to the unroofing of the granite before the end of the Permian.

Lithology

Reid and others (1912) considered it likely that the granite at present exposed lay at no great depth beneath the top of the pluton. Recent mapping of the two main granite types recognised by Brammall and Harwood (1923), namely coarsely porphyritic "giant" (big-feldspar) granite and the markedly less porphyritic "blue" (poorly megacrystic) granite, has largely confirmed this opinion. A study of the proportions of feldspar megacrysts seen in all outcrops shows a continuous variation between the 15 to 20 per cent level typical of Brammall's "giant" granite and the less than 5 per cent level characteristic of "blue" granite. This information, in conjunction with knowledge of their spatial distribution, suggests that the big-feldspar granite is a roof facies of the pluton and not, as Brammall thought, a separate intrusive phase (Edmonds and others, 1968).

The thickness of the big-feldspar facies seems to be very variable. All the granite at surface in the present district belongs to this facies. Only in Blackingstone Quarry [SX 7839 8578], at approximately 304 m above O.D., and in East Wray Quarry [SX 7823 8315], at approximately 228 m above 0.D., is there evidence of gradation down into rock approaching the poorly megacrystic granite that forms the interior of the pluton to the west. Circumstantial evidence from the Blackingstone and Haytor [SX 757 770] areas points to a minimum thickness of about 15 m, an estimate smaller than that given by Edmonds and others (1968, p. 97). Greater thicknesses occur in the valley of the West Dart River [SX 650 736], where trial boreholes for the proposed Huccaby Reservoir penetrated 30 m of big-feldspar granite. Poorly megacrystic granite is present locally at the south-western margin of the intrusion [SX 573 620]; [SX 543 631] and at Doe Tor [SX 542 848], indicating that a megacryst-rich roof facies did not develop everywhere.

In the Newton Abbot district, feldspar megacrysts generally comprise between 10 and 25 per cent of the granite. They lie alongside quartz megacrysts in a hypidiomorphic matrix consisting chiefly of quartz, potassium-feldspar, plagioclase and biotite. Xenoliths are common. Some outcrops also contain sheet-like or thin dyke-like bodies of finer grained aplogranitic material. At a few localities, coarser pegmatitic developments are associated with the sheet-like aplogranites. Rare quartz-tourmaline veins mainly trend east-north-east-west-south-west. Several mineralised east-north-east-west-south-west vein-structures in the granite were worked in the past for micaceous hematite (p. 171).

Joints

Open joints in the granite can be grouped into those that are horizontal or gently inclined (floor joints) and those which are vertical or dip mainly at high angles. Their incidence is more common towards tops of exposures.

Floor joints reflect topography and are due in part to relief of hydrostatic loading. Where the ground surface is essentially flat, they tend to be approximately horizontal. More usually, joints of this sort are seen dipping at angles between 5° and 25° towards the major topographic depressions.

Blyth (1962) divided the steeply inclined joints into early-formed north-south and east-north-east-west-south-west joints related to consolidation of the pluton, and later north-west-south-east and north-east-south-west joints whose origins stemmed from stresses connected with the Variscan orogeny. He envisaged subsequently mainly dextral movements along the north-west-south-east fractures in Tertiary and later times.

The most prominent joints are nearly vertical but others are inclined at angles as low as 30°. Local variations are common and many individual planes lean towards topographic depressions as they are traced upwards. Individual joints may vary markedly from generalised trends and inclinations, and patterns are revealed only by statistical studies. These joints probably reflect the gravitational instability of the south-west England batholith throughout a period of approximately 290 Ma, during which waning horizontal Variscan stresses were gradually superseded by new forces connected with the complex, three-stage opening of the North Atlantic.

Petrography of the granite

The big-feldspar granite of the Newton Abbot district is a mottled grey-black and buff to pale pink rock at outcrop (MR27847), (MR30868), (MR30870), weathering to rusty brown shades especially in areas affected by quartz-hematite mineralisation (MR30869), E41348). Locally, outcrops have a greyish green hue owing to chloritisation and kaolinisation connected with the passage of mineralising fluids (E39854). The freshest exposures occur in Blackingstone (MR27890) and East Wray (MR31473) quarries.

Feldspar megacrysts range in length from 15 mm to over 140 mm with a mean length of about 40 mm. They are chiefly pale pink or off-white orthoclase-perthites. Some consist of white or very pale bluish green plagioclase. A few less than 25 mm long display rapakivi texture (Hawkes, 1967). The proportions of megacrysts present in the granite (Table 6) show continuous variation, although, because of lichen cover and local fluctuation in megacryst distribution, the figures quoted should be regarded as approximate. The only accurate value is that of 9.5 per cent for Blackingstone Quarry which is based on 10 500 counts taken when the stone was being worked.

Alignment of feldspar megacrysts is a feature of many outcrops and may be an indication of original flow texture. Directions are not consistent, nor are they maintained over significant distances in either vertical or horizontal planes.

The large feldspars, and anhedral quartz megacrysts averaging about 12 mm in size (range 5 to 30 mm), are set in a matrix of quartz, potassium-feldspar, plagioclase and biotite. Tourmaline occurs as single crystals, groups of crystals, and in irregular quartz-tourmaline pods. Garnet was not noted but commonly occurs elsewhere in the big-feldspar granite. The mean size of matrix crystals is generally in the range 1 to 3 mm. However, there are both large and small areas of finer grained rock in most outcrops, in which feldspar megacrysts average less than 10 mm in length, biotite may constitute over 10 per cent by volume and in which plagioclase megacrysts commonly predominate over those composed of perthite. Edmonds and others (1968, p. 109) suggested that this finer grained material resulted from contamination of the magma by argillaceous country rock. Gradations in texture and in modal mineralogy between normal big-feldspar granite and recrystallised argillaceous and silty argillaceous xenoliths seem to support this explanation.

Thin sections show the orthoclase-perthite megacrysts to have irregular margins, small projections commonly linking with potassium-feldspar crystals of the matrix in optical continuity (E32356A). In addition to albite exsolution lamellae, the perthites contain small subhedral prismatic crystals of oligoclase rimmed by albite, sinuous patches and rounded anhedra of quartz, and scattered flakes of biotite. In some cases the inclusions show a concentric arrangement, seemingly determined by the structure and morphology of their host (E32356). Diffuse patches of lamellar-twinned plagioclase (probably mainly oligoclase) scattered through some perthites show a common optical orientation (E32356).

Plagioclase megacrysts also have irregular margins. The crystals show lamellar twinning. Some are concentrically zoned, with calcic oligoclase-sodic andesine in the interior grading outwards to sodic oligoclase and, in a few instances, to albite (E42059). Most plagioclase crystals contain inclusions of quartz, plagioclase of more sodic composition, and biotite. In some, biotite inclusions are abundant (E42059), (E33068).

Quartz megacrysts consist of clusters of anhedral grains joined by secondary growth (E41348). The component parts usually extinguish separately, although some groups show common crystallographic orientation (E32356). Inclusions of biotite and plagioclase commonly lie along the original boundary zones of component anhedra. Strain shadows are characteristic of all large quartz crystals.

Matrix feldspars are subhedral and share many features with the megacrysts. However, the matrix perthites contain more patches of uniformly orientated lamellar-twinned and, in some cases, chequer-twinned oligoclase (E33068), (E32356A), (E42059).

Biotite appears as ragged crystals whose morphology indicates corrosion during the growth of neighbouring quartz and plagioclase. The flakes are commonly in clusters and contain abundant small inclusions of apatite and zircon, the latter generally surrounded by haloes of radioactive particle damage. Larger grains of accessory apatite are scattered throughout the matrix, although most occur near to biotite (E32356A). Accessory opaque oxide (probably mainly magnetite) is also generally found near black mica.

Other accessory minerals appear to be mainly of secondary origin. Finely divided sericite occurs as a sparse alteration product in most plagioclase crystals. In several cases flakes of muscovite have developed from sericitic aggregates (E32356A). Muscovite is associated with biotite and in some cases replaces the dark mica; however, certain plumose crystals present in the matrix of one specimen (E32356A) may be of primary origin. Biotite shows sporadic alteration to chlorite and, in rare examples (E32356A), (E32358A), to chlorite and epidote. The most striking accessory mineral is schorlite tourmaline, which partially or completely replaces plagioclase, potassium-feldspar, biotite, muscovite and chlorite. It occurs as isolated crystals, in patches up to 40 mm or more across of tourmaline and quartz, and in irregular veinlets with or without quartz.

Two chemical analyses of big-feldspar granite from Blackingstone Quarry were quoted by Edmonds and others (1968, p. 105). They indicate rock of adamellitic composition.

Aplogranitic rocks and fine-grained granite veins

Aplogranitic rocks are characterised by their comparatively fine grain-size, the rarity of quartz and feldspar megacrysts and a general paucity of biotite. They are pale grey, weathering to shades of buff and pale orange-brown, and occur most commonly as dyke-like and sill-like bodies 25 to 250 mm thick with an average grain-size of 0.2 to 0.3 mm.

Biotite typically accounts for less than 2 per cent of these quartz-sodic plagioclase-potassium-feldspar rocks (MR27824) and may show partial alteration to muscovite. In some cases finely divided tourmaline is more abundant than the micas (MR27834)–(MR27835). One specimen (E6810) contains accessory garnet.

Aplogranites of coarser grain-size form sheets 0.3 to 0.6 m thick; some show signs of grain-size banding. One specimen (MR27849) has a biotite-rich selvage along one junction with the host granite. Although contacts with the latter are generally well defined, the aplogranites do not consistently become finer grained towards their margins.

Some sub-horizontal aplogranitic bodies are over 1 m thick, have an average grain size of around 0.4 mm, and locally contain as much biotite as does the granite. Quartz and feldspar megacrysts are more abundant than in finer-grained bodies, but less common than in the big-feldspar granite. Aplogranitic debris in the woods [SX 811 803] above Aller Farm is of this type.

Some aplogranites contain patchy pegmatitic developments. At the south-west corner of Tottiford Reservoir [SX 8067 8245] a sheet of aplogranite passes up abruptly into a pegmatite consisting predominantly of pale pink , perthite with lesser amounts of quartz and schorlite (MR27825). Similar coarse-grained patches of perthite, quartz and tourmaline, with or without plagioclase, occur in normal big-feldspar granite, for example alongside a forestry track [SX 8183 8184] near Beadon Bridge.

Several small fine-grained granite veins penetrate the country rocks in Whitstone Quarry and along the footpath a few metres south of Gradner Rocks. Other larger apophyses which cut silty hornfelses south of Lustleigh occur as gently dipping sheets 0.9 to 9 m thick, locally parallel to bedding but generally crosscutting. The rock is white to cream coloured and commonly speckled or zoned with tourmaline. In places hydrothermal processes have altered it to white clay with quartz and chlorite. Thin sections show sericitised plagioclase and orthoclase, corroded quartz, a little biotite and accessory tourmaline and topaz. The texture is equigranular. Orthoclase and plagioclase are present in about equal amounts, suggesting an adamellitic composition.

In the railway cutting [SX 7888 8034] north of Packsaddle Bridge two fine-grained granite sheets dip at 30°/135°, parallel to bedding in hornfels; in Packsaddle Bridge Quarry [SX 7889 8024] they cut across bedding. Outcrops in forestry tracks [SX 7895 8009] near Drakeford Bridge show a fine- grained granite–hornfels contact dipping at 5°/075°; small pits and cuttings expose hydrothermally altered chlorite-rich fine-grained granite with xenoliths of tourmalinised hornfels. Tourmalinised fine-grained granite is in contact with hornfels farther west [SX 7873 8002], and in Rudge Wood [SX 7835 7982] fine-grained granite with 13-mm-thick tourmaline-rich bands is traversed by possibly hydrothermally altered zones and by brecciated zones containing hornfels and rotten fine-grained granite.

These fine-grained granitic veins probably represent late-stage intrusions. The aplogranitic bodies in the granite may have had a similar origin but an alternative suggestion (Edmonds and others, 1968, p. 111), based on form, grain-size and general textural anomalies, is that they might represent incompletely assimilated country rock sandstones which in part behaved rheomorphically.

Xenoliths

Xenoliths are fairly common locally in the granite. Most are ovoid and up to 300 mm or more across. All are fine grained (<0.1 mm to 0.25 mm). They consist of quartz, plagioclase (mainly oligoclase), potassium-feldspar (replacing plagioclase), and biotite. Some contain secondary chlorite (after biotite) and sericite/muscovite (chiefly after plagioclase). Accessory constituents include zircon and apatite (not typically contained in biotite), schorlite, topaz, garnet, andalusite and altered cordierite. Variable biotite contents result in colour ranges from pale grey (MR27871) to medium or dark grey (MR27870). Scattered megacrysts of feldspar, quartz and biotite occur (MR27871), (MR27848). A 110-mm-long xenolith in a disused roadside quarry at Knowle [SX 7905 8060] contains two potassium-feldspar megacrysts, one of them 60 mm long.

Quartz-porphyry

Small exposures of quartz-porphyry or elvan [SX 8350 8433]; [SX 8360 8438] occur about 0.5 km south-south-east of Christow Post Office, and in the bed of the River Teign [SX 8410 8446] where a 1.7-m-wide dyke in Combe Shale strikes a few degrees south of east. Thin stringers cut the shales 2 m south of this last outcrop.

The rock comprises megacrysts of white or pale pink feldspar (0.5 X 1.0 mm to 20 X 40 mm in size), rounded areas of clear quartz (0.25 to 15 mm in diameter) and scattered flakes of altered mica (up to 2 mm across) set in a very fine-grained pale grey matrix. The megacrysts are smaller towards the walls of the dyke. A thin section (E40239) shows that the feldspar megacrysts are composed mainly of sodic plagioclase, although most show some partial replacement by potassium-feldspar. In a few crystals in the hand specimen replacement appears to be complete. Most of the megacrysts have euhedral prismatic form, but crystal edges are markedly rounded, and scattered ovoid pieces of feldspar occur. One megacryst shows partial replacement by groundmass constituents. Small amounts of secondary sericite occur in many of the plagioclase crystals.

The quartz megacrysts are commonly rounded and occur singly or in clusters. Most of the altered mica megacrysts are ragged, and several have been disrupted along cleavage traces by replacement growth of groundmass constituents. These micas now consist of chlorite with rare patches of relict biotite. They still contain original inclusions of apatite and zircon, the latter surrounded by haloes of radioactive particle damage. The matrix of the elvan consists of quartz anhedra, small laths of sodic plagioclase showing extensive replacement by potassium-feldspar, and chlorite which has also replaced plagioclase. Locally patches of sericite replace plagioclase. Accessory ilmenite, apatite and zircon are present; some of the apatite crystals are conspicuously large (0.5 to 1 mm long) and show signs of fracturing and rounding.

Intrusive bodies of this type in Devon and Cornwall are generally termed quartz-porphyries. Studies of textures and constituents have led some authors (Stone, 1968; Henley, 1974) to suggest an origin involving gas fluxing, and the form of the megacrysts considered here suggests derivation in a manner other than crystallisation from a homogenous magma. Typical elvan contains 10 to 25 per cent secondary sericite/muscovite, developed chiefly by replacement of plagioclase (Stone, 1968; Henley, 1974; Hawkes and others, 1975); in the Teign Valley elvan, however, sericite is scarce, plagioclase showing instead extensive replacement by chlorite.

The metamorphic aureole of the Dartmoor Granite

The rise in temperature caused by the intrusion of the granite induced internal rearrangement of constituent elements of the country rock, and there is evidence of metasomatism immediately adjacent to the granite. The presence of tourmaline and axinite indicates the introduction of boron, and scapolite the introduction of chlorine and carbon dioxide, while biotite in the dolerites suggests potassium metasomatism. The limit of the thermal aureole was mapped to include all spotted slates, but metamorphic effects on calcareous minerals extend beyond this line. Thus some dolerites and calcareous cherts immediately outside the aureole west of Bovey Tracey show new calc-silicate assemblages. The eastern granite margin is almost vertical and the aureole just over 0.8 km wide.

Argillaceous rocks

At the limit of the aureole the argillaceous rocks are slightly indurated. Towards the granite they show pale spots of sericite and chlorite. Near the granite the quartz of the matrix shows recrystallisation and a little muscovite and biotite has developed. Andalusite (including chiastolite) is present as white or brown crystals up to 6 m long, as locally in the Hyner Shale [SX 8323 8248]; [SX 8320 8134]. Quartz-biotitecordierite-hornfelses occur within 60 m of the granite [SX 8143 7923]; [SX 8280 8154]; [SX 8261 8278]; [SX 8176 8538]. The cordierite forms rounded crystals in a completely recrystallised fine-grained groundmass of quartz, mica and a little feldspar, and is commonly altered to chlorite and sericite. Late-stage hydrothermal alteration of the biotite has also produced chlorite and sericite. Hornfels near the contact in Whitstone Quarry shows tourmalinisation (Busz, 1899).

Cherts

Cherts outside the aureole are commonly grey or black. Within the aureole recrystallisation of the cryptocrystalline silica has produced fine-grained pale grey, cream and white quartz-hornfelses. Calcareous impurities within the cherts have led locally to the formation of calc-silicates. Thin green bands containing diopside, tremolite-actinolite, hornblende and wollastonite are present at Scatter Rock Quarry [SX 8216 8550] and elsewhere.

Calcareous rocks

Slightly calcareous Upper Devonian shales and mudstones which have been altered to calc-flintas occupy much of the aureole. These rocks are hornfelses of cherty appearance and are streaked or banded in green, white, brown and black [SX 8310 8174]; [SX 8320 8242]. Diopside occurs as bands or disseminated grains. Other calc-silicate minerals include pale yellow epidote, colourless zoisite, sphene, and small garnets. In a few specimens axinite has been observed in thin veins. Subsidiary green hornblende is common. The sericite and quartzose groundmass of the original rocks shows replacement by a fine mosaic of xenoblastic quartz and feldspar, and in the most altered varieties biotite may be present.

In the Ilsington area, outside the limit of spotting in the slates, calcareous cherts of the Teign Chert have been converted to green, cream and grey, exceedingly fine-grained calc-flintas; thin sections show granules of epidote and rare actinolitic amphibole in a cryptocrystalline matrix. Such rocks occur in Slipping Stones Quarry (p. 68) and in Lenda Wood Quarry (p. 68). Calcareous cherts within the aureole have been converted to sugary textured pyroxeneplagioclase-hornfelses, as in the railway cutting at Pullabrook Halt (p. 68). Dark green, banded rocks there consist of equant plagioclase and pyroxene, the latter commonly rimmed by amphibole. Quartz and sphene are also present. The crystalloblastic texture has obliterated most of the original features, though pyroxene-rich bands may reflect calcareous horizons.

Tuffs

Recrystallised tuffs have a fine granoblastic groundmass of quartz and a little feldspar, together with chlorite, abundant sericite, and a little biotite. Fragments within this matrix include quartz, showing signs of recrystallisation, and feldspars commonly replaced by sericite. Metamorphosed tuffs occur in Scatter Rock Quarry [SX 8218 8549].

Dolerites

The sequence of metamorphic changes within dolerites is evident at Bottor Rock [SX 8270 8045] and Inn Down [SX 8280 8435]. Farthest from the granite, hornblende occurs replacing chlorite, as narrow fibrous rims replacing the outer zones of augite crystals, and as small needles and flakes replacing the sericitic and chloritic inclusions within feldspars. Biotite is less common, but occurs locally as aggregates of minute flakes commonly replacing chlorite. Apatite and iron oxides are unaltered.

More intensive metamorphism resulted in entire augite crystals being replaced by large sheets of hornblende that still retain ophitic and subophitic relationships with the feldspar. This hornblende is commonly pale green and free from inclusions but may contain small isolated residues of augite. The feldspars are clearer, and their enclosed hornblende more euhedral in form, than in dolerites on the margins of the aureole, and biotite is slightly more plentiful.

Nearer the granite the primary augite has been completely replaced by brown hornblende, and there is more biotite present. The feldspar has recrystallised into clear andesine and labradorite, rather than albite. This replacement of albite reflects addition of lime and alumina and removal of soda. Flett (in Ussher, 1913) suggested the absorption of alkalis in the formation of biotite from chlorite, and release of lime by conversion of augite to hornblende. Iron oxides are less abundant, and are surrounded by granules of brown biotite which may have absorbed part of the iron. Leucoxene and sphene are absent. Apatite remains unaltered.

Dolerites adjacent to the granite tend to be altered to a granular mosaic of hornblende, biotite and labradorite. Pneumatolysis has locally resulted in the development of deep blue tourmaline. Comparable changes are shown by the dolerites between the Narracombe Thrust and the granite. North-west of Higher Brimley, in Eastern Cleave Plantation [SX 7882 7722], porphyritic dolerite outside the limit of spotting in the slates shows changed mineralogy. Pyroxene has been replaced by large actinolitic amphibole crystals. Amphibole also occurs in small sheaves, perhaps after chlorite, and as isolated blebs and crystals following the cleavage of plagioclase laths. The large plagioclase laths show twinning and lie in the oligoclase-andesine range. Fresher stumpy plagioclase crystals point to crystallisation of some new feldspar. Ophitic relationships are preserved. Ilmenite, commonly leucoxenised, epidote and sphene occur as accessories.

In the railway cutting at Pullabrook Halt [SX 7924 8005] crags of black, fine-grained dolerite hornfels shows subophitic texture. Intensely altered plagioclase laths occur. Large pale green amphiboles contain remnants of the original pale lilac clinopyroxene, and very small euhedral amphiboles pepper the plagioclase. Granular epidote is also present. Thin veins of garnet and pyroxene, probably of metasomatic origin, traverse the crags.

The dolerite intrusion north-west of Packsaddle Bridge is faulted against the granite but shows the last stages of the obliteration of dolerite texture. Locally [SX 7846 8030] it exhibits large amphibole crystals, presumably after pyroxene, but they are yellowish brown to dark green hornblendes rather than the pale green variety. Large original plagioclase crystals are fairly fresh but peppered with small euhedral hornblendes. The original ophitic texture is nearly obliterated by a mosaic of small equant plagioclase and hornblende crystals. Quartz and ilmenite occur as accessories. Elsewhere [SX 7846 8080] this dolerite has been completely transformed to amphibolite, with clusters of equant hornblende and fresh labradorite forming a crygtalloblastic texture.

Spilitic lavas

The spilites at Scatter Rock [SX 822 856] have retained their vesicular structures and show metamorphic changes similar to those of the dolerites. Hornblende and a little biotite have formed, and the feldspar has been altered to andesine and labradorite. A little scapolite is present in thin veins. The secondary minerals filling the vesicles show replacement by dark green hornblende, colourless or pale green pyroxene and brown biotite, together with subsidiary epidote and axinite; the last-named mineral indicates metasomatism.

Large crags of dark grey splintery hornfels in Pullabrook farmyard [SX 7924 7955] are altered amygdaloidal spilites. Green and mauve blebs represent the amygdales, and diffuse bands of purple and green fine-grained and coarser-grained material may be interbedded tuffs. In thin section the rock consists of a highly altered turbid mass of randomly orientated feldspar microliths in which are scattered completely sericitised square or rectangular feldspar phenocrysts. Small patches of chalcedony are present in places. Vesicles are generally filled with actinolitic amphibole and calcic plagioclase; more rarely colourless clinopyroxene is also present, but it is usually rimmed or partially replaced by the amphibole. Chalcedony, epidote and quartz also occur in some of the amygdales, many of which contain large crystals in the centre while the edges are rimmed with smaller ones. The thin bands, which may be metamorphosed tuffs, consist of fragments of amygdaloidal metaspilite set in a fine matrix of tiny altered feldspar laths and small decussate amphiboles. Bands of pure amphibole are present in places.

In the railway cutting [SX 7922 8006] north of Pullabrook Halt, immediately north of a metadolerite, occur small crags of very dark green, faintly banded, hornfelsed amygdaloidal spilite which seem to be part of the same horizon as that seen at Pullabrook. Thin sections show the rock to consist of altered feldspar microliths, quartz and a little granular clinopyroxene. The amygdales differ from those at Pullabrook Farm in being normally filled with garnet, clinopyroxene and calcic plagioclase, though one contained actinolitic amphibole. The pyroxene locally forms rims on garnet or plagioclase. The crags are cut by irregular veins and lenses of massive pinkish brown garnet with large crystals or abundant granules of clinopyroxene. In places the garnet is extensively chloritised. Possibly these veins are of metasomatic origin and related genetically to the garnet of the amygdales.

Permian lavas

Lavas which may represent the closing phases of Permo-Carboniferous igneous activity occur at the base of the Permian, rest on Upper Carboniferous strata and dip gently eastwards beneath Teignmouth Breccia. Their main development lies to the north, but two outcrops occur in the present district [SX 8740 8586]; [SX 8786 8303].

The lavas are greenish grey to purple aphanitic rocks containing small crystals of euhedral reddish brown iddingsite, together with a few corroded plates of plagioclase and scattered blebs of quartz, in a fine-grained basaltic ground-mass. The upper parts of the lavas are amygdaloidal, the amygdales commonly being elongated, aligned in swirl structures indicative of flow, and filled with calcite and zeolites.

Commingling of sand and lava suggests extrusion over subaerial waterlogged sands, or perhaps intrusion into the surface layers of such deposits. Veins of sand occur filling cooling cracks within the lavas. Tidmarsh (1932) thought that the lavas were composite, originating in a succession of eruptions separated by short periods of subaerial action. These views have been substantiated by Knill (1969).

Details

Dartmoor Granite

Blackingstone–East Wray–Moor Barton

Blackingstone Quarry [SX 7839 8578] is in big-feldspar granite containing scattered patches of quartz and tourmaline up to 50 mm across. Hematite staining occurs locally on some joint planes. Alignment directions of individual streams of feldspar megacrysts change over distances of a few metres. The major joints strike predominantly 170°–180°, dip 80° east to vertical, and 060°–100°, dip 70°–80° north. A few planes striking within both these azimuth ranges dip respectively east or north at 30°–60°. The floor joints are roughly horizontal, but show dips of 5° to 10° towards topographic depressions to north, north-west and west. At Blackingstone Rock [SX 7865 8560] the granite is similar, but the feldspar megacrysts are more abundant. The main joints strike 160°–170°, dip 80° west-north-west to vertical; a comparatively weakly developed set strikes 060°–100°, mainly vertical but locally dipping at 70° to 80° southward. Floor joints dip at 5° to 15° towards the main topographic hollow which lies to the north-west.

At the edge of Laployd Plantation, Hollowpark Rock [SX 8084 8477] consists of coarse megacrystic granite deeply undercut by weathering. Feldspar crystals are up to 140 mm X 32 mm, and scattered patches rich in schorlite occur, as well as sporadic dark xenoliths. The granite is traversed by a 100-mm-wide vein of aplogranite, inclined at about 30° to the nearby horizontal joints.

East Wray Quarry [SX 7823 8315] is about 12 m deep, in granite with pale pink feldspar megacrysts, mainly less than 50 mm in length, set in a comparatively fine matrix. Major joints strike 160°–190°, dip at 75°–85° east (locally 60°/102°); also 060°–080°, vertical or dipping at 70°–80° to north or south. Floor joints appear to dip south-westwards at 10° or more.

Elsford Rock [SX 7865 8300] shows about 3 m of big-feldspar granite, in which feldspar megacrysts up to 75 mm long are set in a matrix which varies from normal grain-size to a finer one probably indicative of contamination. There are scattered small xenoliths and quartz-tourmaline patches. Two sets of vertical joints are present striking 150° and 050°, and the floor joints dip at 10° to 15° south-westwards, towards the main topographic hollow. Granite exposures south and east of Elsford include an outcrop [SX 7925 8270] 120 m south-east of the farm which shows pale pink feldspar megacrysts up to 100 mm long and dark xenoliths up to 350 mm across. On the east side of the stream 710 m east of Higher Elsford, 3 m of coarse megacrystic granite were seen in what is possibly an old quarry [SX 7998 8332].

At the western end of the Kennick Reservoir dam wall an old quarry [SX 8058 8383] shows megacrystic granite which contains quartz-tourmaline patches and is reddened in places. Near-vertical major joints strike 150°–180° (mainly 180°) and 070°–100°.

Coarse megacrystic granite crops out at the road junction 500 m at 333° from Moor Barton. A 'granite pavement' outcrop [SX 817 839] 100 m east of the junction reveals pale pink feldspar crystals up to 100 mm X 75 mm, and 50 m west-south-west of the junction the granite contains an aplogranite sheet more than 50 mm in thickness. In the vicinity of the junction three sets of steep joints strike 150°–160°, about 070° and about 030°; the floor joints are sub-horizontal or gently inclined to the south-west. Red and green chloritised granite is exposed along the trackway [SX 8171 8345] 150 m west-north-west of Moor Barton.

Caseley Court to Great Rock

Pink weathered granite was noted at the road-bend just north of Caseley Court [SX 786 821], not far from the line of a possible fault trending north-west. The railway cutting [SX 7872 8213] 100 m east-north-east of Caseley Court is some 15 m deep and reveals altered green contaminated megacrystic granite beneath apparently less altered granite. The green hue is due to the widespread development of chlorite; this mineral has partly replaced biotite, with exsolution of opaque oxide which may be in part rutile (E39854). The feldspars show alteration to sericite and kaolinite. Kaolinisation along the joint planes–which are sub-horizontal and sub-vertical, striking 020°, 080° and 140°–has produced rounded blocks, characteristic of spheroidal weathering, near the bottom of the cutting.

Bullaton Rock [SX 7965 8220], 480 m at 288° from Bullaton, consists of granite with feldspar crystals up to 125 mm long and knots of quartz and tourmaline up to 25 mm across, set in a relatively fine-grained matrix. Sets of sub-vertical joints strike 060° and 150°–160°, and a plane striking 070° and dipping at 60°/340° was noted. Floor joints dip at 10° to 15° towards the south-west.

Straightening of the road 550 m east-south-east of Kelly has exposed [SX 7968 8162] megacrystic granite which is traversed by a vein of aplogranite about 75 mm wide. The feldspar crystals in the granite are up to 125 mm long and show rough alignment; biotite is common and there are patches rich in schorlite, as well as dark xenoliths up to 300 mm across, some containing abundant feldspar megacrysts. Joints strike 010°, dip at 75° west; 060°–080° (mainly 075°–080°), dip at 80° north to vertical; and 120°, dip at up to 80° to north-east and south-west. The inclination of floor joints is not clear, but appears to be at 15°–20° to south-west or south-south-west. The granite is partially kaolinised along the joints, particularly the floor joints, and harder cores weather out to form large rounded blocks. Soil of a reddish brown colour fills the vertical joints to a depth of 1 to 2 m, and also some of the floor joints. Some sporadic hematite mineralisation occurs along two zones that strike approximately 085° and dip at 80° to 85° northwards. The granite shows alteration for up to 0.3 m on either side of thin, irregularly developed quartz stringers containing rare crystals of specular hematite. Radiate sheaves of chlorite occur along the margins of some quartz stringers, and in the adjacent granite chlorite has partly or completely replaced biotite, chloritic or sericitic aggregates have replaced plagioclase, and potassium-feldspar has broken down to quartz-sericite aggregates (E41349). In granite next to this altered rock biotite is partially chloritised; plagioclase contains myriads of flakes of secondary sericite and kaolinite, but the potassium-feldspar is relatively fresh (E41348).

Ussher (1913, p. 67) recorded that the road cutting [SX 810 828] north of the dam wall between the two parts of Tottiford Reservoir showed "vertical dykes of fine-grained granite intersecting the porphyritic granite, which in one place contains a mass of pegmatite". These rocks can still be seen, the small pegmatite body displaying well-formed crystals of quartz, feldspar and black tourmaline. Granite has been quarried fairly extensively east-north-east of the lower dam wall at Tottiford Reservoir, probably for the construction of the dams themselves; at one place [SX 8067 8245] this granite contains an aplogranite sheet with a pegmatic development at the top. Along a track [SX 8090 8234] some 250 m east of the lower dam wall, pink megacrystic granite has chloritised bands adjacent to major joints which trend a little north of east; films of micaceous hematite occur on the joint planes.

Beside a forestry track [SX 8183 8184] 300 m at 058° from Beadon a cutting shows megacrystic granite with pods of quartz and tourmaline, a few dark xenoliths and local reddening. Major joints strike 070° and 1600; zones of chloritisation up to 0.6 m wide are associated with the former set, the central fissures carrying in places rosettes of tourmaline, some quartz and patches of hematite. Granite adjacent to one east-north-easterly trending joint consists, with the exception of quartz megacrysts, entirely of secondary constituents (E41360): chlorite, locally intergrown with secondary muscovite and quartz, has replaced biotite; plagioclase has been reduced to aggregates composed chiefly of kaolinite, with lesser chlorite, sericite and quartz; sericite or quartz-sericite aggregates mark the sites of potassium-feldspar crystals. Despite the degree of alteration, original granite texture is preserved. Two pegmatite developments were noted, 0.3 to 0.45 m across and 0.6 to 1.0 m in vertical extent, consisting of potassium-feldspar, rosettes of tourmaline needles and some quartz and plagioclase. Crystals of tourmaline, quartz and potassium-feldspar project into open cavities. An aplogranite sheet in the granite is 0.6 m thick, dips at 15° south-eastward and shows no marked chilling at its upper contact.

Dines (1956, p. 727) recorded that in Great Rock Mine the "country rock of grey granite with porphyritic feldspars is traversed by numerous small veins of tourmaline, alongside which the feldspars are stained red, but in the wall rock of the haematite lodes they are chloritized". At Great Rock [SX 8217 8185], 350 m at 193° from Netton, about 15 m of big-feldspar granite are exposed in crags on the hillside; pink feldspar crystals are set in a normal to fine-grained contaminated matrix, with sporadic small xenoliths. Several irregular aplogranite sills and dykes 0.3 to 0.6 m thick traverse the granite. Major joints strike 165°, dip at 75° east to vertical, and 060°–090°, vertical. Floor joints dip at 10°–15° north-west towards the Beadon Brook. On the south side of the Beadon Brook, some 425 m to the east, crags [SX 826 818] of granite are traversed by a pink aplogranite sheet up to about 0.6 m thick which is inclined irregularly towards the stream.

Gradner Rocks–Shaptor Down–Whitstone

Gradner Rocks [SX 7812 8023], 250 m south of Hisley, are near the granite margin, and here the feldspar crystals are rather fewer and smaller than those seen elsewhere in the district, the matrix ranging from normal to contaminated. Xenoliths up to 150 mm across were observed.

An old quarry [SX 7905 8060] at Knowle, about 0.8 km south-southeast of Lustleigh, is in kaolinised contaminated megacrystic granite in which there are knots of quartz and tourmaline. The granite is traversed by a thin hematitised tourmaline vein striking 170°. One xenolith measures 110 mm X 60 mm, is angular and contains a potassium-feldspar crystal some 60 mm X 35 mm. In Higher Knowle Wood, 330 m at 018° from Knowle, a massive outcrop of granite with pink feldspar megacrysts was noted.

Crags [SX 8088 8122] in a wood on the southern side of the stream north of Shaptor Down are about 3.5 m high and consist of granite with large feldspar crystals and scattered xenoliths. Joints are gently inclined towards the stream. At Shaptor Rock [SX 8095 8085] some 15 to 24 m of big-feldspar granite are exposed in slabs and crags extending down the hillside. Feldspar megacrysts up to 115 mm X 90 mm are set in normal to finer grained (contaminated) matrix. Xenoliths up to 0.3 m across are common, many containing big feldspar crystals. A slab to the west of the main exposure is crossed by a vertical 25-mm aplogranite vein which trends 025°. Vertical major joints at Shaptor Rock strike 085° and 135°, and a third weakly developed set strikes 1800; floor joints generally dip at 5°–10°/210°, but locally at 5°–10°/007°–017°.

In a road cutting [SX 8045 8024] 370 m at 341° from Wolleigh the granite contains tourmaline and biotite, but few feldspar megacrysts, and appears to be partially chloritised. To the east, in an area roughly 300 to 500 m across around Northcombe [SX 8066 8030], non-megacrystic granite predominates over the megacrystic type, but the relationship between the two varieties was not established.

A for [SX 8097 8041] in the wood 680 m at 037° from Wolleigh is 6 m high and consists of moderately megacrystic granite. Crags [SX 8169 8032] 300 m at 282° from Lower Bowden show a flat-lying vein of pink aplogranite up to about 150 mm thick in megacrystic granite. Steep joints in this vicinity strike 180° and 100°, but floor joints cannot be reliably measured.

At Whitstone Rock [SX 8134 7926] some 9 m of big-feldspar granite, with normal to finer grained (contaminated) matrix, are exposed on the hillside. The granite is traversed by many aplogranite sheets 25 to 230 mm thick, which dip at about 20° south. Major joints strike 170°–180°, dip at 80°–88° west; 075°–110°, dip at 70°–85° north; 110°–130°, dip at 75°–80° north-east; and a weakly developed set strikes 020°, dip at 85° north-west. The inclination of floor joints is not clear. In Whitstone Quarry [SX 8145 7923] the contact between granite and shaly hornfels is exposed (p. 87). The granite is of big-feldspar type, with feldspar megacrysts 30 to 35 mm in length and xenoliths ranging from 12 mm to more than a metre across.

Patches of finer grained (contaminated) biotite-rich granite occur, and veins of fine-grained granite penetrate the hornfels. Vertical joints in the granite strike 090° and 170°, but the attitude of the floor joints was not obvious.

Chapter 7 Permo-Triassic

General account

The Permo-Triassic sequence comprises locally-derived breccias and sandstones–deposited in alluvial fans adjacent to a mountain-front and showing variable but predominantly eastward palaeocurrent directions–and a flood-plain sequence of sandstones and mudstones which rests disconformably upon and overlaps the breccias, which has predominantly northerly palaeocurrent directions, and which was derived from a more distant source to the south (Laming, 1966; Henson, 1970, 1971, 1972). The Permo-Triassic succession near Teignmouth and that between the Haldon Hills and Budleigh Salterton are shown in (Figure 16), and a suggested relationship of the south Devon sequence to that of the English Midlands is presented in (Table 7).

In the southern cuvette the Watcombe Breccia is succeeded by the Oddicombe, Netherton and Teignmouth breccias, which together are equivalent to the Teignmouth Breccia of the northern cuvette. All are of local derivation and the formations are distinguished by the composition of the contained clasts. The overlying Dawlish Sandstone resembles the Bridgnorth or Dune Sandstone (Lower Mottled Sandstone) of the Midlands and is regarded as coeval with this and other dominantly aeolian sand formations, and thus of late Lower Permian age (Table 7). The Exmouth Sandstone-and-Mudstone and the Littleham Mudstone, constituent formations of the Aylesbeare Group (Smith and others, 1974), are correlated with the Permian Upper Marls of the Bakevellia and Zechstein basins. The succeeding Budleigh Salterton Pebble Beds occupy a similar position in the Permo-Triassic successions to the 'Bunter' Pebble Beds and are considered to be their lateral equivalent. The Anisian/Ladinian age of the Otter Sandstone has been suggested by the work of Walker (1969) on the rhynchosaurs (p. 101). The Budleigh Salterton Pebble Beds and the Otter Sandstone are equated with the Sherwood Sandstone Group of the Midlands and the Keuper Marl with the Mercia Mudstone Group, the group terminology adopted for the Triassic following Warrington and others (1980).

Lavas at the base of the New Red Sandstone at Dunchideock lie on Carboniferous shales and their age has been determined as 279 ± 6 Ma (Miller and others, 1962); lavas at Killerton, dated as 281 ± 11 Ma (Miller and Mohr, 1964), are underlain by a considerable thickness of New Red Sandstone sediments, as also are similar rocks near Crediton. Thus certain New Red sandstones and breccias in Devon may be of late Carboniferous age, since an age of 280 Ma was adopted for the Carboniferous–Permian boundary by Harland and others (1964). However, the New Red rocks overlie a major unconformity, and, in the absence of palaeontological evidence for tie age of the strata, this surface has been taken as the base of the Permian.

The boundary between the Permian and the Triassic is arbitrarily placed at the base of the Budleigh Salterton Pebble Beds. This follows from the possibility that pebble beds in the lower parts of the British Triassic are related to a single depositional event. However, the Exmouth and Littleham formations overlie the Permian breccia sequences with disconformity and, having abundant metaquartzite grains and northerly palaeocurrent directions, show greater affinity with the basinal flood-plain sequences assigned to the Triassic (i.e. the Budleigh Salterton Pebble Beds and higher formations) than with the breccia successions.

Permian

Watcombe Breccia

The Watcombe Breccia consists predominantly of fine-grained slate and sandstone fragments, with lenses and beds containing coarse- and medium-grained clasts of sandstone, limestone and scattered porphyry. Sandstone lenses, composed of quartz, comminuted slate and sandstone grains, are common; they usually show planar bedding but some units are cross-bedded. Fining-upwards sedimentation units overlying erosion surfaces are present. Platy slate and sandstone clasts usually lie parallel to the bedding. Thick reddish brown mudstones occur locally and Ussher (1877) described a claypit at the head of Watcombe Combe in which mudstones, showing an earthy fracture, conchoidal weathering and intercalated thin sands, were overlain by fine-grained slate breccias. Nowadays the clays are not worked and exposures are rare. The mudstones resemble those of the Exmouth Sandstone-and-Mudstone and Littleham Mud-stone formations; some are micaceous and fissile.

Oddicombe Breccia

The Oddicombe Breccia is characterised by large clasts of Devonian limestone, commonly 0.3 m across and exceptionally up to 1.5 m. Generally over 60 per cent of all clasts greater than 50 mm in diameter are composed of limestone; sandstone may form up to 50 per cent. Slate and sporadic quartz-feldspar-porphyries are also present. The limestone clasts resemble the Chercombe Bridge and East Ogwell limestones south-west of Newton Abbot and west of Bishopsteignton. Abundant fragments occur of purple coarse arkosic sandstone, with enclosed clay clasts, similar to the Ugbrooke Sandstone and presumably derived from not far to the west. Fine-grained breccias of sandstone and slate occur sporadically.

The limestone fragments are usually rounded. Laming (1954, 1966) used degrees of roundness to deduce that fans spread eastwards into a depositional basin from limestone masses at Torquay and Kingskerswell. In the coastal cliffs the limestone cobbles are distinctly grey, their hematitic film having been removed by weathering.

The matrix is hematite-stained silty sand composed of quartz grains, lithic fragments and mud; it forms up to 30 per cent of the rock. Planar-bedded sedimentation units showing graded structure and imbrication of the coarser basal clasts are common. A few units show large-scale cross-bedding. Cross-bedded or planar-bedded sandstones are present locally at the top of graded units.

Sand dykes are present in places, being vertical or near-vertical fissures filled with generally red-stained silty sandstone, each usually restricted to a single sedimentation unit. Possibly the fissures were filled by injection of quicksand from below. Alternatively desiccation and loading following flood deposition may have led to fissuring, and subsequent dewatering of the sediments may have caused fine sands to be washed into the fissures and deposited.

Netherton Breccia

Lithologically the Netherton Breccia cannot be distinguished from the Watcombe Breccia. It oversteps and overlaps the Oddicombe Breccia from Bishopsteignton westwards.

Teignmouth Breccia

The Teignmouth Breccia occupies most of the Permian outcrop. It rests on Devonian and Carboniferous rocks west of the Haldon Hills and along the Teign Estuary, and elsewhere interdigitates with the Oddicombe and Netherton breccias. The rocks range from fine- to coarse-grained, and are usually made up of fining-upwards, planar-bedded sedimentation units overlying erosional bases (Plate 10). The coarser clasts near the bases of the units, commonly show imbrication. The matrix is hematite-stained silty sand or sandy silt, composed of quartz grains and lithic fragments with silt and clay, and forms up to 30 per cent of the rock. Alternating coarse and fine breccias and sandstones in the basal part of the formation pass upwards into more uniform sediments of finer mean grain size. Some planar-bedded sandstones and mudstone lenses occur. Sandstone units 0.2 m to 2 m thick are commonly interbedded with breccia towards the top of the formation and show characteristics of both aeolian and fluvial deposition. Raindrop impressions have been recorded by Pengelly (1864) and Laming (1954, 1966). Sand dykes like those in the Oddicombe Breccia are present locally.

The following rock types have been identified in the breccias:

1. Devonian and Carboniferous sandstones, slates and cherts
2. Soft, reddish brown, medium- and fine-grained sandstones, probably penecontemporaneously derived clasts
3. Fine-grained sandstones, hornfelses, slates, spotted slates and tourmalinised aureole rocks
4. Tourmalinised and untourmalinised igneous rocks, mainly red and purple quartz-feldspar-porphyries. Basaltic lava fragments occur near the basal Permian volcanic rocks; debris of other basic lavas, andesites and spherulitic flow-banded rhyolites occur elsewhere
5. Murchisonite (pink and white perthitic feldspar)
6. Granites and microgranites, scarce and confined to the upper parts of the formation (see also Pengelly, 1862a; Worth, 1890; Scrivenor, 1948)
7. Devonian limestones, mainly in the lower parts near limestone outcrops.

Dawlish Sandstone

The Dawlish Sandstone (Plate 1) overlies and interdigitates with the Teignmouth Breccia and comprises uncemented fluvial and aeolian cross-bedded sandstones with breccia lenses. These lenses are best developed in the upper parts of the formation and are locally aggregated to form units up to 10 m thick. Typically the sandstones are 65 to 80 per cent quartz, up to 20 per cent orthoclase, up to 3 per cent plagioclase and around 10 per cent lithic fragments.

The sandstone sets and cosets of fluvial origin range from 0.2 to 2 m in thickness; they are composed of silty, cross-bedded, medium- to fine-grained sandstone or locally planar-bedded medium-grained sandstone. Contained gravel pavements comprise subangular to subrounded coarse- and fine-grained clasts, 10 to 15 mm in diameter, of sandstones, aureole rocks and quartz-feldspar-porphyries. Laterally persistent mudstone lenses, 5 mm to 25 mm thick and generally showing desiccation cracks, are common. Clasts of mudstone occur in the breccias and along the foreset planes of cross-bedded sets and the bedding planes of planar-bedded units. Many of the grains in these fluvial sets are rounded, indicating the re-working of aeolian dunes by stream action.

The intercalated breccia lenses are from 50 mm to 4 m thick. They become finer upwards from irregular erosion surfaces and commonly pass up into planar-bedded fluvial sands with scattered gravel clasts. Some occupy channels incised into the underlying sediments. In general the thicker the lens the larger the clasts. Lenses up to about 0.3 m thick show slate, sandstone and scattered porphyry clasts 5 mm to 30 mm across in bedded sandstones. Those more than 0.3 m thick show subangular sandstone and porphyry clasts 20 mm to 50 mm across in a silty sand matrix; bedding is absent but the clasts are roughly orientated parallel to the bases of the lenses. The clasts are similar to those of the Teignmouth Breccia but much smaller.

Sets and cosets of aeolian sandstone, 0.5 m to 5 m thick, are distinguished by large-scale dune cross-bedding with long curved asymptotic foresets, pronounced foreset lamination, absence of intercalated breccias and mudstones, and scattered gravel clasts. The sandstones are characteristically slightly better sorted, less silty and more rounded than the fluvial sandstones and contain no mica. Grain-size analyses (Laming, 1954, 1966) showed them to be well sorted and to have a median grain-size range between 0.31 mm (1.70 b) and 0.25 mm (2.15 (I)), close to the ranges of Recent aeolian sands and aeolian units of the Otter Sandstone. Laming also noted the abundance of 'millet seed' grains and the frosted grain surfaces characteristic of aeolian abrasion. However, most samples contain mixed aeolian and fluvially abraded grains. Fluvial sandstones and breccia intercalations commonly truncate aeolian sets. Gravel clasts in pebble layers at the base of aeolian sets show wind-cut facets.

Exe Breccia

The Exe Breccia overlies and interdigitates with the Dawlish Sandstone and is unconformably followed by the Exmouth Sandstone-and-Mudstone. The rocks resemble those of the Teignmouth Breccia. Clasts generally less than 0.15 m across, and usually between 30 mm and 100 mm, lie within hematite-stained sand and silt matrix. Graded sedimentation units are recognisable, and cross-bedded and planar-bedded sandstone sets are common at their tops. Sets and cosets of cross-bedded breccia up to 0.3 m thick occur towards the top of the formation. They comprise breccia clasts 20 mm to 40 mm in diameter in a silty sand matrix; the clasts are concentrated on the foreset planes. Solitary sets are present locally. Sandstone sets up to 3 m thick and showing large-scale dune cross-bedding occur sporadically in the upper parts of the formation; they contain no gravel or breccia and are probably aeolian.

Exmouth Sandstone-and-Mudstone

Ussher (1875, 1906, 1913) recognised but did not map a twofold subdivision of his "Lower Marls with occasional sandstones". The lower, the Exmouth Sandstone-and-Mudstone, is 255 m thick and comprises thick red and green cross-bedded sandstones intercalated within a reddish brown mudstone sequence. Five cyclothems are present in the coast section (Figure 17). In each a sharp, probably erosional, base is succeeded by planar-bedded and large-scale cross-bedded sandstones. The sandstones grade upwards into mudstone units with intercalated poorly sorted sandy siltstone beds. Some of the thick sandstones show mud-pellet conglomerates above their basal erosion surfaces. Laming (1954) suggested that the sandstones were of fluvial origin, their well-rounded grains being derived from aeolian dunes, and that the mudstones were fluvially redeposited loess. Later Laming (1966) and Henson (1970, 1971) concluded that the rocks were the deposits of a flood plain complex: the thick sandstones were formed from channel sands, the thin sandstones represented crevasse splays and levees, and the mudstones were flood-basin deposits of a northward-flowing river system.

Colour is generally linked to grain size. The thick sandstones are medium brown, with a few units of pale olive, and the sand infilling of injection structures is commonly olive green. The calcareous thin silty sandstones and sandy siltstones are pale olive below and mottled reddish brown above. Mudstones are a uniform reddish brown. The thin impersistent sandy silt laminae within the mudstone lenses of the sandstone beds are pale olive. Carus-Wilson (1913) described the occurrence of copper carbonates in green sandstones at Orcombe Point and Sandy Bay.

The sandstones form features and have been exploited in some small, now disused, quarries. On the coast between Exmouth and Straight Point [SY 036 797] the sandstone units generally persist for over 200 m. Most commonly they comprise up to 1 m of poorly sorted planar-bedded mottled green and red sandy silt and silty sand, locally showing convolute lamination. Red and green cross-bedded sandstones, with lenses of mudstone, desiccation cracks, load structures and local evidence of erosion into the underlying mudstones, range from 1.25 m to 3.0 m in thickness. Scattered red and green cross-bedded composite sandstones, generally between 10 m and 20 m thick, show large-scale and small-scale cross-bedding, mud-pellet conglomerates, desiccation-cracked mudstone lenses and load structures. Many thin green silty sandstone lenses occur, but are rarely traceable for more than 10 m. Sandstone units make up 60 per cent of the succession and comprise 86 to 95 per cent sand grade quartz and feldspar grains set in hematite-stained silt and clay. Gravel is very rare, but at Straight Point subangular breccia clasts of sandstone and chert up to 30 mm across occur in the uppermost sandstone bed (bed j, (Figure 17)). Some silty sandstones contain up to 20 per cent silt.

Large-scale cross-bedding is present in the thick sandstone units and in the lower parts of some better-sorted silty sandstones. Sets range in thickness from 0.2 m to 1.0 m in the former and 50 mm to 100 mm in the latter. Foreset laminae are usually less than 5 mm thick. Incipient calcareous cementation occurs sporadically in the coarser parts of the laminae, especially in the bottomsets. Some of the large-scale cross-bedded sets with well-laminated foresets are possibly of aeolian origin. Traces of small-scale cross-bedding are present.

The sandstone grains are subangular to subrounded, the latter common among the coarser fractions. Quartz occurs mainly as clear grains, showing straight or slightly undulose extinction with few inclusions. Composite grains of feldspar and quartz occur sporadically. Orthoclase is the dominant feldspar but microcline and plagioclase are also present. The grains are generally altered, commonly by sericitisation. Some show turbidity caused by vacuolation. The presence together of altered and unaltered feldspar suggests that the alteration is not a product of modern weathering. Meta-quartzite grains are the dominant lithic fragments, but a few pieces of slate or argillaceous sandstone were noted.

Within the thick sandstone units load casts and flame structures occur on the soles of sandstones, convolute lamination is widespread in silty beds and polygonal desiccation cracks (sun cracks) are common in mudstones. There is evidence for reworking of the silty sandstone by burrowing organisms.

Within the silty sandstones and sandy siltstones, the former usually make up the thicker beds, some showing planar and cross bedding, and the latter form the structure-less thinner beds. Poorly sorted angular to subrounded sand grains, of similar composition to those of the thick sandstones, are set in a matrix of illite, with kaolinite, chlorite, secondary calcite and micas (Henson, 1973).

The mudstones comprise angular to subangular silt and very fine sand grains, of similar composition to those of the sandstones, set in hematite-stained clay. A sporadic lamination is disrupted by structures possibly caused by bioturbation. The rocks are poorly sorted and generally contain more than 45 per cent clay. Impersistent sandy laminae may represent sand blown on to the mudstone bedding surfaces from channel bars exposed at low water. The matrix is mainly illite, with a little kaolinite and chlorite (Henson, 1973), scattered birefringent flakes of fine-grained mica and some diagenetic calcite. Small olive-green spots which occur in some mudstones probably represent tiny local concentrations of organic material.

Littleham Mudstone

The Littleham Mudstone is the upper division of Ussher's "Lower Marls with occasional sandstones". It is mainly composed of reddish brown mudstones with intercalated persistent beds of olive-green silty sandstone and sandy siltstone up to 0.6 m thick. About 90 per cent of the rocks are nearly structureless mudstones resembling those in the Exmouth Sandstone-and-Mudstone. The formation has a maximum thickness of 275 m. Good exposures occur on the coast between Littleham Cove and Budleigh Salterton. Inland exposures are largely confined to road cuttings and disused brickpits and marl-pits.

The mudstones are medium reddish brown, commonly with scattered pale green spherical and irregular spots. They comprise grains of quartz, with a few of feldspar, set in hematite-stained silt and clay consisting dominantly of illite, with kaolinite and chlorite. Silt content varies between 30 and 50 per cent, clay between 40 and 70 per cent and fine sand between 2 and 7 per cent. The beds of silty sand are composed of quartz, feldspar and rock-fragment grains, set in silt and clay which have a local calcite cement. Each is usually pale olive at the base and becomes finer and mottled brown and green upwards. Clay-rich laminae within the sandy beds are reddish brown.

The sediments appear to represent fine-grained flood-basin and coarse-grained overbank deposits with no channel deposits. The- sandstones and siltstones have sharp and planar or gently undulating bases, and exhibit small-scale cross bedding. Minute asymmetrical and symmetrical ripple-drift bedding (Walker, 1963) probably resulted from migration of small-scale ripples under conditions of abundant sediment supply. Planar bedding is common at the bases of the sandstones and siltstones. Corrugated lamination also occurs.

Ellipsoidal and stellate concretions, 50 to 200 mm in diameter and flattened parallel to the bedding, are common in the basal parts of the formation just west of West Down Beacon. They comprise dark radioactive and vanadiferous cores with pale olive-green haloes. Metallic elements and minerals reported (Carter, 1931; Perutz, 1939; Ponsford, 1955; Harrison, 1975; Durrance and George, 1976) include native copper, silver, niccolite, rammelsbergite, algodomite, bornite, covelline and malachite. The radioactive minerals include probable coffinite and thucholite-type complexes with secondary tyuyamunite and metatyuyamunite (Harrison, 1975; Durrance and George, 1976). The concretions contain up to about 0.4 per cent e U₃O₈. Organic acids may have brought about the localisation and concentration of metallic elements in the cores, though the ultimate sources of these elements are unknown. The green haloes have been attributed to leaching or reduction of original ferric iron in association with the organo-metallic environment (Harrison, 1975), to circulating groundwater (Bloomfield, 1963), or regarded (Durrance and George, 1976) as representing the primary sediment colour preserved from oxidisation. Substantial disequilibrium in the ²³⁸U decay series at the ²²²Rn position has been detected in these concretions by Durrance and others (1980), and this gives rise to a radon anomaly in stream water around the concretion occurrence.

Triassic

Budleigh Salterton Pebble Beds

The cliffs between West Down and Budleigh Salterton (Plate 11) expose the full succession of the Budleigh Salterton Pebble Beds, the easterly dip bringing successively higher horizons down to beach level. The formation is about 26 m thick on the coast, but a borehole [SY 0481 8823] -north-east of Blackhill Quarry proved 31.4 m. Inland, exposures are generally good only in quarries and pits.

Cobbles, boulders and pebbles form about 80 per cent of the formation, and are set in gravel and silty sand. Beds of cross-bedded sandstone are commonest in the topmost third of the formation, and there is an irregular upward decrease in the size of the larger clasts. Metaquartzite cobbles and boulders are the dominant clasts and are up to 0.45 m in diameter. Cobbles and pebbles of schorl, vein quartz, porphyries and sandstone are also present. Roughly planar-bedded units 0.5 m to 1.0 m thick are recognisable. Feldspar grains in both matrix and porphyries exhibit post-depositional in situ kaolinisation. The Pebble Beds are medium brown, with the sandstone beds slightly paler. Brown staining of the coarse clasts is removed by abrasion or by bleaching in the acid subsoil.

Scoured surfaces occur where cobble beds overlie sands, and are commonly represented by pebble pavements between sand sets. Blocks of mudstone up to 0.7 m across that are locally present above scoured surfaces provide evidence of penecontemporaneous erosion.

Large-scale cross-bedding is common in the sandstone beds, which are conformable on underlying cobble beds and usually show solitary sets. The foreset height of the sets. is between 0.1 m and 0.4 m, but the top of the sets is usually truncated by overlying cobble beds. Cosets are more common in the thicker sandstones in the upper part of the formation. Some of the sandstones are laminated, contain no mica flakes or gravel clasts, and may be aeolian in origin. A few thin mudstone intercalations show desiccation cracks.

Within the cobble beds cross-bedding is more rare. The cross-beds fill large, flute-shaped scours several metres deep and long, and each is a fining-upwards unit of cobbles, pebbles and sand commonly up to 0.2 m thick. Such structures are characteristic of channel infilling by braided rivers (Doeglas, 1962). Large-scale cross-bedding of cobble and sand units not associated with the infilling of channels indicates the construction of braid bars, large bars subject to erosion at their upstream faces and cross-bedded deposition at their downstream faces.

The quartzite clasts rarely show internal structures. Beneath their hematite staining they are generally grey, pale green or mottled reddish brown. Dark liver-coloured quartzites occur rarely and owe their colour to hematite pellicles around the grains, as do the red parts of mottled rocks. Many specimens show the original forms of the grains as hematite pellicles within syntaxial silica rims. Sand forms over 95 per cent of the original material in most cases. Almost all the grains are of clear quartz, with opaque inclusions commonly in rows. Turbid orthoclase, rarely sericitised, is usually the only feldspar present, but a few fine grains of unaltered plagioclase occur. The grains are usually well sorted and sub-rounded. Where the material is poorly sorted, subrounded medium to coarse sand grains lie in a matrix of subangular fine sand. An argillaceous matrix is rare and any original silt matrix has been recrystallised. However, preservation of the original shape of grains indicates that metamorphic recrystallisation has been slight. Probably the quartzites are the product of low-level regional metamorphism. Quartz in veins up to 3 mm wide traverses some of the clasts; it shows undulose extinction, indicating that the rocks were deformed after the emplacement of the veins.

Clasts other than of metaquartzite include rare tourmalinised sandstone and a few of rhyolite and porphyry. Some porphyries have been strongly tourmalinised, and most show alteration by in situ weathering. Campbell-Smith (1963) suggested that porphyries from the Budleigh Salterton Pebble Beds and the Midlands Pebble Beds were derived from the same petrographic province. Possible sources include the quartz-porphyries of Cawsand Bay and Withnoe, certain elvans, and spherulitic rhyolites of the Kings-bridge district. The presence of tourmalinised porphyries and tourmalinised hornfelses suggests derivation from aureole rocks near one of the local granites.

The sands of the matrix and of intercalated beds are reddish brown, owing to a hematite pellicle investing each grain, and they are rarely cemented. They contain little clay material but are otherwise poorly sorted. Gravel clasts are subrounded to rounded, sand grains angular and subangular to sub-rounded. Most of the gravel clasts are of metaquartzite, but many are vein-quartz, some tourmalinised sandstone and a few mudstone. The sand grains are mainly of quartz and orthoclase. The quartz commonly contains inclusions scattered at random or aligned in rows, but some crystals are clear with euhedral forms and re-entrant angles. Orthoclase grains are turbid owing to vacuole formation; cleavage traces are rarely distinguishable but a few are picked out by flakes of sericite. Grains of metaquartzite are abundant and show sub-grains with slightly sutured boundaries and marked undulose extinction. Most feldspar of the matrix was probably derived directly from igneous or metamorphic rocks. Much of the quartz and all the metaquartzite grains could have resulted from disintegration of the cobbles and pebbles, although a good deal of quartz may have arrived with the feldspars. Euhedral quartz and composite grains with sub-grains of quartz, feldspar and biotite suggest derivation from quartz-feldspar-porphyries.

Otter Sandstone

The Otter Sandstone, or Upper Sandstones of Ussher (1913), overlies the Budleigh Salterton Pebble Beds unconformably. It consists of cross-bedded sandstones containing beds of calcite-cemented conglomerate and scattered lenses of mudstone, and it was 118.8 m thick in a borehole at Great Well, near Ottery St Mary.

The formation is exposed in coastal cliffs (Plate 11), old sandpits and sunken lanes. As the sandstones are rarely cemented exposures are rapidly weathered; bare surfaces are bleached and the silt and clay matrix washed out, leaving the sandstone friable and unconsolidated. Sets and cosets of medium- and coarse-grained sandstone showing large-scale cross-bedding and with scattered gravel pavements are interbedded with silty sandstones showing small-scale cross-bedding and planar-bedding and with a few lenses of mudstone. The sandstones are rich in muscovite flakes, and the gravel and coarse sand grains are usually well rounded. Desiccation cracks in mudstone laminae are common and some of the thicker mudstone lenses contain thin green silty sandstones showing load casts, pseudonodules, flame structures, and distortion of lamination.

In the west the sandstones consist predominantly of planar-bedded units which grade upwards from coarse sand and granules to silty medium- and fine-grained sand; some large-scale and small-scale cross-bedding occurs. In the central part of the outcrop large-scale cross-bedding characterises coarse- and medium-grained sandstone sets and cosets which are interbedded with silty, micaceous, fine- and medium-grained sandstones showing large- and small-scale cross-bedding and planar-bedding. East of the district, especially in bluffs flanking the River Otter floodplain, recurring cycles comprise conglomerates passing up through sandstones showing large-scale cross-bedding into sandstone with small-scale cross-bedding.

On the coast, at the junction with the underlying Pebble Beds, a layer of ventifacts lies beneath aeolian sandstones. No such layer is present in a disused sand and gravel pit [SY 0223 8460] on Lympstone Common and the top of the Pebble Beds shows signs of slight reworking.

The sediments making up the Otter Sandstone range from pebble conglomerate to mudstone. However, sandstone predominates, largely composed of quartz, metaquartzite and feldspar grains, with flakes of muscovite in hematite-stained silt and clay matrix. Muscovite concentrated along the bedding planes commonly makes the mudstones and fine-grained sandstones fissile. The conglomerates comprise clasts of metaquartzite, sandstone, vein quartz and mud-stone, set in a matrix of sand. Calcareous cement is much more common in the conglomerates than in the sandstones.

The rocks are usually brown, locally tinged yellow. Mudstone lenses are reddish brown. Exposure to weathering and sunlight has produced bleaching to pale buff. Most of the deeper colours are concentrated in the silt and clay fraction, and washed sand grains are pale orange-brown. A thin pellicle of hematite invests each quartz grain; its absence from the feldspars is due to decomposition of that mineral.

Scoured surfaces are common in the cyclic succession immediately east of the district. The conglomerates which overlie these surfaces are succeeded by sandstones showing large-scale cross-bedding, with mudstone clasts concentrated along the foresets. In both east and west aeolian sandstones appear to be intercalated with fluvial deposits, as they are in the Exmouth Sandstone-and-Mudstone and in the Budleigh Salterton Pebble Beds.

Conditions of deposition and provenance

The Permo-Triassic rocks accumulated in a region predominantly of tropical to sub-tropical semi-arid climate, the Permian being less humid than the Triassic. The breccia formations were deposited as piedmont fans by stream floods and streams with occasional sheet floods in fanhead and midfan areas ((Figure 18); Laming, 1954, 1966; Henson, 1971). The early breccias were deposited in distinct cuvettes. As the sediments accumulated the divides were buried and the fans coalesced and interdigitated, and the later sediments were less affected by local factors.

The Dawlish Sandstone is a fanbase deposit of aeolian and fluvial sands with breccia intercalations, that migrated up-fan over the earlier breccias during a period of increased aridity that affected the whole of the British Isles in the late Lower Permian.

The Exmouth Sandstone-and-Mudstone and the Littleham Mudstone formations were deposited as an alluvial flood-plain complex (Henson, 1970, 1971). Their thick sandstone units have the characteristics of channel deposits forming coarse units at the base of cyclothems. The thin sandstones and silty and clayey sandstones show the characteristics of temporarily suspended bed-load material deposited near the channel margin as coarse overbank deposits, crevasse splays and levees. The mudstones are the fine overbank deposits of poorly sorted silt and clay deposited from suspension away from the channel area. The Exmouth Sandstone-and-Mudstone is typical of low-sinuosity low-braided alluvial environments, flood-basin silts and clays being of approximately equal thickness to the channel deposits. The Littleham Mudstone, composed predominantly of flood-basin deposits intercalated with crevasse splays and levees, represents overbank deposits accumulating away from the site of channel migration. Palaeocurrent analysis of the Exmouth Sandstone-and-Mudstone (Henson, 1971), shows a pattern characteristic of braided channels with a flow direction from the south-west (Figure 19).

The Budleigh Salterton Pebble Beds are extremely coarse and show evidence of deposition from a dense suspension and of great internal reworking, indicating deposition by stream-floods in braided channels on an alluvial fan (Henson, 1971). The succeeding Otter Sandstone is composed of fluvial channel sands, with occasional aeolian sets, occurring either as fining-upward graded units in the lower part or as cyclic fining-upward sequences overlying scoured  surfaces, with no overbank deposits. Deposition was in low-sinuosity, high-braided streams with a palaeocurrent flow from the south (Henson, 1971).

Aeolian dunes produced from the reworking of exposed channel bars are of considerable thickness in places and were formed as restricted dune-fields or isolated dunes. They occur throughout the succession, their chief development being in the Dawlish Sandstone.

Shannon (1927) ascribed all the clasts of the Watcombe, Oddicombe and Teignmouth breccias to local sources. He recorded no typical Dartmoor Granite fragments. However, the heavy minerals of the breccias were found by H.H. Thomas (1909) to include abundant acicular blue tourmaline, a mineral characteristic of the aureoles of the granites of south-west England. The Teignmouth Breccia, Dawlish Sandstone and Exe Breccia contain many fragments of the pink and white perthitic potash feldspar murchisonite. North of Dawlish the murchisonite occurs as detached fragmental crystals, whereas to the south and east it occurs within clasts of porphyry and with pieces of fine-grained groundmass adhering to the crystals. The Crediton Conglomerates (Edmonds and others, 1968) contain murchisonite in their upper parts and are considered to be lateral equivalents of the Teignmouth Breccia.

Dangerfield and Hawkes (1969) reported that granitic and microgranitic pebbles from the Crediton Conglomerates were identical to granitised argillaceous xenoliths exposed within the Dartmoor Granite, and concluded that detritus from the roof of the granite was incorporated into the Permian breccias. Similar pebbles are present in the Teignmouth Breccia. Several early workers recorded granite within the breccias, but this has not been confirmed. A feldspar phenocryst in a porphyry from the Teignmouth Breccia was determined as sanidine (Dunham in Scrivenor, 1948), but not all the feldspar megacrysts are sanidine. Hutchins (1953, 1963), assuming that the murchisonite was sanidine, concluded that it had originated in a crystal tuff. Dangerfield and Hawkes (1969) considered it to be fragments of potash feldspar megacrysts derived from granite and xenoliths in the roof of the pluton. Their study of the distribution of the marginal, big-feldspar granite led to the suggestion that no more than 50 m to 200 m of granite have been stripped off by erosion. Perhaps in Permian times the granite was exposed only at the bottoms of steep gorges through the aureole rocks.

Palaeocurrent directions (Figure 19) deduced from imbrication and roundness in the breccias (Laming, 1966; Hamblin, 1969) suggest derivation from a highland area to the west. Evidence from the Teignmouth Breccia of north-easterly currents may indicate that the mountain front swung eastwards to close the depositional basin south of the present district. Measurements of palaeocurrent directions in fluvial units of the Dawlish Sandstone show a lack of preferred orientation thought to result from their deposition by interdigitating braided streams. In contrast, current directions in the aeolian dune sandstones of this formation point to consistent palaeowinds related to the north-east trade-winds of Permian times, blowing from the south-south-east (Laming, 1966; Henson, 1971) in the present polar orientation.

Within the Budleigh Salterton Pebble Beds the only clear indication of the current flow is the form of the sandstone beds; on the coast at Budleigh Salterton they do not persist far from east to west, whereas in the Blackhill Quarry sections they are extremely persistent from north to south. This suggests the orientation of the stream channels, and flow was probably mainly from south to north as with the adjoining formations. The mixture of possible sources of the Pebble Beds (p. 97) points to tectonic uplift in the source area producing a flood of coarse detritus into a previously gently subsiding basin.

All the sandstones above the Exe Breccia are of fairly constant composition, except that those of the Exmouth Sandstone-and-Mudstone contain some slate fragments and little feldspar. Thus all probably came from the same sources. The typical "common" quartz and predominantly orthoclase feldspar point to derivation from granitic and arenaceous rocks and not to a regionally metamorphosed source.

The richest assemblage of heavy minerals (H.H. Thomas, 1902, 1909) occurs in the Pebble Beds, with staurolite, brown tourmaline, zircon, kyanite and sillimanite, in addition to the more widespread fluorite, anatase, rutile, cassiterite, ilmenite and micas. Garnet is absent from the Pebble Beds of the district, though present in presumed equivalent beds farther north and in the underlying and overlying formations; this may be due to preferential solution by rapidly circulating groundwater. Staurolite, common at Budleigh Salterton, decreases in amount northwards. Cassiterite, rare at Budleigh Salterton, increases northwards. Blue tourmaline is rare at Budleigh Salterton. H.H. Thomas (1902, 1909) suggested that a major source to the south contributed staurolite, brown tourmaline, zircon, kyanite and sillimanite, and that north of Uffculme locally-derived minerals such as blue tourmaline and cassiterite were introduced by rivers draining from the west.

Hence two main sources are indicated for the sediments above the breccias: a southern one which included high-grade metamorphic rocks, and the western Cornubian massif. Slate fragments, tourmalinised aureole rocks and porphyries probably came from the west. The southern source rocks, on the evidence of fossiliferous pebbles and cobbles, ranged from Ordovician to Devonian. The quartzites may be compared with the Gres Armoricain of Brittany and the Carne Quartzite of south Cornwall. The metaquartzite grains, pebbles and cobbles were probably derived from an earlier conglomerate, as the striking degree of rounding in the Pebble Beds suggests at least one phase of recycling.

The sands of the Exmouth Sandstone-and-Mudstone–Otter Sandstone sequence normally contain a mixture of turbid and sericitised feldspars. This mixing of two feldspar alteration types may be explained by the transport of turbid grains resulting from humid weathering in the source area, together with fresh feldspars released by stream erosion into a semi-arid depositional environment, where sericitisation of the unaltered feldspar ensued. The Pebble Beds exhibit extensive kaolinisation in both pebbles and matrix. The porphyry pebbles are particularly strongly affected, and as they now appear too fragile to survive prolonged transport it is evident that the kaolinisation took place after deposition. This kaolinisation is probably mainly due to pore water acting in the recent weathering environment.

Palaeontology

Apart from the important vertebrate remains of the Otter Sandstone, the only indigenous fossils are some plant stem fragments and rare trace-fossils. On the other hand local and exotic derived faunas and floras are relatively prominent.

The vertebrate remains occur near the base of the Otter Sandstone near the mouth of the River Otter at Otterton Point [SY 0773 8194], just outside the map area, and again at the top of the Otter Sandstone at the top of the cliff in Picket Rock Cove, Sidmouth. The original discovery was made at the Otterton locality by Whitaker (1869). It was described as a jawbone and identified as belonging to the reptile Hyperodapedon by Huxley (1869). The Sidmouth locality was discovered by Lavis (1876). It has yielded considerable material attributed to the amphibian Labyrinthodon and a single bone of Hyperodapedon (Seeley, 1876). Metcalfe (1884) described further material from this locality and other sites in the Otter Sandstone and attributed recognisable fragments to Labyrinthodon. Walker (1969 and in discussion of Warrington, 1970) noted that the records of Hyperodapedon from the Otter Sandstone referred to Rhynchosaurus, and that the few identifiable remains indicated a more primitive species than those known from the 'Keuper' Sandstone of the Midlands and Shropshire. The true Hyperodapedon of Elgin he took to occur near to, or just above, the Carnian–Norian boundary, an horizon later than the 'Keuper' Sandstone. He suggested that the 'Keuper' Sandstone fauna of the Midlands was about early to mid Ladinian in age and that the Devonshire material was possibly of mid Anisian age.

The plant remains (exclusive of miospores) amount to no more than equisetalean longitudinally fluted stem fragments, at the top of the Otter Sandstone at Picket's Cliff (Hutchinson, 1879), and similar impressions from just east of Orcombe Point (Laming, 1966).

The trace fossils of the Permo-Triassic include the rather striking ribbon-like form referred to as "annelid burrows" by Pengelly (1864) and briefly described by Laming (1966). Important features are: the meandering course of the ribbon-like traces following the bedding of the breccias; the elliptical cross-section with a concentric structure of reoriented rock fragments–the so-called "meniscus fill"; the apparent dependence of the dimensions of the fossil on the coarseness of the containing breccia; and the intersection and reworking of the earlier traces by later traces. Examples in the coarse breccias of Waterside Cove [SX 895 588] are 10 to 110 mm wide and from 100 to 200 mm long, and contain fragments up to 30 mm long in the characteristic meniscus infill. In the finer-grained Watcombe Breccia at Watcombe Cove [SX 927 673] trace fossils are smaller, being rarely more than 10 mm wide. The form of the burrows and the nature of the contained sediment make it inconceivable that they are traces of annelid activities; most probably they were made by reptiles for occupation or in search of food (Henson, 1971; Ridgway, 1974). Similar structures are seen in the fluvial sandstones of the Exmouth Sandstone-and-Mudstone at Orcombe Point [SY 020 795] and Straight Point [SY 040 797]. Thin ribbons, 5 mm to 20 mm wide; are common on the bedding surfaces and show typical meniscus infill. Less distinct bioturbation traces are visible in the mudstones of the Exmouth and Littleham formations.

Derived miospores of Devonian and Carboniferous age were obtained by Dr G. Warrington and identified by Dr B. Owens as follows: the Exmouth Sandstone-and-Mudstone at Orcombe Point, 108 m south-east of Rodney Steps [SY 0203 7957], yielded a small poorly preserved flora, including Densosporites sp., Dictyotriletes cf. falsus, Dictyotriletes sp., Savitrisporites nux, Savitrisporites sp.and Triquitrites sp.(SAL 508); the same formation 91.4 m east of the steps at Orcombe Point [SY 0216 7951] contained a rich and well-preserved flora of 44 forms (Owens, 1972) including Dictyotriletes cf. reticostus, Hymenozonotriletes lepidophytus, H. pusillites, Lophozonotriletes cf. excisus, L. malevkensis, Punctatisporites irrasus, Retusotriletes incohatus, Vallatisporites ciliaris, V. galearis and Verrucosisporites grumosus [SAL 917]; the Littleham Mudstone at Littleham Cove [SY 0392 8037], 475 m south-east of West Down Farm, yielded a few miospores, which included Apiculatisporis sp., Punctatisporites spp. and Vallatisporites cf. ciliaris. Thus indigenous miospores have not yet been obtained from either the Permian or the lower Triassic strata as far up the succession as at the district boundary in the Otter Sandstone. But a short distance east of the boundary, at Higher Dunscombe Cliff, in the lower part of the Upper 'Keuper' Marl, an assemblage of middle or late Carnian age has been found (Warrington, 1971).

At Weston Mouth six productive horizons yield Triassic microfloras, some associated with derived Carboniferous forms. The four highest are within a 10-m-thick sandstone and mudstone unit which is similar in many ways to the Arden Sandstone of the Midlands. The highest of these horizons has an abundant and well preserved assemblage of late Carnian age which closely resembles an Arden Sandstone assemblage in Worcestershire (Warrington, 1971). The remaining red Upper 'Keuper' Marl overlying the sandstone yielded no miospores, but assemblages of Rhaetian character occur in the overlying Grey Marls 2.4 km east of Seaton (Warrington, 1971).

Details

Permian

Basal unconformity

The basal unconformity is exposed on the south bank of the River Teign near Arch Brook Bridge [SX 9078 7219] and in Petit Tor Cove [SX 9265 6650]. South of the Teign the unconformity shows considerable relief; the infilling of relict valleys in the Devonian limestone by the Watcombe Breccia is clearly seen [SX 870 671] south of Kingskerswell. At Petit Tor Cove a talus slope oflimestone blocks from the adjacent limestone inlier is overlapped by Watcombe Breccia. The boundary of the Whiteway Slate and the Netherton Breccia at the western end of the Teign Estuary dips steeply to the south. At the eastern end of the estuary, where geophysical surveys suggest Devonian slates are within 7 m of the beach surface, the unconformity is probably a plane dipping gently to the east. The altitude of the pre-Permian surface around Netherton and Bishopsteignton suggests the presence of a feature extending east from the highland area into the depositional basin. North of the Teign, the base of the Teignmouth Breccia is exposed overlying Devonian limestones in several sections, for example at Harcombe Quarry [SX 884 819], and overlying Crackington Formation shales north of Kiddens Plantation at [SX 8740 8535]. Hamblin (1969) considered the pre-Permian surface north of Bishopsteignton to have been horizontal in the west and east-dipping in the east (Figure 20), the horizontal surface west of the Haldon Hills and the low summits of the Teign Valley being remnants of the late Carboniferous pediment which extended across the Teign Valley towards Dartmoor (Lowe, 1903). A seismic survey (Henson, 1972) indicated an eastward slope of the mountain front of at least 12° and a pediment infilled by breccias derived from deposits of an earlier erosion cycle.

Watcombe Breccia

West of Kingskerswell, roadside exposures [SX 8601 6839]; [SX 8598 6843] near Brooklands Farm show up to 1.25 m of weathered sandstone and porphyry breccia. Planar-bedded units dip at 14°/195° and a few beds are rich in limestone cobbles. A small overgrown quarry [SX 8698 6865] shows breccia consisting of clasts of limestone up to 0.17 m in diameter and of sandstone up to 0.10 m; planar-bedded units are discernible but their dip is not clear. A roadside cutting [SX 8746 6813] shows the proportion of limestone and the mean size of the clasts to decrease upwards. A cart track [SX 8744 6794] farther south exposes breccia of slate and sandstone with bedding vaguely defined by parallel orientation of the slate fragments. Similar traces of bedding occur in fine-grained slate breccias at Kingskerswell [SX 8777 6811]; [SX 8780 6797]; [SX 8831 6751]; the rocks contain scattered larger clasts of subangular sandstone, some resembling Ugbrooke Sandstone.

A small overgrown quarry [SX 9025 6670] north of Broomhill Plantation shows coarse- to medium-grained sandstone-rich breccia with a few limestone clasts and incipient calcite cementation. Some planar-bedded graded beds are present, dipping at 40°/353°. A few sandstone clasts show evidence of wind faceting.

In the valley south-west of Watcombe Head angular and subangular clasts mainly of slate and sandstone occur in graded sedimentation units at the tops of which thin burrows are common. In Watcombe Cove the dip is 48°/055°. Mudstone beds are exposed in small sections in the valley and at Watcombe Head beneath the Oddicombe Breccia.

Oddicombe Breccia

Watcombe–Coffinswell–Stokeinteignhead

At the base of the cliffs at Roundhouse Point [SX 9268 6666] limestone breccias overlie Watcombe Breccia of slate and sandstone. In Watcombe Cove the basal Oddicombe Breccia is distinguished from the underlying Watcombe Breccia by an increase in variety of clasts and in mean grain size. In the Valley of Rocks [SX 9250 6755] limestone breccias are magnificently exposed in the fault-scarp face. A small quarry at the foot of the scarp [SX 9235 6746] shows a dip of 31°/026° in breccias composed of sub-rounded limestone cobbles and boulders in a matrix of sandstone and slate gravel with sand and mud. Graded sedimentation units are present.

A disused quarry [SX 9166 6759] behind The Avenue, Brunel Estate, shows 4.6 m of graded sedimentation units containing sub-rounded limestone cobbles 50 mm to 150 mm in diameter dipping at 30°/355°. In a road cutting [SX 9100 6790] beneath Great Hill, breccias of limestone, sandstone and slate in a sand and mud matrix contain clasts of Permian sandstone and Ugbrooke Sandstone.

Breccia in a disused quarry [SX 8830 6874] shows limestone cobbles up to 0.25 m in diameter in planar-bedded units dipping at 8°/028°. In another quarry [SX 9050 6888] breccias comprise limestone cobbles up to 0.25 m in diameter in a sand and silt matrix predominantly composed of comminuted slate fragments. A road section extending from [SX 9124 6909] south towards Higher Rocombe Barton shows southerly dips. A hedgerow exposure [SX 9056 6963] shows limestone boulders up to 1.0 m in diameter.

Roadside sections [SX 9074 7002]; [SX 9082 7018] show breccias of sandstone and slate with isolated clasts of limestone up to 0.45 m in diameter; many of the sandstone clasts are of Ugbrooke Sandstone. Roadside sections near Stokeinteignhead show limestone cobbles up to 0.1 m across [SX 9165 7009]; [SX 9176 7024]. A temporary exposure [SX 9098 7095] in similar rocks showed limestone clasts up to 1.2 m.

Netherton to Teign Estuary

In the road section [SX 8900 7172]; [SX 8905 7175] near Manor Farm coarse limestone breccias are interbedded with fine-grained breccias of slate and sandstone. An old quarry [SX 8986 7166] at Westborough contains 3 m of planar-bedded graded units of limestone breccia dipping at 5°/090°. Another disused quarry [SX 9005 7164] in limestone breccia at Coombe Hatch shows horizontal planar-bedded graded units with concentrations of coarse clasts above the basal erosion surfaces becoming finer upwards into sands. A further quarry [SX 9115 7223] shows limestone clasts in a fine-grained matrix of slate and sandstone; some clasts of Ugbrooke Sandstone are present. Similar rocks crop out in planar-bedded graded units along the Teign Estuary between Arch Brook Bridge [SX 9095 7205] and Ringmore [SX 9211 7232]. Dips are variable but generally gentle and southerly. Quartz-feldspar-porphyry clasts and cross-bedding occur locally.

South-west of Bishopsteignton limestone breccias dip at 50°/360° [SX 8992 7289] and are faulted against red Luxton Nodular Limestone. About 200 m to the east similar breccias rest horizontally on shales [SX 9010 7287]. At Luxton's Steps [SX 9033 7282] coarse limestone breccias dip at 4°/180°, and immediately to the north-west [SX 9025 7294] finer breccias dip at 6°/150°. Coarse-grained limestone breccias crop out alongside the estuary [SX 926 729].

Coast section, Maidencombe to Bundle Head

The Oddicombe Breccia forms the undercliff, is coarser than the overlying Teignmouth Breccia and shows prominent, clearly defined, planar, graded beds. Limestone cobbles and pebbles are set in a fine-grained slate and sandstone matrix; a few sandstone and porphyry pebbles are present. The formation dips to north and south in local flexures, at angles generally less than 10°.

Netherton Breccia

Roadside sections at Buckland Barton [SX 8804 7168]; [SX 8805 7176] show fine-grained breccia with the bedding poorly defined by the orientation of slate and tabular sandstone clasts. An overgrown quarry [SX 8827 7157] shows 6 m of similar breccias which lie in well-defined planar-bedded units and dip at 10°/090°; slate and sandstone clasts usually 15 to 20 mm in diameter, with limestone and porphyry clasts up to 0.15 m, are contained in a sandy silt matrix. Coarse- and medium-grained breccias of limestone and sandstone in a small quarry [SX 8834 7172] occur in bedded units 0.2 to 0.4 m thick and dip at 7°/222°. The large clasts, predominantly limestone, are concentrated at the bases of planar-bedded units.

Teignmouth Breccia

Kiddens–Waddon Brakes–Bishopsteignton

North of Kiddens Plantation fine-grained slate breccias rest on weathered Carboniferous shales [SX 8740 8535]. Basaltic lava at the base of an artificial waterfall [SX 8786 8303] at Whiteway House (Woodward and Ussher, 1899) may be in situ and may lie at the junction. Farther south similar breccias rest on Devonian limestone; the junction is exposed in Harcombe Quarry [SX 884 819] and in other limestone quarries [SX 883 812]; [SX 8835 8150]; [SX 885 812]. At Waddon Brakes the breccias rest on Carboniferous rocks. Fine-grained slate breccias are faulted against Carboniferous sandstone in the Roman road cutting [SX 890 789].

Breccias are exposed overlying Devonian limestone in the valleys between Ideford and Luton. Sandstone units occur at the base of the Teignmouth Breccia north-west of Ideford and north-west of Bishopsteignton (Hamblin, 1969). A cross-bedded sandstone lens near Ideford Arch consists of well-rounded quartz grains with little feldspar and no mica and probably originated as aeolian dunes. The sandstone mapped on the southern slopes of Little Haldon reaveals an anticlinal structure between Rowden Copse [SX 902 747] and [SX 909 742]; the sandstone is 5 m thick, composed of well-rounded quartz and feldspar grains with flakes of mica, and interbedded with fine-grained breccias. The roundness of the quartz and feldspar grains suggests aeolian abrasion, the mica content and breccia intercalations fluvial deposition. At Higher Radway Farm [SX 903 745] the sandstone is faulted out and the breccias above abut against Devonian tuff.

Bundle Head to Dawlish

South of Bundle Head [SX 9372 7125] the Teignmouth Breccia forms the upper part of the cliffs above the Oddicombe Breccia. It is largely inaccessible but comprises medium- to fine-grained breccias of sandstone, porphyry, chert and slate with scattered coarser beds containing cobbles of sandstone, porphyry and limestone. North of Bundle Head the Teignmouth Breccia forms the entire cliff section; planar-bedded graded units dip at 5°/030° and 8°/010° in Ness Cove and at 10°/025° at The Ness. At Ness Cottage [SX 9383 7209] 2.5 m of weathered breccia of sandstone and porphyry are exposed and an exposure behind a barn at Teignharvey [SX 9158 7213] shows fine-grained breccia of slate, sandstone and chert with a few isolated clasts of porphyry and scattered rotten soft brown clasts, locally surrounded by an olive-green halo, which may be decalcified limestone.

The cliffs between Teignmouth and Dawlish expose breccias showing planar-bedded graded units with clasts of sandstone, hornfels, porphyry, slate and limestone. Limestone clasts occur mainly south of Hole Head, and are common in the breccias forming this headland and The Parson and Clerk. Cross-bedded aeolian sands in Shell Cove [SX 960 757], Coryton's Cove [SX 961 760] and behind the Parade at Dawlish [SX 962 763] are overlain and cut into by breccias. Aeolian and fluvial sandstone beds are common in the upper parts of the formation. Dewatering structures and imbrication are locally present. Asymmetrical ripple marks on some sand horizons are covered with a thin layer of mudstone. Desiccation cracks are evident in thin mudstone laminae.

Dawlish to Kenn

Road cuttings at Strand Hill [SX 9627 7693] and Stockton Hill [SX 960 771] show 5 to 8 m of fine-grained breccias of slate, sandstone and porphyry in graded planar-bedded units 0.2 to 0.35 m thick dipping northwards.

A stream section [SX 9411 7771] west of Higher Rixdale shows partially cemented, medium-grained breccias of sandstone, hornfels, porphyry and slate in planar-bedded units. Boulders of quartz-feldspar-porphyries occur up to 0.6 m in diameter. Roadside sections [SX 9430 7920]; [SX 9410 7965]; [SX 9386 7978]; [SX 9382 7970]; [SX 9335 7999] and a small disused quarry at Langdon Barton [SX 9322 7921] expose breccias of sandstone, hornfels, porphyry and slate, with planar bedding defined by the parallel orientation of tabular clasts. The breccias are predominantly fine grained but contain no sandstone units. Fragments of murchisonite are abundant.

A section [SX 9451 8076] near Brickhouse Farm shows 1.5 m of breccias of sandstone and porphyry; the cobbles are concentrated at the base of graded units and the dip is 12°/065°. In the roadside [SX 9402 8048] north of Snow's Cottages reddish brown coarse- and medium-grained sandstones dip at 21°/135°; cross-bedding is locally discernible although usually obscured by weathering. A lane [SX 9370 8072] west of Mamhead Gardens cuts through breccias of sandstone, hornfels, porphyry and slate; planar-bedded graded units dip at 10°/130° and fragments of murchisonite are abundant.

A disused quarry [SX 9350 8125] shows 4.5 m of breccias of sandstone and porphyry. Graded planar-bedded units are picked out by concentrations of coarser clasts above erosion surfaces. The planar bedding of the units is defined by the parallel orientation of tabular clasts, and the beds are horizontal.

In a disused quarry [SX 9265 8390] near Lower Thornton Farm 6 m of breccia of sandstone, slate and porphyry with much murchisonite are exposed; planar-bedded graded units dip at 119090°. Roadside sections at Beers Copse [SX 9406 8393] show fine-grained planar-bedded breccia of sandstone, slate and porphyry with abundant murchisonite dipping at 4°/088°; stream-channel washouts are evident and cross-bedded sandstones are present locally.

Dawlish Sandstone

Coast section, Dawlish to Langstone Rock

The Dawlish Sandstone overlying Teignmouth Breccia at [SX 9661 7690] comprises aeolian and fluvial sandstones with breccia intercalations. The aeolian sandstones occur as sets and cosets 2 to 3 m thick with long gently curving asymptotic foresets. The fluvial sandstones are less well sorted, more silty and commonly contain gravel pavements, scattered gravel clasts and desiccation-cracked mudstone lenses; they are generally planar-bedded but also show some cross-bedding. Breccia lenses occur throughout the formation but are most common towards the junction with the Exe Breccia.

Dawlish–Kenton–Blackheath Farm

A temporary section at Gatehouse Farm [SX 9620 7771] showed large-scale cross-bedding in 0.9 m of aeolian sandstone; the grains had typical 'millet seed' form. A section in the driveway of Secmaton Farm [SX 9639 7825] showed breccias of sandstone and porphyry intercalated into cross-bedded fluvial sandstones with thin breccia lenses. Roadside exposures [SX 9467 8091]; [SX 9464 8102] west of Mowlish Farm show aeolian sandstones with breccias, and 2.5 m of planar-bedded and cross-bedded silty fluvial sandstones are exposed [SX 9501 8105] opposite the farm.

At Arch Lodge medium- and fine-grained planar-bedded graded breccia units interbedded with cross-bedded and planar-bedded fluvial sandstones and overlying cross-bedded aeolian sandstones are exposed at the roadside [SX 9379 8158]; murchisonite fragments are common in the breccia lenses.

The road section [SX 9534 8270] near Kenton Cemetery shows fine-and medium-grained breccias overlying cross-bedded silty fluvial sandstones with breccia lenses and cross-bedded aeolian sandstones. In East Town, Kenton [SX 9581 8346], a large barchanoid dune overlies fluvial sandstones with gravel pavements and breccia lenses. It is cut by a west–east semi-circular channel filled by unbedded silty sandstones with gravel at the base and containing gravel lenses with mudstone clasts, and it is truncated above by a coarse-grained fluvial sandstone unit. A section in Fore Street [SX 9605 8323] shows cross-bedded aeolian sets, planar-bedded and cross-bedded silty fluvial sandstones, gravel pavements and breccia lenses.

The road section [SX 9480 8385] near Chiverstone Farm shows reddish brown cross-bedded and planar-bedded sandstones.

Fine-grained breccias interbedded with planar-bedded sandstones were seen in a temporary exposure in Painter's Wood [SX 9728 8285]. In a disused quarry at Discombes [SX 9641 8466] planar-bedded breccias overlie cross-bedded aeolian sandstones. The roadside section at Powderham Arch [SX 9531 8490] shows fine-grained breccias of slate, sandstone and porphyry with abundant murchisonite and interbedded sandstones dipping at 12°/040°.

Roadside sections in Kenn Lane [SX 9372 8520] to [SX 9355 8562] show fine- and medium-grained breccias of sandstone, slate and porphyry in graded units interbedded with planar-bedded and cross-bedded sandstones with gravel pavements. Similar breccias interbedded with planar-bedded sandstones are exposed in a disused quarry [SX 9464 8594] and the adjacent road section.

Exe Breccia

The Exe Breccia is well exposed in the Langstone Cliff and forms Checkstone Ledge, a reef off the eastern edge of Dawlish Warren which is uncovered only occasionally at very low tides. In the cliff coarse- to fine-grained breccias of sandstone, porphyry and slate overlie and interdigitate with the Dawlish Sandstone. Breccias low in the formation contain interbedded sandstones some of which are aeolian; these breccias are generally planar bedded, but those towards the top are commonly cross-bedded.

Breccia of sandstone, porphyry and slate in poorly defined planar-bedded units is exposed in a disused quarry [SX 9784 7833] near Dawlish Warren and in roadside sections around Westwood. In all places the apparent dip is easterly but the true dip is indeterminable. Easter Hill Lane [SX 9710 8115] cuts through planar-bedded graded units of breccia interbedded with planar-bedded sandstones; some murchisonite fragments are present.

Coarse- to fine-grained breccias, cross-bedded silty sandstones with gravel pavements and cross-bedded aeolian sandstones are exposed in cliffs bordering the Exe Estuary at Lympstone [SX 9885 8385]; [SX 9873 8420]. Planar-bedded graded units dip at 9°/057°. The breccias comprise sandstone, porphyry and slate clasts in a silty sand matrix. Murchisonite occurs as discrete fragments and in quartz-feldspar-porphyries.

Breccias of sandstone, porphyry, hornfels and slate clasts are exposed in the relict cliffs behind The Maer [SY 0022 8045] dipping at 8°/084°.

Exmouth Sandstone-and-Mudstone

The Maer to Straight Point

The section between The Maer and Orcombe Point, now poorly exposed, was described by Ussher (1913). Green sandstone is cut by a fault [SY 0155 7990] throwing down to the east, which brings down two beds of coarse white sandstone. A second fault [SY 0162 7989] downthrowing west cuts out the two sandstone beds. A third fault [SY 0174 7985] downthrowing east brings down a red sandstone bed that thickens to 9 m at the east end of the Promenade [SY 0190 7975], where it is thrown down to beach level by a further fault. This red sandstone (bed b, (Figure 17)) forms Rodney Point and Orcombe Point, and between these points it is cut by two small normal faults that have produced a graben. The eastern fault contains a calcareous-cemented fault breccia that forms a pronounced ridge across the beach. The sandstone unit is 10 to 13 m thick, dips at 5°/067° and is cross bedded, medium and fine grained, with desiccation-cracked reddish brown mudstone laminae and beds 4 to 130 mm thick. Some of the sandstones interbedded with mudstones show load casts on their bases and bioturbation structures in their upper parts. Mud-pellet conglomerates are present locally. The lower sets and cosets are in part olive green. The topmost sediments are olive green silty sands with grey mudstone laminae.

Overlying this composite bed b are 2.5 m of mudstone, followed by 0.7 m of green silty sandstone (bed c) [SY 0203 7957] with bioturbation structures in its upper part. Some 6.75 m of reddish brown mudstone with two thin bands of silty sandstone overlie bed c. Bed d comprises about 0.4 m of mottled silty sandstone with local accumulations of granule gravel in the lower part. The base of the bed is a sharp and possibly erosive contact and dips at 5°/097°. The succeeding mudstone is 5.25 m thick and contains a few thin impersistent beds of siltstone and sandstone.

Bed e is of red and green mottled fine-grained sandstone and siltstone, 0.7 m thick, dip 5°/096°, and showing small-scale ripple-drift bedding. It is overlain by 18 m of mudstone with traces of silty sandstones. Bed f is of sandstone and thickens eastwards to 2 m. It is thrown down to beach level by minor faulting and has the following sequence at [SY 0261 7960]:

The overlying 7 m of mudstone with scattered thin silty sandstones are succeeded by bed g, up to 3 m of fine- and coarse-grained sandstones with siltstones and thin mudstones and traces of calcareous cement at the base, dip 8°/103°. Above lie 44.5 m of mudstone with a few silty sandstone beds; some of the silty sandstone bands are cross-bedded and occupy channels cut into the mudstone.

The section is cut by small faults, and bed h may be examined at beach level at the foot of the path to Sandy Bay [SY 0348 7980]. Its base is an irregular surface with load casts. The unit comprises 0.7 m of green cross-bedded sandstone with mudstone partings overlain by 0.24 m of red and green mottled silty sandstone with reddish brown mudstone beds up to 100 m thick. The dip is 5°/105°. Above lie 4.4 m of mudstone containing a single thin bed of silty sandstone.

A series of 9 boreholes was drilled in connection with a sewage disposal works between the High Land of Orcombe and Straight Point. The mudstones proved to be of hard silty clay becoming harder with depth and having low compressibility. The thick red sandstones were weak and friable, but some beds of well-cemented green sandstone were encountered. Slickensided fissures were present in the mudstone but only a few joints within the sandstones.

Bed i (Figure 17), the lower of the two sandy units prominent in Straight Point, is exposed along the west side of the headland [SY 0375 7950]. It dips at 8°/113° and comprises the following section:

Some 10.8 m of mudstone with a few thin silty sandstones underlie bed j, the uppermost unit of the formation, which is exposed on the eastern side of Straight Point as follows:

Exmouth–Lympstone

The following section (equivalent to bed a in (Figure 17)) occurs alongside Plantation Walk, Exmouth [SY 0085 8027]:

The disused brick and tile works in Mud Lane [SY 0015 8210] has been largely filled but around its perimeter a few exposures of reddish brown structureless mudstone with blocky fracture remain. A rubbish-filled brickpit [SY 0115 8220] at Withycombe Raleigh also shows a reddish brown structureless mudstone. A temporary section behind a house [SY 0146 8201] on the south bank of the Withycombe Brook showed 4 m of reddish brown mudstone containing two green silty sandstone beds 0.2 m and 0.35 m thick. Cuttings on the east side of Bradham Lane [SY 0160 8184] show planar-bedded and cross-bedded sandstones with mudstone lenses and mud-pellet conglomerates, dip 9°/054°. The lower part of a bluff by the Withycombe Brook [SY 0189 8212] shows the following beds dipping at 6°/107° and overlying mudstone in the stream: planar-bedded silty sandstone with alternating coarse and fine beds, 0.89 m, overlain by planar-bedded poorly sorted sandstone and granule gravel with mudstone, 0.84 m, cross-bedded poorly sorted fluvial sandstone, 0.4 m, and reddish brown well-sorted cross-bedded probably aeolian sandstone, 3 m.

The base of the Exmouth Sandstone-and-Mudstone is exposed by the Exe Estuary at the end of Sowden Lane [SX 9898 8365], where reddish brown mudstone overlies breccias. Mudstone forms the estuary banks immediately to the south and contains a fewimpersistent beds of silty sandstone. A borehole at Sowden was recorded as having penetrated red marl, 29.8 m, on red sandstone, 21 m, on red marl, 3.3 m, on white sandstone, 0.15 m, on red marl and sandstone, 7 m, on red marl, 10.3 m. Another [SY 0065 8437], near Watton Farm, passed through sandy and pebbly Head, 3 m, into sand, 6 m, on red clay, 27 m, on red sand, 0.6 m, on red clay, 23 m.

The following section [SY 0070 8410] occurs beneath stream gravels 100m east of Watton Farm:

Littleham Mudstone

At Littleham Cove the Littleham Mudstone comprises about 275 m of mudstones within which occur the following successively higher variations. At [SY 0392 8025] two green silty sandstone beds, fining upwards, the lower one 0.4 m thick but thinning eastwards, the upper one coarser with lenses of fine gravel, dip 8°/100°. Vanadiferous radioactive nodules are abundant in the section above these beds. At [SY 0389 8032] green sandstone 0.18 m thick is cross-bedded in its lower part and carries casts ofdesiccation cracks and small scours on its base. At [SY 0393 8041] poorly sorted cross-bedded mottled red and green silty sandstone 0.55 m thick overlies green-spotted mudstone. At [SY 0394 8042] 0.12 m of fine sandstone grades up into 52 mm of mudstone overlain by coarser planar-bedded sandstones. At [SY 0427 8071] a unit up to 1.9 m thick comprises two thin green silty sandstones containing lenses of green mudstone, separated by thin red sandy mudstone. Cross-bedding, symmetrical and asymmetrical ripples and a few desiccation cracks are present. Overlying green-spotted mudstones are succeeded [SY 0428 8072] by 0.4 m of poorly sorted silty sandstone, planar-bedded in its lower part and cross-bedded above. At [SY 0445 8082], a 0.4-m sequence shows olive-green cross-bedded medium- and fine-grained sandstone overlain by silty sandstone with ripples and an upper unit of coarser sandstone and siltstone with load casts on its base.

The disused brickpit [SY 0225 8205] by Salterton Road (A376) shows 30 m of reddish brown mudstone. About halfway up the section on the west side of the pit is a 0.35-m bed of olive-green silty sandstone with mudstone lenses in its upper part. In the southern side of the pit this bed is separated by 1.6 m of mudstone from a higher bed of silty sandstone which is 0.5 m thick, dips at 5°/022° and shows large-scale cross-bedding in its lower part. Both sandy beds fine upwards from probably erosional bases. The beds are not traceable in the northern and south-eastern parts of the pit owing to talus slopes and the effects of minor faulting. A fault trending east–west has apparently brought up the sandstone beds along the southern face. A channel filled with pebble and flint gravel was exposed at the top of the pit during operations to clear the overburden and was apparently directed into the valley lying to the north.

A borehole at Knappe Cross about [SY 019 832] is recorded as having passed through 18.6 m of brown clay into 72.8 m of red marl with subordinate sandstones.

Triassic

Budleigh Salterton Pebble Beds

In the cliffs east of the The Floors [SY 0558 8150] a few clasts of mudstone and increased clay content are the only indications of erosion at the base of the Budleigh Salterton Pebble Beds, which is an almost plane surface. The top of the Pebble Beds is an apparently conformable erosional junction with the Otter Sandstone (Plate 11). About 90 per cent of the cobbles and pebbles in the Pebble Beds are metaquartzites, with a few porphyries, sandstones and cherts. They lie in a matrix of coarse- and medium-grained silty sand, rich in muscovite. Pebble-to-pebble contact is rare. The clasts are well rounded, ellipsoidal, slightly flattened and chatter marked.

The cobbles, pebbles and boulders range from 3 mm to 0.45 m across. In general the size of the clasts decreases upwards, both within the formation and in the vaguely defined coarse bedding units; but boulders are present throughout. Rarely, a conglomerate unit from 0.5 to 1 m thick grades into a sandstone bed. Inclined graded beds of cobbles and pebbles indicate large-scale sedimentary structures such as channel bars. Cut-and-fill channels occur locally. Many of the clasts show a coarse imbrication which suggests transport broadly from the south, with components from the east and west.

Sandstone beds up to 1.5 m thick are most common in the upper part of the Pebble Beds. They are cross-bedded, with conformable bottoms but eroded tops, and usually contain well-developed gravel pavements at the bases of individual sedimentation units. Some mica-free sets with no gravel clasts may be of aeolian origin. A few mudstone lenses are present showing desiccation cracks.

At the top of the formation up to 1.5 m of silty sandstone with beds and lenses of pebble gravel are overlain by up to 70 mm of silty sandstone with pebbles of metaquartzite and vein quartz. The surfaces of these topmost pebbles show wind-cut facets and a desert varnish, indicating exposure of the surface of the Pebble Beds to desert erosion.

In Budleigh Salterton a disused pit [SY 0555 8215] at the north-east corner of West Hill Plantation shows 1 to 2 m of cobbles and pebbles in a fine gravel and sand matrix. A second pit [SY 0583 8203] shows 5 m of conglomerate.

Three small overgrown pits [SY 0443 8214]; [SY 0442 8223]; [SY 0440 8227] on the northern extension of West Down Beacon show up to 8 m of well-rounded cobbles and pebbles set in a sand and fine gravel matrix. The long axes of many of the pebbles are vertical, probably owing to cryoturbation.

Bluffs by a stream at [SY 0459 8348] show Pebble Beds overlying Littleham Mudstone. Small mudstone clasts are common in the lowest 100 mm of the Pebble Beds. Bedding of the Pebble Beds is nearly vertical, suggesting displacement by minor faults. A disused quarry [SY 0330 8407] at the northern end of Withycombe Raleigh Common shows cross-bedded sandstones interbedded with pebble beds. The sandstones contain fine-grained gravel clasts concentrated along the planes of the foresets. The cobbles and pebbles show chatter marks and wind faceting; imbrication is present locally and suggests transport from the south-west.

A disused pit [SY 0230 8445] at the eastern end of Marley Lane shows cobbles and pebbles in an orange-brown sand matrix with intercalated sand beds at the top of the face; imbrication indicates transport from the east. Blackhill Quarry [SY 029 857] is the only working pit in the Pebble Beds of the district. The full thickness of the formation is worked and both east–west and north–south sections are exposed. Thick cross-bedded sandstones, showing channel structures with included lenses of gravel, are common. They are up to 4 m thick and more persistent from north to south than from east to west. The sandstones show gravel pavements and scattered fine gravel clasts along the foresets, with alternating coarse-and fine-grained foresets. Their lower surfaces are conformable but their upper surfaces are truncated by overlying pebble beds. The latter comprise cobbles, boulders and pebbles set in fine gravel and silty sandstone. Throughout the formation the mean size of the clasts decreases upwards, and the degree of sorting increases. Pebble imbrication indicates a transport generally from the south, but with some easterly and some westerly components. Cross-bedding azimuths indicate transport from between south-east and south-west.

Otter Sandstone

In the coastal cliffs [SY 0630 8174] aeolian sandstones overlie the deflation layer at the top of the Pebble Beds at this place. The lowest 100 to 150 mm of the sandstone are bright yellow and white. Above lie cross-bedded reddish brown medium- and fine-grained sandstones containing abundant slate and mudstone clasts. Mudstone lenses and mud-pellet conglomerates are locally present. The beds dip at 5°/108°.

A roadside section at the northern end of Meadow Close [SY 0599 8218] shows horizontal planar-bedded sandstones containing clasts of mudstone and gravel fragments aligned parallel to the bedding. A bed of red mudstone with fine-grained sand and mica occurs within the sandstones. It is up to 100 mm thick, rests on an eroded surface and is overlain by coarse-grained sandstones rich in mica.

About 20 m of sandstones are exposed in the roadside in Dark Lane [SY 0569 8238]. At the base of the section 5 m of reddish brown aeolian sandstones are overlain by 50 to 70 mm of mudstone. Above lie interbedded coarse- and fine-grained micaceous sands with gravel pavements, dip 7°/020°. They are locally cross bedded and contain mudstone as thin beds and clasts. The upper parts of the section comprise reddish brown sandstones showing both cross bedding and planar bedding. A mudstone lens containing thin sandy beds measures 7 m x 0.33 m. Iron pan occurs as irregular sheets or rough spherical bodies. The sandstones contain gravel pavements, of well-rounded quartzite and vein quartz with mudstone clasts. Knowle Quarry [SY 0543 8236] shows 2.82 m of cross-bedded and planar-bedded fine- to coarse-grained reddish brown micaceous sandstones, containing thin beds of siltstone and mudstone. The dip is 5°/061°. Well-rounded grains of fine gravel and coarse sand occur in laminae or scattered throughout the section. At a lane junction [SY 0587 8282] yellowish brown sandstone overlies cross-bedded sandstone. Iron pan shows as sub-horizontal dark brown banding, and one such lens below the cross-bedded set is 0.12 m thick and 4 m long.

A cutting of the disused railway line [SY 0515 8306] south of Shortwood House shows yellowish brown micaceous cross-bedded friable sandstones with thin gravel pavements and scattered fine gravel fragments. Irregular sub-horizontal pan layers are common. Where unweathered the sandstones are reddish brown. Lane cuttings [SY 0519 8346] near Dalditch Apiary expose dark reddish brown sandstones dipping at .8°/093°. Massive coarse-grained sandstone units with gravel pavements and scattered gravel fragments are interbedded with silty sandstones showing traces of cross-bedding. A disused quarry [SY 0570 8385] shows conglomerate 0.15 to 0.2 m thick underlain by yellowish brown sandstone and overlain by 0.5 m of reddish brown sandstone. The conglomerate has a calcareous cement and comprises quartz, sandstone and mudstone clasts set in a matrix of sand and clay.

In a disused gravel pit [SY 0223 8460] 1 to 2.3 m of friable orange-brown sandstones are cross-bedded with few gravel clasts in their lower part and planar-bedded with abundant gravel clasts in their upper part. They rest conformably on Pebble Beds conglomerate.

A trench [SY 0404 8453] 1 m deep by 10 m long exposed pebbly drift on fine- and coarse-grained yellow and brown sandstones with a thin bed of mudstone showing desiccation cracks, all apparently horizontal. Rain gullies on the north side of Wheathill Plantation [SY 0411 8486] show uncemented orange-brown interbedded coarse- and fine-grained sandstones with thin siltstones showing desiccation cracks; iron pan has developed preferentially within the coarser beds.

A disused sandpit [SY 0424 8518] shows planar-bedded, horizontal, interbedded coarse- and fine-grained sandstones showing evidence of channelling and reworking, 3.5 m, overlain by cross-bedded sandstones on planar-bedded coarser sandstones and ripple-marked sandstones, 0.4 m. An irregular vertical zone of iron-pan cementation at the east end of the pit is associated with distortion of the adjacent sandstone beds.

A road cutting [SY 0488 8511] near Hayes Barton shows yellow and pale reddish brown medium- to fine-grained sandstone, 0.1 m, overlain successively by reddish brown silty sandstone, 0.15 m, pale reddish brown sandstone, 0.13 m, and reddish brown silty mudstone with impersistent green laminae containing fine sand and mica; above lie an irregular layer of pan, 0.4 m, and reddish brown weathered sandstone, 1.7 m. A 5-m-deep cutting at the top of Hayeswood Lane [SY 0560 8426] shows silty sandstones with interbedded medium-grained sandstones, overlain by massive sandstone with subordinate silty sandstones.

A road [SY 0610 8554] near Yettington cuts through coarse- and medium-grained sandstones, locally cross-bedded, interbedded with micaceous silty sandstones and containing a calcite-cemented conglomerate, 0.15 to 0.3 m thick, of metaquartzite and sandstone gravel clasts with mudstone clasts in a matrix of silty sandstone.

Another cutting [SY 0616 8545] shows cross-bedded reddish brown sandstones, planar-bedded sandstones and thin mudstone lenses; gravel pavements occur within the cross-bedded sets and a 0.2-m conglomerate bed is present.

A section [SY 0702 8456] beside the A376 road shows reddish brown sandstone beneath gravel. The sandstones are fine to medium grained and cross-bedded, the bases of individual sets showing concentrations of silt-grade material. Thin siltstone beds are present, some of them lying along the foreset planes of the cross-bedded units. Conglomerates up to 120 mm thick are common. Traces of calcareous cement are present and the dip is 6°/ 104°.

A road cutting [SY 0637 8340] shows coarse flint and pebble gravel overlying sandstones dipping at 5°/101°. The sandstones are reddish brown, fine to medium grained, micaceous and rich in feldspar. Thin beds of fine gravel within them comprise subangular to rounded fragments of sandstone, vein-quartz and slate set in a silty sand matrix. Lower in the cutting the sandstones are cross-bedded. Thin mudstone beds show desiccation cracks. Thin beds of silty sandstone show concentrations of mica along the bedding planes. Interbedded coarse sandstones contain gravel pavements and imbrication in the gravels indicates that currents flowed from 230°.

Chapter 8 Cretaceous: Upper Greensand

General account

The term Upper Greensand is used in this account to include all the Cretaceous arenaceous strata in the district. Recent palaeontological work (p. 112) has demonstrated a range from Upper Albian (Lower Cretaceous) to Cenomanian (Upper Cretaceous), and the formation thus includes correlatives of the Cenomanian Limestone of the east Devon coast (Figure 21). During the Late Albian, thin blankets of glauconitic sand were widely deposited beneath rising sea waters on many ancient massifs. The Upper Greensand outliers of the Newton Abbot district are examples of such deposits, and are the most westerly outcrops of Cretaceous strata in England. The marine transgression continued during the Upper Cretaceous, with minor regressive phases, and probably culminated in the total submergence of Devon and Cornwall during the Upper Campanian.

The Upper Greensand of the Haldon Hills ranges up to at least 76 m thick and rests on Teignmouth Breccia (Plate 12). On the low ground of the Bovey Basin between Chudleigh Knighton and Kingsteignton up to 20 m of Upper Greensand rest on Devonian and Carboniferous rocks; around the Decoy Basin, south of Newton Abbot, the formation overlies Devonian and Permian strata. The difference in height between the outcrops on the Haldon Hills and in the Bovey Basin is a result of Tertiary downwarping (p. 146).

The succession consists mainly of unconsolidated and non-calcareous sands containing glauconite, with chert and cherty sandstone, thin clay seams, and shell beds largely made up of fragmentary exogyrine oyster and pectinid shells. A few calcite-cemented sandstones and limestones occur in the Bovey Basin, and it is probable that the whole formation was formerly more calcareous than it is now, though it was always less so than the Upper Greensand of the east Devon coast. The fauna (p. 178) is a largely bottom-dwelling near-shore marine assemblage, including thick-shelled bivalves and corals of littoral type. The presence of cross-bedding, gravel bands, primary glauconite, beds of shell debris and arthropod burrows indicate deposition in relatively shallow water, periodically subject to strong current and wave activity. The coarsening of many of the sediments from Haldon to the Bovey Basin, and the presence of coarse gravel at the base of the local succession at Wolborough, where pebbles of schorl are associated with apparently granite-derived quartz, indicate the proximity of the shoreline to the west and suggest that the Dartmoor Granite and its aureole rocks were being eroded in Upper Greensand times.

Construction of the A380 and A38 dual carriageway roads in 1968–1971 revealed sections on the Haldon Hills which enabled Hamblin and Wood (1976) to identify four lithostratigraphical subdivisions. Correlation is possible with the exposures in the Bovey and Decoy basins (Figure 21), where however there is evidence of basal overstep and shoreward coarsening, and with the succession on the coast between Sidmouth and Lyme Regis (p. 113). Bed numbers in the following account refer to the type section at Woodlands Goyle [SX 902 840] (pp. 115–117).

Telegraph Hill Sands (beds 1 to 6)

The Telegraph Hill Sands are green and red glauconitic sands, generally fine-grained, clay-free, and well-sorted, and containing several layers of chertified sandstone. A basal conglomerate (bed 1) with oyster-encrusted pebbles and scattered corals is present locally. Lenses of shelly indurated sandstone near the base contain chalcedonised bivalves (Basal Shell Bed, bed 3). The Telegraph Hill Sands are not known in the Bovey Basin and were probably not deposited there.

Woodlands Sands (beds 7 to 14)

The Woodlands Sands comprise glauconitic clayey sands and shell beds containing layers of siliceous concretions. At the base is the Haldon Coral Bed (bed 7), a chocolate-brown shell bed which is locally silicified or jasperised and has yielded exogyrine oysters, pectinids, trigoniids, bryozoa and corals (Plate 14); some of the fossils are beekitised. In the Bovey Basin the Haldon Coral Bed is absent and slightly higher beds lie at the base of the sequence; at Wolborough [SX 855 700] a coarse basal gravel, comprising pebbles and cobbles of quartz, schorl, Ugbrooke Sandstone and dark grey shale, indicates proximity to a shoreline. The overlying sediments are green glauconitic fine-grained clayey sands beneath buff non-glauconitic sands. Thin shelly limestones at Wolborough and Kingsteignton are the only calcareous Upper Greensand sediments in the district.

Ashcombe Gravels (beds 15 to 19)

Three beds of sand and gravel, separated by two sand units which show local cross-bedding and contain some clay seams, have been grouped together as the Ashcombe Gravels and are shown on the 1:50 000 map as a gravel horizon cropping out around the Haldon Hills. In the Bovey Basin the Ashcombe Gravels are coarser and less well sorted. Examination of the fraction coarser than 3.36 mm (-1.750) from five samples of gravelly sand from Babcombe Copse [SX 869 766] (p. 118) showed 75 per cent quartz, 4 per cent friable white feldspar grains and 2 per cent silicified shell material; the remaining 19 per cent comprised tourmaline, quartz-tourmaline, quartz-tourmaline-feldspar, pale grey sandstone, dark grey and black hornfels and dark grey slate. The percentage of feldspar is probably underestimated owing to the friable nature of the grains. Individual bands of gravel persist over long distances and indicate deposition in a relatively shallow sea subject to strong current action; their constituents were largely derived from granite. Pebbles and cobbles of vein quartz, vein quartz-tourmaline and metamorphic rock in the Bovey Basin came from the aureole of the Dartmoor Granite.

Cullum Sands-with-Cherts (beds 20 to 22)

The lower part of the Cullum Sands-with-Cherts comprises sands with courses of banded and irregular chert, and with clay seams, quartz granules and kaolinised rock fragments in its basal bed; the higher beds are chert-free sands. Absence of glauconite is characteristic. Sands rich in schorlite were seen overlying bed 22 in the upper quarry of Telegraph Hill, but were not noted at Woodlands Goyle. In the Bovey Basin the best exposure is in Babcombe Copse sandpit (Plate 13).

Subsidiary lithologies

Chert, as distinct from chertified sandstone, is restricted to the Cullum Sands-with-Cherts. Three main varieties have been distinguished (Edwards, 1970b). Banded cherts are tabular, and bedding is represented by light and dark grey alternating bands some 1 to 10 mm thick; cross-bedding and erosion surfaces are common and some specimens show cross-cutting burrow structures. They grade into bioturbated cherts, which are also tabular, with a patchy or mottled appearance due to burrows. Individual banded and bioturbated cherts are generally less than 0.3 m thick and up to 0.6 m in diameter; the largest block seen, at Waddon Brakes [SX 8916 8065], Great Haldon, was over 2 m across. Shelly cherts are irregular in form and largely composed of silicified fossil material; tests of the large foraminiferid Orbitolina are common in some examples. All three categories commonly have a light grey to dark bluish grey core and a thin porous, light brown or white skin. Thin sections show them to comprise a predominantly chalcedonic matrix enclosing detrital grains of quartz and silicified fossil material, together with minor glauconite (commonly limonitised), tourmaline and muscovite. Sedimentary structures reveal that the banded and bioturbated cherts in particular have formed by silicification of an originally clastic sediment, probably detrital limestone (Edwards, 1970a), and not from precipitated silica gel (Tresise, 1961).

Several relicts of impure limestone occur in the Woodlands Sands of the Bovey Basin, notably at Wolborough near Newton Abbot, at Kingsteignton and at Penninn. The limestones around Wolborough are poorly exposed [SX 855 700] some 400 m south of Wolborough church. They are detrital limestones, consisting largely of grains of carbonate and quartz, and include coarse-grained limestones, fine-grained bioclastic limestones, and sorted biosparites (Edwards, 1970b). The coarse-grained limestones have grains generally more than 1 mm across and comprise fine calcirudites or coarse calcarenites on the carbonate grain-size scale of Folk (1962); the discrete carbonate grains or allochems are dominantly abraded skeletal fossils (bioclasts), including echinoderm, algal, foraminiferal, polyzoan and molluscan material. Subsidiary lithologies of grains of sub-rounded quartz, minor tourmaline and glauconite cemented by sparry calcite are sandy biosparites or sandy biosparrudites (Folk, 1959). Where they contain intraclasts (up to 50 mm across) of penecontemporaneously eroded carbonates they are intraclastic biosparites. The fine-grained bioclastic limestones contain detrital angular to subangular quartz grains and pellets of glauconite and are characterised by numerous tests of Orbitolina (Hart, 1971, 1973). The cement is largely clear sparite, but patches of micritic material occur. These rocks are thus sandy glauconitic Orbitolina biosparites. The sorted biosparites are moderately well-sorted with rounded bioclasts and detrital quartz bound together by a clear sparite cement. Bioclasts are commonly silicified in their central parts, especially in the biosparrudites and intraclastic biosparrudites. Some intraclasts also show a central zone of chertification, with carbonate grains and matrix completely silicified.

It is not certain whether these limestone types form a succession. A trench at Wolborough (p. 122) showed the fine-grained biosparites with Orbitolina to be the oldest limestones in the local sequence. The coarse-grained limestones, seen only as blocks about 11 m from the eastern end of the trench, are probably the youngest.

Use of a combined alizarin red S and potassium ferricyanide stain on thin sections demonstrated the cementation sequence: (i) precipitation of non-ferroan dogtooth calcite as rims to allochems and linings to cavities, (ii) deposition of granular ferroan calcite to fill much of the remaining interparticle space, (iii) precipitation of coarse granular non-ferroan calcite in the remaining voids (Edwards, 1979).

In the Kingsteignton area there are blocks of pale greyish green highly calcareous sandy shelly glauconitic limestone [SX 8770 7393] and sandy medium-grained glauconitic limestone [SX 8778 7380], and south of Newton Abbot a trial pit [SX 8775 6929] at the Royal Aller Vale Quarry exposed blocks of sandy shelly calcarenite. Many blocks of light grey sandy calcarenite were noted in the spoil from excavations at Penninn [SX 8716 7050]; treatment with acid showed a sample from one block to have a carbonate content of 58 per cent.

Grains of glauconite, here taken to include all sand-sized green grains, not all of which consist entirely of glauconite (McRae, 1972), are abundant in the Telegraph Hill Sands and Woodlands Sands, less common in the Ashcombe Gravels, and rare to absent in the Cullum Sands-with-Cherts. A sample of Telegraph Hill Sands from Telegraph Hill [SX 9120 8355] showed equal proportions of glauconite and quartz grains, but such a concentration is rare and the green colour of the sediment is commonly due largely to staining of the quartz. Glauconite usually occurs as green to very dark green opaque lustrous rounded and botryoidal sand-sized grains. Glauconitic moulds of fossils, particularly foraminifera, have been noted. Some of the glauconite has weathered to limonitic iron oxides, which impart a reddish brown colour to the sediment. Such weathering is particularly common on Little Haldon, and may have occurred at the same time as decalcification. Some 5.7 m of Telegraph Hill Sands at Summercombe Wood [SX 928 771] are almost entirely bright red; this is the highest outcrop of the member, and seems to have been the most susceptible to weathering. The characteristic absence of glauconite from the Cullum Sands-with-Cherts is, however, unlikely to be a result of weathering, for these sands are light grey and yellowish brown rather than red; Tresise (1961) commented on the antipathy between chert and glauconite.

Boswell (1923) described the abundant heavy minerals that occur in the sands at Woodlands Goyle [SX 902 840] and Smallacombe Goyle [SX 922 768]. He recorded tourmaline, muscovite, andalusite and topaz as dominant, the andalusite being clear white and glassy or pleochroic; staurolite, corundum and kyanite as present in minor amounts; garnet and feldspar as rare. Hancock (1969) listed the percentages of heavy minerals from Great Haldon as tourmaline 94, zircon 2, rutile 2, topaz <1, andalusite <1, staurolite <1, anatase (single grain seen). Whether the heavy minerals came from the Dartmoor Granite, the Cornish granites or, for example, from Brittany, is still in dispute. (Boswell, 1923; Shannon, 1927; Groves, 1931; Smith, 1961b; Hancock, 1969).

Palaeontology

Jukes-Browne and Hill (1900, pp. 225–226) listed fossils from the Haldon Hills below the Cullum Sands-with-Cherts using the work of Downes (1882) and collections in the Geological Museum. They divided the succession into Basement Beds, Middle Beds and Chert Beds, corresponding approximately to the Basal Shell Bed, the Haldon Coral Bed and the shelly cherts of Babcombe and Kingskerswell (see below). A more comprehensive faunal list given by Reid (in Ussher, 1913, pp. 99–101) was not assigned to specific stratigraphical units.

The fauna from the Haldon Hills is listed in Appendix 3, pp. 178–180. Exposures in cuttings on the A38 and A380 roads at Woodlands and Telegraph Hill showed current-drifted accumulations of exogyrine and other oysters at various levels and three main fossiliferous horizons: the Basal Conglomerate and the Basal Shell Bed of the Telegraph Hill Sands, and the Haldon Coral Bed. Each of these beds has a characteristic lithology which allows its recognition in unlocalised museum material. A minor fossiliferous horizon at Summercombe Wood (pp. 178–179) is thought to lie within bed 4 of Hamblin and Wood (1976); another is the massive green-hearted chertified sandstone (bed 5), which has a wide lateral extent. Much fossiliferous material in museums labelled "Haldon" came from weathered grey devitrified shelly cherts ('Kingskerswell cherts') occurring as megaclasts in the Aller Gravel (p. 180); their faunas are somewhat similar to those of shelly cherts at Babcombe Copse (see p. 180) and discussion in Hamblin and Wood, 1976), and are presumed to have been derived from correlatives of the Cullum Sands-with-Cherts.

In the Bovey Basin and adjoining areas the main fossiliferous strata are the Rutitrigonia Bed of Babcombe Copse (p. 119) and its presumed correlatives, the basal shelly limestones seen during construction of the Kingsteignton bypass road, and the lower (orbitoline-rich) limestone at Wolborough. Although lithologically and faunally similar to the Haldon Coral Bed, it is believed that these beds lie towards the top, rather than at the base, of the Woodlands Sands. A minor fossiliferous horizon is the manganiferouscemented bryozoan-rich shell-gravel in the upper part of the Ashcombe Gravels at Babcombe Copse. Loose shelly cherts found on the floor of this pit, but almost certainly derived from the Cullum Sands-with-Cherts, have yielded rare poorly preserved external moulds of Hyphoplites, Mantelliceras, and a turrilitid (Hypoturrilites?) indicative of a Lower Cenomanian age.

The Upper Greensand of the Haldon Hills and its correlatives have hitherto been wrongly attributed wholly to the Upper Albian, largely on the basis of incorrect ammonite records, but it is now clear that the Cullum Sands-with-Cherts are Cenomanian (Hamblin and Wood, 1976). Dating of the lower three members is less certain, and relies to some extent on evidence from other areas.

The fauna from the Basal Shell Bed and that found at Summercombe Wood are comparable with those of the (Upper Albian) Foxmould of the east Devon coast sections, and in particular with the fauna of Downe's bed 10 of the Blackdown Greensand (Downes, 1882), which probably lies within the Hysteroceras varicosum Subzone. The only ammonites so far recorded from the Telegraph Hill Sands are indeterminate hamitids, questionably attributed to Stomohamites, that are suggestive of the Callihoplites auritus Subzone or possibly the top of the H. varicosum Subzone. The absence of the diagnostic inoceramid Birostrina sulcata, which is known from the lower part of the Blackdown Greensand, strongly suggests that the whole of the Telegraph Hill Sands post-dates the Hysteroceras orbignyi Subzone. The overlying Woodlands Sands yield faunas of Cenomanian aspect. Those of the Haldon Coral Bed (Plate 14) include molluscs apparently conspecific with forms from the type Cenomanian; the diverse coral fauna is comparable with Cenomanian assemblages from the type area and with supposed Cenomanian assemblages from Greece and Italy. However, the Ashcombe Gravels correlate with beds yielding an upper Stoliczkaia dispar Zone ammonite fauna (uppermost Albian) on the east Devon coast (Hamblin and Wood, 1976, p. 146), and therefore all the fossiliferous horizons in the Woodlands Sands and their correlatives – including the supposedly Cenomanian orbitoline limestones of Wolborough – must be Upper Albian. It is of interest that the Woodlands Sands have yielded abundant Neithea gibbosa, a species apparently restricted to the Albian, whereas the shelly Babcombe and Kingskerswell cherts are characterised by the Cenomanian Neithea quinquecosta. The Cenomanian aspect of the Woodlands faunas is probably due to an incursion of Tethyan faunas within the higher part of the Albian, for which supplementary evidence is provided by the occurrence of hermatypic corals, rare rudistids and orbitolines.

The shelly cherts at Babcombe Copse and Kingskerswell have yielded small holasterids, Catopygus columbarius and Hyposalenia acanthoides, but echinoids are not a prominent component of the faunas from the Haldon Hills, although loose examples of Pygurus lampas, presumably derived from the Woodlands Sands, have been noted from Woodlands Goyle and Telegraph Hill. Brachiopods are also poorly represented; the small Albian assemblage from the Basal Conglomerate is of considerable interest, but above this level brachiopods, apart from rare small Cyclothyris valves in the Haldon Coral Bed, are important only in the derived Kingskerswell cherts in the Aller Gravel, which have yielded Cyclothyris difformis and C. aff. compressa. A large rhynchonellid from Halldown, near Chudleigh, figured and described by J. Sowerby (1821) as Terebratula dimidiata, although heavily beekitised, does not match any known preservation on the Haldon Hills; it may be an Upper Jurassic Torquirhynchia, possibly of extra-British provenance, rather than an anomalous Cretaceous Cyclothyris of the difformis group.

Regional correlation

On sedimentological and faunal grounds it appears reasonable to correlate the Telegraph Hill Sands with the Foxmould of the east Devon coast (Figure 21) and with the middle part of the Blackdown Greensand, and the Woodlands Sands with the Chert Beds of the coast at Beer. The change from well-sorted, fine-grained, non clayey Telegraph Hill Sands to coarser, less well-sorted and clay-rich Woodlands Sands may reflect the "rapid and fundamental palaeogeographical change" reported by Smith (1961b, p. 324) between the Foxmould and the Chert Beds. The middle of the three gravel horizons of the Ashcombe Gravels (bed 17) probably equates with the "Coarse Band" taken by Smith as the base of his Top Sandstones, and bed 15 with the quartz-rich shell bed overlying a bored pebble horizon at Dunscombe, which is represented by a notch in the higher part of Smith's Chert Beds at Whitecliff. Thus Jukes-Browne and Hill (1900, p. 225) were probably wrong to correlate the chert beds of the Haldon Hills (i.e. the Cullum Sands-with-Cherts) with those of the east Devon coast.

Above the level of bed 17 correlation becomes difficult, although the highest of the Ashcombe Gravels (bed 19) may be equivalent to the quartz-rich basal bed (A1 of Jukes-Browne and Hill, 1903) of the Cenomanian Limestone of the coast; faunally the Cullum Sands-with-Cherts resemble the latter. The loose shelly cherts with at Babcombe Copse have yielded lower Cenomanian ammonites may equate with one or more levels within the Cenomanian Limestone.

Details

Buller's Hill

Buller's Hill Quarry [SX 8821 8465] shows Upper Greensand beneath the Tower Wood Gravel (see p. 128). In the north-west of the pit the former is pale brown and contains horizontal lenticles of tourmaline and fine quartz gravel. Coarse sands at the top of the unit, with quartz and schorl pebbles up to 18 mm in diameter, die out westwards. This sequence correlates with the highest beds seen at Telegraph Hill (p. 117). At the eastern end of the quarry 5.8 m of similar Upper Greensand are exposed. The sands are pale grey, coarse and micaceous, with darker bands rich in tourmaline and with brown iron-stained bands particularly evident in the top 0.6 m. Matrix kaolin has been washed down from the Tower Wood Gravel. Lenticular bands of kaolinitic quartz gravel occur throughout, and there are two courses of lensoid milk-white shelly cherts, 3.4 m and 4.9 m below the top of the formation. Fossiliferous cherts on the floor of the quarry contain Cyclothyris, small exogyrine oysters, Holaster, Thalassinoides burrow systems and some conifer debris. The south-west corner of the quarry showed the following section:

A fault may cut off this corner of the quarry, for the above section appears to be stratigraphically lower than the beds exposed elsewhere in the quarry; it may be equivalent to beds 20 and 21 at Woodlands Goyle.

Deer's Hill–Cullum Goyle

Pathside exposures [SX 8961 8426] south of Deer's Hill show horizontally-bedded dark brown Upper Greensand overlain by flinty and cherty Head. In a nearby quarry [SX 8968 8397] the Upper Greensand is overlain by Head Gravel.

Jukes-Browne and Hill (1900) recorded the following section at the head of Cullum Goyle; bed numbers relate to the section at Woodlands Goyle, below:

An exposure [SX 893 830] in the stream at the head of Cullum Goyle shows 1 m of pale grey and pale brown sand, fine to coarse grained, very micaceous and rather clayey, with three courses of cored cherts (= bed 21 at Woodlands Goyle).

A section [SX 893 828] lower down Cullum Goyle is:

The lowest six beds probably correlate with beds 12 to 14 at Woodlands Goyle (see below), and the top three beds with beds 15, 16 and 17.

Woodlands Goyle

A temporary gully [SX 902 840] in a road cutting near Woodlands Goyle showed the following beds (Figure 22) resting on the Teignmouth Breccia and overlain by the Tower Wood Gravel; it is the local type-section for the Upper Greensand.

This sequence occupies a depression in the top of Permian breccias; the section was measured at its thickest point. On one flank of the depression [SX 9015 8385], beds 6 to 21 were found to have thinned individually, their combined thickness being 7.31 m.

Telegraph Hill

A road cutting [SX 912 836] showed the following section; the bed numbers relate to the above section at Woodlands Goyle:

Two well-known quarries obliterated during the construction of the A380 dual carriageway, were described by Reid (in Jukes-Browne and Hill, 1900). Blocks of the shelly basal conglomerate of the Telegraph Hill Sands occur nearby [SX 9120 8365], and Reid recorded 0.51 m of the basal conglomerate in the lower quarry [SX 9121 8354]. The Ashcombe Gravels were exposed in this quarry in 1966 as follows; bed numbers are as at Woodlands Goyle:

Above were exposed 0.76 m of the basal bed of the Cullum Sands-with-Cherts (bed 20), finely laminated sand with scattered quartz and schorl pebbles up to 20 mm across, and large masses of chert.

Some 1.83 m of unexposed beds separate this section from the overlying Cullum Sands-with-Cherts measured by Reid (in Jukes-Browne and Hill, 1900) in the upper quarry [SX 9113 8342]:

Roadwork sections [SX 9114 8335] showed cryoturbated masses of flint gravel within the topmost bed of the Upper Greensand. This bed comprised grey and brown sands with tourmaline-rich bands and quartz-gravel lenticles up to 37 mm thick, and quartz pebbles up to 8 mm in diameter; it probably correlates with the highest bed seen by Reid.

Harcombe Valley to Ashcombe Cross

At the head of Harcombe Goyle [SX 9103 8230] 0.3 m of brown sandy laminated gravel with iron pan (?bed 19 of Woodlands Goyle) are overlain by 1.8 m of brown coarse clayey quartz sand with chert masses concentrated in the lowest 0.61 m (?beds 20 and 21). Jukes-Browne and Hill (1900, p. 222) recorded hereabouts 3.05 m of similar sands with cherts decreasing in size upwards, overlain by 1.83 m of yellow and grey glauconitic sand (?bed 22). To the west a section [SX 9075 8229] beside a forestry path shows 0.9 m of brown clayey coarse sands (?bed 16) overlain by 0.3 m of sandy gravel (?bed 17).

Between Telegraph Hill [SX 912 835] and Holloway Lane [SX 916 823] three exposures of the base of the formation demonstrate a steady southerly fall of 24.3 m in 1200 m. Blocks of Basal Conglomerate found in a culvert [SX 9256 8134] yielded a fauna which included a number of corals (p. 178).

The base of the Greensand was exposed for 216 m along a forestry track [SX 920 804] in Mamhead Bottom. Some 22.7 m of beds were present between the Teignmouth Breccia and the Haldon Gravels, but only the lowest 2 m of Telegraph Hill Sands were exposed. These were soliflucted and cryoturbated and comprised loose green loamy sands containing blocks of indurated sandstone up to 0.25 m across but no true chert. The lowest 0.3 m was coarser than the sands above and contained blocks of breccia up to 0.22 m across and derived pebbles up to 0.25 m. A wash of exogyrine oyster fragments from the Coral Bed and quartz granules from the Ashcombe Gravels was present below the Head.

Forestry track sections in Grammarcombe Wood [SX 903 800] and [SX 904 799] showed the junction of the Telegraph Hill Sands and Woodlands Sands. The shell bed equivalent to the Haldon Coral Bed contains a diverse fauna of oysters, and locally includes green
jasperised blocks of Coral Bed lithology with many rolled Yaadia and rare corals such as Placosmilia cuneiformis sensu Duncan and Stereocoenia sp.A hard shelly chert (bed 5 of Woodlands Goyle) matches that seen at Teignmouth Golf Course (below). In an old pit [SX 9054 7967] above Ashcombe, presumably that referred to by Jukes-Browne and Hill (1900, p. 223), the Telegraph Hill Sands include slabby lithified green and red glauconitic sandstones (?bed 4 of Woodlands Goyle) with poorly preserved green bivalve moulds, and oyster and pectinid shells. A forestry track [SX 9045 7947] south of the road from Ashcombe Cross to Ashcombe exposes 1.7 m of variegated clayey and gravelly sands overlain by 1.6 m of quartz gravels with sand lenticles and thin clay seams at their top; this is the section whence the name Ashcombe Gravels is derived.

Little Haldon

A forestry track in Summercombe Wood [SX 928 771] shows Telegraph Hill Sands resting on Teignmouth Breccia:

The steepness of many of the bands of lithification suggests possible Pleistocene processes.

Smallacombe Goyle [SX 922 768], the most renowned Greensand section in the Haldon Hills in the 19th century, is now almost completely overgrown. Small blocks of possible Coral Bed containing rolled Yaadia were seen virtually in place at the top of a 0.5-m exposure of Telegraph Hill Sands. Small exposures of Ashcombe Gravels within the quarry show marked pink coloration. The overlying Cullum Sands-with-Cherts are yellow, with cherts yielding poorly preserved Neithea and Limatula. Downes (1882) recognised six beds here. Jukes-Browne and Hill (1900) and Reid (in Ussher, 1913, p. 97) recorded 28m of strata as follows:

These four units appear to correlate approximately with the Telegraph Hill Sands, Woodlands Sands, Ashcombe Gravels and Cullum Sands-with-Cherts.

At Beacon Park Plantation an overgrown quarry [SX 9075 7472] shows red sands with soft and hard sandstone blocks (Woodlands Sands), overlain by 2 m of loose red, orange, greenish yellow and greyish yellow sand containing horizontal bands of gravel with rounded quartz pebbles up to 8 mm in diameter (Ashcombe Gravels). Exposures in a path [SX 9120 7490] show 0.5 m of bright red sand overlain by 0.5 m of Ashcombe Gravels.

The base of the Upper Greensand resting on Permian breccias is exposed in a road section [SX 9208 7496] 280 m at 128° from the Clubhouse of the Teignmouth Golf Club. Farther uphill the junction of the Telegraph Hill Sands and Woodlands Sands crops out in the roadside. The Coral Bed is represented by a shell bed with exogyrine oysters. The top hard bed of the Telegraph Hill Sands (bed 5 of Woodlands Goyle) is here a richly fossiliferous dark green-hearted chert with hollow moulds of exogyrine oysters (Amphidonte?) as at Grammarcombe Wood (above) and earthy moulds of Venilicardia and other heterodont bivalves.

A borehole [SX 9201 7530] on Teignmouth Golf Course proved 8.2 m of flint gravels on 24.7 m (possibly 30.4 m) of Upper Greensand. Reid (in Ussher, 1913, p. 98) reproduced the record of a well on Little Haldon which showed 8.8 m of "Chert-Beds" on 17.8 m of "Greensands"; 0.5 m of rounded gravel at the base rested on Permian rocks. The distinction between "Chert-beds" and "Greensands" may represent the division between Cullum Sandswith-Cherts and the lower members.

Gappah to Kingsteignton

Trenching south of Gappah exposed the sub-Cretaceous unconformity [SX 8627 7725]; brown, greenish grey and grey sands with cherty sandstone and a few rounded quartz pebbles rested on dark grey shales of the Ugbrooke Sandstone.

Babcombe Copse sandpit [SX 869 766] provides the best exposures of Upper Greensand seen in the Bovey Basin (Plate 13). The beds worked are the Ashcombe Gravels and the lower part of the Cullum Sands-with-Cherts, which dip at 5° to 6° westward; the underlying Woodlands Sands are generally considered too clayey to work, although they were exposed in the north-east corner of the pit and in 8 trial holes dug in 1974 (Figure 23). A composite section of trial pit 1 and the adjacent quarry face in the north-west corner of the sandpit is given below. Bed numbers refer only to Babcombe Copse sandpit:

The individual beds were traceable throughout the pit (Figure 23), although varying somewhat in thickness, and in the north-east corner some of the lowest beds thinned out against upstanding Ugbrooke Sandstone. Several trial pits revealed lenses of quartz gravel.

A trench dug in 1971 proved lenticles of pale green jasperised cemented shell masses at the base of bed 4; the relatively poorly preserved fossils from this Rutitrigonia Bed' included common Rutitrigonia and Helicomptus, together with Callistina, Crenella, Limatula, Protocardia, Pterotrigonia, Trigonarca, Avellana and a turritellid. Rare small orbitolines, including conical forms, were also present.

The pinkish white concretions of bed 9 are locally sub-vitreous and some contain exogyrine oysters. This bed is also characterised by convolutions, possibly loading structures or caused by cryoturbation. The shell bed at the base of bed 10 is correlated with bed 15 of the Woodlands Goyle type section. The layer of manganiferous concretions low in bed 11 ranges from a patchily cemented shelly gravel in the south of the pit to large lenticles in the north; it has yielded bryozoa, Barroisia, exogyrine oysters, and Cyclothyris?. Cross-bedding in bed 11 indicates current flow from the north-north-east.

Some 96 m at 216° from the above section (C in (Figure 23)), basal Cullum Sands-with-Cherts rest on the following Ashcombe Gravels sequence. Bed numbers correspond to the above section at Babcombe Copse.

At 38 m beyond this section, in the south-west corner of the pit, the following beds were measured (B in (Figure 23)), their numbers corresponding to the Babcombe Copse section.

Blocks of fossiliferous cherts on the floor of the pit have yielded external moulds of three Cenomanian ammonites (p. 113); they are almost certainly from the Cullum Sands-with-Cherts, perhaps from a bed slightly higher than any now exposed in the pit faces.

East and north-east of Babcombe Copse, boreholes sunk in connection with proposed road works proved: 6.5 m of green and brown clayey sands, pebbly in part and \ with scattered cherts, resting on Upper Carboniferous rocks [SX 8745 7711]; 6.4 m of Upper Greensand, base not reached [SX 8719 7656]; 10.6 m of olive-brown and green pebbly sands, shelly in places, base not reached [SX 8715 7641].

Trial pits [SX 866 760] to [SX 868 754] showed up to 2.4 m of olive-green medium to coarse-grained glauconitic sand overlain by up to 4.0 m of Aller Gravel. Two boreholes [SX 8664 7586] and [SX 8674 7564] showed 4.6 m and 12.2 m respectively of sands overlain by Aller Gravel.

A seismic refraction survey indicated an increase in thickness of the Upper Greensand from about 12.2 m in the central part of the Sands Copse gravel-pit to 20.4 m in the northern part (p. 159).

Route-investigation boreholes east of Sands Copse proved: 6.7 m of brown and green pebbly sand with ?chert horizons, base not reached [SX 8708 7608]; 13.6 m of coarse clayey sand, base not reached [SX 8706 7593].

A now-degraded section [SX 8714 7570] in the lane leading to Combe Farm was recorded and figured by Godwin-Austen (1842) as "showing the base of the [Greensand] deposit, resting on carbonaceous shales, and containing fragments of Culmiferous grit, also Exogyrae, Pecten quinquecostatus, etc., in great abundance". A westerly dip into the Bovey Basin was shown. Godwin-Austen also noted that "below Ponswine Farm, the formation presents fine-grained beds with green earth and numerous characteristic fossils. The deep cutting for the new road affords a good section showing the thick capping of the greensand; in this section are lines of chert, and some very remarkable beds almost entirely composed of Orbitolites". The location of this section is uncertain.

Temporary exposures [SX 8770 7484] 70 m west of Torhill Cottages showed 0.6 to 0.9 m of greenish brown medium- to fine-grained micaceous sands with scattered pebbles. Fissures and hollows in the upper surface of Devonian limestones in Rydon Quarry [SX 8747 7412 and 8749 7389] are filled with pebbly olive-green Upper Greensand. The Greensand sags into one hollow [SX 8746 7396].

Near Humber Lane [SX 8773 7396] 2 m of medium- to coarse-grained slightly micaceous green sands with pebbly bands containing white friable feldspar are overlain by dark olive-green clayey glauconitic sand with rounded pebbles, 0.3 m, and Aller Gravel, 2 m. The junction between Upper Greensand and Aller Gravel is marked by 30 to 40 mm of brick-red clay.

Cuttings for the Kingsteignton bypass [SX 877 739] revealed lower Upper Greensand beds occupying solution cavities within the Chercombe Bridge Limestone.

The following is a composite section:

The lowest two beds probably equate with bed 6 of Wolborough (see below) and with the Rutitrigonia Bed of Babcombe Copse. The top three beds are probably Ashcombe Gravels.

Staplehill–Whitehill

Godwin-Austen (1842) considered that Upper Greensand was present along the southern margin of the main Bovey Basin, at Staplehill and Whitehill. Possible Upper Greensand was augered [SX 8285 7360] near Singmore Farm.

Wolborough to Kingskerswell

Field brash 400 m south of Wolborough church includes shelly calcarenites packed with Orbitolina. A north-east-south-west trench [SX 855 700] revealed the following composite section (A in (Figure 21)). Dips were easterly and mainly slight, but steepened to 32° at one place. Bed numbers refer only to Wolborough.

Augering up to 11 m beyond the north-east end of the trench revealed dark olive-green and reddish brown glauconitic gravelly sand; blocks of cemented quartz gravel occurred as brash, and these materials may represent basal Ashcombe Gravels. All the strata in the trench, including both limestones, belong to the Woodlands Sands and are of Albian age. The lower limestone has yielded poorly preserved moulds of large bivalves including Callistina,

Glycymeris, Trigonarca and several trigoniid genera (Apiotrigonia, Pterotrigonia and Yaadia), numerous turritellid and naticid gastropods, in addition to the ubiquitous Helicomptus and complex serpulid masses. The higher beds contain abundant orbitolines, and Neithea gibbosa and Helicocryptus are characteristic. Beds 4 and 5 correlate with beds 1 to 4 of Babcombe Copse and bed 7 with bed 8 of Babcombe Copse. It is probable that the lower part of the lower limestone, with moulds of Trigonarca and other large bivalves, equates with the lenticular Rutitrigonia shell bed locally developed at the base of bed 4 of Babcombe Copse, and with the basal shelly limestone at Kingsteignton (p. 121).

Temporary exposures [SX 8598 7075] north-east of Wolborough church showed 0.9 to 1.2 m of pale olive-green sparsely glauconitic pebbly sands dipping southward; the pebbles ranged up to 10 mm in diameter and were mainly of quartz with some white friable feldspathic material.

Roadworks at Wildwoods Copse [SX 876 717] showed Upper Greensand overlain by Aller Gravel. A borehole [SX 8754 7177] passed through flinty Head, 2.9 m, on olive-green clayey sand, pebbly in part with ?cherty bands, 8.4 m, on Upper Devonian slates. Other boreholes sunk in connection with road construction proved: 9.6 m of Aller Gravel on 1.8 m of dark green and grey compact clayey sands with scattered cherty bands [SX 8715 7063]; 7.3 m of Head and Aller Gravel on 4.9m of sands [SX 8717 7073].

An excavation [SX 8718 7051] at Penninn showed 2 m of Upper Greensand beneath about 2 m of flinty Head. Micaceous pale green fine- and medium-grained sand was overlain by green pebbly and clayey gravelly sand and gravel. Grain-size analysis showed the lower sands to be comparable with bed 9 at Babcombe Copse (Woodlands Sands), and the gravels with the lower part of bed 10 at Babcombe Copse (Ashcombe Gravels). Blocks of pale grey sandy limestone noted in the spoil from an excavation [SX 8716 7050] resembled limestones from bed 8 in the trench at Wolborough.

The following section was seen behind Newtake Garage [SX 8743 7060]:

Temporary exposures [SX 8805 7043] showed greenish yellow micaceous glauconitic medium- to fine-grained sands overlain by 0.6 to 1.2 m of reddish brown flinty Head. Similar sections were seen in shallow excavations in Square Plantation [SX 884 703] to [SX 885 706].

A quarry-face section [SX 877 695] in Royal Aller Vale Quarry was extended downwards in a trial pit. Beneath Aller Gravel the composite section of Upper Greensand was:

Possibly the dark green beds at the base equate with beds 1 to 4 of Babcombe Copse (Woodlands Sands), and the highest Greensand includes Ashcombe Gravels and the lowest part of the Cullum Sands-with-Cherts.

Examination of spoil and of quarry company logs of a nearby trial pit [SX 8775 6929] suggests that at least 6.7 m of Upper Greensand are present almost immediately below the quarry floor. Blocks of sandy Shelly calcarenite within greenish yellow micaceous glauconitic sands suggest a possible correlation with the Orbitolina limestone and overlying micaceous sands at Wolborough (p. 122)

The presence of an Upper Greensand outlier, capped by Aller Gravel, around Connybear Brake [SX 8908 6934], was indicated by surface debris and exposures in temporary trenches. The base of the formation appears to dip at about 7° south-west.

Between the north end of Harpins Brake [SX 8815 6878] and Daccombe Mill [SX 8874 6779], the Upper Greensand dips at about 12° west beneath Aller Gravel.

Chapter 9 Palaeogene: Eocene flint gravels

General account

The Haldon Hills are capped by three units of flint gravel, formerly called the Haldon Gravels (Figure 24). Of these the Tower Wood Gravel and the later Buller's Hill Gravel are of Eocene age; the third unit is now considered to be Pleistocene (p. 148). Farther west, the (Eocene) Aller Gravel crops out along the eastern side of the Bovey Basin and around the Decoy Basin south of Newton Abbot.

Tower Wood Gravel

The Tower Wood Gravel, named from Tower Wood Quarry [SX 8768 8567] at the northern end of Great Haldon, comprises up to 8 m of unbedded closely packed large unabraded flints in a matrix of clay and a little sand (Plate 15). A basal sandy flint-free bed up to 125 mm thick, containing well-rounded quartz and schorl pebbles, rests on Upper Greensand. Similar gravels occur outside the present district at Combpyne Station, north-east of Seaton, and at Hardown, east of Lyme Regis.

There are several outcrops on the northern and central parts of Great Haldon. Farther south, and all around Little Haldon, unabraded flints in the Head Gravel point to the presence of concealed Tower Wood Gravel, but any outcrop is too narrow to be shown on the 1:50 000 map. Brunsden and others (1976) recorded Tower Wood Gravel sandwiched between Upper Greensand and Aller Gravel in solution pipes in Devonian limestone along the Kingsteignton–Newton Abbot bypass [SX 874 743] to [SX 879 734].

The flints of the Tower Wood Gravel are pale grey inside and white outside, rough and pitted, and not abraded by transport. Their margins have been shattered in situ by frost action. The matrix clay is generally white with local brown staining. Hamblin's (1973b) examination showed some samples containing high-crystallinity 'china clay' kaolinite, probably of hydrothermal origin, with a little disordered 'ball clay' kaolinite and illite; other samples contained a mixture of 'china clay' and 'ball clay' kaolinite similar to that in the overlying Buller's Hill Gravel, whence it may have come by eluviation.

The base of the Tower Wood Gravel may be horizontal, folded (Hamblin, 1972), or gently undulatory possibly owing to periglacial heaving, but there is no piping at the base comparable with that beneath the Clay-with-flints farther east.

Buller's Hill Gravel

The type locality of the Buller's Hill Gravel is Buller's Hill Quarry [SX 8821 8465] in the northern part of Great Haldon. The gravel comprises up to 10 m or more of abraded flint gravels with subordinate clay bodies and sand beds. It occurs on both Great Haldon and Little Haldon, and where locally removed its former presence is suggested by the occurrence of abraded flints in the Head Gravel.

The deposit comprises flints up to 0.3 m across, together with pebbles of vein quartz, schorl, quartzite and thermally altered Carboniferous shale and chert, in a sand and clay matrix. It is pale brown or pale grey, but generally darker than the Tower Wood Gravel. The flints are grey and have been abraded during transport–many bear small interlocking arcuate fractures ('chatter marks') on their surfaces produced by the impact of other cobbles. Pleistocene frost action has shattered many of the flints in situ. The other pebbles are generally smaller, more rounded and less frost-shattered. The megaclasts are in close contact, suggesting late infiltration of the matrix, and the latter is sandier than in the Tower Wood Gravel. Beds of sand associated with the gravels are similar to the sand of the matrix. Hamblin (1973b) found that the clay of the matrix comprised approximately equal amounts of ordered and disordered kaolinite, with more illite than occurred in the Tower Wood Gravel; he concluded that the matrix clay probably represented a mixture of materials reworked from the Tower Wood Gravel and derived from the weathering of Upper Palaeozoic shale; high- and low-crystallinity kaolinites were possibly added from weathering of the Dartmoor Granite.

Sheet-like horizontal bodies of grey or white kaolinitic clay occur within the Buller's Hill Gravel; the largest recorded example [SX 9180 7515] measured 11 m x 8m x 2 m. They comprise mainly disordered kaolinite with a large and variable content of illite and silt, possibly derived from the weathering of shales, and they are less pure than the commercial ball clays of the Bovey Basin. Each clay body commonly has a bed of sand at its base, and some contain pods of flint gravel up to a metre across.

Aller Gravel

The Aller Gravel (Plate 16) contains grey and brown flints which show chatter marks and a moderate degree of abrasion. Other components include vein quartz, tourmaline and quartz-tourmaline rock, Upper Greensand chert, pale grey Lower Carboniferous chert, grey Upper Carboniferous sandstone, ?dolerite and metadolerite, hornfels, and other dark grey fine-grained aureole rocks. All are set in sands with subordinate white and grey silts and clays. The formation normally rests on Upper Greensand and crops out along the eastern side of the main Bovey Basin between Sands Copse [SX 866 760] and Kingsteignton, and more extensively south of Newton Abbot where it reaches its maximum thickness of some 25 m near Aller. North of Newton Abbot both Aller Gravel and Upper Greensand dip west beneath the Bovey Formation. To the south the gravels dip inwards towards the centre of the Decoy Basin.

Small outliers of Aller Gravel and Upper Greensand north-east of Kingsteignton, and at Connybear Brake [SX 8908 6934] near Kingskerswell, are now at greater heights than the main outcrop and may indicate a link with the Cretaceous–Tertiary sequence of the Haldon Hills.

Sedimentation and derivation

Boswell (1923) and Groves (1931) found that the heavy minerals of the Haldon Gravels included abundant rutile, zircon and large grains of tourmaline, plentiful large colourless glassy grains of topaz, moderately abundant andalusite, scarce staurolite, local anatase and rare to very rare garnet and kyanite; single grains of epidote, hornblende and chlorite were recorded. This mineralogy is common to both Tower Wood and Buller's Hill gravels, but heavy minerals are rare in the clay bodies within the latter.

The Tower Wood Gravel is characterised by unabraded flints and probably represents the material remaining after the in situ solution of chalk. The basal bed of sand with pebbles of quartz and schorl is considered to be the residuum of flint-free lower Chalk that was deposited as a basal conglomerate during the westward transgression of the Chalk sea.

A fluvial origin for the Buller's Hill Gravel is indicated by the low degree of rounding of the flints, with chatter marks restricted to the extremities of the horns. Its clays are kaolinitic, more characteristic of river sediments than marine, and the presence of well-ordered kaolinite argues against the later conversion of other clays (Hamblin, 1973a). The coarseness of the Buller's Hill Gravel, its lack of sedimentary structures, and its apparently uniform spread over the Tower Wood Gravel throughout the Haldon Hills, suggested to Hamblin (1973b) the action of sheet floods alternating with conditions of quiet fluvial redistribution and the washing in of the• clay and sand matrix during the dry seasons.

The uniformity of the clay bodies in the Buller's Hill Gravel, and their general distribution over Haldon, suggest that they once formed a continuous sheet overlying the Buller's Hill Gravel (Hamblin, 1973a). They may have been laid down in a coastal lagoon or as river flood-plain deposits: Conversely Green (1974) thought they might have been the products of a tributary to the stream that deposited the gravels; however, the masses of clay and the enclosing gravels have no connecting sedimentary structures, their clay mineralogies are markedly distinct, and the clay bodies lack the granitic heavy mineral suite of the gravels; hence Hamblin (1973a) concluded that the clay bodies and the gravels owed their present relationships to periglacial cryoturbation and solifluction, although the basal sand bed of the clay deposit was probably an original sedimentological feature.

Sedimentation units of the Aller Gravel are generally lenticular in cross-section and bounded by curved to planar erosional surfaces. Marked lateral and vertical changes in grain size are common. In places, several channel forms may be seen resting on one another with erosional junctions; the channel sediments are generally poorly sorted and show local large-scale cross-bedding. Fine-grained deposits are usually restricted to small lenticles of silt or clay and sub-rounded clasts of clay set in gravel or sand (Plate 16). Edwards (1973) related these sedimentary features to deposition by braided rivers, which are characterised by numerous channels of coarse sand and gravel, with rare fine-grained sediment representing deposition in depressions or channels on the flood plain.

Pebbles of the Aller and Buller's Hill gravels came only from the west and suggest river transport from the higher ground of the aureole rocks. Hamblin (1973a) considered that the pebbles of the Buller's Hill Gravel were so abraded as to suggest that they had been recycled. Flints in the Aller and Buller's Hill gravels probably came from residual gravels such as the Tower Wood Gravel.

Palaeontology

The Haldon flints have long been known for the abundance of their derived Chalk fossils, dominantly brachiopods, echinoids and bivalves, superbly preserved as external and internal moulds (steinkerns). Such fossils are less common in the Aller Gravel.

Jukes-Browne and Hill (1904) thought the fossils probably ranged from the Terebratulina Zone to the Marsupites Zone. Their conclusions were repeated by subsequent authors, notably Hancock (1969). Edmonds and others (1969), using preliminary results of the recent survey, suggested a range from Micraster cortestudinarium to Belemnitella mucronata, i.e. the whole of the Senonian. The zones represented in the Haldon Gravels can now be stated with certainty to range from the Micraster cortestudinarium Zone to the upper third of the restricted Gonioteuthis quadrata Zone. Outside this range there is limited evidence for some representation from the Holaster (Sternotaxis) planus Zone at the base of the Upper Chalk, and from the basal part of the broad Belemnitella mucronata Zone. There is no evidence of derivation from the higher part of the mucronata Zone and the Maastrichtian. The apparent absence of fossils of Middle Chalk provenance may be due to preservation failure rather than to non-deposition of sediments of this age over Cornubia, since flint is poorly developed at this level in the Devon outcrops.

The inferred zonal stratigraphy of the parent rocks has been determined largely on the basis of echinoids and brachiopods, supplemented by inoceramid bivalves. Echinocorys and Micraster are particularly useful as indicators of horizon in the Chalk but, except in the case of certain stratigraphically restricted morphological extremes which have received varietal or subspecific names and which are readily recognisable as steinkerns, the bulk of the Haldon examples are generalised innominate forms which provide only limited information. Other fossils, however, have a much more specific biostratigraphical significance and can be used with confidence in building up a stratigraphical framework. The postulated succession from which the Haldon fossils were derived closely matches the standard successions established by Brydone and Gaster in Hampshire and Sussex respectively, as well as successions in Kent.

Similar derived fossils ranging from the H. planus Zone to the B. mucronata Zone occur in the flint gravels of Orleigh Court in north Devon (Rogers and Simpson, 1937), indicating an extensive Upper Cretaceous sea in Senonian times and broadly uniform conditions over the whole of southern England. The topmost Upper Chalk (i.e. Upper Campanian and Maastrichtian) may have been present in

Devon, although no evidence has been found; it occurs in the English Channel (Curry and others, 1971), and has contributed fossils to high-level gravels around Weymouth and in the southern part of the Isle of Wight. The derived fossils of the Haldon Gravels, probably mainly from the Tower Wood Gravel, are listed in Appendix 4, pp. 181–182.

Chronology and correlation

Tower Wood and Buller's Hill gravels

Fossils preserved in the flints of the Tower Wood and Buller's Hill gravels indicate an earliest possible age of Turonian. The upper age limit may be taken at the initiation of the down-warping of the Bovey Basin, from which time the gravels on Haldon would have occupied an upland area, and the lower part of the Bovey Formation is probably Eocene. Fresh Dartmoor detrital minerals in Eocene deposits east of the present district (Groves, 1931) point to transport from the west of Reading Beds and Bagshot Beds; the Buller's Hill Gravel was probably laid down at this time.

Waters (1960a, b) has suggested that the Buller's Hill Gravel was deposited on the Chalk during the Eocene, and that the Chalk dissolved beneath it after mid-Tertiary folding to produce the Tower Wood Gravel. However, this would probably have resulted in mixing of the Tower Wood and Buller's Hill gravels, which has not happened. The clays of the former were probably introduced after the start of the solution of the Chalk, but before deposition of the Buller's Hill Gravel, which has a different clay mineralogy (Hamblin, 1973b). The Tower Wood Gravel is considered to be post-Senonian and pre-Reading Beds and the solution of possibly 200 m of chalk may well have occupied most of this period.

Hamblin (1973a) has argued that there was little or no erosion of the Buller's Hill Gravel before deposition of the associated clay, that no major earth movement had occurred and the clay was probably deposited before the subsidence of the Bovey Basin, and that the clay is Eocene, probably younger than the Bagshot Beds, and possibly as young as the Bracklesham Beds or Barton Beds.

Aller Gravel

Edwards (1973) correlated the Aller Gravel with the Buller's Hill Gravel because they both rested on Upper Greensand and were only 2 km apart, apparently forming the remnants of a single sheet which now dips westward as a result of subsidence associated with formation of the Bovey Basin; also the two deposits were lithologically similar fluviatile deposits, the lack of sedimentary structures in the Buller's Hill Gravel being explained as due to periglacial modification. He later (1976) suggested that, in view of the probable Eocene age of the lower part of the Bovey Formation, deposition of the Aller Gravel might have been in progress at the time of initiation of the Bovey Basin, and might have been progressively restricted to the axis of the newly forming basin, where the gravels continued to accumulate after deposition had ceased on the Haldon Hills and in the marginal areas of the Bovey Basin.

Hamblin (1974), however, considered that the Buller's Hill Gravel did not rest directly on Upper Greensand but on Tower Wood Gravel, and that the Aller Gravel contained a greater variety of exotic pebbles (in particular Greensand chert), much more sand, silt and clay, and exhibited sedimentary structures. He suggested that the Aller Gravel, though Eocene, was entirely younger than the Buller's Hill Gravel and its overlying clay sheet, and probably formed at the very start of the downfolding of the Bovey Basin.

Details

Tower Wood Gravel

Tower Wood Quarry [SX 8768 8567] shows 6 m of gravel with unabraded flints up to 0.3 m in diameter in a matrix of clay with very little sand (Plate 15). The flints are peripherally frost-shattered, leaving a core surrounded by small chips. The large flint cores are generally horizontally aligned. The uppermost 5 m of gravel are white; slight iron staining occurs below. A description by Pickard (1949) suggests a basal bed of gravel with quartz and schorl pebbles, which is now obscured.

A large overgrown quarry [SX 8764 8542], south of Whitepath Clump, shows 2 m of pale brown gravel similar to that at Tower Wood Quarry. The gravel rests on Upper Greensand and may have moved slightly by solifluction. At Haldon Plantation, a landslip in the bank of a large overgrown quarry [SX 8830 8543] shows 2 m of pale brown unabraded gravel on 3 m of brown gravel with abraded flints present at its base; these strata have been inverted by solifluction.

Buller's Hill Quarry [SX 8821 8465] shows both the Eocene flint gravels. The Tower Wood Gravel rests on Upper Greensand and comprises shattered unabraded flints up to 0.3 m in diameter in white clay, with no exotic clasts or abraded flints. In the north-west face, overlying pale brown gravel with scattered abraded flints probably represents soliflucted material. Elsewhere in the quarry, Buller's Hill Gravel rests on up to 7 m of Tower Wood Gravel, and periglacial action has effected some mixing.

Between the crossroads at [SX 8973 8378] and Telegraph Hill, Tower Wood Gravel is exposed in many roadside ditches and earth banks. During route investigations along the A38 road, four boreholes around [SX 9023 8398] penetrated typical Tower Wood Gravel with white unabraded flints in a matrix of white or pale brown clay; thicknesses of 3.6 m to 6.1 m were recorded.

Roadworks at Telegraph Hill exposed Tower Wood Gravel resting on Upper Greensand. One temporary section [SX 9114 8335] showed pale grey gravel with some brown mottling and no abraded flints. Quartz and schorl pebbles in a clayey sand matrix were present at the base of the gravel. Frost heaving had placed bodies of gravel within the Greensand and folded the junction between gravel and Greensand. A borehole [SX 9118 8333] at Haldon Chalet proved 2.4 m of Tower Wood Gravel resting on Upper Greensand.

Tower Wood Gravel 1 m thick was exposed in a quarry [SX 9115 8348] north of Haldon Chalet. A temporary excavation [SX 9105 8310] showed 5 m of pale grey and brown gravel comprising shattered, unabraded flints in a matrix of clay with little sand; brown clay rested on the uneven surface of the gravel. Excavations for the A380 road between The Thorns and Ashcombe Cross [SX 9025 7964] showed sections in abraded and unabraded flint gravels, the latter being more widespread; some mixing of Tower Wood and Buller's Hill gravels had been effected by solifluction.

Buller's Hill Gravel

Great Haldon

A seismic survey at Haldon Plantation indicated 21 m of gravels, probably mostly Buller's Hill Gravel, in a fault trough. Near the site of the old fire station at Buller's Hill [SX 8850 8483] a deep gully at the confluence between two streams shows pale brown gravels with flint, quartz and schorl pebbles in a patchy clayey sand matrix. A structureless mass of pale grey clay within the gravels is mottled brown near its margins where it contains a few flints and pebbles. The east end of the clay body is vertical. The promontory between the streams is entirely of clay. On the south side of the southern stream abraded flints in a sandy matrix, all possibly reworked, are overlain by 0.6 m of closely packed abraded gravels capped by sand. At the eastern end of this section lies a small patch of clay similar to that of the larger clay body.

Buller's Hill Quarry [SX 8821 8465] shows Buller's Hill Gravel on Tower Wood Gravel, with some periglacial mixing. The former is pale greyish brown with frost-shattered chatter-marked flints and exotic pebbles in a matrix of sandy clay which comprises about equal amounts of ordered and disordered kaolinite with a little illite. In the east face about 3 m of Buller's Hill Gravel rest on 1 m of Tower Wood Gravel; the junction has been folded and disturbed by periglacial forces which have moved large tabular unabraded flints upwards to stand vertically in the Buller's Hill Gravel.

Five boreholes drilled along the west side of Haldon Race Course in connection with A380 roadworks proved a maximum of 5.8 m of Buller's Hill Gravel [SX 8977 8379]. A borehole about [SX 897 835] at the race course is recorded as having shown 11 m of gravels resting on Greensand; this is the second greatest thickness of Eocene strata recorded on the Haldon Hills.

Buller's Hill Gravel was extensively exposed near Kenton Hill [SX 9060 8177] to [SX 9085 8200] during widening of the A380 road. A temporary exposure [SX 906 817] showed 3 m of pale brown gravel comprising abraded flints and exotic pebbles in a matrix of sandy clay; a boulder of schorl 0.25 m across proved to be the largest exotic seen on Haldon. Three bodies of pale grey structureless clay were present within the gravels, one [SX 9058 8170] measuring 12 m X 3 m.

An overgrown pit [SX 9170 8036] south of Zigzag Wood shows Tower Wood Gravel and Buller's Hill Gravel, the latter comprising poorly rounded flints and pebbles of quartz, schorl and aureole-rock in brown sandy clay. Other pits, as at [SX 9208 8075], and numerous forest tracks between Kenton Hill and Mamhead Bottom, expose fluvial flint gravels.

Little Haldon

A large shallow pit [SX 920 767] above Smallacombe Goyle exposes 3 m of flint gravel, probably soliflucted. The flints are closely packed and abraded. Some were shattered and dispersed before solifluction, others merely cracked in situ. All lie in clayey sand with flint chips.

A borehole [SX 9196 7529] at Teignmouth Golf Course proved 8.2 m of flint gravels on Greensand. Excavation [SX 9180 7515] near the clubhouse showed up to 3 m of Buller's Hill Gravel with large bodies of clay and small pods of sand. The gravels were uniform, mottled pale brown and grey, containing closely packed abraded flints up to 0.3 m in diameter and quartz, schorl and quartzite clasts in a sandy clay matrix. A few cobbles possibly of poorly abraded Greensand chert were recorded, but no abraded flints. The clay bodies comprised structureless pale grey clay with little silt and virtually no iron staining. The largest body measured 11 m X 8 m X 2 m thick and had a nearly horizontal top. The upper metre of this clay in the north face of the excavation contained four pods of gravel up to 375 mm across, possibly introduced by cryoturbation, and loose flints were present in the top 0.3 m. Rotten flints were noted around the margins of the clay bodies. Five lenses of brown horizontally bedded sand, each around 0.6 m long and 0.3 m thick, occurred in the gravels.

Aller Gravel

Sands Copse–Kingsteignton–Milber

Gravels are exposed in a shallow elongated pit [SX 8653 7595] to [SX 8685 7540]; exposures in the central part show up to 3.0 m of coarse flint gravel with some cross-bedded coarse sands. The gravels comprise flint, quartz and tourmaline rock, with some Greensand chert, Lower Carboniferous chert and Upper Carboniferous sandstone; white, rounded clay clasts are common. Poorly abraded cobbles of Greensand chert range up to 0.5 m in diameter. Finer-grained, more distinctly bedded flint gravel overlies the coarse gravels.

Pits dug to prove gravel reserves in the Sands Copse area showed the following thicknesses of Aller Gravel: 1.4 m resting on Upper Greensand [SX 8654 7594]; 4.0 m, 3.6 m and 3.4 m respectively without reaching the base [SX 8679 7573]; [SX 8669 7564]; [SX 8667 7563]; 4.3 m containing a 0.7-m clay body [SX 8680 7543]. Two boreholes showed 4.6 m [SX 8664 7586] and 6.1 m [SX 8674 7564] of Aller Gravel, but the gravel thicknesses may be exaggerated.

Shallow pits farther east [SX 8703 7583] to [SX 8697 7532] show Aller Gravel and gravelly Head. Two boreholes showed 3.8 m [SX 8699 7549] and 5.2 m [SX 8700 7518] of Aller Gravel without reaching the base. North-east of Higher Sandygate, a small disused pit [SX 8711 7538] exposes 1 m of coarse flint gravel with a matrix of yellowish green sand. To the south, immediately west of the A380 road, an overgrown pit [SX 8706 7524], 7.0 m deep, contains Aller Gravel. In its south-east corner 1.8 m of brown gravelly coarse sand, with clasts of abraded flint up to 0.2 m across, quartz, schorl, and sparse Greensand chert, dip at 5° west. The north-west face shows 4.6 m of pale greyish brown gravel, with blocks of flint and Greensand chert up to 0.3 m across, together with vein quartz, schorl, Upper Carboniferous sandstone, and dark grey hornfels. At the north end of the pit, the gravels contain two lenticles, one of white silty clay, the other flat-topped and of pale brown sand. Old workings for flint gravels occur in Stony Copse [SX 8712 7521]; [SX 8714 7503].

The Sandygate Borehole [SX 8672 7507] passed through 17.4 m of red-mottled clays and sands of the Bovey Formation into flint gravels. Roadworks [SX 8704 7489] in the Colley Brook area showed up to 5 m of coarse gravel with clasts up to 0.3 m across, dominantly of flint, with Greensand chert, vein quartz and schorl, in a matrix of reddish brown coarse angular flinty gravelly clayey sand. Some fine gravel occurred, with clasts of quartz in a matrix of reddish brown coarse clayey sand, and also some pale greyish green soft medium to coarse angular sand, and pale brown cross-bedded very coarse angular sand. Boreholes proved the following thicknesses of Aller Gravel without reaching its base: 4.6 m [SX 8704 7492], 9.0 m [SX 8705 7491], 4.9 m [SX 8705 7490], 9.4 m [SX 8707 7494], 7.6 m [SX 8708 7492] and 6.4 m [SX 8706 7485].

North-east of Kingsteignton, gravels and Upper Greensand occupy pockets and fissures in Devonian limestone. One fissure was proved in a borehole [SX 8782 7373] to be 26.8 m deep.

Roadworks along the A380 bypass between Rydon Quarry [SX 874 741] and St Oswalds [SX 879 735] showed up to 4 m of Aller Gravel resting patchily on Upper Greensand, and occupying fissures in Chercombe Bridge Limestone. Some 2.4 m of flint gravels and sands [SX 8787 7409] enclosed a 0.6-m body of white sand. South of Humber Lane 4 m of Aller Gravel rest on Upper Greensand [SX 8787 7370]. At the north end of the limestone quarry [SX 8845 7415] at Bickleyball, mixtures of Upper Greensand and Aller Gravel fill a deep fissure in the Devonian limestone.

Aller Gravel on Upper Greensand at Wildwoods Copse was proved by a borehole [SX 8754 7177] to be 2.9 m thick.

Roadworks boreholes in the Penninn area proved: 1.2 m of soil and reddish brown clay on 6.1 m of Aller Gravel, on Upper Greensand [SX 8717 7073]; 0.9 m of soil and clayey Head on 7.6 m of Aller Gravel [SX 8716 7069]; 0.9 m of soil and sandy Head on 5.2 m of Aller Gravel [SX 8720 7066]; 9.6 m of Aller Gravel on Upper Greensand [SX 8715 7063]; 1.7 m of soil and flinty Head on 4.4 m of Aller Gravel [SX 8739 7144].

Ussher noted overgrown workings [SX 8760 7083] for flint gravels at Buckland. A small disused pit [SX 8736 7072] near Penninn shows the following section in Aller Gravel:

Wolborough–Kingskerswell

On the south slopes of Wolborough Hill, coarse sands and gravels with lenses of white clay and a little flint dip at 25° south [SX 8606 7056]. Old pits [SX 8535 7034] 130 m west-south-west of Wolborough church show 2.4 m of flint gravels resting on or faulted against Upper Devonian slates; Woodward and others (1900, p. 440) recorded a 9-m-high face of gravels and sands dipping at 35° to 45° south-east. At the southern edge of Decoy Brake, disused pits [SX 8596 6952] expose 3.0 to 4.6 m of flint gravels dipping at 12° to 15° north-north-east.

The Royal Aller Vale Quarry [SX 877 693]–[SX 876 696]–[SX 879 695]–[SX 879 691] and the Zigzag Quarry [SX 8786 6898] to [SX 8795 6910] expose about 25 m of abraded flint gravels and sands with subordinate silts and clays (Plate 16). The sediments, which show marked lateral and vertical variation, are dominantly boulder to pebble gravel and gravelly sand to coarse sand. Medium- and fine-grained sands are uncommon, and silts and clays rare. Erosional surfaces, channel structures and cross-bedding are present. Many beds, especially fine-grained ones, are lenticular with sharp erosional junctions. Elsewhere lateral changes in lithology occur gradationally. Chalk flints abound, showing a moderate degree of abrasion; vein quartz, tourmaline and schorl rock are also abundant. Greensand chert is fairly common. Other constituents include Palaeozoic rocks, especially Lower Carboniferous chert, Upper Carboniferous sandstone, dark grey hornfels, and a few pebbles of tuff and dark grey igneous rock. The gravels in both quarries dip at 4° to 10° west. A trial pit [SX 8810 6936] to the east of the main workings at Royal Aller Vale Quarry proved 9.1 m of Alter Gravel without reaching its base.

An overgrown pit [SX 8908 6934] at Connybear Brake shows 3.0 m of brown gravel with 0.2-m abraded flints together with vein quartz and Palaeozoic rocks. The depth of the pit suggests the presence of at least 5 m of gravel. Disused gravel workings [SX 8905 6920] occur to the west-south-west.

Disused pits [SX 8813 6862] in Harpins Brake expose poor sections in abraded flint gravel; Ussher recorded westerly dips here. Farther south-south-east a small disused pit [SX 8853 6808] shows 1.5 m of gravels with abraded flints and horizons of coarse sand.

Chapter 10 Palaeogene: Bovey Formation

General account

The Bovey Formation, which underlies the low ground (generally below 61 m above OD) between Bovey Tracey and Newton Abbot, comprises Tertiary kaolinitic clays, sandy and silty clays, silts, lignites and sands. The sediments occupy a partially fault-bounded trough termed the Bovey Basin (Figure 25), which comprises a main basin, extending 11 km north-west from Newton Abbot, together with the smaller Decoy Basin south of Newton Abbot. The Bovey Basin, in common with the Petrockstow Basin in central Devon and the offshore Stanley Bank Basin east of Lundy Island, lies on the line of the Sticklepath–Lustleigh Fault Zone and owes its origin to subsidence within this zone (p. 146).

Fasham (1971) estimated that the maximum thickness of the sequence was 1245 m, and Vincent (1974) 1066m (p. 159), but only the top 300 m or so are known. The formation is underlain along its eastern outcrop and south of Newton Abbot by the Aller Gravel and the Upper Greensand; elsewhere it is faulted against or rests on Devonian, Carboniferous or Permian rocks. Brunsden and others (1976) recorded Bovey Formation clays and lignites preserved in solution pipes in cuttings for the Kingsteignton–Newton Abbot bypass [SX 874 743] to [SX 879 734], confirming that the original extent of the formation was significantly greater than the surviving outcrop. Along the eastern part of the main basin a considerable amount of data is available from borehole records and workings for ball clay, for the Bovey Basin is the main source of ball clay in the United Kingdom. The term 'ball clay' derives either from an early method of working in which the clay was cut into approximately, 0.25 m cubes or 'balls' weighing about 15–17 kg, or from the implement, known as a tubal, used to dig the clay. In modern usage it refers to a fine-grained highly plastic kaolinitic sedimentary clay, the higher grades of which fire to a white or near-white colour in an oxidising atmosphere (Highley, 1975). The clays are worked in open pits and inclined adits between Chudleigh Knighton and Newton Abbot, and at Stover Park [SX 844 741] and Ringslade–Mainbow [SX 844 728]. Siliceous clays and sands, from the upper part of which ball clays are obtained, are succeeded by lignites and brown clays including high quality ball clays, which are in turn overlain by sands with clays. Subordinate clay facies at Chudleigh Knighton and Ringslade, on the northern and southern margins of the main basin, also yield ball clays. Gravelly deposits at Staplehill [SX 819 738] are tentatively included in the Bovey Formation, as are siliceous pebbly and conglomeratic sandstones north-west of Bovey Tracey.

Subdivisions shown on the published map are given below, to the left of a revised classification which draws largely on the work of Edwards (1970b, 1976) and Vincent (1974):

Middle Bovey Formation

Staplehill Gravel Member

The Staplehill Gravel, which crops out along the western part of the southern margin of the main basin between Blackpool Wood [SX 813 740] and [SX 832 737] near Singmore Farm, comprises pale grey gravels and sands with clasts of quartz, Carboniferous chert, dark grey igneous rock, sandstone, Greensand chert, and rare flint.

Reid (in Ussher, 1913, p. 115) noted "at Staplehill excellent sections of the marginal deposits [of the old lake], banked against a steep slope of slate and dipping at an angle of 50° to the north. A large sand-pit on the south side of the high-road shows a basement-bed of coarse gravel, consisting largely of quartz, grit, chert and igneous rocks, with which occur masses of Greensand-chert, some Chalk-flints and a few pieces of Devonian limestone. This gravel passes under 30 feet or more of coarse granitic sand, which in the road-cutting seems to dip northward under white pipe-clay with nearly vertical bedding." At Ringslade Claypit Staplehill Gravel on Aller Gravel dips north beneath Ringslade Clay; this suggests possible correlation with part of the Lappathorn Member, which rests on Aller Gravel near Sandygate.

Lappathorn Member and Abbrook Clay-and-Sand Member

Along the eastern side of the Bovey Basin the lowest members of the succession are the Lappathorn Member and the Abbrook Clay-and-Sand Member. The former is about 80 to 100 m thick (Vincent, 1974); it is characterised by 'pinks' — siliceous clays, sands and silts, mottled red, pink and brown by ferric oxide, and containing in places small ironstone spherules. The Abbrook Clay-and-Sand comprises 40 to 50 m of non-carbonaceous high-silica clays and sands, generally grey but with subordinate brown clays; thin clay breccias occur locally; silt-free and sand-free clays are rare, as are lignites. The junction between the two members is exposed in White Pit [SX 859 755]. Good sections are scarce but a borehole [SX 8464 7606] near Twinyeo Farm (Figure 28) passed through the Southacre Clay-and-Lignite into the top 44 m of the Abbrook Clay-and-Sand. The latter was about 70 per cent grey plastic sandy and silty clays and 30 per cent sands; some pale brown clays and carbonaceous clays were present. The rarity of lignites was illustrated in another borehole [SX 8472 7615], in which 37 m of beds contained only a single 0.3-m lignite horizon and 7.6 m of fawn and carbonaceous clays.

The Lappathorn Member rests in different places on the Aller Gravel, the Upper Greensand and Carboniferous rocks. The Sandygate Borehole [SX 8672 7507] passed from 'pinks', through red-mottled brown very fine-grained sands, into flint gravels which are considered to represent the Aller Gravel.

Southacre Clay-and-Lignite Member

The Southacre Clay-and-Lignite, named after the Southacre area of Preston Manor [SX 855 754], comprises 30 to 67 m of lignites and carbonaceous brown and black clays, with subordinate grey to greyish brown silty clays and rare sands (Plate 17). The carbonaceous clays are high-kaolin, low-silica, very white-firing types, which include some of the most valuable ball clay beds in the Bovey Basin; they are well exposed in several open pits between New Bridge Claypit [SX 847 765] and Pinsent Claypit [SX 861 730].

Lignites and brown clays at the disused Blue Waters Mine (previously known as the Bovey Coal Pit) [SX 811 769] in the north-west part of the main basin were thought when surveyed to be equivalent to those at Southacre. However, Vincent (1974) has shown that the Southacre Clay-and-Lignite passes laterally into the Chudleigh Knighton Clay (Figure 25), and that the clays and lignites in the Blue Waters Mine area are a younger sequence, the Brimley Member.

The most striking feature of the Southacre Member is the lignite (Plate 17); associated carbonaceous clays range from pale brown to dark chocolate brown or black and contain lignitic material either as fragments set in clay or in colloidal form. These clays are generally 'smooth', free from quartz grains of silt or sand size. A few beds of sand and silty grey clay occur. The 'Parks Seam', 6 m below the top of the member in the Preston Manor area, is up to 4.3 m thick and comprises pale grey clay, silty in places, with laminated parts which are locally brecciated. It forms a marker horizon characterised by well-ordered kaolinite, abundant siderite and low alkali content. Rootlets are common in the bed, and also occur in others lower in the sequence. Brecciation is seen in the Parks Seam and in other beds, for example in the upper part of the member in the Pinsent Claypit.

Chudleigh Knighton Clay-and-Sand Member

Clays, silts, sands and a .lignite horizon extending from Bradley [SX 829 776] to south-west of Chudleigh Knighton are shown on the published map as underlying the Southacre Clay-and-Lignite, but Vincent (1974) has demonstrated that they are lateral equivalents (Figure 25). Clay-pellet breccio-conglomerates, present throughout, consist of angular to sub-rounded clasts of grey clay up to 40 mm across in a matrix of similar clay; they are best seen on weathered surfaces, where the clasts protrude from the matrix (Edwards, 1976, plate 4A). A number of sandy units occupy channels cut in underlying clays.

Sands become commoner north-westwards, and in the Bradley area the clays remain only as a thin, southward-dipping unit cropping out north of the sandy facies. To the south-east the sequence is more lignitic, and in the Clay Lane area [SX 844 768] it includes a number of thin lignite beds associated with brown clays, which are intermediate in character between the Chudleigh Knighton Clay-and-Sand and the Southacre Clay-and-Lignite.

Ringslade Clay Member

Ball clays are worked along the southern margin of the main basin north of Highweek in the Ringslade Claypit [SX 846 726] and in the Mainbow Mine [SX 844 728], and the exposed sequence has been termed the Ringslade Clay (Edwards, 1976). The sediments comprise grey and brown smooth clays and silty clays, sands, and a lignitic horizon. Several rootlet horizons are present. Bristow and Hughes (1971) estimated that up to 150 m of the Bovey Formation lie below the main ball clay sequence in the Ringslade–Mainbow area, and the Blatchford Sand closely overlies it.

A variant of pale grey, commonly silty clay, mottled purple, red, yellow and grey, has been termed the Ringslade Mottled Clay; much of the mottling is due to fragments of Devonian slate incorporated at the time of deposition.

Boreholes north-east of the workings at Ringslade show that the Ringslade Clay passes laterally into the Southacre Clay-and-Lignite (Figure 26).

Upper Bovey Formation

Vincent (1974) divided the Blatchford Sand of the published map into seven members (p. 130) lying discordantly one upon another; most are of sand and clay, but the Twinyeo and Brimley members include substantial thicknesses of lignite. His Blatchford Sand Member comprises over 100 m of locally gravelly grey sands, with subordinate silty and sandy clays, restricted to the southern part of the main basin, and the term is used in this sense in the account that follows. The sands are mainly of angular to subangular quartz with some tourmaline and quartz-tourmaline; when washed they are similar to the sand isolated after china clay production, and they were probably derived from granite. In the south these sediments rest directly on the Southacre Clay-and-Lignite, but farther north sands and clays of the Stover Member intervene. At Stover Park 4.8 m of brown ball clays with a lignitic horizon occur in the Stover Member: one clay bed contains angular blocks of pale greyish fawn clay in a matrix of dark brown clay, but in other parts of the bed the paler clays have a wispy diffuse relationship to the darker clay (Edwards, 1976, plates 3 and 4B). Elsewhere in the bed complex breccias occur, with angular particles of clay, lignitic clay and lignite.

Woolley Grit Member

The Woolley Grit comprises up to 24 m of hard grey siliceous coarse-grained sandstones and conglomeratic sandstones that crop out in two small partially fault-bounded outliers north-west of Bovey Tracey. Exposures in the first of these, near Wolleigh House [SX 804 797], were considered by Groves (1929) to be Carboniferous or Permian, or of igneous origin. The second outlier, in Higher Knowle Wood [SX 793 809], was studied by Blyth and Shearman (1962), who noted cross-bedding in some crags [SX 7921 8103], together with poorly preserved coniferous woody fragments indicating an age between late Cretaceous and late Tertiary. Thin sections show poorly sorted aggregates of angular quartz and tourmaline grains, with some orthoclase, set in a fine-grained ?chalcedonic matrix.

There are no known deposits of the Woolley Grit associated with the sediments of the Bovey Basin. However, north of Chagford, a chalcedonic conglomerate comparable with the Woolley Grit (Edmonds and others, 1968, p. 150) is associated with silts and clays similar to those of the Bovey Formation and this deposit has been included in the Bovey Formation.

Palaeontology and age

The Bovey Formation has been variously referred, on palaeontological evidence, to the Eocene, the Oligocene and the Miocene (e.g. Heer, 1862; Gardner, 1879–1882; Reid and Reid, 1910; Reid and Chandler, 1926). Chandler (1957, 1964) has emphasised the long time range of many Tertiary plant species; she worked on material from the Blue Waters Mine, Heathfield Claypit (also known as Candy's Pit), and from "a pit at Kingsteignton", and tentatively concluded that all was of Oligocene, perhaps Middle Oligocene, age. The presence of Microdiptera parva and Brasenia ovula indicates that the Bovey Formation is not older than the Highcliff Sands or Bournemouth Marine Series. Chandler (1964, pp. 78–79) recorded Potamogeton tenuicarpus, Rubus microspermus and Stratiotes websteri, and noted that they first appear in the Oligocene, the first two in the Hamstead Beds, the last in the Upper Headon Beds. She considered that although the beds in the Blue Waters Mine were likely to be Oligocene, in view of the considerable thickness of Bovey Formation below, Eocene sediments might well be present at depth; similarly the formation might range into the Miocene. Dr G. C. Wilkinson (personal communication, 1979), studying pollen from a borehole [SX 839 758] near Heathfield which penetrated 185 m of upper Bovey Formation on 69 m of Southacre Clay-and-Lignite on 51 m of Abbrook Clay-and-Sand, recorded the Eocene indicators Anacolosidites and Pompeckjoidaepollenites from below 290 m; the beds above were lower and middle Oligocene.

Palaeoecology, sedimentation and derivation

Sedimentation in the Bovey Basin is now believed to have taken place mainly on river flood plains and to a lesser extent in lakes. There is evidence for shallow-water deposition with intervals of sub-aerial exposure, and subsidence and sedimentation must have kept pace for long periods.

Rootlet horizons, indicating in situ growth, have been observed in the Southacre Clay-and-Lignite, the Ringslade Clay and the Stover Member. Most of the material preserved in the lignites is, however, from land plants which grew outside the depositional area (Chandler, 1964). The bulk of the lignites in the Southacre Clay-and-Lignite are composed of Sequoia couttsiae, derived from Sequoia forest clothing the slopes around the basin, and ferns, such as Osmunda, are also present. Chandler (1957, p. 74) recorded "Caricoidea (Cyperaceae), Myrica, Microdiptera, and Lysimachia. True aquatics are represented by Stratiotes, Potamogeton and Brasenia. Climbing plants are represented by vines and Rubus. Trees and plants such as Nyssa, the Lauraceae, Symplocos, Carpinus, Magnolia and Meliosma probably overhung the water and dropped their fruits into it." Dr G. C. Wilkinson (personal communication, 1979) noted that the Bovey Formation (Abbrook Clay-and-Sand Member to Heathfield Member) was dominated by the fern spore Polpodiaceaesporites, which implied damp or humid conditions, the tricolporate pollen Tricolporopollenites, which was associated with trees, shrubs and herbs of broad affinity, and the conifer pollen Pityosporites, which was generally considered to represent the flora occurring at a higher altitude than the typical flood-plain flora and which might be dispersed over long distances. Representatives of the flood-plain flora were Nyssapollenites and Salixipollenites, which were similar to the pollen of modern plants that grow under moist, humid conditions. He reported that Polyvestibulopollenites (pollen similar to that of modern Alnus or alder) did not occur below the Stover Member. A warm and swampy environment as represented by the conifer pollen Inaperturopollenites (similar to that produced by modern Taxodium, the swamp cypress, and deposited near to the place of growth) was especially evident above the base of the Stover Member. The simple monocolpate morphology of Arecipites and Monocolpopollenites was considered to be characteristic of palm pollen which, together with other micro-floral evidence and comparison to distribution maps of modern floras, suggested that the climate at the time of deposition was relatively frost-free. The overall concept was one of a ubiquitous flood-plain flora dominated by shrubs, herbs, trees and ferns, in association with other wet-loving plants and a stable 'upland' coniferous vegetation. There was no significant botanical variation throughout the time of deposition of the Abbrook, Southacre, Twinyeo, Stover, Brimley and Heathfield members.

Clay breccias occur within the Abbrook, Southacre, Chudleigh Knighton and upper Bovey Formation sequences. Best and Fookes (1971) considered that brecciated silty clays from the Broadway Claypit probably resulted from desiccation of sediment laid down in temporary lakes on a flood plain. The clasts in clay-pellet breccioconglomerates, well seen at Chudleigh Knighton, are somewhat abraded and may have been transported for short distances. The mixing and brecciation of ball clays in the Stover Member at Stover Park (p. 133) may have been brought about by foundering and slumping, possibly attributable to fault movements contemporary with deposition (Best and Fookes, 1971, p. 231).

The sand units within the clays point to fluviatile conditions. Some, such as those in the Abbrook Clay-and-Sand and in the Chudleigh Knighton Clay-and-Sand, show erosional bases cut in underlying clays and are attributable to channelling by streams on a flood plain.

Blyth and Shearman (1962) considered the Woolley Grit to be a fluviatile facies of the Bovey Formation. The sediment was possibly deposited as a fluviatile gravel in a ravine. The outcrops lie close to a wrench fault extending northwestwards from Bovey Tracey, and presumably a Tertiary river followed lines associated with this fault.

The Bovey Formation was probably deposited in two main phases. Analogy with the lower part of the Petrockstow Basin sequence (Freshney, 1970) suggests that the concealed lower Bovey Formation may include gravels, sands, silts and very silty clays. The second phase began when the Abbrook Clay-and-Sand was deposited; faults (except for the western margin fault) became less active and the sediments spread out over a wide area of older rocks to the north and east of the main basin and into the Decoy Basin.

It has generally been considered that the sediments of the Bovey Basin were derived from decomposition of the

Dartmoor Granite into kaolinitic clays and quartz grains (e.g. Key, 1862, Ussher, 1913; Scott, 1929; Groves, 1931). However, Maw (1867) thought that the clays were the insoluble residue of chalk, and Jones (in discussion of Brackenbury, 1931, pp. 269–270) suggested a derivation from metamorphosed sedimentary rocks. Bristow (1968) argued for derivation of the Tertiary sediments of the southern part of the Petrockstow Basin from the weathering mantle developed on Carboniferous rocks during the lower Tertiary. He discounted a granite source on the grounds that the kaolinite formed by hydrothermal alteration of feldspar has a well-ordered lattice and relatively large crystals, whereas the kaolinite of ball clays usually has a disordered lattice and is extremely fine-grained; coarse-grained quartz sand with tourmaline such as is produced as a by-product of china clay working is lacking at Petrockstow, and most ball clays in the world are unrelated to kaolinised granite.

The Bovey Formation sediments were probably derived both from granite and from a weathering mantle developed on a range of other rocks. The upper Bovey Formation includes considerable thicknesses of coarse-grained quartz sands with tourmaline, pointing to a granite source, and associated clays have well-ordered kaolinite and are similar to china clays. The known strata below the upper Bovey Formation lack such sands and include ball clays with disordered kaolinite; they may have been largely derived from a weathering mantle developed on Palaeozoic and possible younger rocks, although it is probable that the granite contributed some material.

Mineralogy and chemistry

The description of the ball clays in this section is largely based on data supplied by Watts Blake Bearne and Co plc and E.C.C. Ball Clays Ltd.

Boswell (1923) recorded the following detrital minerals from the Bovey Formation, in approximate order of abundance: quartz, limonite, tourmaline, zircon, rutile, muscovite, ilmenite, anatase, cassiterite, kyanite, chiastolite and andalusite, glauconite, staurolite, orthoclase, topaz, brookite, monazite, hornblende, epidote and garnet. Groves (1931) noted that nearly every zircon from the Bovey Basin was of Dartmoor zoned type, and that practically the whole Dartmoor suite of minerals was represented. He recorded some yellow sphene, comparable to sphene from the Dartmoor Granite.

The ball clays are composed dominantly of kaolinite, micas and quartz. Carbonaceous material occurs either as discrete fragments of lignite coarser than 0.1 mm, or as a colloidal coating to clay and mica particles. Minor constituents include anatase, siderite and marcasite. The marcasite (known locally as 'mundic') occurs as nodules and in fibrous form replacing lignite; it is fairly common in some beds and is sporadic in the Preston Manor 'White Pit' [SX 860 757].

Kaolinite in the ball clays usually has a disordered lattice and is very fine grained. The plasticity of the clays, one of their most important commercial properties, is related to this fine grain size, to the presence of colloidal carbonaceous material, and possibly to the presence of other clay minerals such as montmorillonite (Highley, 1975). The early classification of the south Devon ball clays as 'Black', 'Dark Blue' or 'Light Blue' simply recorded their raw appearance. Following on the work of Holdridge (1956), the ball clays from the eastern part of the Bovey Basin were classified by Watts Blake Bearne and Co. into four groups distinguished by differing values of irreversible thermal expansion (Figure 27). These groups can be expressed in terms of kaolinite, mica and quartz content (Table 8). In (Figure 27) the mineralogical compositions of the ball clays are plotted on a ternary diagram and show separation into four fields. The thermal expansion curves for Group 1 clays are closest to the kaolinite field; the differing curves of Groups 2, 3 and 4 clays can be explained in terms of the greater mica and quartz contents. Mean chemical analyses of the four groups are shown in (Table 9).

The ball clays of the Bovey Formation are briefly described below according to their stratigraphical occurrences:

Abbrook Clay-and-Sand Member

Group 3 and Group 4 clays are present. The latter (the Siliceous Ball Clays) include the 'Stoneware Clays' of older classifications (Scott, 1929); they are non-carbonaceous clays with free quartz, stratigraphically the lowest clays worked in the eastern part of the basin, and they occur in the lower part of the member. Group 3 clays (Light Blue Ball Clays) are non-carbonaceous, with silica averaging 55 per cent and alkalis totalling 3 per cent; they occur within the upper part of the member.

Southacre-Clay-and-Lignite Member

Group 1 and Group 2 clays are present. Group 2 clays (Dark Blue Ball Clays) occur in the lower part of the member and have a carbon content of between 0.5 and 2.5 per cent. Silica averages 50 per cent and alkalis 2.6 per cent, reflecting the mica content. Group 1 clays (Extra White-firing Ball Clays) occur interbedded with lignites. They are low in silica, Fe203 and Ti02. The alkali content is the lowest of any of the groups, at around 1.7 per cent; the carbon content ranges from 0.2 to 4.0 per cent.

Ringslade Clay Member

Analyses 7 to 9 of (Table 10) are of clays from the Ringslade Claypit and analysis 6 is of a clay from Mainbow Mine. No. 6 is a non-carbonaceous clay with fairly high iron and titanium contents from just above the main lignitic horizon. Nos. 7 and 8 are so-called 'Blue Clays'. No. 9 is the Ringslade Mottled Clay, the high iron content of which is reflected in the reddish brown fired colour.

Upper Bovey Formation

Analyses 1 to 3 of (Table 10) refer to white low-carbon clays from the basal part of the upper Bovey Formation at the Broadway Claypit. Low figures for alkalis (1.64 to 1.78 per cent) reflect low mica contents. Two seams in the Stover Member are worked in the small open pit at Stover Park; one is carbonaceous clay with up to 6.6 per cent of carbon (analysis 4), the other is siliceous (analysis 5).

Details

Middle Bovey Formation

Staplehill Gravel Member

Small ponds [SX 8123 7394] to [SX 8135 7395] in the southern part of Blackpool Wood show poor exposures of white soft sand and clayey sand. Sections exposed in 1972 in a road cutting through the Staplehill Gravel at Blackpool Wood [SX 814 741] were recorded by Gouldstone (1975). The base of the cutting showed 50 to 200-mm-thick seams of grey kaolinitic clay, grey clayey silt and gravelly sand, together with very coarse-grained gravel containing fragments of black Palaeozoic chert, in association with fine-grained quartz-tourmaline gravel with feldspar. The sequence was orange-stained in places.

Between Blackpool [SX 813 740] and Lower Staplehill [SX 822 739] hummocky ground provides evidence of widespread old workings in gravels and sands. Woodward and others (1900, p. 426) noted a sandpit ?[SX 8220 7376] showing "highly-inclined beds of sand and gravel, dipping towards the Bovey valley at an angle of about 35°, and resting against Devonian Slates. The gravel yielded fragments of veined grit, chert, igneous rocks, and more rarely Greensand Chert and Chalk flints". North of Lower Staplehill, an old pit [SX 821 742], now filled in, is said by local residents to have exposed about 12 m of sands and gravels with thin clay beds.

At Ringslade Claypit [SX 846 726], 3 km east-south-east the main Staplehill Gravel outcrop, Devonian slates are overlain by 1 m of abraded flint gravel (Aller Gravel) beneath pale grey gravel with quartz, Carboniferous chert and igneous rocks, probably Staplehill Gravel. The gravels dip northward beneath Ringslade Clay.

Lappathorn Member and Abbrook Clay-and-Sand Member

New Bridge–Abbrook–Kingsteignton

Boreholes [SX 8493 7651] to [SX 8500 7683] in the Blue Marshes area east of the River Teign and north of New Bridge [SX 849 764] showed a south-west-dipping sequence of 46 m of white clays (commonly sandy), sands, clayey sands, and pink-mottled clays. Rare brown clays were present among the upper beds. Other boreholes [SX 8464 7606] to [SX 8507 7633] showed clays and lignites resting on Abbrook Clay-and-Sand, dipping at 10° south-west (Figure 28).

Ball clays of Group 3 (Tight Blue') are worked in the Rixeypark Mine (Watts Blake Bearne (WBB) No. 11 Adit) [SX 8507 7638]. Rixeypark Claypit [SX 852 762] exposes the 'Light Blue' sequence and overlying beds. The pit contains grey silty and sandy clays and smooth grey plastic clays. Grey sands and sideritic nodules occur in places. Two thin seams of dark brown plastic lignitic clay show brecciation, with angular pieces of pale grey and dark brown clay up to 10 mm across set in a dark brown clay matrix (Edwards, 1976, plate 1B). A borehole [SX 8534 7608] near Horsemills Copse passed through 3.9 m of grey and brown plastic clays on grey sandy clays and sands. Immediately south of the old Horsemills Copse Clay Works, boreholes [SX 8547 7581] to [SX 8551 7586] showed south-west-dipping grey sandy clays and sands with about 3 m of pale grey plastic clay, the 'Light Blue' (Group 3) clay.

Clays were formerly worked in open pits and adits at Horsemills Copse [SX 8537 7604] to [SX 8553 7592]; at present (1981) clays of Groups 2 and 3 are mined from an adit (WBB No. 5) [SX 8549 7587] to the south, WBB No. 6 Adit at the same site being closed. Boreholes [SX 8557 7593]; [SX 8551 7606] showed up to 24.4 m of grey clay, fine grey sand, and grey sandy clay. Boreholes [SX 8574 7595]; [SX 8568 7599]; [SX 8562 7599] in the eastern part of the Horsemills Copse area proved up to 22.9 m of grey clays with pink-mottled clays and yellow and grey sands (Lappathorn Member); the first and second boreholes entered mottled clays at outcrop, but the third passed through 13.7 m of grey clays into mottled clays. Sands overlying the pink-mottled clays and trending north-north-west between the second and third boreholes were proved in another borehole [SX 8570 7593] which showed 13.7 m of grey fine-grained sand with traces of lignitic material. Grey clays overlying the sands were formerly worked in the pits here; they are interbedded with grey sands and grey sandy clays.

The Preston Manor White Pit [SX 859 759] to [SX 859 753] contains grey sandy and silty clays, silts, sands, and grey plastic clays. A bed of fawn and lignitic clay dips at 17° south-west. Nodular marcasite is fairly common at some levels. The worked clays belong to Group 4 and are siliceous with much free quartz. Boreholes [SX 858 757] to [SX 858 758] sunk in an area which is now the northern extension of the open pit showed up to 42.7 m of grey clays and sands, the clays ranging from smooth to sandy. Some dark brown lignitic clay and fawn plastic clay are present. The most easterly of the boreholes [SX 8582 7584] passed into red- and yellow-mottled siliceous clays (Lappathorn Member).

A borehole [SX 8591 7589] showed, between depths of 5.5 m and 27.4 m, red-mottled clayey silt similar to that in the Sandygate Borehole (p. 129). Another [SX 8601 7590] showed 14.3 m of pink-stained pale grey sandy clay.

Between the eastern margin of the White Pit and John Acres Lane, shallow boreholes [SX 8606 7569] to [SX 8595 7581] showed grey silty and sandy clays, commonly with red mottling.

Old pits [SX 8607 7536] to [SX 8635 7480] are shown by boreholes to have worked grey plastic clays, commonly containing a good deal of sand, in a sequence of grey silty and sandy clays and sands. Boreholes [SX 8607 7507] to [SX 8623 7518] penetrated grey plastic clays with grey sandy clays which locally passed into 'lumping sand'.

The John Acres Lane Quarry [SX 862 751] exploits the same sequence as that worked in the Preston Manor White Pit. Shallow boreholes between Preston Mills [SX 8605 7483] and Higher Sandy-gate [SX 8671 7520] showed the Abbrook Clay-and-Sand to consist of grey sandy clays, plastic clays and sands, and a few brown clays, and the Lappathorn Member to comprise red- and pink-mottled grey sandy clays and sands, with grey sands and sandy clays. Boreholes west of Higher Sandygate reached a hard layer at depths of 26.5 m [SX 8662 7516] and 15.8 m [SX 8671 7520] beneath sands and siliceous pink-stained grey clays; this layer probably represents the junction of the Lappathorn Member and the Aller Gravel. The Sandygate Borehole [SX 8672 7507] showed: alluvial deposits 7.5 m, on red- and pink-mottled grey clays and silty clays passing down into red-mottled brown very fine-grained sand 7.0 m; at the base medium- to fine-grained greyish white quartz sand 2.7 m thick, rested on coarse-grained flint gravel, which was penetrated for 1 m and was probably the Aller Gravel.

In the Bowling Green area south-east of Preston Mills and north of the Ugbrooke Stream, a borehole [SX 8627 7464] showed 51.8 m of grey silty clays, clays and sands; the lowest 10.7 m included beds of red-mottled grey very silty clay. A pond [SX 8634 7441] west of Abbrook Farm is probably the site of a working noted by Scott (1929, p. 24).

A borehole [SX 8604 7422] west-north-west of New Cross passed through all the Southacre Clay-and-Lignite into 29.6 m of Abbrook Clay-and-Sand, which comprised grey silty clays and sandy clays and sands, with beds of brown and fawn clays and thin lignitic clays in places (Figure 29). In this borehole the boundary between the Southacre Clay-and-Lignite and the Abbrook Clay-and-Sand is drawn where light chocolate or fawn-brown clays and lignites pass downward fairly abruptly into grey sands or grey sandy and silty clays.

A disused flooded pit [SX 8636 7305] to [SX 8638 7270] was worked for Stoneware Clays; Scott (1929, p. 26) gave sections. A borehole [SX 8662 7301] proved slightly clayey, coarse- and fine-grained grey sand to 24.4 m; another [SX 8656 7286] showed 13.7 m of grey sandy clay, pink-mottled in places, resting on 18 m of grey and yellow sand. The disused flooded Zitherixon Claypit [SX 8652 7275] to [SX 8650 7243] was opened in 1874 and worked for Stoneware Clays. Boreholes [SX 8620 7263] to [SX 8656 7267] proved grey siliceous and plastic clays, brown clays and sands; pink-stained clays were encountered at the base of one borehole [SX 8633 7264] and in two others [SX 8644 7266]; [SX 8656 7267].

Newton Abbot–Aller

A borehole [SX 8629 7146] near the River Lemon in Newton Abbot showed alluvial deposits, 5.4 m, on presumed Abbrook Clay-and-Sand coarse-grained sand, 12.1 m, and brown clays, 2.5 m. Jukes-Browne (in Ms) mapped white clay in trenches at the junction of East Street and Devon Square [SX 8636 7109], and Ussher (in Ms) noted at least 36 ft (11 m) of clay at the railway station [SX 8678 7112].

Ball clays were formerly worked in the Main Quarry at Decoy [SX 865 703], now a lake and recreation area, where an eastward dipping sequence of grey and brown sands, pink-mottled grey clays and Stoneware Clays was exposed. Scott (1929) recorded that the Stoneware Clays at Decoy dipped at 14° to 24° east, occurred in lenticles, and were typically mottled and siliceous (66 to 80 per cent SiO₂).

A disused water-filled claypit (South Quarry) [SX 8689 6958] to [SX 8702 6932] shows 1.5 m of white clay in the eastern face; the shape of the workings indicates that the clays strike at 155°.

Southacre Clay-and-Lignite Member

New Bridge to Southacre

Scott (1929, pp. 27–29) recorded a pit and shafts in the 'Whiteware Group' of ball clays on both sides of Clay Lane, near New Bridge. Brown clays and lignites are exposed in a pit [SX 847 764] 200 m west of New Bridge. The sequence contains a 2-m-thick bed of grey silty clay with prominent rootlets comparable to the Parks Seam (see below) farther south. Another grey silty clay 4 m below is 0.6 m thick, contains small rootlets, and is brecciated. Several old shafts of New Bridge Mine, closed in 1956–57, were located [SX 8461 7635]; [SX 8468 7631]; [SX 8473 7624]; [SX 8469 7608].

A cross-section plotted from boreholes [SX 8464 7606] to [SX 8507 7633] (Figure 28) shows the Southacre Clay-and-Lignite resting on Abbrook Clay-and-Sand and dipping at 10°/225°. The Southacre Clay-and-Lignite Member comprises lignites, pale brown and dark chocolate-brown clays, and very carbonaceous clays. A borehole [SX 8464 7606] penetrated 2.1 m of sand at 21.3 m, and proved 67.1 m of Southacre Clay-and-Lignite, thus confirming the northward thickening implied by dips at Longmarsh and in the Twinyeo-Horsemills area. 'Dark Brakes' (Group 2 clays) are worked from the WBB No. 11 Adit [SX 8507 7638] at Horsemills Fields, which also works Group 3 clays.

The Longmarsh Claypit [SX 852 758], now disused (1981), lies about 0.5 km north-west of the Southacre Claypit and exposes lignites and brown clays dipping at 10°/240°. A section measured in 1969 showed 17.5 m of brown clays, lignitic clays and lignites, of which lignites totalled 7.3 m (42 per cent), overlain by 2.9 m of silty clays containing a conspicuous yellow-weathering pale grey clay, the Parks Seam, which is locally laminated and brecciated, and contains well preserved leaf material. It was from the bottom of Old Longmarsh Pit in 1957–58 that the first experimental adit mine (No. 1) [SX 8518 7598] was driven by Watts Blake Bearne and Co. plc (Rolt, 1974, p. 95). Clays of groups 2 and 3 are worked from an adit (Preston Manor No. 5) [SX 8549 7587], another adit (Preston Manor No. 6) at the same site now (1981) being closed.

The large open Southacre Claypit [SX 855 754] west of the Chudleigh–Kingsteignton road shows brown ball clays and lignites which are overlain by upper Bovey Formation at the south-west margin of the pit. Boreholes in the area show that the Southacre Clay-and-Lignite is some 60 m thick, although the base is not sharp.

Immediately south of the Preston Manor White Pit, Group 2 clays are extracted from WBB No. 4 Adit [SX 8596 7518]. The clays immediately above the main lignite were worked from WBB adits Nos. 1, 2 and 3, which are now disused.

Teigngrace–West and East Gold Marshes

The Teigngrace Borehole [SX 8492 7382], drilled in 1917–18, proved Southacre Clay-and-Lignite from a depth of 162.8 m to the bottom of the borehole at 203.3 m; the log is reproduced in Scott, 1929, pp. 10–13.

In 1981, work was in progress to remove upper Bovey Formation sands and clays overburden from the site of a new claypit [SX 859 744] at Denistone, planned to work Group I clays from the Southacre Clay-and-Lignite.

A cross-section (Figure 29) plotted from boreholes [SX 8585 7426] to [SX 8629 7411] shows 38 m of lignites and brown clays in about equal proportions. One borehole [SX 8589 7425] in which lignites made up 54 per cent of the sequence, penetrated two seams 4.9 and 6.1 m thick separated by a 0.8-m clay bed. A bed of pale grey sandy clay occurred immediately beneath the main lignites.

North of Gallows Cross, brown clays of the Southacre Clay-and-Lignite, overlain by upper Bovey Formation, are worked in the Broadway Claypit [SX 860 737] and two adits, Nos. 7 and 8; the beds dip at 11°/275°. A section measured in 1969 (Edwards, 1976, fig. 4) showed 14.1 m of lignites and lignitic brown clays overlain by clays and sands of the upper Bovey Formation. Lower Marsh Claypit [SX 861 734], south of Gallows Cross, shows 2.4 m of brown lignitic clays and lignites.

Boreholes between Ringslade and Teignbridge Crossing [SX 856 733] (Figure 26) reveal a synclinal structure; Southacre Clay-and Lignite underlies Blatchford Sand and passes west into the generally lignite-free Ringslade Clay.

Pinsent Claypit (N.A.C. [Newton Abbot Clays] Claypit) [SX 861 730] is bounded on the east by the River Teign and on the west by the disused Stover Canal. It exposes up to 27.4 m of brown clays and lignites (Plate 17) overlain by white and grey sands, sandy clays and silty clays in the lower part of the Blatchford Sand; the beds dip at 10° westward.

In the West Gold Marshes area, clays in the upper part of the Southacre Clay-and-Lignite are worked from the No. 10 Adit of Watts Blake Bearne and Co. plc. The main drive descends on a bearing of 330° from the mine head [SX 8577 7238] to pass beneath Sandford Orleigh Farm. This area has been worked in the past from vertical shafts but none is now in use.

Boreholes between West and East Gold Marshes show Blatchford Sand on 38 m of Southacre Clay-and-Lignite sediments. Towards the western margin of the basin, the latter pass into lignite-free mottled clays, grey clays, and sands (West Gold Mottled Clay), known only from boreholes, which are overridden by up to 10.4 m of Devonian slates.

Newton Abbot–Aller

At Decoy, brown and black clays with lignites rest on Abbrook Clay-and-Sand and dip eastward. Main Quarry [SX 865 703] is in Stoneware Clays (Scott, 1929, p. 31) and disused pits and shafts lie to the east. A borehole located as "Main Quarry No. 3, near Vale Road", proved 42.1 m of Blatchford Sand on 18.9 m of Southacre Clay-and-Lignite. Magazine Lane Quarry [SX 8665 7007] to [SX 8671 6990], was mapped by Jukes-Browne, the orientation suggesting a strike of 155°. The deepest shaft was 45.7 m; a 4.6-m of lignite was present beneath the ball clays. Strahan and others (1920, p. 19) recorded a section in the Decoy Shaft about [SX 867 698].

Beds transitional between the Southacre Clay-and-Lignite Member and the Chudleigh Knighton Clay-and.Sand Member

Brown clays with some lignites around Clay Lane, south-east of the Chudleigh Knighton Claypit, are transitional between the Southacre Clay-and-Lignite and the Chudleigh Knighton Clayand-Sand. They are extracted from the Clay Lane Claypit [SX 844 768]. Boreholes [SX 8437 7680] to [SX 8449 7690] showed seven lignites within 38.1 m of brown and fawn lignitic clays. These beds are underlain by at least 45.7 m of sands and sandy clays (?Abbrook Clay-and-Sand). Old clay workings [SX 8445 7690] to [SX 8470 7665] have been proved by boreholes to have been in brown plastic clays.

Chudleigh Knighton Clay-and-Sand Member

Near Lower Bradley Claypit [SX 8272 7775] to [SX 8278 7761], now water-filled, the spoil consists of greyish white coarse-grained clayey sand and dark brown smooth clay. A 29-m borehole [SX 8281 7777] east of the pit showed grey and brown sandy clays and sands, with gravels and blue clays (possibly Carboniferous rocks) below 25.0 m.

The claypit formerly known as Middle Field Quarry [SX 8285 7785] has been filled with spoil. A borehole [SX 8293 7773] to the south proved grey sand and clayey sand, 26.0 m, brown very sandy clay, 0.5 m, brown clays, 3.1 m, on lignite, 0.3 m. The brown clays were the seams worked and they dipped at 22°/190°. A disused claypit [SX 8297 7783] to [SX 8314 7777] shows only spoil of grey coarse-grained angular sand and grey to white sandy and silty clay; the axis of the pit trends west-north-west along the strike of the clay beds.

Boreholes on Knighton Heath [SX 8376 7727] to [SX 8396 7713] showed grey and greyish fawn clays, some plastic with traces of fine-grained sand and others very sandy. Sands occurred at intervals throughout the sequence, and brown clays with a thin lignite were penetrated at around 40 m. In one borehole [SX 8391 7717] the lignite bed was 0.3 m thick and lay at 40.5 m; it was overlain by 4.9 m of clay, generally free from sand, dark brown at the base and becoming fawn and grey upwards.

Underground mining in the Chudleigh Knighton area was discontinued in 1967. An open pit [SX 842 771] shows purplish brown smooth clays, grey silty clays, grey silts, and coarse-grained grey sands. Horizons of clay breccio-conglomerate within the sequence consist of angular to sub-rounded particles of grey clay up to 40 mm across set in a matrix of similar clay. Siderite-cemented sand (sandrock) is present at some levels. A 79-m borehole [SX 8409 7707] passed through clays, silty clays, sandy clays and sands, locally pink-stained in the lower parts; a lignite bed and associated brown and fawn clays between depths of 31.7 and 32.2 m mark the junction with the underlying Abbrook Clay-and-Sand Member. The beds dip south-west to pass beneath lignites and brown clays. A north-north-west-south-south-east fault which trends through the open pit to Twinyeo [SX 846 760] has a downthrow of about 25 m to the east-north-east.

Ringslade Clay Member

Mainbow Mine [SX 844 728] exploits ball clays that dip gently northward; the main clay sequence is underlain by lignitic clays and sands. Clays from the upper part of the Ringslade Clay are worked in the open Ringslade Claypit [SX 846 726] and are probably equivalent to those in the Mainbow Mine; they comprise purplish grey and pale grey smooth clays with patches of lignitic material, grey silty clays, and mottled clays with fragments of Devonian slate. Dips seen in the pit are unreliable owing to subsidence associated with old mine workings. In the eastern face of the pit, Bristow and Hughes (1971) noted apparently undisturbed Devonian slates resting on the ball clays. Boreholes in the area proved 39.6 m [SX 8405 7278], 31.1 m [SX 8412 7279] and 11.0 m [SX 8470 7256] of slates on clays. Bristow and Hughes also recorded that old mine workings south of the present pit had penetrated Devonian slates on vertical clay beds, and they considered that the slates had been thrust northwards over the Ringslade Clay (see p. 146).

North of the ball clay workings, boreholes [SX 8418 7300] to [SX 8423 7274] showed Blatchford Sand on Ringslade Clay. The latter comprised pale grey and fawn smooth and silty clays with scattered sands overlying a lignitic horizon; sand and sandrock in one of the boreholes [SX 8423 7272] probably represent the sequence beneath the lignitic beds.

Upper Bovey Formation (undifferentiated)

Bovey Tracey–Heathfield–Twinyeo–Southacre

El Oro boreholes Nos. 4 [SX 8069 7692], 3 [SX 8086 7693], and 1 [SX 8098 7688], drilled in 1947 by British Lignite Products Ltd, respectively proved 56.4 m, 76.2 m and 65.5 m of lignites and clays with scattered sands in the upper parts.

The Blue Waters Mine [SX 811 769], 1.6 km south-south-west of Bovey Tracey, shows in the western face poor sections in clays and lignites that are included in the Brimley Member of Vincent (1974). Pengelly (1862b) measured three sections, the thickest of which showed 38.1 m of clays and lignites with rare sands, and recorded a dip of 121°/215°; Heer (1862) identified the fossil plants collected (Ussher, 1913). Key (1862) recorded a dip of 9° to the south-east and Woodward (in MS) a dip of 5° to the south. Pengelly (1862b) noted that the lignite was also worked in a tunnel extending 190 fathoms (347 m) on a bearing of 295° from the western part of the pit.

Scott (1929, pp. 15–18) and Strahan and others (1920) recorded that Bovey Tracey Borehole No. 1 [SX 8111 7703], in the Blue Waters Mine, proved brown clays and lignites, 35 m, on grey sands and clays, 3 m; Bovey Tracey No. 2 [SX 8120 7666] showed grey clays and sands with a few brown clays, 29.9 m, on lignites and brown clays, 46.7 m, on grey sands and clays, 28.4 m. Devon Industries Borehole No. 3a [SX 8193 7667] proved sandy clays with lignite, 79.9 m, on lignites and lignitic clays, 60.9 m, on generally sandy clays, 25.9 m.

A borehole [SX 8264 7591] west of Heathfield penetrated 83.8 m of interbedded sands and sandy grey clays with subordinate brown clays and lignites. The disused Heathfield Claypit [SX 831 762] showed, when working (Scott, 1929, p. 32), 18.0 m of generally white clays with 0.9 m of lignite and lignitic clay; Woodward (in Jukes-Browne, 1909) noted a dip of 8° to west-south-west. The Heathfield Borehole [SX 8322 7613] (Ussher, 1913, pp. 111–112) was sunk in the eastern part of the pit and proved 59.4 m of grey sands, clays and sandy clays, with rare brown clay and two lignite seams, overlying 79.6 m of brown clays and lignites. The 'Heathfield Shaft' [SX 8365 7586], south-east of Heathfield Station, showed a similar succession.

In the southern part of Knighton Heath, boreholes proved lignites and clays resting on sediments possibly of the Chudleigh Knighton Clay-and-Sand; one [SX 8368 7688] showed superficial gravel, 8.2 m, on lignites and brown clays with rare grey clays, 14.9 m; another [SX 8377 7699] pale brown and brownish black lignitic clays and lignites, 18.6 m, on sands and sandy clays with scattered brown lignitic clays, 42.7 m; and a third [SX 8384 7686] brown clays and lignites with sporadic grey sandy clays and sands, 27.4 m, on grey clays, sandy clays and sands, 3.4 m. Between Jews Bridge [SX 8386 7639] and Twinyeo Farm [SX 8456 7608], shallow boreholes showed lignites, brown clays and grey clays, with scattered sands.

South of Longmarsh Claypit [SX 852 758] boreholes [SX 8504 7539] to [SX 8529 7565] showed grey sandy clays and sands, with some carbonaceous clays and lignites, resting on Southacre Clay-and-Lignite; one borehole [SX 8504 7539] proved brown clays and lignites between 7.6 m and 15.9 m depth, and the junction with Southacre Clay-and-Lignite at 45.7 m.

Boreholes in the Ventiford area about [SX 850 750] showed up to 53.3 m of upper Bovey Formation; one [SX 8498 7490] proved 7.3 m of alluvium on 40.6 m of sediments, comprising 57 per cent sands, 32 per cent grey sandy clays, 9 per cent brown sandy clays and grey slightly sandy clays, and 2 per cent brown clays and lignites. Immediately south of the Southacre Claypit boreholes [SX 8554 7514] to [SX 8557 7515] showed south-west-dipping upper Bovey Formation, comprising grey sandy clays, clays and sands, and fawn clays and sandy clays, resting on Southacre Clay-and-Lignite.

Halford–Stover Park–Teigngrace

Only the spoil tips remain of an abandoned working [SX 8105 7456] near Halford. Scott (1929, p. 18) noted that clays in disused pits "near Liverton" were of no use for pottery clays, but were occasionally used in the manufacture of saggars (cases of fireproof baked clay used to enclose pottery during firing). A borehole [SX 8161 7458] near Benedicts Bridge showed 3.3 m of drift on 31.1 m of grey sandy clays, clays and sands.

A small pit [SX 844 741] 0.5 km west-north-west of Teigngrace church contained unusual structures (see p. 133 and Edwards, 1976, plates 3 and 4B) and the following section in the Stover Member, measured in 1970:

The Teigngrace Borehole proved 162.8 m of grey and white sands, sandy clays and clays, with a few fawn and carbonaceous clays provisionally allocated to the Stover Member by Vincent (1974), resting on presumed Southacre Clay-and-Lignite.

Boreholes [SX 8585 7426] to [SX 8604 7422] proved up to 34.7 m of west-dipping grey sands, clayey sands, sandy clays and clays, with a few beds of fawn silty clay (Figure 29). One of the boreholes [SX 8589 7425] showed the sequence to comprise 65 per cent sands including clayey sands, 34 per cent grey sandy clays and slightly sandy clays, and 1 per cent brown clays.

North-west of Teign Bridge, a borehole [SX 8543 7379] showed 74.2 m of sand washed out during drilling, on 23.9 m of grey sandy clays and clayey sand with some carbonaceous brown clay, lignite and brown very silty clay, on 47.4 m of Southacre Clay-and-Lignite.

In the Broadway Claypit [SX 860 737], upper Bovey Formation overlying Southacre Clay-and-Lignite comprises white silty clays and coarse-grained to gravelly angular quartz sands which commonly have a clayey matrix (Edwards, 1976, fig. 4). Three beds of white clay occur just above the lignite which marks the top of the

Southacre Clay-and-Lignite; these and lower clays are worked in the pit. Best and Fookes (1971, p. 21) noted that a grey silty clay from this pit was composed of sub-rounded fragments of dark fine-grained clay 1 to 4 mm across set in a pale grey silty matrix sporadically streaked with limonite.

Ringslade–Decoy

In the area north-east of Ringslade, boreholes showed upper Bovey Formation resting on Southacre Clay-and-Lignite and dipping north-east. The greatest thickness of upper Bovey Formation proved was 137.2 m [SX 8515 7330]. A number of unlocated boreholes sunk near Forges Cross [SX 8430 7326] showed up to 57.6 m of yellow sands, grey sands, grey clayey sands and scattered grey clays resting on up to 12.8m of Southacre Clay-and-Lignite.

Between Ringslade, Blatchford and Teignbridge Crossing (Figure 26) up to 103.6 m of Blatchford Sand, comprising quartz sands with grey sandy and silty clays towards the base, occupies the core of a syncline.

The Blatchford Sand in the southern part of the main Bovey Basin has been shown by boreholes to be dominated by thick sequences of quartz sands. Pinsent Claypit [SX 861 730] shows the lowest beds, grey sands and sandy and silty clays, resting on Southacre Clay-and-Lignite.

At Decoy, south of Newton Abbot, a borehole located as "Main Quarry No. 3, near Vale Road" proved 42.1 m of sands, clayey sands and sandy clays on Southacre Clay-and-Lignite. The Decoy Shaft about [SX 867 698] passed through 17.7 m of presumed Blatchford Sand (Strahan and others, 1920, p. 19).

Woolley Grit Member

In Loxter Copse, 3.5 km north-west of Bovey Tracey, the Woolley Grit is well exposed as a line of crags up to 24 m high [SX 7921 8103], which Blyth and Shearman (1962) considered to be a fault scarp trending east-south-east–west-north-west. The rocks are grey hard coarse-grained sandstones and conglomeratic sandstones. Cross-bedding is present in the lower part of the exposure and there is some evidence of channel structures. The beds appear to dip at about 5° southward, but probable movements of joint-bounded blocks make this figure unreliable. Blocks of Woolley Grit are abundant in Loxter Copse, and the boulder train extends almost to Wrayland Barn [SX 7907 8148]. In Higher Knowle Wood the Woolley Grit caps the highest part of the ridge.

The southern occurrence of Woolley Grit extends as a ridge from Forder [SX 801 800] to a place close to Southbrook Farm [SX 809 790]. Surface blocks of grit are common.

Chapter 11 Post-Variscan structure

Permo-Triassic structure

The fold belts formed in south-west England during the Variscan orogeny were eroded during Permo-Triassic times to give a post-orogenic ('molasse') red-bed sequence which rests unconformably on folded Devonian and Carboniferous rocks. The Permo-Triassic formations west of the Exe Estuary generally dip at up to 30° to the east; those east of the estuary dip at 3° to 8° east. Thus younger formations crop out successively towards the east. This eastward dip is probably in part original, particularly in the breccia formations west of the Exe which formed as piedmont fans spreading downslope from an upland area in the west. Such fans typically show original dips of less than 5°, with angles of up to 10° confined to the fanheads (Blissenbach, 1954). Dips of between 10° and 30° are usual in the breccias of the present district, which are mid-fan and fan-head deposits. It seems clear that the Permo-Triassic rocks, together with overlying Jurassic rocks farther east, were tilted eastwards and bevelled by erosion before the extensive transgression of the Upper Greensand sea in Albian and Cenomanian times.

The base of the Permian in the vicinity of the Haldon Hills is generally a mature erosional surface of low relief; no evidence was found of a Permian ridge beneath the Haldon Hills as postulated by Laming (1966, 1968).

At the extreme northern end of Great Haldon the sub-Permian surface dips north-east, at an unknown angle and possibly steeply; dips within the Permian rocks suggest that this inclination is of tectonic origin. The surface farther south, to the southern end of Little Haldon, was considered by Hamblin (1969) to have been essentially horizontal before being block-faulted and taking on an easterly dip. However, a seismic survey (Henson, 1972) showed that the unconformity dips at least 12° east beneath the Haldon Hills and flattens eastwards (Figure 20).

Laming (1965) related folding of the New Red rocks to movements which affected strata up to Wealden age near Weymouth but not the Upper Greensand of the Haldon Hills. However, folding in the latter has since been recorded by Durrance and Hamblin (1969). Hamblin (1969, pp. 40–41 and fig. 1.2) postulated a monoclinal axis trending about 160° [SX 881 823] to [SX 902 752] to explain the horizontal nature of the base of the Permian in the west, its easterly dip beneath the Haldon Hills, and the gentler easterly dips of the New Red rocks farther east. He suggested that this monocline was of Permian age, resulting from the rising of the Dartmoor upland to the west and the sinking of the Permo-Triassic sedimentary basin to the east.

From Ashwell [SX 899 745] to Coombe Cottages [SX 916 740], the dip of the base of the Permian changes from about 5° west to 5° east. From the Bishop's Palace [SX 915 743] to Bishopsteignton the dip is 6° southward. Thus there appears to be a dome beneath the southern end of Little Haldon. A sand lens mapped within the Teignmouth Breccia at Higher Radway Farm [SX 906 744] follows the anticlinal structure in the base of the Permian.

Faults within the Permo-Triassic rocks are readily recognised at the coast, and in inland areas of contrasting lithologies. Elsewhere, as in the Teignmouth Breccia west of the Exe Estuary, faults are locally difficult to distinguish and the density of faulting is probably greater than shown on the map. Fault trends are not well marked. However, east of the Exe there is a northerly trending set, and over much of the area traces of an east–west set. A large fault along the line of the Exe Estuary was thought by Godwin-Austen (1842) and Murchison (1867) to have controlled the form of the lower Exe valley and to separate breccias to the west from sandstones and mudstones to the east, but Henson (1972) considered that there was no evidence for such a fault.

Cretaceous structure

Durrance and Hamblin (1969) concluded from seismic evidence that variations in the thickness of the Upper Greensand of Great Haldon were related to folding and faulting during deposition. The Permian–Cretaceous surface showed a number of basins elongated on east–west axes. Contours on the base of the Greensand also revealed a larger depression, centred about 600 m south-east of Harcombe [SX 889 813] and about 4 km across, which may be a large westward-plunging trough or part of a large basin slightly elongated on an east–west axis. Minor basins show a regular en échelon distribution within the large one. A west-south-west–east-north-east step which runs through the north side of Buller's Hill throws down the base of the Greensand by 10 m south and may represent a degraded fault line of Cretaceous age.

The folding of the Cretaceous rocks of Haldon post-dates the Greensand transgression and is pre-Senonian, since it affects the base of the Greensand but not the Eocene gravels derived from the in situ solution of Senonian chalk. The view of Durrance and Hamblin (1969) that the folding took place during deposition of the Greensand was confirmed during roadworks on the A38 at Woodlands in 1971; a cutting through one of the basins showed sixteen beds each thin at the margin and thickening towards the centre. Folding may have continued from late Albian to early Cenomanian times and possibly, since Greensand younger than any so far exposed may be preserved in the 'Harcombe basin', into mid-Cenomanian times. Intra-Cretaceous movements took place in south-east Devon at similar times (Smith, 1957, 1961a, 1965; Drummond, 1970; Hart, 1971).

The form of the sub-Greensand surface beneath Great Haldon was thought by Durrance and Hamblin (1969, p. 86) to have been produced by simple north–south compression and 'dimpling' of the Permian–Cretaceous cover more or less independent of the Devonian–Carboniferous basement. This mechanism is comparable to that suggested by Smith (1965) for folding of the Mesozoic cover in the Beer–Seaton district.

Tertiary structure

The south-western part of the district, between Lustleigh and Newton Abbot, is traversed by a major north-west–south-east trending dextral wrench fault system, the Lustleigh Fault, which forms part of the Sticklepath–Lustleigh Fault Zone. This is one of several such faults that cut the Cornubian peninsula and which may be Variscan structures rejuvenated in Tertiary times (Blyth, 1962; Dearman, 1963). The most recent major movements on the Sticklepath–Lustleigh line took place during the Palaeogene, and most activity ceased before deposition of the upper part of the Bovey Formation. However, minor activity continues and Dearman (in discussion of Blyth, 1962, p. 451) related an earth tremor in 1955 to movement within this zone.

The Sticklepath–Lustleigh Fault Zone is associated with deep, sediment-filled, Tertiary basins at Bovey and Petrockstow and east of Lundy Island. Reid (in Ussher, 1913) considered that the north-west–south-east faulting was post-Middle Oligocene in age and contemporaneous with the formation of the Bovey "lake-basin". He regarded the Bovey Basin as having originated in a rift valley bounded by parallel north-west–south-east faults.

The fault zone north-west of Bovey Tracey is delimited on its south-western side by the main Lustleigh Fault, which runs along the deep narrow valley of the River Bovey (Plate 18), entering the present district in the vicinity of Lustleigh. Immediately west of the district this fault may be seen to have displaced the granite boundary by about 1.3 km. It is presumed to continue south-eastwards along the axis of the basin.

The western margin of the main Bovey Basin between Lower Down [SX 790 785] and Penn Wood [SX 805 746] is a fault. Fasham (1971) calculated that the Palaeozoic–Tertiary contact dipped at 63° to the east and estimated a vertical throw of some 700 m. This boundary fault appears to be normal, with no evidence of wrench movement at its southern end. It probably operated continually during the sedimentation of the Bovey Formation and may be a splay of the Lustleigh Fault which continued to move after the parent fault had become inactive.

The fault zone is bounded on its north-eastern side by the Bovey Tracey Fault, which probably forms the eastern margin of the Woolley Grit near Forder and may have been active both during and after the deposition of the grit. This fault displaces the granite margin by about 1.9 km; it trends south-east through Bovey Tracey to disappear beneath the alluvium of the River Bovey and the Tertiary beds of the Bovey Basin.

Smaller faults within the main fault zone include one which runs south-east from Lustleigh along the valley of the River Bovey and shows a dextral displacement of about 0.36 km, and a minor sinistral wrench fault east of Lower Hisley. Two west-north-west–east-south-east fractures south of Lustleigh appear to be normal faults.

The Narracombe Thrust may have acted as a wrench fault in Tertiary times.

South of Newton Abbot, a north-west–south-east extension of the Lustleigh Fault is thought to separate Bovey Formation in the west from Aller Gravel to the east. This fault line may be projected through Torquay into Tor Bay, where Laming (1969) has established the presence of a north-west–south-east fault zone.

On the southern margin of the main Bovey Basin Devonian slates rest on Tertiary sediments in the Ringslade–Mainbow area and around [SX 857 724], north of Knowles Hill, Newton Abbot (pp. 142, 143). Bristow and Hughes (1971) suggested a thrust structure which represented rejuvenation of the Variscan thrust recognised at Bickington and Holne, and that movement on this Tertiary thrust was related to dextral wrench movement on the Lustleigh Fault. However the Bickington and Holne thrusts are two distinct structures well north of the Ringslade–Mainbow area and generally trending at a considerable angle to the southern margin of the main basin, and it is more likely that Bristow and Hughes's evidence points to local low-angle reverse faulting or to rotational slip on the steep-sided margin of the basin.

There is little evidence from the southern margin of the basin west of Mainbow of the relationships of Palaeozoic and Tertiary rocks. Reid (in Ussher, 1913, pp. 115–116) suggested that the Staplehill Gravel rested with a steep unconformable contact on Devonian slates, and steep to vertical dips indicate the possible presence of faults. Fasham (1971, p. 127) considered that the base of the Bovey Formation in the south-west part of the main basin dipped at about 40° down to 1200 m below OD.

The present outcrop pattern shows the Decoy Basin almost separated from the main basin. It is possible to imagine the southern margin of the main basin flanked by an east–west ridge of Devonian rocks in the Newton Abbot area which caused the proto-Teign river draining the main basin to turn eastwards. Periodically the river may have breached this ridge, flowing across the fault-bounded col of the Lustleigh Fault Zone and depositing sediments in the depression of the Decoy Basin. Continuity between the basins must have been well established in later Bovey Formation times, since the two sequences are apparently similar. Later, with the evolution of the present Teign drainage system, the hypothetical ridge may have been planed down to flood-plain level but it may still exist at depth.

The eastern and northern margins of the main basin are unfaulted, and the beds dip gently inwards. Fasham (1971, p. 127) suggested that in the north-east the contact between Bovey Formation and older rocks is inclined at about 10° near the surface but steepens to 35° below 600 m.

Downwarping associated with the Bovey Basin is reflected by the relationships of the Cretaceous and Tertiary rocks of the eastern margins with those of the Haldon Hills. Between Ugbrooke Park [SX 877 777] and the Teign Estuary the Upper Greensand rests mainly on Devonian and Carboniferous rocks and dips at 5° to 7° westward. Its base rises to some 120 m above OD east of Kiln Brake [SX 875 771], and that of the nearest Upper Greensand of the Haldon Hills is at about 190 m. There are similar disparities in level in the case of the Eocene flint gravels. Subsidence has given the Upper Greensand–Eocene strata of the basin a westerly dip, and erosion has removed their connection with the similar sequence of the Haldon Hills. South of Newton Abbot, the Upper Greensand and Aller Gravel of the Decoy Basin are also at a low level; they are disposed about a central outcrop of the Bovey Formation and dip inwards beneath it.

Minor folds of Tertiary age in the Haldon Hills have been correlated with the tectonic subsidence which formed the Bovey Basin (Hamblin, 1972). They include an east–west syncline considered to cross Little Haldon and the southern end of Great Haldon, and periclinal structures farther north on Great Haldon thought to represent the tightening of pre-Senonian folds known in the Upper Greensand.

Evidence from boreholes and resistivity surveys (Henson, 1971) shows that the base of the Budleigh Salterton Pebble Beds generally dips uniformly eastward but shows small ridges and troughs with east–west and north-west–southeast axes. These minor folds may reflect early Alpine movements.

The Tertiary beds of the Haldon Hills and Bovey Basin are affected by faulting. In the northern part of Great Haldon, two major faults show throws of around 20 m and at least four minor structures have been detected by seismic survey. Durrance and Hamblin (1969, p. 81) related these structures to local tectonic adjustment during formation of the Haldon Gravels in Eocene times. Over the remainder of Great Haldon, with only one exception, faults trend north-north-west–south-south-east or north-north-east–south-south-west. All are high-angle dip-slip faults, with throws of 5 to 11 m and no consistency in direction of throw. They may be tears brought about by compressive stress from north and south (Anderson, 1951); possibly they are rejuvenated Variscan structures. All the Eocene beds are faulted, and it is suggested that the faults form a system of accommodation fractures contemporary with the movements which formed the Bovey Basin.

The Bovey Formation sediments are cut by small-scale faults trending between east–west and north-north-west–south-south-east and with throws of up to 24.4 m. The trend swings from west-south-west–east-north-east in the southeast of the main basin to south-west–north-east and south-south-west–north-north-east farther north; this suggests a gravity fault pattern related to subsidence of the basin.

Chapter 12 Pleistocene and Recent

General account

Whether or not south-west England was overriden by Pleistocene ice is a matter of debate, but certainly much of the region lay for long periods within the periglacial zone adjacent to the ice sheets. Variable and extensive deposits of Head and Head Gravel formed by the downslope mass movement of water-logged material over frozen ground beneath. Windblown silt (loess) became largely mixed with the Head, but has been separately mapped on the Haldon Hills as "loam". Deposits in caves in Devonian limestone attest the presence of early man. The most complete sequence, dating from the Hoxnian interglacial, is that at Tornewton, near Torbryan. Terraces of river gravel at various levels represent stages in the denudation history of the area. Alluvium covers the present-day flood plains of the main streams, and minor alluvial fan deposits occur on the margins of Dartmoor. Muds, silts, sands and gravels deposited in the estuaries of the Teign and Exe rivers constitute marine and estuarine alluvium, and other marine-estuarine sediments occur along the coastline on beaches and tidal flats. A few areas of windblown sand are present on the coast, the largest at Dawlish Warren. Several small patches of sand and gravel between the Bovey Basin and the Haldon Hills are of uncertain age and origin.

Head

The term 'Head' was used by De la Beche (1839) to denote superficial deposits resting on raised beaches in Devon and Cornwall. It has since been widely employed to describe masses of locally derived rubble in clay and sand, moved downslope by solifluction under Pleistocene periglacial conditions. In this memoir Head is taken to include also rock debris originating by in situ weathering, and Recent colluvium, since it is not always possible to make distinctions. Thus a mantle of Head, generally only a metre or so thick and whose nature reflects that of the local rocks, obscures much of the underlying geology. For reasons of clarity the only such deposit shown on the published 1:50 000 map is the Head Gravel of the Haldon Hills.

Decomposition of the Dartmoor Granite has given rise to coarse yellow to brown sandy debris surrounding fragments and boulders of granite. In some places the deposits are more or less in situ, grading down into weathered granite. Locally, for example in the Caseley Court railway cutting [SX 7872 8213] and beside the road [SX 7968 8162] near Kelly, weathering of the granite has taken spheroidal form, as kaolinisation along joints isolated relatively unaltered rounded blocks. Such blocks may have given rise to some of the boulders which litter the granite surface. In places where Head on the granite has moved down slope by solifluction, it commonly grades laterally into alluvial deposits in the valley bottoms.

Considerable thicknesses of Head have accumulated on the steeper slopes on Devonian and Carboniferous rocks. Slate fragments in silty clay mask the slates, and variable stony deposits the aureole rocks, cherts, and sandstone-rich parts of the Crackington Formation. Of the Permo-Triassic formations, only the Budleigh Salterton Pebble Beds have given rise to significant Head deposits and the thickness of these is variable.

Waters (1966, p. 125) described gullies filled with Tertiary cobbles and frost-shattered chips and flakes of flint on the eastern flank of Great Haldon. This Head Gravel crops out in a belt immediately downslope of the Eocene gravels (Figure 24). Typically it comprises a structureless or poorly stratified mass of abraded flints, together with quartz, tourmaline, aureole rocks, and a variable but low proportion of unabraded flints, in a matrix of sand and clay. The Head Gravel was derived largely from the Buller's Hill Gravel and to a lesser extent from the Tower Wood Gravel, according to the relative availability of the two Eocene gravel facies in different areas. Mineralogically the matrix clay lies between the clays of the Buller's Hill Gravel and those of its included clay bodies, and Hamblin (1973a, p. 475) suggested that it originated by the mixing of these sources. He considered that some sorting of the Head Gravel had occurred during solifluction, and much clay had been washed out.

The Bovey Formation is generally overlain by up to 9 m of gravelly, sandy, and clayey superficial deposits of Pleistocene and Recent age, which have in the past been collectively termed Head (e.g. Pengelly, 1862b). They comprise Head (possibly locally incorporating some loess), colluvium, and fluviatile and estuarine deposits. Most of the material is probably of fluviatile origin. The superficial deposits of the Bovey Basin have been described by Gouldstone (1975).

A sheet of pale brown clay and loam up to 2 m thick ("loam" on the published 1:50 000 map) mantles gravels on the Haldon Hills. Its constitution varies slightly according to the proportion of clay in the gravel beneath. "Loam" overlying Tower Wood Gravel is largely composed of brown sticky clay, but where it rests on Buller's Hill Gravel it is a brown silty loam; both facies contain frost-shattered chips of flint. Comparable deposits occur also along the eastern margin of the Bovey Basin and south of Newton Abbot.

Harrod and others (1973) described similar thin silty drifts on the Devonian limestone plateau (where silty drift up to 1.5 m thick floors narrow dry valleys and thinner layers rest on interfluves), the east Devon plateau, the Budleigh Salterton Pebble Beds, the Dartmoor Granite, the Haldon Hills, and locally in the Bovey Basin. They noted that the silts were non-calcareous and mineralogically uniform, and suggested they had been deposited by winds from the east, and might correlate with the loess at Pegwell Bay, Kent (Pitcher and others, 1954). Locally derived sand and stones in the drifts were presumed to have been mixed with the aeolian silt by rainwash, frost or biological means.

Zeuner (1959) recorded phases of solifluction in northern France, followed by loess deposition and then by weathering, corresponding to tundra, steppe, and temperate forest. The sequence of Head Gravel followed by brown clay and loam (loess) on the Haldon Hills, may similarly represent tundra followed by steppe conditions.

Angular flint drift up to 0.3 m thick, comprising frost-shattered flint chips in a matrix of sand and forming a black soil, rests on Haldon Gravels and brown clay and loam alike. It may be of late Pleistocene age or younger. Gelifraction of the flints in the subjacent Haldon Gravels would have produced such a deposit, as would removal of clay and silt from the brown clay and loam. In the latter case, the weathering might have taken place in the third of Zeuner's climatic zones, the temperate forest; the Haldon Hills were probably covered by temperate forest until the Bronze Age, as shown by a brown earth profile commonly developed in the brown clay and loam (Mr B. Clayden, personal communication).

Caves and cave deposits

Two main groups of caves occur within the district, centred upon Torbryan and Chudleigh. In both areas the caves are developed in Chercombe Bridge Limestone. Other isolated caves scattered through the Devonian limestone outcrop (Cullingford, 1962) generally lack cave deposits. Pleistocene deposits with mammalian faunas are present in a number of the Torbryan and Chudleigh caves, and those at Tornewton Cave [SX 8172 6738] represent the most complete sequence of stratified beds of Pleistocene age yet discovered in any British cave. Sutcliffe (1969) has noted that Devon lay south of the fluctuating Upper Pleistocene ice and was characterised by overlapping cold and warm mammalian faunas. Northern mammal species (including woolly mammoth, woolly rhinoceros, and reindeer) spread southward during glacial stages; warm species (including straight-tusked elephant, narrow-nosed rhinoceros and hippopotamus) spread northward during the interglacials. The cave deposits at Tornewton and at Kent's Cavern, Torquay, preserve the evidence.

Palaeolithic and Mesolithic cultures in Devon have been discussed by Rosenfeld (1969). A Middle Palaeolithic site is known at Cow Cave [SX 8646 7867], near Chudleigh, while Upper Palaeolithic cultures, richly represented in Kent's Cavern, have been recorded at Three Holes Cave [SX 8155 6747], Torbryan, and at Tramp's Cave [SX 8667 7868], Chudleigh. Mesolithic cultures were more widespread in Devon, although the area remained thinly populated. Evidence at Three Holes Cave points to a late Mesolithic age (Rosenfeld, 1964).

River terraces

The major streams of the district are bordered locally by terraces of river gravel. Scattered terrace remnants occur along the Teign and Bovey rivers but the evidence is clearer in the Exe and Otter valleys, where five terrace levels have been distinguished. The denudation chronology of the River Exe has been studied by Kidson (1962). Henson (1971) correlated terraces east of the Exe with those of the Otter at the same level; the highest (5th) terrace differs from the lower ones in lacking cobbles of quartzite from the Budleigh Salterton Pebble Beds.

Alluvium

The main areas of alluvium in the district are along the Teign and Bovey rivers, especially in the area between Newton Abbot and Bovey Tracey where the flood plain is up to 0.6 km wide. Tributaries of the River Exe, and many other minor streams, are bordered by narrow alluvial strips. The nature of the alluvial sediments varies according to the local geology and the catchment areas of the streams.

Sand and gravel of unknown age

Several patches of sand and gravel occurring between the Bovey Basin and the Haldon Hills were shown on the previous edition of the Newton Abbot Sheet as Bovey Beds or as Valley Gravels (Ussher, 1913, p. 120). Some are now distinguished as Upper Greensand or Aller Gravel, but six patches remain as sand and gravel of unknown age.

Marine and estuarine alluvium

Extensive areas of estuarine alluvium are present along the rivers Teign and Exe, whose mouths are marked by spits at Teignmouth and Dawlish Warren.

Teign Estuary

The estuary of the River Teign maintains a fairly constant width of 0.5 to 0.75 km over much of its 7-km length. The following account and (Figure 30) are based largely on the work of Mr R. S. Nunny of Exeter University.

Towards the mouth of the estuary, 'The Salty' [SX 935 726] is an area of muddy gravelly sand extensively mantled by colonies of the mussel Mytilus edulis. Finer material brought down by the ebb tide is here mixed with coarse material from the beach and offshore area brought in by the flood tide. Coarse-grained marine sands are uncommon west of Shaldon bridge and largely confined to the channel floors where tidal flows are concentrated. The upper estuary is characterised by coarse sediments brought in by the river and found on channel floors, dark organic clays that settle out in still waters over the mudbanks, and gravelly deposits derived from the valley slopes that accumulate as narrow 'beaches' along the shores.

The main channel meandered during early sedimentation in the estuary, forming bands of coarse sand within the muds. During the past 100 years, however, it has been artificially confined by dredging.

The maximum extension of the estuary and marsh was apparently during the Bronze Age, as shown by a radiocarbon date of 3332 ± 70 B.P. from Pinsent Claypit (Radiocarbon, 1974) and by Bronze Age relics from Zitherixon Marshes (Pengelly, 1883).

Durrance (1971) established the presence of a buried channel whose base sloped from −7.8 m above OD at the head of the estuary to –20.5 m above OD at Teignmouth, an average gradient of 1 in 470. The channel is probably floored by Devonian slates for much of its length, although the Teignmouth Breccia underlies the Teignmouth foreshore.

Boreholes at the head of the estuary, drilled in connection with the A380 bypass road, passed through channel deposits into Devonian slates at between −6.6 and −7.8 m above OD. The channel was filled with medium to coarse gravel to about −5.0 m above OD, overlain by silt and fine sand capped by 0.6 to 1.3 m of modern clay. On the northern margin of the channel, traces of peat occurred within the silts, and a bed of peat was recorded between −4.4 and –3.5 m above OD.

The levels recognised in the buried channel correspond well with those in the younger (late Devensian) channels of the Exe (Durrance, 1969b). The absence of an earlier period of channelling in the Teign corresponding to the early Devensian channels of the Exe was considered by Durrance to be due to the less vigorous erosion of the Teign, the older channels possibly being confined to an area seaward of the present coastline.

The cyclical development of the Teignmouth spit was described by Spratt (1856) and Robinson (1975). In the first stage a hook of shingle and coarse sand extended 1 km from The Ness. It was then breached by a channel close to The Ness, and replaced by two banks called the Inner and Outer Poles which migrated to lie partly across the position of the former channel. During the third stage the hook from The Ness reformed, displacing the main channel to the east, and the Inner Pole Sand was driven on shore. This cycle of events has been repeated several times.

Exe Estuary

That part of the Exe Estuary which lies within the present district is 7 km long and ranges in width from 1.5 km in the north to 2.5 km in the south. The alluvium forms large sand banks with mega-ripples, and extensive mudflats with colonies of Mytilus. Gravel horizons occur locally, but the alluvium is generally soft and unconsolidated. Dr J. M. Thomas of Exeter University has drawn a map (Figure 31) showing the generalised distribution of modern sediments in the estuary. The sediment boundaries have been superimposed on a base map showing the shape of banks in the estuary in 1963, since when there have been minor changes.

Dr Thomas's data show that the muds and sandy muds are reduced black organic sediments, as are certain of the muddy sands. A thin reddish brown oxidised surface layer is generally present. Much of the mud is pelletal and probably composed largely of faecal material from bivalves, particularly from Mytilus. The sands of the lower estuary, fine to coarse but predominantly medium-grained, are probably marine and comparable to the beach sands of The Warren and Pole Sand. They grade into both muds and gravels. Gravels and gravelly sands are present towards the mouth of the estuary on the Great Bull Hill Bank and east of Dawlish Warren. Most of the clasts in the gravels are red-stained, fairly well rounded Palaeozoic fragments; the coarsest gravels, however, east of Dawlish Warren and in the eastern part of Great Bull Hill Bank, contain more flint and Greensand chert, and it is possible that they represent the top surface of the channel-fill sediments of late Pleistocene age recognised by Durrance (1969b).

Durrance (1969b) demonstrated the presence of buried channels of the Exe between Exmouth and Dawlish Warren. Early channelling, to a base over 49.6 m below OD, cut into New Red Sandstone and was followed by gravel infilling of these channels and then their partial re-excavation to a base over 27.6 m below OD. Terraces formed during stillstands within the early regression occur at depths of 22.6, 29.0, and 33.6 m below OD, and within the later regression at 3.4, 8.0, 11.3, 14.7, 19.6 and 24.7 m below OD. Both sets of channels trend north-west–south-east parallel to minor faults seen at Langstone Rock and Orcombe Point, suggesting the possibility of fault control. Channelling probably resulted from falling sea level during the early and late Devensian, indicating a middle Devensian age for the gravels. Sea levels must have stood at much below 49.6 m and 27.6 m below OD respectively during successive periods of channelling.

Dawlish Warren is a double spit 2 km long extending from the western side of the mouth of the estuary north-eastwards towards Exmouth, leaving a channel about 500 m wide (Figure 31). Changes in Dawlish Warren since the 18th century have led to anxiety that erosion might eventually result in the destruction of the spit (Peacock, 1869; Kidson, 1964), an important coastal defence and tourist attraction. The Warren formerly consisted of two belts of low sand hills, the Inner and Outer Warrens, separated by a tidal creek, the Greenland Lake. The Outer Warren has been subject to continuous erosion and has been largely destroyed; the Inner Warren is older, shorter and stable. About 1932, tidal flow into the Greenland Lake was prevented by retreat of the dunes of the Outer Warren, and the 'lake' became a shallow vegetated depression (Kidson, 1964).

The bulk of the sediment of Dawlish Warren has been derived from the cliffs of Permian sandstone to the west. Holme (1949) quoted data which indicated that Warren Point [SX 990 802], the northern part of Great Bull Hill Bank, and the northern section of Pole Sand were gravel, while the southern part of Great Bull Hill Bank and of Pole Sand, Dawlish Warren and the Exmouth shore were sand. The large embayment in the north of the Warren south of Exe Bight [SX 988 801] was sand. Mud and muddy sand predominated north of the Inner Warren, west of Exe Bight. However, surface sediment distribution is likely to vary, both seasonally and over longer periods.

Kidson (1964) recorded the following log of a borehole [SX 9838 7913] sunk in 1949 at the margin of Greenland Lake, the base of the drift being at 9.4 m below OD:

Kidson mentioned other boreholes on the southern edge of the Inner Warren which revealed sand, similar to that making up the present Outer Warren, resting on beach shingle at 0.3 to 0.6 m above OD. Certain of the cores showed intercalations of clays and muds, and Thomas (1971, p. 209) noted that similar intertidal muds deposited on the floor of the Greenland Lake were exposed below high water mark east of Dawlish Warren [SX 9887 7914] and showed the empty shells of mud-dwelling lamellibranchs still in their life positions.

Dawlish Warren is generally considered to have been built by an easterly drift of sediment across the mouth of the estuary from Langstone Rock.'Subsequent erosion of the spit was at first thought to be due to a decrease in the supply of sediment, resulting from the construction of the railway between Newton Abbot and Exeter in 1849 which isolated the sediment source. However, Martin (1872, 1876, 1893) showed that the Warren was wasting before the building of the railway. Kidson (1950, 1964) envisaged a gradual diminution in the overall supply of sediment along the south coast since the Flandrian transgression, and noted that Pole Sand decreased from 85 to 72 hectares between 1945 and 1954. Durrance (1969a, p. 93) observed that the Admiralty Chart for 1960 showed both Pole Sand and Great Bull Hill Bank to have increased in area since 1945, that the low-water shape of The Warren might also have increased in area, and that the net loss of sand from The Warren had been small. Nonetheless that part of The Warren which is above high-water mark is diminishing in area, and Durrance (1969a) suggested that this might be due to periodic increased erosion associated with long-term rising sea-level and short-term oscillations in sea-level. He considered that the development of the spit might be related to eddies from the main tidal stream. Kidson (1964) saw no evidence for a rise in sea-level over the last two or three centuries.

Henson (1971, p. 313) noted that the composite surface of Exe Breccia and late Pleistocene gravels which was inundated by the Flandrian transgression lay above the low-water level of ordinary spring tides, and might have acted as a trap for cobbles and pebbles soon after its submergence. If so, the lag gravel could have accreted, acted as a trap for windblown sand, and initiated the development of The Warren. Between Langstone Rock and Warren Point the composite surface is at 11.6m below OD, and it is considered that the main channel of the Exe cut through here until comparatively recently. Durrance (1969b) noted that the Exe runs parallel to late Pleistocene channels beneath Dawlish Warren. It is probably re-excavating a buried Pleistocene channel and will become stabilised provided The Warren is not breached at its most vulnerable place, west of Warren Point and south of Exe Bight.

Marine beach deposits and tidal flats

The majority of beaches in the district are only a few tens of metres wide or less. More extensive are the intertidal banks of sand, comparable to beach sand, which occur to seaward of the mouths of the Teign and Exe estuaries. The sediments of beaches and tidal flats range from sand to coarse gravel; their typical red coloration reflects derivation from local Permo-Triassic rocks.

Blown sand

Unconsolidated yellow fine sands, transported and deposited by wind, are present at Dawlish Warren and at The Maer, south of Exmouth. On Dawlish Warren the remains of the largely destroyed low sand hills of the Outer Warren have been partly stabilised by the planting of tree-lupins (Lupinus arboreus) and by other conservation measures. Kidson (1964) noted that the erosion of the Outer Warren on its seaward side was accompanied by deposition of wind-blown sand on its landward slopes, and (1950) that the Lower Warren was composed of sand blown from the Outer-Warren.

The Maer is bounded on its southern side by low grassed dunes of blown sand [SY 003 802] to [SY 012 799]. Kidson (1964, p. 181) noted that heavy erosion of Dawlish Warren coincided with wind-blown sand covering the sea-front road at Exmouth, and considered that the sand dunes of The Maer had formed in this manner.

Siliceous blocks of unknown age

Large red cherty blocks up to 2 m across occur near the boundary between Permian and older rocks in the areas of Ideford, Well [SX 880 770], Whiteway Barton [SX 885 752] and Lindridge House [SX 897 758]. They are particularly evident in an orchard behind Underhays Farm [SX 880 771], near Hamblecombe Lane [SX 8882 7703], and north of Hestow Farm at Luckaberry Copse [SX 8899 7632]. At the last two localities the blocks have been re-arranged in rough circles, but probably they are too heavy to have been moved far.

Ussher (1913, p. 120) noted that the blocks near Well appeared to be associated with Tertiary deposits. Hamblin (1969, p. 41) considered that the blocks were the remains of a siliceous duricrust developed beneath a sub-Permian pediplain before accumulation of the breccias. Alternatively they may represent silicification of Devonian limestones, since such limestone is present at Well and beneath the Permian east of the fault that runs between Ideford and Whiteway Barton.

In the Bickington area, cherts comparable to those in the Whiteway Barton–Ideford area occur as crags and blocks [SX 8077 7332] to [SX 8087 7357]; [SX 7955 7247]; [SX 7971 7254]. The cherts are orange, pink and brown, with vuggy quartz-lined cavities, and they show brecciation. They were considered by Riddolls (1970b) to have formed by replacement of Chercombe Bridge Limestone after tectonism; the area is characterised by a strong zone of thrusting, and the chert outcrops are bounded by thrusts.

Landslip

Upper Greensand has slipped over Permian breccias on Little Haldon at Smallacombe Goyle [SX 922 768], where the lower slopes of the Greensand are marked by three slip scars, and on Great Haldon, where there is a more extensive slipped mass [SX 880 845]. Landslips are also present on the coast west of Budleigh Salterton, as in The Floors [SY 0558 8150] where small slipped masses, largely of Pebble Beds material, obscure the base of the cliffs. Small landslips have also occurred on a few steep slopes underlain by Crackington Formation sandstones and shales.

Denudation chronology

Of the surfaces of low relief recognised at various altitudes in south-west England, all products of early Tertiary to Pleistocene erosion, those above the '690-foot' (210-m) surface are generally considered to have formed under subaerial conditions, and the lower surfaces are thought to be of marine origin.

The '1000-foot' (305-m) surface is well seen around Moretonhampstead and may extend into the north-west corner of the present district (Ussher, 1913). Gregory (1969) mapped surfaces at around 800 ft (244 m) on the Haldon Hills and on the flanks of Dartmoor, and a notch at 690 ft (210 m) on the southern flank of Little Haldon and around the western border of the district has been taken to mark the shore-line of the early Pleistocene sea. However, areas of low relief at around 430 ft (131 m) have been related to the Pliocene platform so well developed elsewhere in south-west England (Ussher, 1913, p. 117).

Reid (in Ussher, 1913) considered the '800-foot' platform to have been cut probably in Miocene times. However, Hamblin (1969) has shown that the flat top of the Haldon ridge is an Eocene sub-aerial surface associated with the formation of the Eocene gravels, albeit considerably modified by mid-Tertiary folding and faulting and later erosion, and by solifluction during the Pleistocene. He correlated it with the '1000-ft' (305-m) platform of east Devon (Green, 1941; Waters, 1960a).

Kidson (1962) correlated the '690-ft' (210-m) surface with his Westermill stage of the River Exe. Green (1949) and Orme (1964) regarded the promontories of Luscombe and Harcombe as part of this surface. Simpson (1964) considered that this supposed shore-line was the exhumed Upper Cretaceous shore-line. The recent survey has shown (Hamblin, 1969) that benches at various altitudes along the base of the Greensand have been developed by spring-sapping, and that the 210-m notch at the southern end of Little Haldon was formed in this way.

Deposits of sand and gravel between the Bovey Basin and the Haldon Hills may be related to the '430-ft' (131-m) Pliocene surface (Hamblin, 1969).

Ussher (1913, p. 108) noted a 131-m platform around Lower Down [SX 793 784], 2 km west of Bovey Tracey; a similar feature farther north extends west and south-west from [SX 788 798] to straddle the Lustleigh Fault. The origin and age of these platforms is uncertain.

The drainage pattern of Devon is thought by some to have been superimposed from a Cretaceous cover, and its north–south elements to have developed as a result of early Tertiary tilting to the south. Jukes-Browne (1904) considered that drainage in and around the present district was easterly, with the upper part of the present River Teign continuing eastward to the Exe and the east–west stretch of the present River Dart flowing east via Bickington to join the present River Teign near Newton Abbot. Green's (1949) conclusions about the Dart were similar to those of Jukes-Browne. By middle Oligocene times the rivers developed on Cretaceous and Eocene rocks of the easterly-sloping plain had probably cut down into the granite, removing detritus which went to form the Bovey Formation. A southerly tilt probably occurred in middle Oligocene times, extending the southward-flowing subsequents and killing off those flowing northward.

Much of the present drainage pattern of the district must date from the formation of the Bovey Basin. The north-west–south-east course of the River Bovey is closely controlled by the course of the Lustleigh Fault Zone.

Five river terrace levels distinguished in the Exe and Otter valleys afford evidence of changes of base level during the Pleistocene. The presence of buried channels in the Teign and Exe estuaries (Durrance, 1969b, 1971) testifies to low sea levels during the Pleistocene; in the Exe, periods of channelling are correlated with low sea levels of early and late Devensian times. Subsequently the rising waters of the Flandrian drowned both estuaries.

Superficial structures

Superficial structures, for the most part of periglacial origin, have been described from scattered localities in the district.

Most striking are those affecting the Bovey Formation, particularly the Southacre Clay-and-Lignite. Dineley (1963) described dome-like and anticlinal structures, commonly circular or quadrate in plan, and considered that uneven expansion of permafrost into unfrozen beds may have produced unequal pressures leading to the formation of diapir-like structures. The disturbed zone is related to the present ground surface, and the contortions extend downward no farther than about 8 m. They are truncated upwards by the late gravels of the Teign. Straw (1974) considered that the contortions were most likely of Wolstonian or Anglian age.

Fossil ice wedges have been recorded in the Budleigh Salterton Pebble Beds (Gregory, 1969, plate 8); similar structures occur in the top of the Aller Gravel [SX 878 693].

Details

Head

Head on Dartmoor Granite

In Blackingstone Quarry [SX 7842 8578] up to about 1.2 m of soil and granitic debris were seen to rest on the granite. At Laployd Barton, a bank behind a barn [SX 800 859] showed 1.5 m of weathered granite in which there was much black tourmaline and micaceous hematite. Beside a forest track near East Wray, cuttings [SX 7823 8282] and [SX 7886 8273] revealed some 1.8 m of granite debris containing boulders of coarse megacrystic granite up to 1.2 m across; along the tracks west and west-north-west of [SX 7886 8273] 0.6 to 1.2 m of similar Head were exposed. Beside a forest road at [SX 8012 8336] 0.9 m of Head were observed. South of Lustleigh a bank at the side of a levelled site [SX 7840 8086] was in 1.8 m of granite debris. At Beadon an excavation for a barn [SX 8142 8163] showed about 0.3 m of soil resting on 1.5 m of Head, consisting of angular fragments of granite up to 0.4 m across set in a matrix of fine granite debris.

In Loxter Copse [SX 7915 8120] there are many tumbled blocks of Woolley Grit, which have been mapped as an extension of the outcrop to the south. Although they are probably not far removed from their place of origin, they present a disturbed aspect and might be classed as Head.

Head on Devonian and Carboniferous rocks

Head is widespread on the eastern slopes of Dartmoor. Near Shuttamoor, hornfels debris is over 5.5 m thick in a forestry road cutting [SX 8263 8283]. In the middle Teign valley Head is present only as a thin veneer, rarely well exposed, but east of the River Teign it is thicker as the in situ sandy weathering product of sandstones and shales of the Crackington Formation, locally incorporating angular flints derived from the Haldon Gravels.

The Crackington Formation around Lower Down [SX 793 785] is mantled by Head, particularly in the east–west-trending valleys. The steep slopes of the Bovey valley from the western edge of the district to Pullabrook [SX 793 795] are covered by locally thick hornfels-rich Head. Similar Head on the south-eastern slopes of the bluff in the meander loop of the River Bovey west of Pullabrook merges into the Head deposits covering the north-western part of the Bovey Formation. In 1964 excavations at a potential reservoir site [SX 7844 7921] showed 1.0 m of compact deep reddish brown silty clay containing rock fragments overlain by fragments of grey hornfels and reddish brown quartzite set in a little ochreous brown clay, 0.4 m, and dark bluish grey stony clay with brown patches and lenses and numerous fragments and angular blocks of hornfels and quartzite, 1.0 m.

Between Ilsington and Rora Wood [SX 801 748] little Head is present. The steep western and northern slopes of Rora Wood, and the steep slopes of Penn Wood [SX 804 746], are mantled by Head that is probably thick locally; Rora Wood and Penn Wood are ridges of Mount Ararat Chert, and the Head consists of angular chert debris.

The Devonian and Carboniferous rocks between Chudleigh and Newton Abbot carry little Head. In the limestone areas south-west of Newton Abbot, yellowish red to brown silty head, incorporating possible loess, is up to 1.5 m thick in the valleys but thinner or absent on the interfluves (Harrod and others, 1973).

Head on Permo-Triassic rocks

Thin weathered derivatives from the Upper Greensand and Haldon Gravels mantle the Permian breccias around Great Haldon. Harcombe Valley contains little drift, Cullum Goyle [SX 888 823] and Deadman's Combe [SX 882 840] rather more. Extensive areas of drift occur around and north of Waddon Brakes [SX 895 801]: at [SX 8936 8007] 3 m of Greensand sand; at [SX 8940 8070] 3 m of bedded flint and Greensand drift in a dry river course. A stream section above Higher Dunscombe [SX 893 791] shows Head down to 161 m above OD, dying out towards the spurs to north and south.

Up to 1 m of Head derived from the Upper Greensand and Haldon Gravels skirts Little Haldon, becoming thicker in the south-west.

Between the Haldon Hills and the River Exe, Head on the Permian breccias is of variable thickness and rarely mappable. A deposit of flint and Greensand chert overlies Teignmouth Breccia west of Splatford Farm [SX 905 858].

Elsewhere on the Permo-Triassic rocks Head is widespread but patchy. It is mappable only at the foot of the ridge formed by the Budleigh Salterton Pebble Beds, where the reddish brown mudstones of the Littleham Mudstone are mantled by up to 1.3 m of bleached cobbles and pebbles. The outcrop is typically around 400 m wide, but widens northwards to 1.5 km on the west side of Blackhill Quarry [SY 029 857]. Up to 1 m of Head derived from the Pebble Beds is exposed in drainage ditches south of Squabmoor Reservoir [SY 0325 8438]; the cobbles are washed clean of their iron-staining and sand and silt matrix. Farther downslope from the ridge, cobbles and pebbles are abundantly scattered over the outcrops of the Exmouth Sandstone-and-Mudstone and Littleham Mudstone.

Head on Upper Greensand and younger strata

Within much of the Bovey Basin it is difficult to distinguish Head from old fluviatile deposits. The north-western part of the basin is mantled by angular clasts of Palaeozoic rock (commonly Upper Carboniferous sandstone) in silty clay. The Chudleigh Knighton Clay-and-Sand at the Chudleigh Knighton Claypit [SX 842 771] is overlain by 3.0 to 4.6 m of unbedded gravels, with cobbles up to 0.3 m in diameter of abraded granite, Upper Carboniferous sandstone, schorl, quartz, and aureole rocks, set in a brownish white sandy and clayey matrix. The gravels are probably fluviatile deposits which have undergone solifluction. Along the southern margin of the main part of the basin, slaty Head has spread down in places from the steep slopes of Devonian and Carboniferous rocks to rest on the Bovey Formation clays. Drainage ditches [SX 8080 7456] showed angular fragments of chert and dark igneous rock set in brown-stained white silty clay. Farther north the Head contains more sandstone clasts derived from the Upper Carboniferous, and at [SX 8057 7526] near Liverton it comprises purple slate fragments set in white and brown clay. At Ringslade Claypit [SX 846 726], the Ringslade Clay is overlain by 0.6 to 1.5 m of parallel-orientated slate fragments with scattered blocks of spilite.

Along the eastern margin of the Bovey Basin and south of Newton Abbot, the Upper Greensand and Aller Gravel are overlain in places by up to 3 m of reddish brown Head, which may be clayey, sandy or gravelly. It is comparable at some localities to the brown clay and loam of the Haldon Hills, and may similarly incorporate a loessic component. However solifluction has introduced shattered abraded flints and other clasts. The deposit, 1 m thick, rests on Upper Greensand at Babcombe Copse [SX 870 766]. In boreholes between there and Higher Sandygate it was sandy and gravelly, and it rests on Aller Gravel at [SX 8706 7524]. At Higher Sandygate Nursery, a trench [SX 8696 7530] showed 0.6 to 0.9 m of reddish brown silty clay with flint chips, rounded vein quartz, and schorl pebbles. South of the River Teign, reddish brown flinty clay rests on Upper Greensand and Aller Gravel. Excavations near Square Plantation [SX 8840 7030] to [SX 8850 7060] showed 0.6 to 1.2 m of reddish brown silty clay containing scattered flint cobbles and chips, resting on Upper Greensand; similar Head was seen in temporary exposures at Milber [SX 8805 7043] and east of the Centrax factory [SX 8827 7087], where it was 1.5 m thick.

Head Gravel is present on both Great Haldon and Little Haldon, but thickest on the former. At Buller's Hill Quarry [SX 8821 8465], Eocene gravels show signs of solifluction and contain mixtures of abraded and unabraded flints. East of Wild Ditches, an overgrown quarry [SX 8874 8279] shows pale grey, slightly brown-stained solifluction gravels composed of abraded flints, tourmaline and quartz in a dense clayey matrix. The topmost bed, up to 1 m thick in the northern part of the quarry, consists of frost-shattered flint chips, stained dark brown peripherally, together with quartz and tourmaline pebbles in a dark brown sandy matrix.

North of Veitch's Plantation, a quarry [SX 8931 8432], now filled in, showed gravels containing a mixture of abraded and unabraded flints. A degraded quarry [SX 8964 8397] south of Deers Hill shows about 2 m of unabraded and abraded flints with exotic pebbles in a sandy matrix, resting on Upper Greensand.

A quarry [SX 8982 8244] south of Haldon Race Course shows 2 m of uniform coarse abraded flint gravel, containing a few unabraded flints and exotic pebbles of quartz and tourmaline, in a matrix of slightly iron-stained clayey silt.

West of Harcombe Plantation, a face in the centre of a quarry [SX 9040 8200] shows a sand body interbedded with flint gravels. The lower gravel contains closely packed abraded flints, with quartz and tourmaline pebbles up to 75 mm in diameter, set in coarse sand with much angular flint and tourmaline. A few unabraded flints are present. The upper gravel is similar, but with no unabraded flints, a more clay-rich matrix, and many flints smaller than 25 mm. The small amount of clay, the presence of a concordant sand body, and the position low down in a river valley, point to some Pleistocene re-sorting in a stream or small pond.

The eastern carriageway of the A380 road between Telegraph Hill [SX 912 834] and Harcombe Plantation [SX 909 819] cuts through abraded flint gravels beneath 1 to 2 m of brown clay. A temporary excavation [SX 9105 8310] on Telegraph Hill showed brown clay resting on Tower Wood Gravel. At least 3 m of abraded gravels are exposed in a large overgrown quarry [SX 9175 8157] east of Kenton Hill. West of Colleywell Bottom, a quarry [SX 9203 8145] showed up to 2 m of mixed abraded and unabraded flint gravels. Construction of the A380 road between [SX 900 805] and Ashcombe Cross [SX 902 796] revealed Head Gravel containing masses of Upper Greensand; the degree of mixing varied.

On Little Haldon, all the gravels show some signs of solifluction and it is difficult to trace the junction between Eocene and Pleistocene gravels. Along the east side of Teignmouth Golf Course [SX 919 757] to [SX 920 751] thick Head Gravel spreads down to 222 m above OD, 24 m below the crest of the hill. A large shallow pit [SX 920 767] above Smallacombe Goyle shows up to 3 m of Head Gravel which has probably moved downhill for at least 200 m. The flints are close-packed, abraded, and show all stages of frost shattering from cracks to wide dispersal of fragments. The matrix of the gravel is clayey sand, made up of flint chips and quartz grains. About 1 m of brown loam overlying the gravel contains scattered frost-shattered flint chips and is capped by 0.3 m of angular flint drift.

Caves and cave deposits

Torbryan caves

The Torbryan caves comprise ten small caves in Chercombe Bridge Limestone, several of which have been excavated and have yielded Pleistocene fossils. Seven caves occur in the face of a low cliff on the south-west side of the dry Torbryan valley: Torcourt Cave [SX 8184 6723]; unnamed cave [SX 8183 6723]; Tornewton Cave [SX 8172 6738]; the Old Grotto [SX 8169 6738] (with which are associated the remains of a medieval building); unnamed cave [SX 8169 6738]; Three Holes Cave [SX 8155 6747]; Broken Cave and rock shelter [SX 8152 6749]. On the northern side of the valley are Pulsford Cave [SX 8122 6777] and Levaton Cave [SX 8103 6797]. Rectory Cave [SX 8177 6752] is in the rock cliff behind Torbryan Old Rectory.

Tornewton Cave was excavated by J. L. Widger between 1870 and 1890 (Walker and Sutcliffe, 1967), and by Sutcliffe and Zeuner between 1953 and 1960 (Sutcliffe and Zeuner, 1962). The earliest deposits are thought to be interglacial (Hoxnian), and are unfossiliferous laminated clays deposited by water flowing through the cave; the possible frost-brecciation of the upper layers of the clays indicates the onset of colder conditions. A succeeding cold climate (Wolstonian) is indicated by the 'Glutton Stratum', and the cave at this time was a den occupied by bears, glutton (wolverine), reindeer and other animals. The overlying 'Hyaena Stratum' indicates the milder climate of the Ipswichian interglacial, when the cave was a hyaena den; remains of warm species such as hippopotamus, narrow-nosed rhinoceras and fallow deer have been recovered. Subsequently the Devensian ice sheet advanced southward, and the 'Reindeer Stratum' contains remains of reindeer and woolly rhinoceros, as well as the first evidence of Man. The youngest deposit dates from the historic period.

Three Holes Cave was excavated in the 19th century and yielded Upper Palaeolithic material. Later excavations by Rosenfeld (1964) produced evidence of Upper Palaeolithic and Mesolithic cultures. The Creswellian material occurred in the upper level of a scree deposit with a poor Ursus arctos fauna, and was sealed by stalagmite. An overlying sterile level was succeeded by a Mesolithic microlithic assemblage.

Pulsford Cave was excavated in 1955–56 by Cornwall and Zeuner, who found a bone of rhinoceros but no habitation levels. Levaton Cave was discovered in 1950 by Mr W. A. Cheesman; the rodent and amphibian fauna, comprising vole, field mouse, frog, and ?toad, were assigned by Carreck (1957) to the first half of the Devensian.

Lemonford Cave

Lemonford Cave [SX 792 723] lies in a quarry between the A38 and A383 roads near their junction, and has yielded bones of Devensian age (Sutcliffe, 1969, p. 70).

Chudleigh caves

Some 20 caves, mostly small, have been found in Chercombe Bridge Limestone around Chudleigh. The best known are Pixies Hole main entrance [SX 8654 7867], Chudleigh Cavern [SX 8649 7877], and Cow Cave (Cow Hole) [SX 8646 7867]. Pixies Hole is located in Chudleigh Rocks alongside the Kate Brook valley, and has 265 m of passages and three openings; Buckland excavated one of the openings in 1823, and noted a prehistoric hearth with charcoal, pottery and flint knives, and, beneath stalagmite, hyaena, deer and bear remains. Further information about early excavations is given by Pengelly (1873).

Chudleigh Cavern was noted by Ussher (1913, p. 118) as affording good examples of stalagmitic encrustations but as not having yielded animal remains. Cow Cave is a more usual linear cave, in which remains of bear have been found. Rosenfeld (1969, p. 133) noted six flakes of probable Middle Palaeolithic age. Recent re-excavation yielded only one flake, from a lower deposit separated from an overlying reindeer-bearing bed by fractured stalagmite.

Tramp's Cave (Tramps Hole) [SX 8667 7868] is a small rock shelter in the Chudleigh Rocks. Evidence of an Upper Palaeolithic backed blade industry, with obliquely blunted blades, has been obtained from a cave earth containing Bos, Equus and Genus elaphus. The cave is now covered with rubble from Palace Quarry.

River terraces

Rivers Bovey and Teign and tributaries

Erosion has proceeded more rapidly on the sedimentary formations adjoining the granite than on the granite itself, resulting in stream captures. It seems possible that the stream which feeds the Kennick Reservoir and the upper part of the Tottiford Reservoir formerly continued its south-easterly course beyond [SX 813 827], but that its waters were captured by a tributary of the Trenchford Stream in the valley now occupied by the lower part of the Tottiford Reservoir (now Trenchford Reservoir); this stream, in turn, was diverted twice north of Beadon to join the Beadon Brook, a tributary of the River Teign. Similarly, the Wray Brook may once have flowed south-eastward over the col at Slade Cross [SX 799 812], but it was probably captured by a tributary of the stream which rises south-west of Sanduck [SX 7685 8364], 3 km north-west of Lustleigh; this stream is itself a tributary of the River Bovey. The profiles of the Sanduck Stream, the Wray Brook, and a stream which rises in Stonelands Waste, 0.8 km east-north-east of Wolleigh, are graded to a base level 18 to 27 m higher than at present, but the rejuvenated River Bovey shows no corresponding knick point.

At East Wray the discontinuous alluvium of the Wray Brook is bordered by a low terrace up to 50 m wide on both banks. At two places a second, slightly higher, terrace has been formed. South of Kelly the narrow alluvium of the tributary is flanked by flat areas up to 150 m wide, which seem to represent the lower terrace. This terrace is again present on both sides of the alluvium at Lustleigh, where gravelly soil was noted. West of Knowle [SX 790 805] the surface of the low terrace is about 1 m above the alluvium.

A terrace of the River Bovey is present between Forder [SX 8003 7998] and [SX 8114 7874], west of Bovey Tracey, its base is some 6 to 9 m above the adjacent alluvium and its surface slopes gently to the south-east. The terrace is dissected at Southbrook Farm [SX 8085 7900] by a small tributary of the Bovey. There are numerous overgrown gravel pits in the woods west of Southbrook Farm and Ussher (1913, p. 121) described 1.5 to 1.8 m of coarse brownish sand made up of granitic and Carboniferous detritus, resting on coarse gravel.

Ussher (1913, p. 121) recorded terrace-like features about 10.7 m above the alluvium of the River Teign near Spara Bridge, east of Christow church and also east of Court Barton. Several patches of terrace gravel are present in the Kate Brook Valley.

Ill-defined possible low-level terrace remnants related to the Teign and Bovey rivers occur in the Bovey Basin. Superficial deposits resting on the lignitic sequence at the Blue Waters Mine [SX 811 769] were described by Pengelly (1862b) as "Head", comprising 1.8 to 3.4 m of clasts of granite, metamorphic rock, carbonaceous grit, and trap, with a very few of flint and chert, set in sandy clay. Both angular and rounded stones were present. A section at Bovey Heathfield (Pengelly, 1862, p. 1031; 1883, p. 370) showed the following beds:

Leaves in the white clay 2.7 m below the surface were of willow resembling Salix cinera, S. repens and S. ambigua, and of the arctic birch Betula nana. They were considered by Heer to be much younger than the lignite beds and to indicate a climate colder than the present one (Pengelly, 1862b, p. 1031). Large roots were found just beneath the leaves.

On the east bank of the Teign between Chudleigh Knighton and Newton Abbot relationships are mostly obscured by old clay workings. Possible terrace deposits at around 30 m above OD occur between Sandslade Copse [SX 857 770] and Fosterville Wood [SX 859 764]. A section in the disused Heathfield Sandpit [SX 8590 7655] shows quartzitic sands and fine gravels and white and brown clayey sands (possibly a coarse facies of the Abbrook Clay-and-Sand), sharply overlain by 0.6 to 1.2 m of coarse reddish brown ?terrace gravels with subrounded clasts of Upper Carboniferous sandstone and other Palaeozoic rocks. In Sandslade Copse, 3 m of clayey gravel with scattered flints and clasts of shale, sandstone and granite were exposed in a small disused pit [SX 8568 7697]. Exposures of up to 2 m of reddish brown fine gravels were seen in the copse and in the wooded ground east of Gappah Brake, for example at [SX 8618 7714].

Gouldstone (1975) reported that excavation of a new claypit [SX 862 751] in the John Acres Lane area revealed up to 9 m of poorly sorted pebbles and cobbles, dominantly of flint, set in orange and grey sandy silts with subordinate grey silty clay. The deposit is locally unbedded, but usually shows thinly bedded orange and grey silty sand and sand, with parallel laminations of silt and clay up to 20 mm thick and some interbedded pebbly silt, overlain by silt and pebble-rich silt. The top of this terrace deposit forms an almost level surface at 22.9 to 23.8 m above OD.

A possible terrace feature on the east bank of the Teign near Preston [SX 857 749] to [SX 859 741] is at 15.2 to 17.3 m above OD (Gouldstone, 1975). Traces of a similar feature are present on the opposite bank between Ventiford and Teigngrace. A borehole [SX 8579 7455] near Preston showed 2.13 m of water-worn gravel consisting of black chert, subangular black slate, rounded milky quartz, and yellowish white angular quartz, with fragments of granite up to 50 mm in diameter. Another borehole [SX 8587 7434] showed 3.81 m of poorly sorted gravel consisting of angular flint pebbles up to 0.1 m in diameter, black angular Palaeozoic mudstone, and rounded flint cobbles up to 0.15 m in diameter, in a matrix of orange-brown silty clay. These gravels are similar in many respects to those in the John Acres Lane area.

A narrow terrace flanks the east bank of the Ugbrooke Stream between Higher Sandygate and about [SX 8654 7578]. A shallow elongate gravel pit [SX 8665 7547], now disused, showed grey gravel with small angular and subrounded flints and vein quartz, 2 m, overlain in places by brown clay, 0.3 m, on which rested brown gravels banded with thin beds of laminated brown clay and fine sand, 1 to 2 m.

Elsewhere in the main Bovey Basin, the Bovey Formation is overlain by superficial gravelly deposits probably largely fluviatile in origin, and there occur scattered and ill-defined flats which may be terrace remnants (Gouldstone, 1975).

A small isolated patch of gravel [SX 856 722] at Sandford Orleigh is some 30 m above OD. Terraces on both banks of the River Lemon at Newton Abbot [SX 854 713] and [SX 851 708] to [SX 868 714] have bases rising to about 15 m above OD.

South of Newton Abbot, sections showing superficial deposits on Bovey Formation are few, but at the disused South Quarry [SX 8701 6936] 1.2 m of gravel, with clasts of vein quartz, flint, Permian sandstone and Greensand chert, together with coarse reddish brown gravelly sand, rest on white clay.

Terrace gravels are present between the head of the Teign Estuary and Bishopsteignton on the north bank, and between the head of the estuary and Coombe Cellars on the south bank. A railway cutting [SX 8763 7235] at Hackney shows 1.8 m of gravel. Temporary exposures [SX 8777 7228] immediately north of the Hackney Channel showed 1.0 to 1.5 m of gravels resting on purple slates at about 8 m above OD. The contact was flat, and the slates immediately below showed curvature of the cleavage due to soil creep. Components of gravel included schorl rock, sparse flint, greyish green slate, greyish green sandstone of Ugbrooke Sandstone type, Carboniferous chert, granite, and weathered dolerite. North-east of Hackney, thin terrace gravels [SX 881 726] lie at about 23 m above OD. A terrace extending from south of Ware Barton [SX 884 729] to south of Forder Cross [SX 897 734] rises to about 12 m above OD. An extensive terrace occurs south of Bishopsteignton; 3 m of gravel are exposed in the river cliff at Flow Point [SX 9088 7287], and a borehole [SX 910 733] proved 6.1 m of gravel (Rhodes and Dineley, 1957). The terrace rises gently northwards to about 46 m above OD, and may be composite in origin. On the south side of the estuary, from Buckland Point [SX 883 722] to Arch Brook Bridge [SX 909 721], patches of terrace gravel occur at up to 9.1 m above OD and between 15.2 and 22.9 m above OD.

Dawlish Water

River gravels alongside Dawlish Water comprise 3 to 5 m of breccia clasts, subangular flints and Greensand cherts. Patches of gravel above Coryton's Cove [SX 958 762] and north of the Smugglers Inn [SX 9568 7538] are composed of similar materials.

River Exe and tributaries

Higher terrace gravels on the western side of the River Exe are composed predominantly of flints and Greensand cherts with breccia clasts. Lower terraces contain also some Budleigh Salterton Pebble Beds material. Gravels developed beside the River Kenn around Willsworthy range from 0.5 to 1.0 m thick and are mainly composed of pebbles of flint and Greensand chert. At Langstone Rock a cliff section [SX 9787 7807] to [SX 9723 7755] shows 2.5 to 5 m of gravels, composed of subangular flints, Greensand cherts, derived Pebble Beds material and breccia clasts set in red silty sand, overlain by 1.5 to 2 m of red silty sand. Lenses of sand, locally cross-bedded, occur in the gravels, which rest on an irregular surface of Dawlish Sandstone and Exe Breccia. Cobble-sized clasts are commonly concentrated towards the base of the terrace.

Terraces on the eastern bank of the Exe are correlated with those of the River Otter. They comprise 2 to 4.5 m of cobbles and pebbles from the Pebble Beds and subangular to subrounded flints set in a red sandy matrix. Extensive outcrops occur near Lympstone, Exmouth and at the High Land of Orcombe.

River Otter

Terraces of the River Otter have been distinguished at the following levels: 1st Terrace 0.15 m above OD; 2nd Terrace 15 to 30 m above OD; 3rd Terrace 30 to 46 m above OD; 4th Terrace 61 to 76 m above OD; 5th Terrace 91 m above OD. The highest terraces, occurring at Hayes Wood [SY 055 845] and Tidwell Mount [SY 063 837], are composed of slightly abraded angular flints produced by frost shattering, Greensand cherts and indurated Greensand cobbles and pebbles, set in a red sandy matrix containing flint shards from the shattered flints. Sand lenses with pebble pavements are present in places. A small quarry [SY 0637 8370] on Tidwell Mount shows 3.75 m of gravel. The material of these terraces is not far from its source, as abrasion is minimal. It has probably been derived from the Cretaceous outliers to the east of the Otter valley and their former western extensions.

The lower terraces, occurring at East Budleigh, Budleigh Salterton, Little Knowle and West Down Beacon, contain cobbles and pebbles derived from the Pebble Beds in addition to subangular flints and Greensand cherts, all set in red silty sand. Sand beds occur locally within the gravels, and the terraces range from 0.75 to 3 m thick. The cobbles and pebbles from the Pebble Beds are washed clean of their red staining and show no chatter marks. Some sections show the gravels to overlie an irregular surface of Otter Sandstone.

Alluvium

Rivers Bovey and Teign and tributaries

Streams on the Dartmoor Granite tend to originate in broad hollows floored by peaty stony silt. Farther downstream the alluvium is generally 40 to 60 m wide, but locally up to 150 m. In the valley of the Wray Brook south of Caseley Court the alluvium forms a broad flat area some 400 m by 250 m; narrow alluvium borders the tributary flowing from the south-east, from an 80-m-wide spread east of Wrayland. Reid and others (1912, p. 64) suggested that two hanging valleys joining Wray Cleave from the north-east might have been formed at a time when the Wray valley was filled with snow or stagnant ice. A similar obstruction farther downstream near Lustleigh might have resulted in the production of the broad alluvial expanse referred to above. Other hanging valleys join the Wray Brook valley at Elsford [SX 7912 8274] and Bullaton [SX 8011 8205]. Torrents pouring down from Elsford have given rise to an alluvial fan at the foot of the valley, consisting of material up to large boulders in size. A smaller but otherwise similar fan was noted at East Wray [SX 781 827], also on the south-west-facing slope of the Wray Brook valley.

The alluvium of the River Bovey north-west of Bovey Tracey is generally gravelly. That of the River Teign north of Chudleigh Knighton is up to 250 m wide and mainly comprises poorly sorted gravels and coarse sands.

The flood plains of the Bovey and Teign are widest on the low-lying ground of the Bovey Basin, and up to 0.6 km wide between their confluence and Newton Abbot. Sections of alluvium are seen in many claypits. At Etsom Claypit [SX 852 762] 2.7 to 3.0 m of brown alluvial gravels rest on Abbrook Clay-and-Sand. Cobbles up to 0.3 m in diameter are present at the base of the gravels. The clasts include much Carboniferous detritus (notably Upper Carboniferous sandstone and Lower Carboniferous chert), together with hornfels, granite, quartz-tourmaline rock, and a few flints with brown patinas. Similar gravels in the Teignbridge Claypit [SX 860 737] are 4.6 m thick and contain some cobbles of Ugbrooke Sandstone.

At the Pinsent Claypit [SX 861 730], the gravelly alluvium of the Teign is extracted as a by-product of ball clay production. Gouldstone (1975) recorded the following sequence overlying Southacre Clay-and-Lignite:

The basal gravel contained logs of oak (Quercus)[SX 8589 7308] which gave a radiocarbon age of 3332 ± 70 years B.P. (Radiocarbon, 1974, p. 249). Pengelly's (1883) archaeological finds in superficial gravels at Zitherixon Quarry about [SX 865 725] were ascribed to the early Bronze Age, which agrees well with the radiocarbon dating.

Excavations for the A380 bypass road at Colley Brook [SX 8707 7483] showed the following section:

Three streams draining northward into the Teign Estuary are flanked by narrow alluvial flats of silts and clays, with some gravels containing breccia-derived clasts.

River Lemon and the area south and west of Newton Abbot

The River Lemon, which runs from west of Bickington to join the Teign near Newton Abbot, is flanked by a narrow alluvial flat, as are also its tributaries the Kester Brook and Barham Brook. The alluvium of the Kester Brook around and upstream of Longstone Bridge [SX 7957 7136], formerly worked for alluvial tin (p. 172), comprises gravel and sand with boulders of local spilite, limestone and slate. A borehole [SX 8629 7146] in Newton Abbot showed 5.4 m of alluvial deposits on presumed Bovey Formation sands.

The ground south of Newton Abbot is drained by the Aller Brook, whose alluvial flat is up to 200 m wide; the stream is small and apparently a 'misfit'. Pengelly (1862b) recorded a section in the old claypit at Aller, where up to 6 m of gravel rested on Bovey Formation clays.

Tributaries of the Rivers Exe and Otter

The alluvium of the minor streams draining east from the Haldon Hills comprises silts and clays, locally with gravels consisting of flints, Greensand cherts and breccia-derived clasts.

East of the River Exe, narrow tracts of alluvium along the Watton Brook (Lympstone), Withycombe Brook, Littleham Brook and their tributaries comprise silts and clays, with gravels containing cobbles and pebbles from the Budleigh Salterton Pebble Beds. The flat ground of The Maer [SY 008 801], south of Exmouth, is largely underlain by alluvial sand and gravel, with dunes of blown sand on the seaward side.

The alluvium of the small streams draining east to the River Otter contains many cobbles from the Pebble Beds ridge.

Sand and gravel of unknown age

Poorly exposed deposits of sand and gravel occur at about 500 m south of Ugbrooke House [SX 875 775], near Colmansford Bridge [SX 893 764], north of Lindridge House [SX 899 760], and at Humber Lane [SX 897 753]. They comprise flints and chert, with a little Permian material, in a dark grey or brown soil. Augering showed brown sand and brown clayey sand in the Lindridge House and Humber Lane deposits, which rest on benches at around 120 m above OD; both benches are cut into Permian beds and are unrelated to the underlying geology, although the southern one merges southwards into an exhumed sub-Permian surface cut into Devonian limestones. Both deposits are backed by concave changes of slope, and it is possible that they are related to the widespread 430-ft (131-m) marine platform of late Pliocene or early Pleistocene age. However, the flints are largely unabraded, and the gravels may be solifluction deposits which have come to rest on earlier benches. Alternatively, they may be either river terrace gravels, formed at a time when ground levels in the Bovey Basin were very much higher, or correlatives of the Aller Gravel. The Colmansford Bridge deposit, which rest on Devonian limestone and Permian breccias, is adjacent to an alluvial flat and may be redistributed terrace material.

Spreads of gravel south-west of Ideford at Hamblecombe Lane [SX 890 772], and east of Luton [SX 905 767], lie between 90 m and 120 m above OD. The Luton deposit rests on Permian breccias at a level comparable to the Lindridge House and Humber Lane deposits. The Luton and Hamblecombe Lane drifts comprise flint, chert and local Upper Palaeozoic material; they may be Head Gravel, although the Hamblecombe Lane deposit could be a redistributed terrace. The deposit south of Ugbrooke House reaches a height of 120 m in its southern part, and in this respect is comparable to the Lindridge, Humber and Luton deposits.

Marine beach deposits and tidal flats

Beach sand and gravel have accumulated in small coves beneath the steep cliffs of Teignmouth Breccia south of Teignmouth; Watcombe Beach [SX 9263 6734] is indented along the outcrop of the softer Watcombe Breccia. Banks of intertidal sand off the mouth of the River Teign, such as the East Pole Sand [SX 943 723], are associated with the development of the Teignmouth spit. Between Teignmouth and Dawlish Warren long narrow spreads of beach sand and gravelly sand are punctuated by headlands between The Parson and Clerk [SX 961 747] and Old Maid Rock [SX 962 761], and at Langstone Rock [SX 980 780]. A belt of beach sand and gravelly sand flanks the seaward side of Dawlish Warren. Much of what was once the Outer Warren is represented by a bank of intertidal sand, which extends some 300 m seaward of the present high water mark and is connected to the east with the intertidal Pole Sand (Durrance, 1969a) .

Between Exmouth and Straight Point [SY 037 794] is a beach of sand and gravelly sand about 100 m wide. Durrance's (1969b, p. 178) seismic survey indicated 1.5 to 3.6 m of sand between Exmouth Dock and Orcombe Point, resting variously on Permian rocks, clays and late Pleistocene gravels. In the east the beach deposits are largely derived from the Budleigh Salterton Pebble Beds; they consist predominantly of quartzite cobbles together with some pebbles of material derived from superficial gravels.

Storm gravel beach deposits occur between Littleham Cove [SY 039 803] and Budleigh Salterton. They are best developed east of [SY 046 809], owing to the plentiful supply of cobbles from the Pebble Beds, and two clearly defined berms trend along the beach.

Chapter 13 Geophysical investigations

The Newton Abbot district has been covered by regional gravity surveys, and more detailed studies have been made of the form of the Bovey Basin. Seismic surveys by the Institute of Geological Sciences and the Geology Department of Exeter University were intended to resolve specific local problems.

Gravity surveys

The Bouguer anomaly map produced by Bott and others (1958) during their regional gravity survey showed a gradient associated with the steeply plunging eastern margin of the Dartmoor Granite. Superimposed on this was a negative anomaly of about −6 mGal, attributed to the low-density sediments filling the Bovey Basin. Assuming the density contrast between the Tertiary sediments and the surrounding Palaeozoic sediments to be 0.7 g/cm³, the authors postulated a basin depth of 200m. Fasham (1971) made a more detailed examination and, after removing the regional gradient, arrived at a gravity anomaly of −15 mGal. Using a model which assumed that the clay density increased with depth, as in the Petrockstow Borehole (Fenning and Freshney, 1968), Fasham predicted a maximum thickness of 1245 m; this figure would have included the Bovey Formation and any underlying Aller Gravel or Upper Greensand. However, Vincent (1974, p. 94) commented that deep boreholes drilled in the central part of the Bovey Basin had shown that substantial thicknesses of lignite were present, but in the Petrockstow Basin lignite was of minor importance. Such thicknesses of relatively low-density material led Vincent to estimate that the maximum depth of the Bovey Basin was about 1066m.

Seismic surveys

A seismic reflection and refraction survey between the Haldon Hills and Colaton Raleigh revealed (Henson, 1972) the form of the Permo-Triassic basin to be typical of a pediment with a mountain front adjacent to the upland source area to the west. There is no geophysicaal evidence to support the existence of a large fault along the line of the River Exe. Transmission velocities and thicknesses deduced for the formations recognised in the survey are listed in (Table 11).

Durrance and Hamblin (1969) carried out a refraction survey of the Haldon Hills and calculated velocities of 520 to 760 m/s for Haldon Gravels, 1280 to 1680 m/s for Upper Greensand and 2800 to 3000 m/s for Teignmouth Breccia. This survey revealed variations in thickness of the Upper Greensand, from 16 to 84m, which were attributed to a period of intra-Cretaceous folding and faulting. Fault movements during the formation of the Haldon Gravels were thought to have resulted in depressions in which as much as 21 m of Tower Wood Gravel accumulated.

Fasham (1971) reported that the shape of the Bovey Basin indicated by gravity studies was broadly confirmed by the results of an unpublished seismic refraction survey conducted by Mr W. N. Ridley Thomas, but that interpretation was complicated by the facts that seismic velocities in the Bovey Formation increased with depth and the dip of the underlying refractor was variable.

A refraction survey of the northern part of the Sands Copse area, on the eastern margin of the Bovey Basin, was conducted by Mr E. M. Durrance and Dr R. A. Edwards to determine the thickness of the Upper Greensand and Aller Gravel and to investigate the nature of the sub-Cretaceous floor. Seismic velocities were: Aller Gravel about 380 m/s, Upper Greensand 1006 m/s, sub-Cretaceous rocks (Devonian/Carboniferous/Permian) about 2950 m/s. The seismic traverse showed that the Upper Greensand thickened from about 12.2 m in the central part of the Sands Copse Pit to 20.4 m in the northern part. The overlying Aller Gravel was everywhere thin, ranging from less than 1.2 m in the south to 3 m in the north. The sub-Cretaceous floor showed an apparent dip to the north of about 4°.

Durrance (1969b, 1971, 1974) reported on a number of refraction surveys to locate buried channels in the Exe and Teign estuaries. Velocities determined are summarised in (Table 12). Velocities within the Exe Breccia may exceed those in the Devonian slate of the Teign Estuary. The presence of low-velocity horizons within the Permian may mean that the depth to its base suggested by Henson (1972) is a maximum.

The surface of the Permian rock beneath Dawlish Warren dips from 0.6 m below OD at the western end to 20.6 m below OD near Warren Point; farther east it crops out at low water mark at Checkstone Rock, Checkstone Ledge and Maer Rocks. This surface is cut by several north-west–south-east channels, possibly to below 49.6 m below OD. These channels were subsequently filled with gravel and re-excavated to a base below 27.6 m below OD. The buried channel of the River Teign drops from 7.8 m below OD at the head of the estuary to 20.5 m below OD at Teignmouth. It is floored by Devonian slate for most of its length, but by Permian breccias near Teignmouth.

Clarke (1970) used continuous seismic reflection techniques to determine the thickness of offshore Quaternary sediments and the topography of the bedrock. He located a sinuous channel extending southwards from the present mouth of the River Exe, and reported slight evidence for subsidiary channels to the east, but not to the west. Mr E. M. Durrance (personal communication) has observed that Pleistocene sediments with velocities used by Clarke to calculate thicknesses have now been recognised in the lower Exe Estuary and that Clarke's thicknesses for the Pleistocene must be regarded as minimum values.

Chapter 14 Economic geology

Soils and land use

The soils of the present district and of that to the north have been mapped and described by the Soil Survey of Great Britain (Exeter and Newton Abbot, sheets 325 and 339; Clayden, 1971). The following brief account owes much to these studies.

The Devonian slates in the south-west of the district have given rise to generally well-drained brown earths, the most widespread being fine loamy or silty brown soils. Farming is mixed, with the emphasis on dairying. The associated Devonian limestones have soils of variable depth, also classified as brown earths. Much of the limestone country, particularly steep slopes and rocky ground, is wooded, and the remainder is largely given over to pasture.

Soils on the Carboniferous rocks, particularly in the Teign Valley, are closely related to site conditions (Clayden, 1964). Loamy brown earths are present on well-drained steep slopes, but the gentler slopes have surface-water gley soils and more weathered gleyed brown earths. Loamy soils overlie the Ugbrooke Sandstone. Farms are generally small, and farming affected by steep slopes and poor drainage; grass and dairying predominate, and oak woodland occurs on many of the steeper slopes.

The granite of the north-west has given rise to ochreous brown soils, with peaty gley soils capping the alluvium of the small valleys. Most of this land is farmed; potatoes and corn are common but grass is dominant. There is a considerable area of rough grazing and woodland, particularly on steep or boulder-strewn slopes. The catchment area of Tottiford and Kennick reservoirs is largely under coniferous woodland.

The hilly areas bordering the Dartmoor Granite (the "Dartmoor Foothills" of Clayden, 1971, p. 99) include aureole rocks and Devonian and Carboniferous shales and cherts; the soils are mostly brown ochreous and stony and are known locally as 'woodstone soils'. Much of the area is under grass, with some arable cultivation on gentler slopes.

The red soils on the Permo-Triassic rocks were broadly divided by Clayden (1971) into those on breccia hills, sandstone lowlands, marl lowlands, and the pebble beds ridge. The Permian breccias west of the River Exe and south of the Teign Estuary are characterised by well-drained gravelly loams and clay loams, varying with the nature of the local breccias. Mixed farming is predominant, with some horticulture around Stokeinteignhead and Dawlish; woodland is generally of small extent. The sandstone lowlands lie on Dawlish Sandstone west of the Exe and on Otter Sandstone east of the Exe. Their well-drained sandy soils support a wide range of agricultural and horticultural crops and about half the land is arable. Farms here and in the breccia hills tend to be larger than elsewhere in the district. The marl lowlands lie east of the Exe Estuary, on Exmouth Sandstone-and-Mudstone and Littleham Mudstone; they carry mostly gleyed brown earths, commonly with a thin top pebbly layer derived from the Budleigh Salterton Pebble

Beds. There is a long tradition of dairying on the marl lowlands, and grass yields can be very high. The ridge formed by the Pebble Beds is characterised by infertile, mainly coarse, loamy and gravelly soils; patches of silty drift give rise to more stone-free silty loams. The area is mainly heathland, used for recreational purposes and some scattered plantations of conifers occur.

The Upper Greensand and Eocene flint gravels of the Haldon Hills are capped by podzolised soils. There is practically no agricultural land on the Haldon plateau, much of which is given over to plantations of conifers; the Exeter Forest of the Forestry Commission covers 1860 ha (4600 acres), mainly on Great Haldon. The Haldon Hills also serve as a recreational area.

The Bovey Basin carries a wide range of soils and shows diverse land use. The central part is occupied largely by surface-water gley soils developed on the Bovey Formation; elsewhere, peaty gley soils and gleyed brown earths occur. The Upper Greensand and Aller Gravel along the basin margins are covered by humus-iron podzols and ochreous brown soils. Farmland is interspersed with areas of wood, scrub and heath, and workings for clay, sand and gravel. An extensive area of woodland around the Great Plantation in the centre of the basin constitutes the main part of the Bovey Forest. The agricultural land is dominantly given over to permanent grass or long leys, with a small amount of arable cultivation. The Upper Greensand and Aller Gravel, where not quarried or built over, carry widespread wood and scrub.

The alluvial soils of the river flood plains are of limited extent; they are commonly liable to flooding and are largely under permanent grass.

Water supply

Annual rainfall over most of the country around Teignmouth and Newton Abbot averages between 760 and 1015 mm (30 to 40 inches). Slightly more falls on the higher ground west of Hennock and Christow, and it is here, on the granite, that the main reservoirs of the district have been constructed. A variety of geological formations and rock types dictates differences in the rate of infiltration of groundwater, but in general the prospects of obtaining large supplies from wells and borehbles increase eastwards from the Palaeozoic slates and shales across the Permian breccias to the Triassic sandstones.

Rainwater readily penetrates weathered granite, but enters the fresh rock scarcely at all except along fractures. Small supplies of soft water are therefore available within perhaps 10 m of the surface, but they are unlikely to exceed about 0.5 litres per second (Us) or 400 gallons per hour (gal/h).

The Devonian and Carboniferous rocks, apart from the limestones, consist mainly of slates and shales. Infiltration is probably slight and intergranular movement of groundwater neglible, even within the sandstones of the Crackington Formation. However the rocks have been much fractured by tectonic movements and it is probable that small to moderate supplies are available from fissures. Unfortunately the locations of the fissure systems are not predictable and in any case these are likely to be of small extent; thus water will move rapidly within them but will be replenished only slowly. Available evidence suggests that the Devonian slates generally yield slightly more than the Carboniferous rocks, probably owing to more extensive fracturing, but the maximum recorded is only 1.4l/s (1100 gal/h) from a borehole 76 m deep.

The Devonian limestones, in contrast, are commonly massive and locally penetrated by extensive solution channels. Underground flow is rapid and the rocks constitute minor aquifers. Their potential as sources of supply has not been fully utilised; thus recorded yields of 0.3 and 0.9l/s (240 and 700 gal/h) from boreholes near Bickington and Kingsteignton reflect not the possible supply but the capacity of the pumps in use.

Infiltration over much of the Bovey Formation is retarded by argillaceous sediments but may amount to about 300 mm per year over the areas of outcrop of the sandy beds above the Southacre Clay-and-Lignite. Inflow of water from sand has been recorded in various boreholes for ball clay, but generally without any note of quantities; however a 16-m borehole north of Sandygate struck Aller Gravel at the bottom, and the estimated inflow from the gravel was over 1.2l/s (1000 gal/h).

Yields from wells and boreholes in the Permian breccias are variable. Up to 6.3l/s (5000 gal/h) have been obtained from depths to 100 m. A few holes have failed, but in general it appears that small yields of a few hundred litres per hour may be expected within 100 m of the surface throughout the breccia outcrop.

The Upper Greensand and overlying gravels of the Haldon Hills are in hydrological continuity with one another and with the Permian breccias below. They mainly occur capping high ground and are highly permeable, so that water infiltrates quickly; it then traverses the unsaturated zone to a rest level usually far below surface. Yields up to 0.4l/s (300 gal/h) of soft water have been obtained from depths of 33 to 84 m.

Near and north of Dawlish around 60 m of sandstones of the Dawlish Sandstone are intercalated within the Permian breccias. Small yields comparable to those from the breccias farther west are common, but two boreholes 53 and 46 m deep a short distance north of Dawlish yielded 3.3 and 4.5l/s (2640 and 3600 gal/h) respectively. A 31-m borehole at the western end of Dawlish Warren passed through 14 m of blown sand and alluvium into Permian sandstone and breccia; pumping at 0.9l/s (700 gal/h) lowered the water level by 4.8 m, recovery taking two minutes.

East of the River Exe the Exmouth Sandstone-and-Mudstone has provided up to 5.1l/s (4000 gal/h) from boreholes up to 90 m deep. By contrast yields for the Littleham Mudstone, in which thin sandstones and silt-stones form the only permeable strata in a predominantly argillaceous succession, range only up to 0.6l/s (500 gal/h) from holes of similar depths.

The mudstones are overlain by the Budleigh Salterton Pebble Beds and the Otter Sandstone, which contain the principal groundwater resources of the district. These two Triassic formations together constitute a good aquifer. Rainwater infiltrates and moves generally down-dip to the east to pass beneath the Mercia Mudstone Group ('Keuper Marl') beyond the present district. Two shallow wells at Kersbrook, just north of Budleigh Salterton, were 3.8 and 5.5 m deep and yielded 17.7 and 22.7l/s (14000 and 18000 gal/h) respectively, pumping for nine to ten hours per day.

No appreciable quantities of groundwater are available from Drift deposits which, except for the modern alluvium, are rarely more than a metre or two thick. Old river gravels, most extensive on the Permian and Triassic rocks, store water temporarily and thus increase infiltration into underlying strata. The alluvium of the Exe, Teign and Bovey rivers could yield small supplies to shallow wells, but there would be risk of pollution.

Current estimates of population growth CA strategic settlement pattern for the south-west, H.M.S.O., 1974) suggest an increase of 33 per cent by the year 2001 in the Newton Abbot–Exmouth area, together with some light industrial development. Groundwater resources within the district are largely confined to the Triassic rocks in the east, with smaller amounts of hard water probably available from the Devonian limestones; only small local supplies are likely from the hydrological unit Eocene–Upper Greensand-Permian breccias. Surface water storage will continue at Kennick and may be augmented by the use of old claypits in the Bovey Basin and by increased abstraction from rivers. There is little prospect of large inland reservoirs. Potential estuarial storage is available in both the River Teign and the River Exe, but it seems more likely that any large future demands for water will be met from outside the district.

Ball Clay

Great Britain produces between one-quarter and one-third of the world output of ball clay, and some two-thirds of this production is obtained from the Tertiary beds of the Bovey Basin between Newton Abbot and Bovey Tracey. Ball clays (defined on p. 130) are valuable raw materials used in the pottery and other ceramic industries, their chief function being to provide plasticity and increase the strength of the unfired ware. They are used extensively in the manufacture of ceramic whitewares such as earthenware, tableware, wall and floor tiles, sanitary ware and electrical porcelain. Other applications include use as a bond in refractories, as a reinforcing filler in rubber products, and as a coating on granulated fertilisers to prevent caking.

Ball clay is marketed in a variety of forms, only about 5 per cent now being sold as lump clay direct from the workings. Some 80 per cent of the clay produced is shredded, when it is cheaper and easier to handle and transport, and is in a convenient form for blending, milling and calcining. The remainder is dried and pulverised. Some ball clay is calcined in rotary kilns, the resulting material being less reactive and used chiefly in the manufacture of refractories. Most potters use blends of clay to get the properties they require. In the past they mixed their own, but now the clay companies produce blended clays to meet a variety of requirements.

The complex history of the south Devon ball clay companies has been outlined by Rolt (1974, pp. 39–68). The Romans are thought to have been the first to exploit the Devon clays, but working seems to have lapsed until the beginning of the 17th century, when the clays began to be dug sporadically for the manufacture of tobacco-pipes. In consequence, they were commonly known as pipe-clays or cutty clays. At first used only locally, south Devon ball clays were later sent to other parts of the country and there appears to have been an early export trade, for Rolt (1974, p. 24) recorded an Act of 1662 forbidding the export of pipe-clay to foreign countries. The first recorded shipment was of 20 tons in 1700. In 1785 over 10 000 tons of clay were shipped from Teignmouth, and by the end of the 19th century ball clays were being exported to many parts of Europe and to the United States.

Ball clay production from the Bovey Basin between 1973 and 1980 was at a level of around 450 000 to 490 000 tonnes per year; some 70 per cent of the output is exported, much of it from Teignmouth. The clay is shipped mainly to Scandinavian, Central European and Mediterranean countries, and to Rotterdam for trans-shipment to more distant markets.

Scott (1929) noted eight firms working clays in the Bovey Basin. The two present producers are Watts Blake Bearne and Co. plc and E.C.C. Ball Clays Ltd (formerly Hexter and Budge (E.C.C.) Ltd). In the main basin ball clays were being extracted from thirteen open pits and seven adit mines in 1981 (Table 13), and some 40 seams ranging in thickness between 1 m and 6 m were being worked. In the 1970s the exploratory activities of the companies were directed towards the relatively unknown central and western parts of the main basin in efforts to determine the extent of reserves.

Clays have not been worked from the Decoy Basin since 1965, but in the past they were worked south of Newton Abbot at Decoy [SX 865 703], South Quarry [SX 869 695] and Aller (exact location uncertain). In the main basin north of Newton Abbot disused pits are situated at Halford [SX 8105 7456], Heathfield [SX 831 762] and Bradley [SX 827 777], and at numerous localities along the eastern crop between Chudleigh Knighton and Newton Abbot.

Scott (1929) distinguished two broad groups of ball clays along the eastern part of the main basin and in the Decoy Basin; a 'Stoneware' group and an overlying 'Whiteware' group. The present four-fold grouping is based on the irreversible thermal expansion of the clays. The properties of the four groups are outlined below; their stratigraphical location and chemistry are described in Chapter 10.

Group 1: Extra white firing Ball Clays:

Group 1 clays include the highest quality clays produced from the Bovey Basin. They are characterised by exceptional whiteness when fired at over 1200°C, due chiefly to low contents of Fe₂O₃ and TiO₂, and to a lesser degree to carbonaceous contents between 0.2 and 4.0 per cent. These clays have an exceptionally high alumina content, and are the most refractory ball clays occurring in this country. The unfired strength (modulus of rupture) of ball clays is in part a function of the amount of colloidal carbonaceous matter present, and this relationship is evident in the Group 1 clays. The carbonaceous clays have the finest grain size of the group, with 20 to 30 per cent below 0.2 microns equivalent spherical diameter (e.s.d.), compared to 15 to 20 per cent in the non-carbonaceous clays.

Group 1 clays occur in the Southacre Clay-and-Lignite and are worked from six pits between Newton Abbot and New Bridge, and from three adit mines, two in the Broadway Claypit (Broadway Nos. 7/8) and the third the large mechanised West Gold Marshes Mine (No. 10 Adit) at Sandford Orleigh (Table 13). There are numerous old workings. In the Decoy Basin white-burning clays were extracted from presumed Southacre Clay-and-Lignite near Decoy (Scott, 1929, p. 31) and probably from a pit near Aller (Key, 1862).

Work began in 1980 on a new claypit at Clay Lane [SX 844 768], planned to work a sequence of ball clays transitional in character between the Chudleigh Knighton Clay-and-Sand and the Southacre Clay-and-Lignite.

Group 2: Dark Blue Ball Clays:

Group 2 clays (roughly equivalent to the 'Dark Blue Clays' of older classifications) are somewhat carbonaceous slightly siliceous clays which fire white between 1000°C. and 1100°C, and off-white at higher temperatures. They are the strongest extracted from the Bovey Formation, with unfired strengths (moduli of rupture) of up to 140 kgf/cm² recorded from the more carbonaceous varieties. This is related to their very fine grain size, with 27.4 per cent less than 0.2 microns e.s.d The strength, toughness, plasticity and excellent bonding power of Group 2 clays make them particularly fitted for mechanical production of earthenware, and sanitary ware manufacturers value their excellent casting properties and high unfired strength.

Group 3: Light Blue Ball Clays:

Group 3 clays are non-carbonaceous somewhat siliceous clays, firing white between 1000 and 1100°C, but off-white to ivory at higher temperatures. They are very plastic and, although lacking the high unfired strength of Group 2 clays, are more thixotropic and hence useful in the liquid slips used in the manufacture of cast sanitary ware.

Group 4: Siliceous Ball Clays:

The siliceous ball clays are the 'Stoneware Clays' of older classifications. They are non-carbonaceous with high percentages of free quartz. Fairly high contents of Fe₂O₃ and TiO₂ give fired colours off-white below 1100°C and buff above this temperature. These are the coarsest-grained clays, and they are used in stoneware and as refractory bonding clays. In the past they have been worked along the eastern outcrop between Preston Manor and Newton Abbot, and south of Newton Abbot at Decoy. They are now extracted from a large open pit (White Pit) near Preston Manor and from a pit at John Acres Lane [SX 862 751] which was opened in 1973.

Other ball clays:

A variety of ball clays is obtained from an open pit at Ringslade and from an adit mine at Mainbow, both on the southern margin of the main Bovey Basin. They include non-carbonaceous clays of good plasticity, firing pale cream; strong and plastic 'Blue Clays' and the aingslade Mottled Clay', which fires reddish brown owing to its very high, iron content. At the Teignbridge Claypit, clays are extracted from the upper Bovey Formation immediately above its contact with the underlying Southacre Clay-and-Lignite. These clays are white, somewhat siliceous, coarse grained and non-carbonaceous, with generally white firedcolours. At Stover Park a thin sequence of ball clays within the upper Bovey Formation is worked in a small open pit; two seams are dug, one providing a very strong plastic carbonaceous clay, the other a siliceous buff-firing moderately strong clay. Other clays within the upper Bovey Formation were formerly extracted from the Heathfield Claypit; they were of the 'Stoneware' type, and were•mainly used in the manufacture of bricks, drainpipes, tiles, paving blocks and some stoneware (Scott, 1929, p. 32). There are few records of clays having been extracted from the western part of the Bovey Basin, but Scott (1929, p. 18) noted that clays near Liverton, although of no value as pottery clays, were occasionally used for the manufacture of saggars.

The open-cast method is the most economical way of working clays, and is employed down to about 35 m where the overburden ('head' or 'heading') does not exceed about 9 m. The overburden is first removed by dragline or scraper and the clay seams are dug by dragline or face-shovel. In the Southacre Claypit a rotary bucket wheel excavator is in use. In most cases the clay is removed from the pit by dumper lorries.

An early method of extraction, used extensively for Stoneware Clays, was the 'square pit' ('four-sides') method. The pits were 6 to 7 m square and some 12 m, sometimes up to 30 m (Brackenbury, 1931, p. 241), deep, with timbered sides. A heavy spade and a mattock-like tool called a 'tubill' (or 'tubal') were used to cut the clay into 'balls' about 0.25 m square and weighing 13 to 22 kg. At first the balls were thrown to the surface by a chain of men standing on ledges ('eaves'), cut at vertical intervals of some 1.5 m in the side of a pit, and using spiked instruments called 'poges' or 'pops'. Later the clay was raised in buckets lifted by cranes or in tubs running along inclined tramways.

Clay seams too deep for economic open-cast working are extracted by underground mining. In the Bovey Basin all such mining is now from adits (drift mines) and the trend is towards a fairly small number of large long-life mines. The clay seams are followed and extracted down-dip along steel-arched or concrete-lined roadways, commonly for about 600 m. Secondary levels are driven out laterally from the main drive, usually for up to 200 m. The clays are cut from the face by machine or by pneumatic spade.

Until the 1960s the common form of access to underground workings was by vertical shafts. Scott (1929) noted that the Whiteware Clays were generally extracted from shafts, and the Stoneware Clays from Tour-sides' pits. The early shafts were rarely more than 30 m deep (Scott, 1929, p. 4), less than some modern open pits. Brackenbury (1931, p. 267) noted the average life of a vertical shaft mine 15 to 45 m deep as 2 years. Levels were driven about 30 m from the shaft bottom. The Heathfield Shaft' [SX 8365 7586], sunk in 1924–25, was 91 m deep, 3.6 m in diameter and cast-iron-lined; it appears to have encountered difficulties owing to the plasticity of the surrounding clays, and little working took place.

Lignite

The Bovey Formation contains by far the most extensive lignites in Great Britain. Although now generally regarded as a waste by-product of ball clay extraction, the lignites have in the past been worked for fuel and considered as a source of organic chemicals, particularly montan wax. Recent applications appear to have been limited to the use of small quantities in the manufacture of fertiliser.

The major known occurrences are in the Southacre Clay-and-Lignite, which crops out along the eastern side of the main basin between Newton Abbot and New Bridge. The lignite beds occur interstratified with brown clays and some silty clays and sands. A thick lignite bed, or complex of beds, is known locally as the 'Big Coal'. The Southacre Clay-and-Lignite also crops out in the Decoy Basin.

Occurrences at Denistone and Southacre typify the lignite deposits along parts of the eastern crop of Southacre beds. At Denistone a borehole [SX 8604 7422] showed, between depths of 20.4 and 32.9 m, a total of 11.7 m of lignite which included some clay bands and clayey lignite. In the Southacre area a borehole [SX 853 753] showed 6.4 m of lignite with clay partings between depths of 49.7 m and 56.1 m.

The Heathfield Borehole proved 79.6 m of clay and lignite, possibly of the upper Bovey Formation, beneath 59.4 m of upper Bovey Formation sands and clays; lignite beds up to 10.5 m thick were recorded. The Heathfield Shaft showed lignite between 80.1 m and 86.2 m depths.

Lignite-bearing beds of the Brimley Member crop out near Blue Waters Mine. Pit sections and boreholes to the south have shown massive lignites (Wynne, 1947, fig. 2), and this is the only place where large-scale commercial extraction has been practised. Several boreholes have been sunk in the area (p. 143). The Bovey Tracey No. 4 Borehole [SX 8161 7661], sunk by a German company prior to the First. World War, is typical: lignite (recorded as "coal") and brown clay were proved from a depth of 22.2 m to 60.6 m, some 43 per cent of the sequence (16.4 m) being composed of lignite in beds between 0.2 m and 2.3 m thick.

The use of the lignite as fuel dates from the early 18th century, but its use in domestic grates has been limited by the unpleasant smell given off during burning (Pengelly, 1862b; Strahan and others, 1920). Fox (1948, p. 130), however, noted that it burned quite satisfactorily in conjunction with wood logs or other fuels. Pengelly (1862b) mentioned the use of lignite as fuel in a pottery and in some poorer cottages near the Bovey Coal Pit (Blue Waters Mine). Lignite extracted as a by-product of ball clay production was also burned locally (Strahan and others, 1920). There is a record of 172 tons of lignite being dug from Hexter and Budge's New Cross Mine in two months during the coalminers' strike of 1926.

Blue Waters Mine was working at the time of Pengelly's (1862b) report, but Woodward (in MS) noted that by 1875 the use of the lignite had ceased. No reliable figures for output are available until 1853. The following output figures are quoted in the 1871 Coal Commission report; 1853, 18 633 tons; 1856, 3850 tons; 1857, 5660 tons; 1867, 1368 tons; 1868, 1383 tons. Immediately before the First World War, the lignites were being worked by a German company, apparently for the extraction of bitumen. Ussher (1913, p. 136) noted that the lignite had recently been worked for the production of gas to generate electricity for use in the clay mines. The pit was closed from 1914 until 1945, when it was drained (Wynne, 1947) and extraction of lignite was commenced by British Lignite Products Limited. Wynne (1947, p. 334) noted that by October 1947 monthly production had reached 15 000 tons. In 1948 output was around 500 tons per day, and it was estimated that there were 50 000 tons of saleable lignite to every acre within a "considerable radius of the . . . workings" (Fox, 1948, p. 127). The calorific value of the lignite was regarded as about half that of coal, and the lignite was sold at between £1.50 and £2.05 per ton (Fox, 1948) for use as fuel for boilers, central heating plant, etc. The main seam was about 4.6 m thick, and extraction was from underground drives and stalls which were kept within the lignite to help prevent roof falls. Inclines were driven obliquely down-dip and headings along the strike. Opencast methods of extraction were also used.

Ussher (1913, p. 136) gave figures for the calorific values of two lignite samples from the Bovey Tracey area: for one sample 7907 British thermal units per pound (Btu/lb), 18.4 megajoules per kilogram (MJ/kg), as received, and 11 340 Btu/lb (26.4 MJ/kg) when air dried; and for the other 8393 Btu/lb (19.5 MJ/kg) and 10 534 Btu/lb (24.5 MJ/kg). Cawley and King (1946, p. 18) gave a value of 11 500 Btu/lb (26.7 MJ/kg) for a "typical" sample of lignite on a dry ash-free basis, Wynne (1947, p. 329) quoted about 6000 Btu/lb (13.9 MJ/kg) for freshly mined lignite from Blue Waters Mine, and Fox (1948, p. 129) 7570 Btu/lb (17.6 MJ/kg) for lignite used in boiler tests.

Fox (1948) described the use of mixtures of local lignite and sub-bituminous coal, each with high ash and moisture contents, for steam raising. One boiler was fired exclusively with pulverised lignite, and no serious difficulties were encountered in operation of the plant. Plans existed (Wynne, 1947; Fox, 1948) to make lignite briquettes, with an estimated calorific value of over 10 000 Btu/lb (23.2 MJ/kg), but it is not known whether production commenced.

Attempts to extract waxes from the lignites began with the German company which operated Blue Waters Mine just before the First World War (Cawley and King, 1945, p. 237), when a number of boreholes were made to assess reserves. Strahan and others (1920) noted that of nine samples of lignite and lignitic clay examined for their wax content, most yielded too little to be of commercial interest. . A wartime reassessment in 1941–43 (Cawley and King, 1945) indicated that the lignites of the Bovey Basin contained up to 5 per cent of crude wax soluble in benzene; it was thought that there were considerable obstacles to commercial production, "the strata being both discontinuous and intermingled with clay". Vahrman (1967, p. 406) noted that to be of value as a source of montan wax a brown coal must yield 10 to 15 per cent of an extract containing 60 to 90 per cent of the yellow or pale brown waxy substances, and that woody or resinous brown coals were of little value, even if the yield of extract were high.

The richest source of wax quoted by Cawley and King (1945) was a concentration of 5 per cent crude wax in the Big Coal; the associated beds of clay yielded about 1 per cent wax, and the 15 m of overlying interbedded lignite and clay an average of 2.3 per cent. The crude lignite wax contained 30 to 40 per cent of resinous material soluble in ethyl acetate. The properties of waxes from the Bovey Formation lignites are shown in (Table 14). An analysis of one sample of wax gave the following percentages by weight: ash 0.4; carbon 79.2; hydrogen 11.2; sulphur 0.6; nitrogen 0.2; and oxygen (by difference) 8.4 per cent (Cawley and King, 1945, p. 238).

The lignite remains a potential fuel for power stations. Other possible uses include gasification (Ussher, 1913, p. 136); Wynne (1947, p. 335) gave a reported calorific value of about 450 Btu/ft³ (36.9 MJ/m³) for town gas produced from Bovey lignite. A wide variety of chemical and liquid fuels have been produced from lignite deposits elsewhere in the world; 90 per cent of the German wartime production of synthetic liquid fuel was from lignite, mostly by hydrogenation of the low-temperature tar (Vahrman, 1967, p. 410).

There are no published modern analyses of Bovey Formation lignites; the available older analyses are shown in (Table 15).

Construction materials

Sedimentary rocks

In the past the Newton Abbot district supported an extensive quarrying industry over the area of outcrop of the Palaeozoic rocks. Numerous pits of small and medium size in the Devonian limestones demonstrate the extensive use of this material in the construction industry. Ornamental pink and grey limestone was formerly worked from Ransley Quarry [SX 8443 7018], and good roadstones were obtained from Ingsdon Hill [SX 8120 7340], Chercombe Bridge [SX 8351 7100], Grealy [SX 8804 7922] and Chudleigh [SX 8676 7868] quarries.

The remains of limekilns beside many of these quarries are evidence of the widespread production of burnt lime for use as fertiliser. Today, Devonian limestone is worked on a large scale at Stoneycombe [SX 8620 6714] by E.C.C. Quarries Limited. Most of the limestone is sold as aggregate, and other uses of the crushed rock include the manufacture of concrete blocks, ready-mixed concrete, coated stone, concrete aggregates and railway ballast.

The Upper Devonian Gurrington Slate furnished green and purple roofing slates from quarries at Wickeridge House [SX 7870 6970] and at Down [SX 7936 7051]. However, its resistance to weathering is less than that of Welsh or Delabole slate, and it was used only locally. Waste or 'rag' slate from the workings was used for the construction of farm buildings and walls.

Of the Carboniferous rocks, sandstone and chert have been extracted from small quarries and used locally for road metal and building stone.

The harder beds of the Permian breccias, especially those rich in small or medium-sized limestone fragments, have been extensively used locally as building stone. Small pits, all now disused, are scattered across the outcrop. Weathered breccias and the Triassic mudstones were used to make cob, a walling material composed of straw and mud thoroughly mixed wet and allowed to dry; cob is surprisingly resistant to weathering so long as the top surface is kept dry and the wall is undamaged.

A brick and tile industry was formerly developed on the mudstones of the Exmouth and Littleham formations. The last operational brickworks closed in 1969, not because the reserves were exhausted or the quality of the bricks were inferior, but because it could not compete with the huge-scale operations in the Jurassic clays of the Midlands. Mudstone beds of the Watcombe Breccia were used in the production of a terra cotta ware, known as 'Mier Ware'.

The Permo-Triassic rocks also furnished agricultural material; limestone cobbles from the Oddicombe Breccia were burnt as a source of lime, and 'marl' from pits in the various mudstones was applied to sandy soils.

The Budleigh Salterton Pebble Beds are worked at Blackhill Quarry [SY 029 857] by E.C.C. Quarries Limited, for aggregate, building sand, the production of ready-mixed concrete and pre-cast concrete products. The gravel is dug by mechanical excavator, then crushed and washed. A full range of sizes is separated by screening. Sand obtained during the washing is retained, but the silt is rejected.

Igneous rocks

Igneous rocks which have been important as sources of construction materials include the Dartmoor Granite, dolerites ('greenstones') and spilitic lavas. Roadside quarries in the granite provided road metal, blocks for dry stone walling and stone for house building. A larger working at Blackingstone, which provided ornamental and building stone, closed in 1969.

Greenstone outcrops were of greater importance as quarry sites. The good qualities of these hard, uniformly textured rocks have been recognised by the construction industry for many years. Numerous disused quarries exist along the length of the Middle Teign Valley, for example those in albite-dolerite at Trusham [SX 850 812] and Ryecroft Copse [SX 8432 8475]. The only active working in greenstone in the district is that of the Amey Roadstone Corporation at Crockham Quarry. After being blasted from the face, the rock is crushed and screened to produce a range of products, notably aggregate which is used for concrete production and road building. Large quantities of pre-cast concrete products are produced from the finer fractions of the crushed rock.

Spilitic lavas of Devonian and Carboniferous age have also been quarried for road metal. The sites at Scatter Rock [SX 822 856], near Christow, and Chipley [SX 8070 7218] are noteworthy.

Sand and gravel

The Upper Greensand formerly supported several sandpits; large workings flourished at Tower Wood [SX 8768 8567], Telegraph Hill [SX 9113 8342], Babcombe Copse [SX 869 766] and Smallacombe Goyle [SX 9220 7680]. At present the only production is from the upper section of Sands Copse Pit [SX 8674 7564], owned by Kingston Minerals Limited.

The Haldon Gravels have been widely exploited for road-building materials, and their outcrop is ringed with small workings. The only remaining active quarry is a small working [SX 9203 8145] near Colleywell Bottom, which produces calcined flints for the pottery industry. A larger quarry at Buller's Hill ceased operations in 1966.

Sands and gravels in the Bovey Basin are worked from the Aller Gravel, especially at the Royal Aller Vale Quarry and Zig Zag Quarry, south of Newton Abbot. A substantial production of sand and aggregate, from Tertiary strata and alluvial overburden, results from exploitation of the ball clay.

Metalliferous mining

Metalliferous mining has played a significant part in the industrial history of the Newton Abbot district. Vein systems associated with the Dartmoor Granite have variously yielded lead, silver, zinc and micaceous hematite, with smaller amounts of copper, spar minerals and iron ores (Figure 32). Small tonnages of manganese ores have been obtained from thin beds and nodular horizons in the Lower Carboniferous cherts.

Production of base metal ores flourished in the 19th century, but the ore field became moribund before the beginning of the 20th; mining of micaceous hematite, however, continued until the closure of Great Rock Mine in 1969. Details of all the more important workings are given by Dines (1956, pp. 725–749). Small-scale operations to recover cassiterite from alluvium have been tried in recent years (p. 172).

Lead, silver and zinc

Almost all the production of lead, silver and zinc resulted from exploitation of the Teign Valley lodes. These comprise a north–south mineralised zone, cross-cutting the Palaeozoic formations, which extends for over 8 km from 500 m south-west of Teign Village [SX 8348 8058] to 1250 m east of Bridford [SX 8290 8655] and beyond towards Dunsford. This zone has been explored by the following mines, now marked by spoil dumps: South Exmouth Mine [SX 8355 8080], Hennock Mine [SX 8365 8161], Frankmills Mine [SX 8363 8208], Wheal Exmouth [SX 8376 8302], Wheal Adams or Reed Mine [SX 8362 8365], Aller Mine [SX 8344 8395], and Bennah Mine, formerly Christow Silver Lead Mine [SX 8335 8465].

The mineralised structures consist of two main veins with subsidiary and branch lodes, all restricted to a belt of country which is probably little more than 200 m wide. From South Exmouth Mine to Aller Mine the two main veins, West Lode and East Lode, run for a distance of just over 3 km; stretches where the lodes were not worked reflect those zones in which the ore grade was too poor or the structure too narrow to be worth developing. Two short branch lodes occur at Frankmills Mine and several small veins to the west of West Lode at Aller Mine. From Aller Mine northward to Bennah Mine only traces of mineralisation have been found. Farther north, Bridford Barytes Mine [SX 829 866] produced "in excess of 370 000 tons" of barytes (Vipan, 1959, p. 346).

The lodes run almost due north from South Exmouth Mine to the northern end of Wheal Exmouth, whence they trend north-north-west. At Bridford the strike swings round still farther to become nearly north-west. The lodes are roughly parallel to the margin of the Dartmoor Granite, and approximately coincident with the outer limit of the metamorphic aureole. Along the length of their outcrop they maintain a fairly constant distance apart, although they tend to converge slightly with depth. Their attitude is essentially vertical, with angles of dip everywhere steeper than 70°.

The chief ore mineral of the veins is galena, which occurs, commonly with sphalerite, in a matrix consisting variously of quartz, baryte, siderite and clay 'gouge' or 'fluccan'. At the South Exmouth, Hennock and Frankmills mines the gangue minerals seen on the dumps are predominantly baryte and siderite, or dark brown limonitic alteration products of the latter mineral. To the north, at Wheal Exmouth and Wheal Adams, in the Carboniferous cherts and shales, there is more quartz and less siderite. Brecciated fragments of country rock enclosed in quartz or, more rarely, in siderite are common at each locality. A body of iron-stained barytes up to 7 m wide, with some quartz and siderite, forms a cliff-like feature 120 m long to the south-west of Beckhams Tree Cross [SX 8360 8331]. It appears to be the East Lode of Wheal Exmouth, which dips steeply east.

Specimens collected from the mine dumps show the form of occurrence of galena to vary from well-developed cubic crystals, commonly a centimetre or more across and clustered in aggregates, to a dissemination of fine crystals in the gangue; stringers of the mineral in quartz and wallrock are also present. Sphalerite occurs as blebby, commonly rather rounded aggregates of a distinct deep red to reddish brown colour, sometimes referred to as 'ruby blende'. Minor amounts of pyrite and chalcopyrite have been noted, particularly at South Exmouth and Frankmills mines; at the latter site fluorspar, cerussite and stibnite have also been recorded. Schmitz (1973), in a historical survey of the Teign Valley mines, noted the sale of small parcels of copper ore from Wheal Exmouth and Wheal Adams, and the Mining Journal of 1857 recorded a grey silver-copper ore which assayed 2390 g/tonne (78 troy oz/ton) silver. This mineral was probably tetrahedrite, the presence of which in material collected from the dumps at Wheal Exmouth has been confirmed by Mr P. R. Simpson. Nickel and cobalt ores in the veins of Wheal Adams were also noted by Schmitz (1973), quoting a Mining Journal report of 1853.

Little is known concerning the distribution of ores within the lodes. Examination of the available mine plans shows that the pattern of stoping was patchy and irregular; presumably the galena was more or less confined to ore shoots of limited size and irregular shape. Vipan (1959, pp. 350–351), quoting an unpublished report by Sir Kingsley Dunham, suggested stratigraphical control of mineralisation in the Bridford area. Cherts and mudstones of the Teign Chert have been preferred to the Ashton Shale and Hyner Shale as host rocks, and Vipan drew attention to the possibility that a virgin orebody might exist between Bridford Mine and Wheal Exmouth. It may be that early trials in this area were incorrectly placed and that future prospecting could lead to more positive results.

Silver, recovered from the smelted lead, formed an important part of the mines' revenue, and some 12 032 258 g (386 852 oz) are known to have been produced from the Teign Valley mines in the 19th century. The distribution of this element is of some interest. Schmitz (1973), quoting the Mining Journal of 1848, stated that at Wheal Adams the silver values increased with depth in the following manner: 276 g/tonne (9 troy oz/ton) at the 18-fathom level, 490 g (16 troy oz) at the 28-fathom level, 643 g (21 troy oz) at the 40-fathom level and 796 g (26 troy oz) between the 40-fathom and 50-fathom levels. Figures published by Hunt (Mineral Statistics 1857–1880) suggest a similar trend at Frankmills Mine, where the silver content of the ore rose from 118 g/tonne (4 troy oz/ton) in 1860 to 804 g/tonne (26 troy oz/ton) in 1868. As the workings were deepened throughout this period it may be that the silver content increased with depth, and perhaps this form of zoning is a general feature of the lobe system.

Provis (1874, p. 70) gave figures for the silver content of the lead concentrates from Wheal Exmouth and Frankmills Mine as follows:

• Wheal Exmouth
• No. 1 concentrate 82.5 per cent Pb, 358g/tonne (11.75 troy oz/ton) Ag
• No. 2 concentrate 47.0 per cent Pb, 274 g/tonne (9 troy oz/ton) Ag
• Frankmills
• No.1 concentrate 70.5 per cent Pb, Mine 762 g/tonne (25 troy oz/ton) Ag
• No.2 concentrate 60.5 per cent Pb, 770 g/tonne (25.25 troy oz/ton) Ag

In each case the lower grade lead concentrate was relatively enriched with respect to silver. This would seem to indicate that part at least of the silver was held in a separate mineral phase, rather than being entirely in solid solution in galena, and that this separate phase was more amenable to concentration by the dressing process responsible for the lower grade concentrate. Tetrahedrite could well be a suitable argentiferous species contributing to this effect.

(Table 16) summarises the mineral production of the Teign Valley lodes and includes the output of the small Silverbrook Mine at Ilsington [SX 7890 7583], the only other producer of lead, silver and zinc in the district. Here, in country rocks of the Combe Shale and Teign Chert, two north-east-trending lodes which converged at depth yielded a small quantity of galena and a larger amount of zinc blende. The total mineral production was probably considerably greater than is indicated in this table; no records are available of earlier working, although there is some evidence of activity in the 18th and 19th centuries.

Interest has been shown, from time to time, in the Middle Teign Valley as a prospect for future development. While it is unlikely that economic quantities of base metals could be extracted on their own, the present world demand for barytes, used as a component of drilling mud, has revived the need for reassessment of potential reserves.

Copper

Apart from the small amounts of copper produced at Wheal Exmouth and Wheal Adams, the only mine which worked this metal was Yarner or Yarrow Mine [SX 7829 7836] at Yarner Wood. Dines (1956, p. 732) stated that 2337 tonnes of 3.75 per cent copper ore were produced between 1858 and 1865. The principal ore mineral is chalcopyrite, which occurs with arsenopyrite and pyrite in a quartz-chlorite vein.

Manganese

Manganese ores were worked spasmodically at several localities in the district from the end of the 18th century until about 1875. The mines are located on both sides of the Middle Teign Valley at Doddiscombsleigh [SX 8450 8635]; [SX 8530 8608]; [SX 8583 8615] and Ashton [SX 8562 8510]; [SX 8406 8433]; [SX 8535 8420]; [SX 8485 8335], at Riley [SX 8448 8024] east-south-east of Hennock and at Stancombe [SX 8013 7393] near Bickington.

The deposits appear to be restricted to the Teign Chert, and trend parallel to the strike of the host rocks. True lode structures are not present; the ore occurs as beds and nodules interstratified with the cherts and as impregnations in the chert bands. The beds can be traced for distances varying from 450 m to 800 m, but whether they pinch out or are continuous from one set of workings to another is a matter for conjecture, since exposure is too poor to justify any firm conclusion. Mineralogically the ores consist of pyrolusite, psilomelane, wad, rhodonite and manganite. Both the manganese ores and the jasperised cherts associated with them show brecciation and quartz veining.

It appears that the ores were worked by adits or pits wherever a rich patch was found, but never to a great depth. Very little is known of the production from these workings, most of which appear to have been little more than trials. Riley Mine is recorded as having produced 120 tonnes of manganese ore in 1875, and an operation called the Christow Manganese Mine (Schmitz, 1973, pp. 13, 118), probably the working at Hill Copse, [SX 8406 8433], sold 2500 tonnes of ore between 1829 and 1841. The ores were mainly used as decolourising agents in glass-making, and as reagents in the preparation of chlorine, used for bleaching (Collins, 1912, p. 273).

Micaceous hematite

Micaceous hematite, known locally as 'shining ore', is the principal component of a group of veins in the granite near Lustleigh and Hennock. The material is a soft, grey, unctuous variety of specular hematite which, owing to its foliaceous nature and fine particle size, leaves an almost oily smear on the skin when handled. De la Beche (1839) stated that the ore was sold as a writing sand at £3 to £8 per ton. The main period of production extended from the latter part of the 19th century to the closure of the largest working, Great Rock Mine, in 1969. Throughout this period the ore was mainly used as the base of a type of anti-corrosion paint particularly suitable for the preservation of metal. Other industrial applications of micaceous hematite include its use in the manufacture of welding rod coatings and in certain types of specialised lubrication (Michell, 1944).

The ore occurs as irregularly developed interconnecting thin veins, within steep northerly dipping zones of highly altered granite; the alignment of these zones varies from east-north-east–west-south-west to east–west. Individually, the veins range in width from a few millimetres to over a metre, the workable stretches being of limited extent and patchy distribution. Underground the lodes are observed to be slightly irregular in trend, and commonly displaced by minor north–south faults (Henson, 1956) with throws varying from a few millimetres to a metre or more. Vertically slickensided footwalls are common along the veins.

Specimens examined by Dr J. R. Hawkes were composed predominantly of finely divided micaceous hematite, with irregular patches, veins and stringers of quartz, scattered patches of sericite and muscovite, aggregates of pyrite crystals, and accessory schorlite. In colour and appearance the ore varies from place to place; at Hennock it is soft, of very fine particle size, and has a distinct purple tint, but near Lustleigh it is harder, steel grey and more massive, containing a high proportion of coarse specular hematite. All varieties of the ore produce a distinctive cherry-red streak. Pyrite, typically in well-formed pyritohedra in places 25 mm across, is most common in the lodes at Great Rock Mine. Martin (1895) recorded that a few small clusters of cassiterite crystals were present in the hematite lodes of the Lustleigh area.

Massive quartz leaders are absent from the lodes of the Hennock–Lustleigh mines, and it seems from examination of thin sections (E42041), (E42042), (E42043) that the hematite may have resulted largely from replacement of sericitic and clay constituents developed along zones of intense hydrothermal alteration in the granite. This could explain why the hematite found at Great Rock and in some neighbouring mines is so much finer grained than that developed in the more massive quartz-tourmaline lodes of the Birch Tor and Chagford district to the north.

The zones of hydrothermally altered granite adjacent to the hematite veins are commonly several metres wide. Kaolinisation is usually evident and the development of chlorite close to the vein walls has imparted a distinct green colour to the rock. Micaceous and specular hematite are commonly disseminated in the wall rocks. A specimen of completely altered big-feldspar granite (MR27805), recovered from the dumps at Great Rock Mine, has retained its original texture. In addition to secondary kaolinite, sericite, chlorite and quartz, this specimen contains dispersed micaceous hematite and pyritohedra of pyrite. At Great Rock Mine the easternmost parts of some workings at the lower levels pass out of the granite into contact-metamorphosed sediments, where the lodes are little more than stringers and no attempt has been made to carry development farther to the east.

Most of the mines which worked micaceous hematite in the area were small, many of them little more than trials. Only Great Rock and Kelly mines maintained production into the second half of the 20th century. Details of the mine workings are given in Dines (1956, pp. 725–727), but locations and such production figures as are available are set out in (Table 17), together with some additional notes. Mr M. Atkinson of the University of Exeter has kindly supplied details of dates of working, drawn from Mineral Statistics. There are no records of production for periods of working earlier than the late 19th century.

The mines listed in (Table 17) began as independent enterprises, but by 1913 the Ferrubron Manufacturing Co., owners of Great Rock Mine, had gained control of all workings. It appears that the harder ore of Hawkmoor, Plumley and Shaptor mines was considered less suitable for paint manufacture, and so production was concentrated on Great Rock and Kelly mines. The closure of Great Rock Mine in 1969 was due to lack of reserves and rising costs of working.

Preparing the ore for use as a pigment was a relatively simple procedure, which varied from a basic pattern only in detail. Details of the various dressing processes are given by Michell (1945). Usually the material was washed over a screen to separate the soft, fine hematite, after which the coarser material was hand sorted to reject oversize granitic waste. The remaining coarse ore was lightly stamped or crushed in a ball mill to break down the more resistant aggregates. Separation of hematite from sandy waste was then achieved by various combinations of tables, jigs and sluice boxes. After settling in tanks the dressed ore was dried and put in containers for sale. At Great Rock Mine it was necessary to calcine the final product at low red heat to remove traces of sulphide, which is undesirable as an impurity in paint. A typical analysis of micaceous hematite produced in Great Britain for use in paint manufacture is given by Johnstone and Johnstone (1961, p. 442):

Fine quartz, white mica and clay minerals constitute much of the impurities.

Other iron ores

Between 1872 and 1880, 185 tonnes of spathic iron ore (siderite) were sold from Frankmills Mine; an analysis quoted by Ussher (1913, p. 130) gave an average iron content of 38.20 per cent. Ores variously termed limonite and brown hematite were sold from the micaceous hematite mines to the extent of some 1027 tonnes between 1897 and 1907. These were probably extracted from the weathered parts of the lodes where hydration of the micaceous hematite had occurred. Wolborough Mine [SX 8528 6989], mapped in East Ogwell Limestone, produced 1260 tonnes of brown hematite in 1870 and 1874; this material was apparently used as ochre in the preparation of paint.

Tin

Traces of cassiterite occur in the micaceous hematite lodes near Lustleigh (Martin, 1895), but the Dartmoor Granite in this district does not carry workable tin deposits. There is evidence of early trials by tin streamers and miners, but no records of these operations survive. In recent years low-grade alluvial deposits of the River Teign and its tributaries have been prospected and worked.

Sites 200 m west and 200 m south-east of Gale Farm [SX 7945 7150], in the Kester Brook valley, yielded tin from a strip of alluvium about 30 m wide and 3 m deep. The alluvium consists predominantly of silt and clay with large angular cobbles and boulders of Devonian slate, limestone and volcanic rock; well-rounded particles up to coarse gravel size of hornfels, granite, tourmaline, vein quartz and tin ore occur sparsely in the sand.

The sediment was excavated by a dragline and fed into a unit which washed and disaggregated the material prior to removal of the coarse fractions on a set of vibrating screens. Sand and gravel sizes were separated from clay and silt by a hydrocyclone classifier and fed into two fixed screen jigs. The crude, heavy concentrate thus obtained was removed from the site and upgraded by further jigging and tabling. After settling the slimes in a pit, into which the coarse fraction was periodically tipped, the ground was levelled, the topsoil restored and the plant moved on. This operation ceased late in 1975.

Mr J. H. Walbeoffe-Wilson has supplied details of his earlier working at Longstone Bridge [SX 7957 7136], 200 m south-east of Gale Farm. The alluvium was worked to a depth of about 2 m and yielded between 0.38 and 0.63 kg/ m³ cassiterite. The particles of tin ore varied in size from 10 mm to fine sand, none being recoverable from the finer fractions. Examination of a number of the coarser particles showed them to consist of aggregates of yellowish brown to dark reddish brown cassiterite with various amounts of quartz, tourmaline and chlorite. After extraction of the tin ore, the site was returned to agricultural use.

Cassiterite occurs in the superficial deposits of the Bovey Basin. It is usually concentrated towards the base of the beds in coarse sand and gravel. At the Pinsent Claypit, the alluvium overlying the Tertiary deposits was treated in a plant producing tin concentrate, sand and gravel. The tin-bearing sediment, which carried from 0.2 to 2.0 kg/ m³ cassiterite, was selectively removed from the pit and screened. Material coarser than 5 mm diameter was graded for sale as gravel and aggregate, and the finer fractions were pumped to a hydrocyclone which removed silt and clay. The clean sand fraction was then treated in fixed screen jigs, which produced a primary concentrate containing about 30 per cent of cassiterite; tailings went to the sand stockpile. Rewashing of the primary concentrate in the jigs upgraded the final product to about 60 per cent cassiterite, the impurities consisting mainly of hematite, tourmaline, feldspar and ilmenite.

The Aller Gravel has also been examined as a low-grade source of cassiterite (Scrivener and Beer, 1971). Sands from the Royal Aller Vale Quarry and Sands Copse pit contain up to 1.37 kg/ m³ cassiterite in the 25 to 150 B.S.S. mesh size fraction. This size fraction, however, is a small part of the bulk material available and no commercial exploitation has been attempted.

Appendix 1 Devonian conodont faunas

Lists of stratigraphically significant conodont forms are given below. Identifications by Dr S. C. Matthews (Waters, 1975) and Dr B. W. Riddolls (1970b) have been checked and revised by Dr Matthews. Bar forms, which have no bearing on questions of age, have not been examined in detail. None of the icriodids have been identified to specific level. Specific determinations of polygnathids are given only if they have an influence on the estimate of age. The faunas have been dated (see (Table 3)) by reference to Ziegler's (1962, 1971) zonal scheme, noting the modifications of Sandberg and Ziegler (1973).

Within each heading the localities are listed from west to east. Specimen registration numbers are given after the locality details. Each identifies a 32-cavity slide. All of the material studied is stored in the Devonian collections of the Institute of Geological Sciences.

Nordon Slate

1 Small pit [SX 7989 6669] 201 m south-south-west of Lee Farm, near Broadhempston. AD 801. Polygnathus kluepfeli, P. cf. xylus, Icriodus sp., bars. Early Givetian.

2 Disused quarry [SX 8051 6649] 410 m north-east of Broadhempston church. AD 802. Palmatolepis transitans, P.? cf. disparalvea ((Plate 6), figs. 16, 17), Palmatolepis? ((Plate 6), fig. 19), Polygnathus asymmetricus, P. cristatus ((Plate 6), fig. 18), P. decorosus, P. cf. dengleri?, P. pennatus, Schmidtognathus sp., bars. Upper hermanni-cristatus Zone or lowermost asymmetricus Zone–early Frasnian.

3 Small pit [SX 8083 6682] at Waterford Cross, 457 m south-west of Coppa Dolla. AD 803. Ancyrodella cf. rotundiloba?, Ancyrodella?, Palmatolepis cf. transitans, Polygnathus decorosus, Icriodus sp., Spathognathodus sp., bars. asymmetricus Zone–early Frasnian.

4 In woodland [SX 8102 6746] 90m west of Collacombe Bridge. AD 805. Polygnathus linguiformis, P. varcus ((Plate 6), figs. 20, 21), Polygnathus sp., bars. Probably varcus Zone–late Givetian.

5 At [SX 8137 6699] 320 m south-east of Coppa Dolla. AD 804. Polygnathus pseudofoliatus, P. cf. xylus, Ozarkodina bidentata, Icriodus sp., bars. kockelianus (unrestricted) Zone–late ensensis Zone–late Eifelian or early Givetian.

6 Road cutting [SX 8160 6925] near Heathfield Farm. AD 806. Icriodus sp., bars. No worthwhile age estimate possible.

7 Exposure in field [SX 8185 6993] 183 m south of West Ogwell Convent. AD 807. Polygnathus linguiformis, P. cf. varcus. Probably varcus Zone–late Givetian.

8 Dainton Quarry [SX 8510 6597] 732m south of Dainton. AD 808. Polygnathus linguiformis, P. varcus, Ozarkodina brevis, bars. varcus Zone–late Givetian.

9 Disused quarry [SX 8753 6698] at junction of Huxner Road and Whilborough Road, 2.4 km east of Dainton. AD 809. Polygnathus linguiformis, P. varcus, Icriodus sp., bars. Probably varcus Zone–late Givetian.

Kingsteignton Volcanic Group

10 Pit [SX 8871 7315] north-east of Ware Barton. AD 3769. Tortodus kockelianus, P. linguiformis, P. sp., bars. kockelianus (unrestricted) Zone–late Eifelian.

11 Quarry [SX 8968 7375] north of Forder Cross. AD 3770. Polygnathus costatus patulus, P. cf. robusticostatus, P. sp., bars. Upper patulus to C. costatus Zone–early Eifelian.

Denbury Crinoidal Limestone

12 Disused quarry [SX 8203 6912] 366 m north-west of Denbury Post Office. AD 813. Polygnathus sp., bars. No worthwhile age estimate possible.

13 At [SX 8208 6691] 119m north-east of Torbryan church. AD 814. Polygnathus sp., bars. No worthwhile age estimate possible.

14 Deer park [SX 8240 7040] south-west of Chercombe Bridge. AD 812. Polygnathus linguiformis, bars. No worthwhile age estimate possible.

15 Temporary exposure at Chercombe Bridge [SX 8331 7113]. AD

810. Polygnathus cf. linguiformis, P. sp., Icriodus sp.No worthwhile age estimate possible.

Pulsford Limestone

16 Disused quarry [SX 8077 6795] 247 m south-south-east of Pulsford. AD 819. Polygnathus linguiformis, P. cf. varcus, Icriodus sp., bars. Probably varcus Zone–late Givetian.

17 Exposure [SX 8078 6790] 6 m above footpath 293 m south-south-east of Pulsford. AD 818. Polygnathus branti, P. linguiformis, P. cf. varcus, Icriodus sp., bars. varcus Zone–late Givetian.

18 At [SX 8132 6738] 320m north-east of Coppa Dolla. AD 817. Polygnathus decorosus, P. linguiformis, P. varcus, bars. Probably varcus Zone–late Givetian.

19 At [SX 8149 6727] 411 m east of Coppa Dolla. AD 816. Polygnathus linguiformis, P. cf. varcus, Icriodus sp., bars. Probably varcus Zone–late Givetian.

20 Pit [SX 8158 6690] 400m west of Torbryan church. AD 815. Polygnathus cf. varcus, P. sp., Icriodus sp., bars. Possibly varcus Zone–late Givetian.

Chercombe Bridge Limestone

21 Exposure [SX 8321 7089] 210m south-west of Chercombe Bridge. AD 820. Polygnathus linguiformis, P. sp., Ozarkodina cf. bidentata, bars. Possibly late Eifelian.

22 Wood House Quarry [SX 8903 7384] west of Bishopsteignton. AD 3771. Polygnathus linguiformis, P. varcus, P. sp., Ozarkodina brevis, Spathognathodus sp., Icriodus sp., bars. varcus Zone–late Givetian.

East Ogwell Limestone

23 Emblett Hill Borehole [SX 8399 7065]. At 12.5 m depth. AD 841. Polygnathus asymmetricus ovalis, P. decorosus, Spathognathodus sp., bars. asymmetricus Zone–toIα/β.

24 Disused quarry [SX 8413 7002] 366 m east of East Ogwell. AD 811. Polygnathus decorosus, P. varcus, P. sp.Probably varcus Zone–late Givetian.

25 Disused Ransley Quarry [SX 8443 7018]. AD 821. Ancyrodella nodosa, Ancyrognathus triangularis, Palmatolepis subrecta, Polygnathus webbi, P. sp., Icriodus sp., Nothognathella sp., Belodella sp., bars. triangularis Zone to lower gigas Zone–mid Frasnian (toIγ) .

26 North of Ransley Quarry [SX 8448 7036]. AD 822. Ancyrodella curvata ((Plate 6), figs. 1, 2, 7), Ancyrognathus triangularis ((Plate 6), figs. 3–6), Palmatolepis gigas, P. subrecta ((Plate 6), figs. 10, 11, 13–15), Polygnathus ancyrognathoideus, P. webbi, P. sp., Icriodus sp., Nothognathella sp., bars. Lower gigas Zone–mid Frasnian (toIγ).

27 Small disused quarry [SX 8518 6971] north-west of Abbotskerswell. AD 824. Polygnathus linguiformis, P. varcus, bars. Probably varcus Zone–late Givetian.

28 Disused Grange Quarry [SX 8520 6860] near Abbotskerswell. AD 825. Polygnathus cf. linguiformis, P. varcus, Icriodus sp., bars. Probably varcus Zone–late Givetian.

29 East face of Wolborough Quarry [SX 8524 7041]. AD 823. Polygnathus linguiformis, Spathognathodus brevis ((Plate 6), fig. 30), bars. Probably varcus. Zone–late Givetian.

30 Disused quarry [SX 8560 6755] in Gotemhill Wood. AD 828. Polygnathus decorosus, P. varcus, Spathognathodus brevis, bars. varcus Zone–late Givetian.

31 Exposure [SX 8585 6835] in Slade Lane, 329m south-east of Court Grange. AD 826. Polygnathus linguiformis, P. cf. varcus, Spathognathodus brevis, bars. Probably varcus Zone–late Givetian.

32 Thinly bedded limestones in Stoneycombe Quarry [SX 8620 6704]. AD 827. Polygnathus latus ((Plate 6), fig. 29), P. aff. latus ((Plate 6), fig. 24), P. linguiformis ((Plate 6), figs. 22, 26, 28), P. linguiformis cooperi ((Plate 6), fig. 25), Tortodus variabilis ((Plate 6), figs. 23, 27), Ozarkodina bidentata, Icridus sp., bars. Late ensensis Zone–early Givetian.

33 Kerswell Down Quarry [SX 8735 6768]. AD 829. Polygnathus decorosus, P. linguiformis, P. varcus, P. sp., Ozarkodina brevis, bars. varcus Zone–late Givetian.

34 Quarry [SX 8777 7473] south of Torhill Cottages. AD 3772. Ancyrodella rotundiloba [juv .], Polygnathus spp. Probably low in to I.

35 Quarry [SX 8820 7500] south-west of Whiteway Barton. AD 3773. Ancyrodella cf. rotundiloba, Palmatolepis sp., Polygnathus sp., Icriodus sp., bars. Possibly asymmetricus Zone–toIα?

36 Strongs Cover [SX 8880 7519] east of Whiteway Barton. AD 3774. Polygnathus webbi, P. sp.No worthwhile age estimate possible.

Luxton Nodular Limestone

37 Emblett Hill Borehole [SX 8399 7065]. At 3.76m to 4.57 m depth. AD 840. Palmatolepis distorta distorta, P. glabra pectinata, P. glabra prima, P. gracilis gracilis, P. marginifera, P. minuta minuta, P. perlobata subsp., Polygnathus ex gr. "nodocostata", Spathognathodus sp., bars. marginifera Zone–near toII/toIII boundary.

38 Rydon Ball Farm Borehole [SX 8436 6929]. At 97.23 m to 98.45 m depth. AD 836. Palmatolepis distorta, P. minuta minuta, P. perlobata sigmoidea?, Polygnathus sp., Scaphignathus velifer, bars. Lower or middle velifer Zone–toIIIα/β.

39 Rydon Ball Farm Borehole [SX 8436 6929]. At 151.49 m depth. AD 837. Palmatolepis crepida crepida, P. glabra prima, P. minuta minuta, P. quadrantinodosa quadrantinodosa, P. quadrantinodosalobata, P. cf. regularis, P. rhomboidea, Icriodus sp., bars. The youngest forms present indicate the lower marginifera Zone–high toll. Note: The fauna appears to be a mixed one. It includes forms which indicate the crepida (toIIα) and rhomboidea (low toIIβ) zones.

40 Rydon Ball Farm Borehole [SX 8436 6929]. At 156.97m to 157.07 m depth. AD 838. Palmatolepis glabra lepta, P. glabra pectinata, P. marginifera, P. minuta minuta, Polygnathus sp., Polylophodonta sp., Icriodus sp., bars. marginifera Zone–near toII/toIII boundary.

41 Rydon Ball Farm Borehole [SX 8436 6929]. At 191.16m depth. AD 839. Palmatolepis glabra lepta, P. glabra pectinata, P. glabra prima, P. marginifera, P. minuta minuta, Polygnathus spp., Spathognathodus sp., Icriodus sp., bars. marginifera Zone–near toII toIII boundary.

42 Temporary exposure in a trench [SX 8483 6895] north-west of Abbotskerswell. AD 832. Palmatolepis glabra subspp., P. gracilis gracilis, P. marginifera, P. minuta schleizia, P. perlobata schindewolfi, P. quadrantinodosa inflexa, Polygnathus sp., Spathognathodus cf. strigosus, bars. Lower marginifera Zone of Sandberg and Ziegler (1973)–high toII.

43 Exposure [SX 8508 6942] 0.8 km south of Newton Abbot Hospital: Kiln Wood beds. AD 831. Ancyrodella curvata, A. nodosa, A. sp., Ancyrognathus triangularis, Palmatolepis delicatula delicatula, P. delicatula clarki, P. gigas, P. subrecta (including forms transitional toward P. quadrantinodosalobata), P. triangularis, P. sp., Polygnathus decorosus, P. webbi, Icriodus sp., bars. The youngest forms present indicate the middle and upper Palmatolepis triangularis Zone–Ziegler's (1971, chart 4) post-toIδ unit. Note: The fauna is a mixed one. It includes forms which indicate the Ancyrognathus triangularis or lower gigas Zone, i.e. toIγ.

44 Exposure [SX 8526 6985] 270 m south of Newton Abbot Hospital. AD 830. Ancyrodella nodosa, Ancyrognathus triangularis, Palmatolepis punctata, Polygnathus decorosus, Polygnathus sp. [indet.], Icriodus sp., bars. Ancyrognathus triangularis Zone or low gigas Zone–toIγ.

45 Trench [SX 8763 7712] at Well Farm. AD 3794. Palmatolepis distorta distorta, P. glabra lepta, P. cf. helmsi, P. minuta schleizia, P. perlobata schindewolfi, P. marginifera, P. rugosa ampla, Polygnathus sp., Spathognathodus bohlenanus, S. strigosus, bars. Probably styriacus Zone–near toIV/toVboundary. In this case, forms such as P. glabra lepta and (most clearly) P. marginifera must be regarded as reworked.

46 Trench [SX 8764 7711] at Well Farm. AD 3793. Palmatolepis distorta manca, P. glabra lepta, P. gracilis gracilis, P. gracilis subsp., P. minuta minuta, P. perlobata schindewolfi, P. perlobata sigmoidea, P. rugosa subsp., Polygnathus sp., Spathognathodus stabilis, S. sp., bars. Low and mid styriacus Zone–near toIV/toVboundary. Some indication of reworking: P. distorta manca is younger than P. minuta minuta and P. glabra lepta.