Content and licensingview original scan buy a printed copy

Geology of the country around Montgomery and the Ordovician rocks of the Shelve area. Memoir for 1:50 000 Geological Sheet 165 with part of sheet 151 (Welshpool) (England and Wales)

By R Cave and B A Hains

Bibliographic reference Cave, R, and Hains, B A. 2001. Geology of the country around Montgomery and the Ordovician rocks of the Shelve area. Memoir of the British Geological Survey, Sheet 165 with part of Sheet 151 (Welshpool) (England and Wales).

British Geological Survey

Geology of the country around Montgomery and the Ordovician rocks of the Shelve area Memoir for 1:50 000 Geological Sheet 165 with part of sheet 151 (Welshpool) (England and Wales)

By R Cave and B A Hains

London: The Stationery Office 2001. © NERC copyright 2001 First published 2001. ISBN 0 11 884552 7. Printed in the UK for The Stationery Office TJ4998 C6 07/01.

The grid used on the figures is the National Grid taken from the Ordnance Survey map. (Figure 2) is based on material from Ordnance Survey 1:50 000 scale maps, numbers 125, 126, 136 and 137. © Crown copyright reserved. Ordnance Survey Licence No. GD272191/2001

Acknowledgements

The account of the Ordovician, Llandovery and Wenlock rocks was written by Dr R Cave; Precambrian, Přídolí and economic geology was written by Dr B A Hains. Both are authors of the Ludlow and Quaternary sections. Dr Cave wrote the structure chapter which was helpfully reviewed by Dr W T Pratt. This memoir has been written by the authors in their retirement, at their own expense and on a non-contractual basis.

The palaeontology and its assessment for the memoir have been provided largely by Drs D E White and A W A Rushton, with the assistance of Mr S P Tunnicliff and Dr J A Zalasiewicz. Dr Cave is responsible for part of the palaeontological account of the Ordovician and Llandovery rocks. Reports on the palynology of some Ordovician and Silurian rocks were provided by Drs H F Barron and S G Molyneux.

The Department of Industry (1982 to 1984) funded a geological survey of the mineralised Ordovician strata cropping out in the area around Shelve, which lies partly within the Montgomery district and partly the Welshpool district (Sheet 151) to the north; this was carried out by Dr B D T Lynas.

The regional gravity data have been intensified in a survey by Mr R Carruthers and localised seismic, gravity and magnetic surveys have been conducted by the Regional Geophysics Group (BGS) and by University College, Cardiff, mainly in the north-east part of the district. The account of contemporary seismicity was provided by Mrs M E A Ritchie and geochemical data compiled by Mr R J Merriman.

The memoir was compiled and edited by Dr Audrey A Jackson. Figures were produced in BGS Keyworth.

Notes

Preface

Although rather neglected by visitors who tend to focus their attention on the more majestic scenery of the Snowdonia National Park farther west, the Montgomery district encompasses some delightful countryside. Precambrian, Ordovician and Silurian rocks crop out in the region and there are extensive glacial deposits.

The older rocks are host to lead, zinc and barite mineralisation and have been worked intermittently since Roman times. During the course of this survey, the Department of Industry (now the Department of Trade and Industry) funded a geological survey of the West Shropshire (Shelve) mining field as part of its Mineral Reconnaissance Programme of the 1980s. A summary of the results of this work have been included here.

The Ordovician rocks present a complex association of sedimentary, volcanic, volcanogenic and intrusive rocks. The detailed stratigraphy and palaeontology presented here provide a valuable addition to the work of BGS farther west in the Snowdon, Harlech and Cadair Idris districts.

Silurian rocks dominate this part of the Welsh Borderland and occupy most of this district. The memoir details the change in lithofacies of the Silurian rocks from west to east, tracing the changes from a basin and basin-margin depositional environment to a shelf environment. An interpretation of the conditions of deposition and palaeontology is given, as well as a comprehensive review and update of the wider palaeogeographical implications.

The account of the Quaternary geology in this memoir now updates that of the more general accounts of earlier workers published in the 1930s and provides essential information for planners working in the area.

An interpretation of the gravity, magnetic and seismic data is included. The Bishop’s Castle earthquake with its epicentre in the district, was one of the largest to affect onshore UK this century and an account of historical and recent seismicity is included.

David A Falvey, PhD Director, British Geological Survey, Kingsley Dunham Centre Keyworth, Nottingham, NG12 5GG

Geology of the country around Montgomery and the Ordovician rocks of the Shelve area — summary

This memoir describes the geology of part of the Welsh Borderland. Precambrian and Ordovician rocks of the Shelve crop out in the north, but Silurian rocks underlie most of the district providing an insight into the development and infill of the Lower Palaeozoic Welsh Basin. Younger rocks are not represented in the district, though there is an extensive cover of superficial Quaternary deposits.

Precambrian rocks crop out in the north-east, in the inverted limb of a north-east-trending syncline. The oldest rocks are Uriconian (650 my) lavas and volcaniclastic rocks. The Longmyndian Supergroup is thought to be younger. It consists of a thick sequence of conglomerate, sandstone and siltstone. Evidence from adjacent areas suggests that folding and faulting of these Precambrian rocks predate the Cambrian.

The Welsh Basin was a depocentre in which a great thickness of sediment accumulated throughout Ordovician and Silurian times. The early history of the basin is very different from the later history; it is appropriate to consider the Lower Palaeozoic record as the evolution and successive extinction of two separate basins, one of Ordovician age and one of Silurian age. The development of both basins was greatly influenced by movements in the basement, and a north-north-east-trending lineament, the Severn–Tywi Lineament, running through Welshpool to Llandovery, played a significant role.

Ordovician rocks crop out in the Shelve area and Forden Inlier. During the Tremadoc, uniform conditions prevailed over a wide area and a great thickness of sediment accumulated in a shelf-sea environment. This is the Shineton Shale Formation. The base is faulted out against Precambrian rocks; the top passes up apparently unconformably into Arenig rocks, although a brief period of emergence may have occurred for the lowest sandstones of the Arenig are interpreted as near-shore deposits. The succeeding Arenig, Llanvirn and Llandeilo deposits, attest to a deepening marine environment. In the Shelve area, a thick sequence of volcanic rocks, lavas, tuffs and tuffites, are largely Llanvirn and partly Caradoc in age. Associated intrusive rocks are mainly basic in composition. Late Ordovician rocks are missing from the district. This is largely a result of pre-Silurian erosion, but may be partly a result of non-deposition, in very late Ashgill times, coincident with the nadir of a global marine regression.

Marine conditions were re-established during Llandovery times. An extended transgression reached the west of the district in Rhuddanian times and the entire area was submerged in the late Aeronian and Telychian times. Early Silurian rocks are thus of a shallow marine facies, but by Wenlock times a submarine incline had formed which introduced a basinal environment to the western portion of the district while leaving a shoaling or ‘shelf’ sea in the east. In Ludlow times, clastic detritus prograded across the shelf from the south-east, infilling the basin. By Přídolí times both shelf and basin were reduced to a uniformly flat, brackish to fluvial coastal plain.

The period of time between the Přídolí and the Quaternary is not represented by strata in the district, due to erosion or nondeposition. Extensive Quaternary deposits are of glacial, periglacial and postglacial origin.

The West Shropshire (Shelve) mining field has been worked intermittently since Roman times for lead, zinc. barite and calcite, but at present there are no significant commercial works within the district.

The district is notable for a number of earthquakes, of relatively high magnitude by UK standards, and a brief history of recent seismic events is included. In addition, an interpretation of the gravity and magnetic surveys of the district is presented.

(Table 1) Geological succession in the Montgomery district.

(Front cover) The Ordovician rocks of Shelve; the dolerite sill of Corndon Hill forms the highest ground seen on the skyline. View looking east from Montgomery Castle [SO 2215 9680] (GS 431). Photographer Audrey A Jackson.

Chapter 1 Introduction

This memoir describes the Montgomery district, which is the area covered by the British Geological Survey 1:50 000 Series Sheet 165, Montgomery, and the continuation northwards of the Shelve outcrop of Ordovician rocks on Sheet 151, Welshpool (Figure 1). Montgomery is an ancient administrative town of the old county Montgomeryshire which grew up around King Henry III’s castle, but modern urban development has mostly passed it by. The topography of the district is varied; most is cultivated upland (Figure 2). The highest point is Cilfaesty Hill, in the Clun Forest, which rises to 528 m OD. Almost as high, but a much more prominent feature, is Corndon Hill (513 m) formed from an Ordovician doleritic intrusion.

Manufacturing within the district is concentrated upon Newtown, encouraged by its status as a ‘New Town’ granted in 1970. Traditional industry was agriculture and ancillary manufacturing of woollens. Bricks, tiles and drain pipes were fired from local Quaternary clays, while building stones were quarried from several local formations, both Ordovician and Silurian. Roadstone was produced from quarries around Shelve, particularly from the late Ordovician igneous intrusions, while the Cwm Mawr picrite is famous as a source for Bronze Age stone implements.

Geological history

Widely separated periods are represented in the rocks of the district (Table 1). The first occured during the Precambrian. The second ranged through most of the Ordovician and all but the earliest Silurian. After that there is no record of the geological history of the district until Pleistocene times.

The oldest rocks exposed are Uriconian and are in the order of 650 million years (Ma) old. They represent a period of intense volcanism; both acidic and basic lavas were extruded and interbedded with a sparse amount of volcanigenic sand. The contact with Longmyndian rocks is obscured, but although it has long been thought that the Longmyndian rocks overlie the Uriconian and are thus younger, of late Precambrian age, it is possible that they are in part coeval. The Longmyndian represents a period of rapid accumulation of clastic sediments in a fluviatile environment. Initially, these were low-energy deposits and probably distant from the source, but, later, sands and gravels predominated. Their provenance lay, either primarily or secondarily, in a magmatic arc. (Pauley, 1990a; Woodcock and Pauley, 1989). The possibility that some of the early muds were tuffitic suggests contemporary volcanism in the vicinity.

No deposits remain to represent the history between the late Precambrian and the early Ordovician. However, shallow-marine Cambrian rocks are present nearby in the Church Stretton district, and indicate a marine invasion on to a landscape of some relief, consisting of steeply dipping Uriconian and probably Longmyndian rocks (Greig et al., 1968). Although direct evidence is lacking, it is reasonable to conclude that the major synclinal structure in the Precambrian rocks was formed prior to the Cambrian Period. Indeed the upward coarsening of the Longmyndian terrestrial deposits is an indication of increasing, or advancing, tectonic activity. Nearby, at The Wrekin, the lower Cambrian Wrekin Quartzite rests unconformably on the Ercall Granophyre, intruded into Uriconian rocks about 560 Ma (Cope and Gibbons, 1987). This unconformity indicates the extent of the tectonism in this relatively brief ante-Cambrian period.

Following the early Cambrian marine transgression, it is probable that the district remained submerged throughout the Cambrian and early Ordovician (Tremadoc). The Shineton Shale was deposited as marine mud with occasional incursions of fine sand. Unusually uniform conditions covered a wide area during this epoch and great thicknesses of sediment accumulated in a shelf-sea environment, which supported faunas of dendroid graptolites and inarticulate brachiopods and trilobites.

Within the district, the junction between Arenig and Tremadoc strata shows no indication of an angular unconformity. Thus deposition may have been continuous, but the earliest Arenig sandstones are interpreted as nearshore deposits and brief emergence is therefore a possibility at this time. The succeeding Arenig and Llanvirn deposits, the Stiperstones Quartzite, Mytton Flags, and Hope Shale formations attest to a deepening marine environment with detrital sources becoming more distant. Thus it is probable that in late or post-Arenig times deposition extended across the line of the Pontesford–Linley Fault on to, at least, parts of the Long Mynd.

Mud deposition continued throughout Llanvirn times while trilobite and graptolite faunas of a deep- water or outer-shelf affinity thrived. Like the Llanvirn rocks of the Builth–Llandrindod Inlier, the mudstones of this district are interbedded with thick volcanic and volcanogenic sequences. In the succeeding Llandeilo Epoch, the volcanism waned and deposition of mud was interrupted only by brief periods of silty and calcareous sedimentation as global sea level continued to rise. It is clear that during the Llanvirn and Llandeilo epochs relatively deep marine shelf conditions persisted along a north–south tract to the east of the Tywi–Severn Lineament as far east as the position of the Pontesford–Linley Fault, and probably farther. How this rapidly sinking tract or subsidiary basin at Shelve related to the Welsh Basin on the west is obscure. It probably extended some distance to the south-west of Builth and north of Shrewsbury and in the south it was separated from the Welsh Basin by a marked slope (Cave and Rushton, 1996). In the north there appears to have been no such slope. Arenig to Llandeilo rocks are present only west of the Pontesford–Linley Fault, and there is no sedimentological evidence to suggest that the area which now lies immediately east of the fault was land at that time, providing a source for the sediments farther west. It is concluded, therefore, that the full sequence extended to these eastern areas too, but the position of the shoreline is not known.

At the end of the Llandeilo Epoch a tectonic event had a marked effect on deposition in the Shelve area and widespread influence in Wales. The area of marine deposition around the Welsh Basin was extended into parts of Anglesey (Bates, 1972) and to south Pembroke. Lime was introduced into the muddy sequences of South Wales and calcareous turbiditic muds into the basinal area of north Pembroke. Near Shelve, movement on the north-east-trending Pontesford–Linley Fault effected footwall uplift on the east with the consequent removal of the Tremadoc and probably other Ordovician deposits on that side. Marine circulation was affected at this time for there was a change in the chemistry of the sediments. The poorly oxygenated deposits of the Llanvirn and Llandeilo were replaced in Caradoc times by burrowed muds with thin sands, of higher oxicity. In parts of the Welsh Basin, notably along the Tywi Anticline and in South Wales, evidence of early Caradoc oxygenation is absent, and while the water may have been deeper in the area around the anticlinal axis, it is unlikely to have been so in South Wales.

Coincidental with the end-Llandeilo tectonism there was an abrupt coarsening of the Shelve sediments to produce the Spy Wood Sandstone. The initial influx of the coarse clastic sediment was of the same age as the earliest unconformable Ordovician deposits east of the Church Stretton Fault, but older than the similarly unconformable Ordovician deposits on the footwall (eastern side) of the Pontesford–Linley Fault. Parts, or all, of the Long Mynd, between Church Stretton and Shelve, therefore, must have remained subaerial during the deposition of the Spy Wood Sandstone.

Following the early Caradoc influx of coarse siliciclastic sediment, mud deposition returned and persisted through the Harnagian and Soudleyan. Frequent traction currents introduced thin beds of fine sand from the north-east. Two resurgences of volcanism produced rapid accumulations of volcanogenic sand, pyroclastic breccia, tuffite and rhyolitic lava in Soudleyan times. At Montgomery and Stalloes, blocks and boulders of andesite were transported by mass-flow, from a centre like that of the Breiddens area to the north, to form large isolated bodies enclosed within thick accumulations of mud.

No subsequent Ordovician sediments have been preserved and thus the remainder of Ordovician history is obscure. However, it is almost certain that marine deposition continued, producing a upward-coarsening, clastic sequence until late Longvillian times.

In the late Ashgill (Hirnantian), two nearly synchronous events were complementary in their effects upon erosion and deposition. One was a brief period of tectonism which inverted the Shelve–Llandrindod sub-basin, transforming it into the western fringe of the Midland Platform. The other was the Gondwanan glaciation which effected a global drop in sea level. Both events thus subscribed to the withdrawal of marine conditions from the platform margin of the Welsh Borderland. Erosion progressively stripped the emerging areas of Ashgill and earlier Ordovician deposits, laying bare late Ordovician intrusions. Large quantities of detritus were transported rapidly to the margin of the Welsh Basin, where they were unstable, and thence by slump and mass-flow to the deeper parts.

The Silurian Period commenced during a marine advance from the glacial low-stand. Encroaching from the west, the sea reached the line of the Severn where it halted temporarily in late Rhuddanian times. The Severn Valley coincides with major faults, and it is possible that movements on these permitted accumulation of Rhuddanian sediments on the west while footwall uplift of the east contributed coarse debris to local conglomerates.

Not until late Aeronian times did the marine advance cross this Severn ‘threshold’, whence its passage eastward was rapid. By then the terrain was a peneplain and only the Long Mynd to the east rose above it. For the remainder of the Llandovery, the district served as a broad, largely oxygenated, marine shelf. The Llandovery transgression had been a prolonged process. Initially it was sustained by the glacio-eustatic rise, but later phases were probably tectonically driven, influenced locally by the major faults of the district. The most tangible result of the Taconic event within the district is the angular unconformity between Llandovery and Ordovician strata. Westward the disparity of their dips diminishes, as seen in the district to the north, and in mid-Wales the Welsh Basin fill is devoid of late Ashgill folding. At the close of the Llandovery, a major change affected depositional environments. In both the marine outer-shelf and the Welsh Basin, they became anoxic and apart from two or three minor interludes, remained so until mid-Ludlow times.

In the Wenlock Epoch, the shelf-sea deepened and the turbiditic basin to the west extended its influence into the western half of the district introducing first the Denbigh Grits (Penstrowed Grits Formation) and the Nantglyn Flags. Then, during the early Ludlow, fine sand and silty mud, probably brought in turbidity currents, dispersed largely north-westwards across the outer shelf. In general, the currents indicate an origin in the region of the Church Stretton Fault near Presteigne, but by mid-Ludlow times these had waned sufficiently to allow laminated hemipelagite to be the dominant deposit. It was then that graptolites disappeared prematurely from the Welsh ‘province’, for reasons not understood. After this, marine oxicity returned and, in a low-energy shelf regime, laminar silt accumulated to form the Knucklas Castle Formation. This shelf sea was probably still fairly deep, below storm-wave base. The sediments were bioturbated, but preserve sparse remnants only, of both pelagic and benthic faunas. Perhaps the conditions of deposition differed little from those of the underlying top part of the Bailey Hill Formation except in the level of marine oxicity.

The final stage in the history of the marine shelf arrived with storm sheet sands and muds and a prolific brachiopod fauna characteristic of Ludlow deposits throughout the Welsh Borderland. In this district they formed the Cefn Einion Formation.

By Přídolí times the sedimentary regime of the whole region was reduced to a uniformly very shallow epeiric sea. This was succeeded by a flat, mainly fluvial cuvette, which experienced periodic, marine-driven, brackish incursions. Within this cuvette, thick deposits accumulated which form the red, green and grey mudstones and sandstones of the Clun Forest Formation in a generally upward-coarsening sequence.

No deposits remain to represent the interval between the Silurian Period and the Quaternary when, during Pleistocene times, the district was overwhelmed by ice. Till, deposited from the melting of the ice, was spread liberally over the terrain, and valleys were further aggraded with melt-out detritus and lake sediments.

History of research

There was much early interest in the geology of the district, particularly in the Ordovician rocks around Shelve. G H Morton, C Lapworth and W W Watts were pioneers in the study of these rocks in the latter part of the 19th century, and Watts continued through into the first quarter of the 20th century. W F Whittard produced many papers culminating in a comprehensive account of the stratigraphy; this was published posthumously in 1979 under the editorship of W T Dean. Whittard’s field-slips, at a scale of 1:10 560, are preserved in the Natural History Museum.

More specialised studies have also been undertaken, for example Blyth (1938; 1944) studied the petrography of the igneous rocks, Dunham and Dines (1945) described the baryte deposits and Dines (1958) produced a comprehensive account of the West Shropshire mining field.

Most of the research on the Silurian sequence is more recent, and was conducted separately from the research on the Ordovician rocks. Studies by Allender (1958) on the Bishop’s Castle area, Holland (1959) in the Knighton area and Earp (1938, 1940) in the Clun and Kerry areas are fundamental to the understanding of the Silurian strata of the district. Evans (1957) worked in the area north-east of Bishop’s Castle. More recently, research by the authors and others has been concentrated on geological processes and of particular relevance to the district is the sedimentological study of the Bailey Hill Formation of Ludlow age exposed over a wide area of eastern Wales. Likewise Woodcock, of Cambridge University, is involved in a study of the possible extension of the Pontesford–Linley Fault south-westwards from its outcrop in the north-east of the district beneath Silurian cover rocks.

The glacial geology of the district was reviewed in the extensive account of Dwerryhouse and Miller (1930). A more specific study of the Church Stoke area was conducted by Rowlands (1966), and most of the district was covered by Brown (1971).

Chapter 2 Precambrian

Rocks of Precambrian age occupy the north-eastern corner of the district. Their outcrop is limited to the west by the Pontesford–Linley Fault and to the south they are overlain unconformably by Silurian strata (Figure 3). They are divided into the Uriconian Volcanic Group and the Longmyndian Supergroup, the latter largely of sedimentary origin.

A review of work carried out on the Precambrian rocks has been given by Greig et al. (1968, pp.35–37) and Dunning (1975). More recently Toghill and Chell (1984) have summarised the Precambrian stratigraphy. Langford and Lynas (1985), Cave et al. (1985) and Hains and Langford (1985) have given detailed accounts of the Precambrian of the district. Pauley (1990a, 1990b) has described the sedimentology, structural evolution and tectonic setting of the Longmyndian and has also (1991) considered its relationship to the Uriconian.

The relationship between the Longmyndian and the Uriconian in this district, as elsewhere, is not certain and has been under discussion for many years. James (1952) excavated a section at Chittol [SO 3493 9496] and stated that Longmyndian green tuffs (which he included in the Wentnor ‘Series’) were unconformable on the Uriconian, the whole sequence being inverted. Pauley (1991) included the tuffs in his Linley Formation and suggested that the Uriconian was thrust over them. James (1956, p.326) also excavated a section [SO 3659 9677] about 600 m north-north-east of Cold Hill Farm which he also interpreted as an unconformity. At this locality, however, there is cleavage in the mudstones and brecciation at the junction which suggests possible structural complications. During this survey, Longmyndian/Uriconian contacts which are undoubtedly faulted have been excavated at two sites [SO 3421 9357] and [SO 430 9393] in Linley Big Wood.

Strike-trend analysis (Lynas and Langford, 1985) of the junction between the Uriconian and Longmyndian [around 348 950] indicates a planar surface dipping north-east at 30°. The strike trend of the surface parallels that in the Longmyndian, with an angular discordance of 20 to 30°. Such a surface could be interpreted as either a fault or an unconformity.

On the evidence available it is not possible to be certain about the Longmyndian/Uriconian relationship. Many of the contacts are undoubtedly faulted, but the evidence at other locations is not conclusive. Even though there are structural complications associated with the sections previously described as unconformities, an unconformable relationship cannot be ruled out. An unconformity, if it existed, close to a major structure such as the Pontesford–Linley Fault would act, in all probability, as a plane of weakness during post-Precambrian movements and show signs of structural disturbance.

Uriconian Volcanic Group

The Uriconian Volcanic Group in this district comprises a complex sequence of extrusive and intrusive acidic and basic rocks, generally highly altered and commonly silicified, which lie between the Pontesford–Linley Fault to the west and the Longmyndian to the east. These outcrops and those to the north-north-east at Pontesford Hill and Plealey (Sheet 152) are the ‘Western Uriconian’ of a number of authors, so called to distinguish them from a similar, and probably coeval, volcanic suite (‘Eastern Uriconian’) which occurs along the Church Stretton Fault Zone on the eastern side of the Longmyndian outcrop.

The rocks of the group have a characteristic magnetic anomaly, and magnetometer traverses have assisted the geological survey by allowing more precise definitions of mapped contacts with adjacent non-magnetic sedimentary rocks. Additionally, the magnetic susceptibility of the basaltic rocks is significantly higher than that of the rhyolites and this has given an indication of the relative proportions of these main components of the group. Textural evidence suggests that the majority of the volcanic rocks are of extrusive origin.

There are two separate areas of Uriconian rocks (Figure 3), outcropping to the east of the Pontesford–Linley Fault. The northern area gives rise to a distinct ridge along most of its length.

In the southern area, the basalts are fairly uniform, and the typical rock is seen [SO 3370 9260] south of Oldmore’s Wood. It has a normal basaltic texture (E58347), is dominated by strongly altered plagioclase and chlorite and contains small opaques. A weakly developed trachytic texture is visible in basalt (E59366) from Linley Big Wood [SO 3441 9353] and pyroxene pseudomorphs are present. Some of the basalts are vesicular, as for example in Chittol Wood [SO 3482 9495] (E58360). These are fine grained, in places microlithic but rarely feldspar-phyric, and were almost certainly extrusive lavas.

The rhyolites are invariably devitrified, recrystallised and silicified. They were evidently glassy, and may contain feldspar phenocrysts. These phenocrysts may be dominantly plagioclase (E59372), (E58357) but in most cases alteration is so intense as to mask their original composition. The matrices are dominantly micro- to crypto-crystalline. A sample from west-north-west [SO 3420 9314] of Linley Hall shows flow-banding in both hand specimen and thin section (E58345) defined by aligned feldspar microlites in devitrified glass with rare K-feldspar phenocrysts; an analogous banded variety also occurs nearby [SO 3419 9322]. Strike and dip of both rhyolites are similar to those recorded from nearby volcaniclastic sedimentary rocks within the group. Some rhyolites are spherulitic, as for example from Oldmore’s Wood [SO 3358 9267] (E59361), and these primary structures are almost pure silica.

Beside a track [SO 3448 9352] in Linley Big Wood is an outcrop of sedimentary rocks which are taken to be part of the group. The section was enlarged by excavation (January 1984) and exposed basalt in contact with the sedimentary rocks. At 0.25 m from the contact, the basalt (E59363) is amygdaloidal with a microlithic matrix full of altered plagioclase, and was probably originally very fine grained. The indurated immature sandstone (E59364) at the contact is strongly altered; it contains subrounded quartz, lithic and plagioclase clasts, which show no sign of strain. Thin Fe-rich bands within the sample may represent altered mudstone. The contact appears to be discordant, and one metre from the junction the sandstone is blocky and fine grained with thin mudstone laminae. The igneous rock is probably intrusive. These sedimentary rocks have a relatively high magnetic susceptibility, unlike those of the Longmyndian, and are thus considered to be part of the Uriconian. Similarly magnetic sedimentary rocks occur east-north-east [SO 3377 9335] of Upper Bent and south-east [SO 3400 9360] of Heath Mynd.

In the northern area of Uriconian rocks, altered basalts are well exposed along the ridge [SO 366 969] to [SO 368 972] south-east of The Knolls. They are granular and fine grained with much calcite, chlorite and euhedral plagioclase laths (E59157), (E59161), (E59166). The rocks are significantly magnetic, and the results of a magnetometer survey combined with the clear field relations with surrounding rocks indicate a sheet-like form, probably an intrusion. The junction of the basalt with the surrounding rhyolite is best seen alongside a track [SO 3671 9708] where very fine-grained, dark olive-green, red-mottled basalt is in contact with a silicified rhyolite. The contact is near vertical and locally trends west to east. The junction can also be positioned accurately farther north-east [SO 3680 9720] where a phenocrystic rhyolite (E59172) occurs within a few metres of comparatively unaltered granular basalt (E59173).

South-east of the summit of the ridge the junction of another basaltic body (E59179) with rhyolite is visible [SO 3672 9703] close to a track where there is a marked change in colour of the overlying soil and rock debris. Two hundred metres to the south-south-west [SO 3663 9684] the basalt shows less alteration (E59162) with some of the crystal characters of the plagioclase laths still visible, and apparent remnant pyroxene phenocrysts.

Surrounding the basalt intrusions are a variety of rhyolites, generally highly silicified, altered and recrystallised. Silicification can be so severe that the true nature of the original rock is completely obscured. For example, two samples (E59178), (E59179) from [SO 3726 9762] are composed of cryptocrystalline quartz with abundant brown iron-staining, some fragmental strained quartz phenocrysts and rare granular quartz veins. In other parts of the rhyolite outcrop the original texture of the rock is preserved, even though the mineralogy is altered. For example, a specimen (E59174) [SO 3683 9725] has spherulites, subhedral alkali feldspar phenocrysts and flow banding. Silicification is confined to quartz veins, but the crypto-crystalline matrix is chloritised and possibly epidotised. A sample (E59169) from [SO 3671 9708] has an almost cryptocrystalline (formerly glassy) matrix with small euhedral sericitised alkali feldspar phenocrysts.

The north-eastern part of this Uriconian area is poorly exposed, but at two localities the rock is a granophyric microgranite, mineralogically the same as the rhyolites. If the microgranite is intrusive then it is likely that the rhyolites are associated extrusives. The most northerly outcrop [SO 3758 9777] shows a rock dominated by granular quartz with some perthite surrounding the crystals. Some 500 m to the west-south-west [SO 3714 9753] the microgranite contains large sericitised alkali feldspars with perthitic intergrowths and common interstitial quartz (E59177). Both rocks are medium to fine grained and granular.

A magnetometer survey over the microgranite intrusion gave a relatively high magnetic reading suggesting the presence of some unexposed basic rocks and perhaps more complex geology than appears on the map. Details of the magnetometer survey are held at BGS Keyworth office.

Longmyndian Supergroup

Only the south-western corner of the Longmyndian outcrop falls within the district. It is limited to the north-west by the Uriconian Volcanic Group and the Pontesford–Linley Fault, and to the south by unconformable Silurian rocks (Figure 3). The major part of the Longmyndian outcrop, including the Long Mynd itself, lies within the Church Stretton district (Sheet 166) to the east and the Shrewsbury district (Sheet 152) to the north-east.

The Longmyndian is folded into a major syncline trending north-eastwards; the whole of the sequence within this district lies on the western, inverted, limb of this syncline. The classification of the Longmyndian adopted in this account is shown in (Table 1). Correlation with the formations which crop out on the eastern limb of the Longmynd Syncline are shown in (Table 2).

An account of the historical development of the classification of the Longmyndian stratigraphy is given by Greig et al. (1968). At that time, the Longmyndian was divided into the Wentnor Series and the Stretton Series, and each in turn was divided into a number of groups. Dunning (1975) reclassified the Wentnor and Stretton series as groups and their subdivisions as formations. Toghill and Chell (1984) used this classification and introduced the term Longmyndian Supergroup for the whole of the sequence. Modifications to formations by Cave et al. (1985) and Pauley (1991) are considered in the section on the Linley Formation.

The sediments of this supergroup were probably deposited in a shallow-marine or deltaic environment. Their depositional environment has been discussed by a number of authors. Based on a detailed study of the petrography of the rocks, Greig et al. (1968, pp.56–73) concluded that the sequence was paralic in origin, showing similarities of tectonic setting, lithology and thickness with the Old Red Sandstone of Scotland. They considered that the sediments were deposited in a narrow depression or crustal trough controlled by major basement fractures, probably along the lines of the Pontesford–Linley and Church Stretton fault zones. Lithic fragments and pebbles within the rocks indicate derivation from two distinct sources: the Uriconian Volcanic Group and a metamorphic complex similar to the Mona Complex and the Rushton Schists.

A deltaic origin has been proposed by Baker (1973), while Pauley (1990a) has described the supergroup as an upward-coarsening, progradational sequence, in which a turbidite facies at the base (lower part of the Stretton Group) passes up through a subaqueous delta facies into alluvial floodplain and braided alluvial facies in the upper part of the Stretton Group and in the Wentnor Group. Palaeocurrent data indicate a derivation from the south, between east-south-east and west-south-west, and therefore the Uriconian Volcanic Group, the main source of the sediments, lay in that direction rather than directly adjacent to the area of deposition. Pauley considers that the Longmyndian and Uriconian are at least in part coeval, and that later strike-slip faulting juxtaposed the Longmyndian and Uriconian rocks.

Stretton Group equivalent

Linley Formation

This name was introduced by Pauley (1991) for a succession of greenish grey, rarely purple-grey, mudstones, siltstones and very fine-grained to fine-grained sandstones, with local lapilli tuffs, which occurs in a number of narrow outcrops between the Uriconian Volcanic Group and the Bayston–Oakswood Formation. Neither base nor top is exposed, and all junctions appear to be faulted.

Originally (Lapworth and Watts, 1910), the beds making up the Linley Formation appear to have been included with the Western Uriconian. James (1952, 1956) included them with the Bayston–Oakswood Formation (Table 2) although they are lithologically quite distinct from the purple sandstones and conglomerates of that formation. Cave et al. (1985) and Langford and Lynas (1985) placed them in the Portway Formation of the Stretton Group which immediately underlies the Wentnor Group on the eastern side of the major syncline. However, there is no direct evidence of a sedimentary contact between these beds and the Wentnor Group and the contacts may all be faulted. Also, the Portway Formation differs from the Linley Formation in that it contains purplish red mudstones, siltstones and fine- to medium-grained sandstones with erosional bases and included mudstone clasts. The Linley Formation is predominantly greenish grey and does not contain such sandstones. For these reasons, Pauley (1991) considered that these beds on the western limb of the syncline could not be compared with the Portway Formation on the east and he erected a new formation, the Linley Formation, to include them; he concludes that they are most nearly lithologically comparable with the upper part of the Burway Formation or the lower part of the Synalds Formation, both included in the Stretton Group (Table 2).

Details

The Linley Formation is seen in three separate areas: in the eastern side of Linley Big Wood [SO 342 935] and along the valley of the River West Onny, around and to the north of Chittol [SO 3493 9496] and to the west and north of Cold Hill Farm [SO 3635 9625].

In Linley Big Wood, finely banded mudstone and siltstone are exposed in a track [SO 3421 9357] adjacent to the Uriconian, and the excavated contact revealed several crush zones and bedding nearly parallel to the contact. The mudstone is pale green and may be tuffitic. On the same track, an excavation [SO 3430 9393] revealed pale mudstones in faulted contact with the Uriconian. Here, the contact is a series of anastomosing faults with crush zones up to 0.16 m wide in the Longmyndian. Along the western banks of the River West Onny [SO 345 941] are sections whose stratigraphical position is less certain. They comprise delicately laminated and well-bedded purplish grey mudstone and siltstone with an apparent transition to poorly bedded, medium-grained, purplish brown sandstone. These lithologies are not typical of either the Linley Formation or the Bayston–Oakswood Formation. The mudstones and siltstones show way-up criteria; for example, at one exposure [SO 3450 9416] 1 mm-thick laminae show clear inverted grading and another section [SO 3443 9417] shows ripple crests (wavelength 2 to 2.5 mm) in a siltstone which point downwards indicating inversion. The sequence is repeated by faulting farther south along the river [SO 3457 9392].

Exposure at Chittol [SO 3493 9496] is now confined to finely laminated colour-banded grey to green tuffite with graded bedding indicating inversion. The Uriconian-Longmyndian junction was excavated by James (1952, 1956) at this locality and provided evidence for his suggestion of an inverted unconformity between these two groups (see p.4).

In the northernmost area, laminated green-grey mudstones, and siltstones or tuffites, are seen [SO 3627 9633] near Cold Hill Farm. Farther west [SO 3613 9626] similar mudstone appears to be faulted against conglomerate, with the mudstone strongly cleaved near the fault plane. About 600 m north-north-east of Cold Hill Farm (formerly Coldyeld Farm), cleaved blue-grey and green-grey mudstones, and siltstone or tuffite is exposed [SO 3659 9677]. Here the junction with the Uriconian was excavated by James (1956) and interpreted as an inverted unconformity. However, cleavage in the mudstone and brecciation on the junction suggests possible structural complications.

Wentnor Group

Bayston–Oakswood Formation

This formation has a broad outcrop, some 1.5 to 2 km wide, limited to the south by unconformable Silurian rocks between Linley and Norbury and extending to the eastern margin of the district, east of Squilver.

It comprises sandstones and conglomerates with rare thin mudstone and siltstone beds. The sandstones are dull purple to purplish brown, and coarse to medium grained. They are composed of subangular to subrounded quartz and feldspar grains; they are grain supported, poorly sorted, massive, and show little evidence of bedding. Mudstone clasts are common and the sandstones are locally pebbly. There appear to be at least four impersistent conglomerate horizons. In these the predominant clasts are quartz, with some volcaniclastic rocks. Clasts are commonly subrounded and 20–30 mm in diameter, but may reach 120 mm locally. Detailed petrographical descriptions of conglomerates from this formation in the Church Stretton district to the east are given by Greig et al. (1968, pp.57–62). Elsewhere these conglomerates have been mapped as continuous beds (Greig et al., 1968; Pocock et al., 1938), but within this district they appear to occur in discrete channels.

Numerous outcrops of medium-grained sandstone and conglomerate can be seen [SO 3558 9312] east-north-east of Squire Hall. Subrounded to subangular clasts up to 20 mm across are common. Bedding is absent or poorly developed and in places is shown by a gradation from conglomeratic sandstone to conglomerate. Clast abundance in the conglomerate is variable and the matrix is grain-supported and generally depleted of fines. The dominant clast is pink iron-stained quartz, commonly with a white centre; other clasts include igneous and volcanic rocks.

The gradation from sandstone to fine-grained conglomerate can be seen at Nurton [SO 3528 9361]. Here, scattered outcrops of massive medium-grained purple sandstone grade into fine-grained conglomerate with clasts up to 4 mm in diameter. Nearby [SO 3514 9368], there is a purple conglomerate with poorly sorted subangular to subrounded clasts up to 10 mm in diameter. Also there are clasts of dark purple mudstone or siltstone up to 40 mm across. The gradational nature of the sandstone–conglomerate junctions makes accurate distinction of conglomerate horizons subjective and difficult.

The sandstones are well exposed in a quarry [SO 3526 9495] near Beach Farm. They are mainly coarse-grained purple sandstones, in places conglomeratic with clasts up to 5 mm in diameter. Dark colour-banding indicates irregular bedding planes. Within the sandstone are oblate purple-brown mudstone fragments up to 50 mm long. The surfaces of these clasts are pitted, caused by compaction of sand grains into the mudstone. The fragments are roughly parallel with bedding, and are probably locally derived from the break-up of thin, barely compacted, mudstone beds. Nearby, in Chittol Wood [SO 3477 9469] within a coarse-grained conglomerate, a sandstone shows inverted grading and also a 1 m inverted channel.

Near Cold Hill Farm [SO 3636 9620] are numerous exposures of massive clast-supported conglomerate. The clasts (up to 100 mm diameter) are poorly sorted, subrounded to angular, with coarser pebbly beds giving a crude stratification.

Bridges Formation

This formation has a triangular outcrop extending north and east from Norbury to the edge of the district (Figure 3). It comprises purple mudstone, siltstone and fine-grained sandstone, commonly well laminated. There are no coarse-grained horizons.

It is well exposed around Norbury [SO 364 928] and an excavation [SO 3633 9275] at Hall Farm showed inverted purple-red to red-brown mudstone, siltstone and sandstone lying unconformably beneath Silurian Pentamerus Sandstone. Siltstone beds, 0.04 to 2.40 m thick, alternate with mudstones, 0.03 to 0.13 m thick. The thicker siltstones may be slightly sandy or grade into sandstone and are characterised by mudstone flakes. There is an inverted channel in sandstone and siltstone, at least 1 m long by 0.5 m deep.

Just north of Norbury, along a minor road [SO 3652 9327] to [SO 3656 9339] north of Mount Pleasant, a number of outcrops reveal interlaminated purple mudstone and siltstone with flat bedding and dark grey fine-grained sandstone with included siltstones up to 2 mm thick. Some of the bedding planes in the sandstone show very small muscovite flakes with a preferred orientation. More massive, but fine-grained, sandstones occur west of this area marking a gradation from the coarser underlying Bayston–Oakswood Formation.

Two small sections [SO 3781 9171]; [SO 3784 9174] in the River East Onny in purplish grey interlaminated siltstone and very fine-grained sandstone are situated close to the presumed position of the axis of the Longmyndian syncline.

Chapter 3 Ordovician: Introduction and Tremadoc to Llandeilo

Introduction

There are about 5.5 km of Ordovician rocks at outcrop in the Shelve area (Figure 4) and an unquantified but thick sequence in the Forden and Montgomery areas. They were marine sediments, mainly argillaceous, with large volcaniclastic and volcanic intercalations. Magmas of basic and intermediate compositions intruded these deposits in late Ordovician times.

Five of the six Ordovician Series are represented in the district — Tremadoc, Arenig, Llanvirn, Llandeilo and Caradoc (Table 3). The Tremadoc Series is distinctively different from succeeding series in comprising rocks, mainly mudstones formed in varying oxic to anoxic environments, which are free from volcaniclastic incursions and which coarsen upwards at the top, ending probably in a non-sequence. Thus they are the reflection of a shallowing environment which the Arenig Series reverses and it is argued that the early Arenig Stiperstones Quartzite was the initial shoreface deposit of the Shelve Ordovician ‘marine basin’. Llanvirn and Llandeilo rocks are thus largely dysaerobic mudstones with major volcanic and volcaniclastic components in the Lower Llanvirn.

At Llandeilo, about 100 km to the south-west, Williams et al. (1981, p.673) assessed the lithological variety of the Llanvirn Series and concluded that there was no need to invoke rapid changes of sea depth to account for the cyclic recurrence of coarse deposits within finer flags and shales. He viewed the sequence there as a product of sublittoral to intertidal conditions with three regressive (upward-coarsening) cycles due to cyclic increases in overall marine energy.

The Shelve rocks are approximately on strike with those of Builth and Llandeilo and Willams’ general assessment of continuous accommodation of sediment in a shallow marine basin is seen as applicable to Shelve also. Marine conditions were less energetic, and more tranquil, probably deeper water permitted a more stable marine-water profile. However, shallow-marine volcanic basins, as Shelve was, can be subjected to rapid changes of water depth, but they are localised to the volcanic centres. Such changes induce transient submarine slopes and the various transportation regimes that accompany them, and it is this type of environment that is envisaged at Shelve in Llanvirn times.

The Llandeilo epoch spans 5 to 7 million years and is the shortest of the Ordovician epochs. Nevertheless it is distinctive, particularly in its faunas. The pendent Didymograptids had disappeared abruptly, leaving the biserial and earliest Dicellograptid graptolite stocks to dominate the plankton. Trilobites evolved rapidly during this epoch and provide a biozonation to stand alongside the two graptolite biozones of ‘Glyptograptus’ teretiusculus and the greater part of the Nemagraptus gracilis. However, despite the sequence being unbroken and well exposed there are still uncertainties in defining the base of the Caradoc.

The Caradoc Series also consists largely of mudstones, but of a more oxygenated marine environment; two separate units of volcanogenic sandstones allow a subdivision into formations.

Shineton Shale Formation

The Geological Survey map of 1850 placed the ‘shales and sandstones’ that lay below the quartzite of the Stiperstones in the Silurian (b1). The beds were named Shineton Shales by Callaway (1877) who recognised lithological and stratigraphical similarities with the rocks near the village of Sheinton near Much Wenlock. Biostratigraphical proof of their equivalence was lacking, thus leading Lapworth (1916) to erect the name Habberley Shales after the village situated on the outcrop. Later, Stubblefield and Bulman (1927) found palaeontological proof of their Tremadoc age and reinstated Callaway’s decision to call them Shineton Shales, and eventually Whittard (1979) followed. The British Geological Survey persisted with the local name Habberley Shales Formation; Pocock et al. (1938) used it for the Shrewsbury district and in consequence it was used on the 1:25 000 Series map ‘The Shelve Ordovician Inlier’ (BGS, 1991), but for the 1:50 000 map the synonym Shineton Shale Formation was preferred. Concurrently with the recent BGS survey, Fortey and Owens (1992) made a detailed study of the higher Tremadoc beds. They concluded that these beds are sufficiently different, and mappable, from the rest to warrant a separate identity under the name Habberley Formation [as distinct from Habberley Shale Formation or Habberley Shales of earlier accounts], limiting the Shineton Shale Formation to the beds below. The BGS did not map this distinction and thus this account follows the traditional usage of the formational name. To avoid possible confusion it is suggested that the upper division may be made the Habberley Member of the Shineton Shale Formation.

The outcrop of the Shineton Shale extends from about 1 km north of Bromleysmill [SO 3323 9181] to the northern edge of the district near The Knolls [SO 3635 9742] and beyond to Pontesbury where it is concealed by Coal Measures. The formation is thrown down against Precambrian rocks on the east by the Pontesford–Linley Fault and on the west it is overlain by the Stiperstones Quartzite probably wholly of Arenig age and probably with a non-sequence between them.

The full thickness of the formation in this outcrop is not known, but about 900 m are present near Habberley.

Because of the Pontesford–Linley Fault the base of the formation is not present at outcrop. The lowest parts exposed are graptolitic dark grey mudstones and siltstones; the middle part consists of grey and greenish grey blocky siltstones and shales with layers of micaceous laminated siltstone. The topmost 150 to 200 m are lithologically more diverse and more arenaceous (Figure 5). They include blocky laminated grey siltstones and shales, beds of sandstone, green-grey laminated siltstones and mudstones and at the very top up to 4 m of dark grey mudstone with flaggy, micaceous siltstone layers. Burrows commonly disturb the silt laminae, especially in the upper portion where there is close lithological similarity to the Mytton Flags (Cave et al., 1985, p.5). Re-identification of Whittard’s (1931) Asaphellus homfrayi as comparable with A. graffi enabled Fortey and Owens (1992) to show that these topmost beds are of youngest Tremadoc age and biostratigraphically higher than the sub-Caradoc Shineton Shale near Sheinton (Lapworth and Watts, 1894, p.310) and they constitute most, or perhaps all, of their Habberley Formation.

The base of the Habberley Formation is defined at the junction between black, flaggy, micaceous siltstones and underlying soft, yellow-green-weathering shales with interbedded sandy beds of turbiditic origin, as in Linley Big Wood (Fortey and Owens, 1992, fig. 1, p.555). The thickness of the Habberly Formation is given as perhaps 500 m (Fortey and Owens, 1992, p.557) although in Linley Big Wood, where its outcrop is constrained by a top and base, there does not appear to be space to accommodate much more than half this thickness.

The bulk of the Shineton Shale was deposited on an open marine shelf, well oxygenated most of the time, but occasionally dysaerobic and anoxic (Fortey and Owens, 1992); this vacillation continued into the late Tremadoc when there was a general increase in sand, as noted by Whittard (1931) and seen in Linley Big Wood, producing a deposit closely comparable with the Mytton Flags (Cave et al. 1985; Fortey and Owens, 1992). It is probable that the depositional environments of the two periods were also very similar, the former indicating a shallowing, regressive sea, and the approach of the littoral sands of the Stiperstones Quartzite, the latter indicating a subsequent progressive deepening. However, the sedimentary profile from the Shineton Shale, through the Stiperstones Quartzite, into the Mytton Flags, is not symmetrical and even; the base of the Stiperstones Quartzite is sharp and conglomeratic. Thus a brief interval of erosion may have preceded its deposition, so removing the last deposits of the shallowing cycle below.

Biostratigraphy

Fossil collections derived from the outcrop have been re-examined by Dr A W A Rushton during this survey. The record by Whittard (1931) of Asaphellus homfrayi from the top part of the formation was based upon a fragmentary pygidium that appears closer to A. graffi, which derives from the ‘Schistes a Gâteaux’ in the Montagne Noire, France, where it lies close to the Tremadoc–Arenig boundary, as recognised there. Asaphellus homfrayihas, however, been found in beds at a slightly lower level, about 180 m below the Stiperstones Quartzite, in Linley Big Wood, by Fortey and Owens (1992, p.555), suggesting correlation with the S. pusilla Biozone of the type Shineton Shale. Whittard’s (1931) record of Shumardia pusilla cited from the top part of the formation with A. homfrayi is not upheld. The specimen, when developed, proved to be a bradoriid ostracod very like a species from the Micklewood Beds of Gloucestershire, reported to be of late Tremadoc age. Fortey and Owens’ (1992) records of Angelina sedgwickii and Peltocare olenoides provide good evidence that the highest part of the Shelve outcrop is younger than the highest Shineton Shales of the type area.

Dictyonema flabelliforme [now Rhabdinopora flabelliformis] and ‘Clonograptus’ tenellus [now Adelograptus? tenellus] were recorded by Stubblefield and Bulman (1927) from low horizons in the Shelve outcrop of the formation. These identifications are confirmed, thus upholding the Tremadoc dating and indicating correlation with the lower part of the Shineton Shales of the type area. Other correlations are detailed in the following pages.

It has not been possible to identify the Tremadoc–Arenig boundary but it has been taken at the base of the Stiperstones Quartzite (Figure 5).

Details

Several of the best exposures of the lower part of the formation occur in Habberley Brook between Gittinshay Wood [SO 3858 9988] and Brook Vessons Farm [SJ 3959 0122] where the formation consists of graptolitic dark grey shales and siltstones, commonly disturbed in proximity to the Pontesford–Linley Fault. Crushed greenish grey mudstones from this section [SJ 3938 0100] yielded sponge spicules, Eurytreta sabrinae, Adelograptus? [fragment] and R. flabelliformis anglica. This fauna is similar to that of the lower part of the type Shineton Shales, approximately the ‘Transition Beds’ of Stubblefield and Bulman (1927). It was from just west of Brook Vessons Farm, and thus from beds some metres higher in the succession, where Stubblefield and Bulman (1927, p.116) found R. flabelliformis in Brook Coppice and A. tenellus with ‘a large bellerophon’ in a tributary to Habberley Brook. These fossils indicate the flabellformis and tenellus biozones respectively.

In the middle part of the outcrop, grey and greenish grey contorted shales and blocky siltstones are exposed. Near the Knolls, behind a house [SO 3633 9730], several metres of grey to dark grey fissile siltstones and streaky siltstones of Mytton Flags appearance are exposed. They fracture along rough surfaces which are coated with brown iron oxides. Other exposures along this part of the outcrop occur in Gatten Plantation [SO 3743 7885] and in the upper reaches of Habberley Brook near The Hollies Farm [SO 3790 9904].

The top part of the formation is exposed in several places. The most southerly exposure is in an estate road [SO 3387 9412] to [SO 3388 9407] on the eastern flank of Heath Mynd where some 20 m of pale green-grey shales and siltstones occur about 20 m below the base of the Stiperstones Quartzite. Further exposures [SO 3416 9442] to [SO 3414 9446] occur in a forest track near Coppice Dingle, and 200 m to the north Drs Fortey and Owens collected A. homfrayi from mudstones in a track side exposure [SO 3418 9459] about 180 m below the Stiperstones Quartzite. The best exposure of the top part of the formation occurs along a forest track in Linley Big Wood in about 180 m of beds below the base of the Stiperstones Quartzite [SO 3406 9493] (Figure 5). It exposes green-grey laminated siltstones and mudstones interbedded with flaggy, bioturbated micaceous siltstones and fine sandstones. Some of the sandstones are feldspathic and poorly laminated; in parts the rocks have an appearance similar to that of the Mytton Flags (Cave et al., 1985). A log of these beds together with the overlying Stiperstones Quartzite is given in (Figure 5). The fauna collected from this section consists of Lingulella sp., Angelina sp. [lacking genal spine], Asaphellus cf. graffi and Leptoplastides sp.

Farther north, the upper part of the formation is exposed around Upper Vessons Farm [SJ 3866 0203], where 15 m of blocky, laminated grey siltstones and shales are exposed behind a building. There are similar beds in the brook [SJ 3866 0211] and in an old level [SJ 3881 0217] where several lingulids were obtained.

A forest track [SJ 3906 0339] at the north end of Eastridge Wood lies alongside a fault and dolerite intrusion and exposes green, purple and grey shales and siltstones for some 150 m to the north-west [SJ 3892 0344], just short of the contact with overlying Stiperstones Quartzite. The beds in this section may have been affected by contact metamorphism. At the lower end of the section [SJ 3906 0339] the beds yield Lingulella bella. Just north of the fault, grey to green shales and blocky micaceous siltstones with some laminated fine-grained sandstones are exposed on another forest track [SJ 3913 0341] to [SO 3905 0350]. These beds are probably stratigraphically lower than those near the intrusion, but dips are variable, and some are directed towards the south-east, so that their exact horizon is not certain. These beds [SJ 3905 0350] contain Lingulella bella. An exposure of similar beds [SJ 3915 0362] occurs near a disused reservoir. About 130 m higher and some 50 to 60 m below the Stiperstones Quartzite an exposure by a track [SJ 3914 0373] yields the trilobites Angelina sedgwickii and Peltocare olenoides (Fortey and Owens, 1992); these species are known from the highest Tremadoc strata in the Portmadoc area, North Wales, and P. olenoides is recorded from correlative beds in the Lake District (Rushton, 1988). Beds within 10 m below the Stiperstones Quartzite, are exposed in a forest roadside [SJ 3893 0362] and again about 100 m to the north [SJ 3896 0372] immediately east of a quarry near Granham’s Moor Farm. The former exposure yields L. bella, Rioceras sp., A. cf. graffi and trilobite (Calymenacean?) fragments. These sections expose beds comparable with those in Linley Big Wood, and the section close to the quarry is that described by Whittard (1931, p.324). The latter section exposed the junction between the Shineton Shale Formation and the overlying Stiperstones Quartzite which was redescribed by Lynas (1985a) (Figure 5). There is no angular discordance at the contact and no faunal evidence for a break, but the sedimentological evidence does point to the presence of a minor non-sequence. Like that in Linley Big Wood the section shows a broadly upwards-coarsening sequence, again suggestive of marine shallowing.

Stiperstones Quartzite Formation

The formation was first mapped by the Geological Survey as Quartz Rock and shown as a separate unit on the 1-inch map of 1850. Lapworth (1887a, b) referred to it as the Stiper Group, and in 1894 Lapworth and Watts coined the name Stiperstones (Quartzite). Subsequent terminology is summarised in the table drawn up by Whittard (1979) who accorded member status to the Stiperstones Quartzite. More recently it has been changed to a formation (BGS, 1991; Lynas, 1985a) on the basis of its lithological integrity and marked difference from rocks above and below.

The formation forms a feature which can be traced from south to north across the district (Figure 4). It has created the most strikingly stark skyline of the district with its isolated rugged pinnacles bearing evocative names like Black Rhadley, Nipstone Rock and Devil’s Chair. The base is exposed in Granham’s Moor quarry [SJ 3896 0372] where grey-white medium-grained quartzites (Bed 6, (Figure 5)) are separated from greenish grey, fine-grained sandstone of the Shineton Shale Formation below by 20 to 30 mm of grey clay (Lynas, 1985a).

The thickness of the formation at Linley Big Wood in the south is about 240 m. However, based upon the width of the outcrop, it differs greatly from place to place without any obvious trend. Whittard (1979, p.12) believed it to be between about 168 m in the south and 198 m in the north, but at Black Rhadley Hill [SO 3428 9550] in the south and Poles Coppice [SJ 3915 0455] in the north it is as much as 270 to 280 m thick. Nearer the centre of the outcrop, just south-east of Blackmoorgate [SJ 3776 0100] it appears to be only about 150 m thick.

The Stiperstones Quartzite is a hard, almost white, quartz arenite of even grain size. Secondary overgrowths of clear silica have welded the rock into an orthoquartzite. It is sparsely feldspathic in parts, but pebbles are present and in some beds they are common. The clasts are well rounded, up to 20 mm across, and consist of quartz, quartzite, indeterminate fine-grained volcanic rocks and welded tuff. Tabular rip-up clasts of mudstone up to 80 mm long also occur. Some beds are graded, with dark lithic grains and granules present in the lower parts. The base, seen only in two excavations, is sharp and rests with contrast on mudstones with fine-grained sandstones and siltstones. A thin basal conglomerate containing clasts similar to those above was recorded by Whittard (1931).

The formation is massive to well bedded with parallel, wavy and lenticular beds typically 0.10 to 0.50 m thick and exceptionally up to 1.5 m thick. Wave amplitude is generally 0.10 m. Faint cross-lamination is outlined by darker grey partings in some exposures. The quartzite is well jointed and the outcrop is characterised by craggy exposures rising from a surface draped with large angular blocks.

Pale grey to white quartzite dominates most of the formation, but towards the top the bedding is thinner and interbeds of dark grey siltstone are common, creating a passage into the Mytton Flags Formation. In lower parts of the formation there are a few thin layers of pale green clay and soft sandstone. These clays may be bentonite and if so record distant volcanism.

The formation was probably a shoreface deposit of well sorted, quartz sand, but its provenance is obscure for the immediate substrate consisted of fine-grained Tremadoc silt or siltstone. A land area of older, Longmyndian, rocks had almost certainly been exhumed not far distant, implying late Tremadoc or early Arenig tectonic movements, perhaps on faults. Such movements could account for the sharp lithological contrast between the formation and Shineton Shales below and also have allowed the possibility of a non-sequence.

Biostratigraphy

Bioturbation is a common feature of the formation. Vertical burrows, up to 100 mm long and 25 mm wide (Whittard, 1979), have been reported at certain horizons. Many are U-shaped burrows of Diplocraterion and trails of Cruziana, filled with sand coarser than the host rock. Such burrows indicate shallow water deposition. Apart from a few inarticulate brachiopods, and possible algal and sponge remains, the only fossil found in the formation is the trilobite Neseuretus grandior which typifies the Neseuretus community (Fortey and Owens, 1978) that characterised Gondwanan, shallow marine, arenaceous deposits of Arenig age. Fortey and Owens (1987, p.238) have synonymised N. grandior with N. ramseyensis from the Ogof Hên Formation (Lowest Arenig, Moridunian) of Ramsey Island, south-west Wales.

Details

Sporadic exposures along a forest track in Linley Big Wood (Figure 5), traverse a full thickness of the formation, of some 240 m thick, base at [SJ 3406 9493].

Whittard’s excavation [SJ 3896 0372] (1931, p.324) and Granham’s Moor quarry nearby [SJ 3895 0372] expose lower parts of the formation and the junction with the Shineton Shales (Figure 5).

One of the best exposures is the quarry [SJ 391 046] in West Poles Coppice where 100 m of medium- to coarse-grained quartz arenite are displayed. Pebbly beds occur, containing well rounded clasts up to 100 mm across and there are a few partings of green-grey soapy shale (possibly bentonitic) less than 0.05 m thick. Individual beds of arenite are typically 0.10 to 0.15 m thick, with a maximum of 1 m; they tend to pinch and swell. Broad ripples occur on a surface in the north-east part of the quarry and have an amplitude of 0.10 to 0.15 m with wave-length about 0.8 m. Other, less obvious ripples occur on the east side of the quarry, through many metres of strata, all with an alignment north-south. Intra-bed laminae are almost absent, but pale green lithic clasts occur together with a scatter of irregular striations and burrows 10 mm in diameter.

The rock is much fractured though there are some regular joints. The fractures are almost invariably occupied by bituminous hydrocarbons, and are commonly coated by a yellow sulphur-like compound. In a thin section of the typical quartz arenite (E59503), the rock is completely sutured by secondary quartz, and thus the name ‘quartzite’ may be appropriately applied to the formation. The clasts are dominantly quartz, many with a curious cross-hatched strained extinction, and dusted with fine inclusions; a few pelitic and polycrystalline quartz grains also occur. The lithic clasts of the pebbly facies (E59504) include foliated psammites, pelites, strongly sheared coarse quartz arenites, equigranular mosaics of quartz-feldspar and strongly sheared and altered feldspathic rock. The grain sizes are variable, making a poorly sorted rock, and grain shapes are difficult to determine, due to silicification, except for the larger, exotic clasts. The rock here is a pebbly lithic quartzite.

Mytton Flags Formation

These flags were first mapped by the Geological Survey (1850) as part of the ‘Llandeilo Flags and Sandstone’ a grouping which in fact embraced all the Ordovician sedimentary rocks of the Shelve area except the Stiperstones Quartzite. In 1887, Lapworth introduced the Ladywell Group as the basal unit of these rocks and the formation is equivalent to its lower half. Lapworth and Watts (1894) were the first to distinguish the formation as an entity and they gave it the name Mytton Flags. Since then there have been many groupings and subdivisions of the formation (Lapworth and Watts, 1910; Lapworth, 1916; Watts, 1925; Whittard, 1931 and following references). Lapworth (1916) for instance, registered a top division of the formation which he called the Tankerville Flags and Shales, Gravels and Shelve Church Beds, a middle division composed of the Ladywell and Snailbeach Grits (Flags) and Shales, and a basal division, the Lord’s Hill Beds, from which Watts (1925) removed the Lord’s Hill Beds. Each of these terms was pertinent to certain parts of the outcrops, and although Whittard (1955) reinstated many of them temporarily, they cannot be mapped and the names have lapsed into informal and imprecise usage. This memoir retains Lynas’ (1985a) formational status. Correlation with other areas in Wales is shown in (Table 3).

The main outcrop (Figure 4) occupies a linear tract on the west side of the Stiperstones Quartzite from the River Camlad near Roveries Bridge [SO 3268 9188] to the northern edge of the district at the White House [SO 3571 9768]. This outcrop is broken by many faults, including the Stiperstones Fault, a strike fault on its western side, which cuts out the higher parts of the formation. The outcrop extends northwards to near Pontesbury [SJ 400 060], where it is overstepped by Carboniferous rocks. In this latter area the outcrop is up to 1250 m wide and more fully representative of the formation; both top and bottom are present west of the Stiperstones and the thicknesses of the formation is estimated at 500 to 900 m. The type section is located in Mytton Dingle [SJ 367 005], where a great deal of the thickness is exposed along the northern side (see Details).

A second outcrop has been mapped across the high ground of Shelve Hill, in the core of the Shelve Anticline. It extends between Appletree [SO 3229 9690] and the Batholes [SJ 3411 0062] just north of which it is faulted against Hope Shales. There is no base within this outcrop and thus no firm proof of identity, but faunally and lithologically it is dissimilar from any formation other than Mytton Flags.

The Mytton Flags commonly consist of thinly inter- bedded pale grey siltstones and fine sandstones with medium to dark grey argillaceous and micaceous siltstone; mudstone is present sporadically. Pale grey siltstone-sandstone layers, less than 10 mm thick, are laminated; the laminae consist of angular quartz grains, with some feldspar, commonly about 0.03 mm across but may be up to 0.1 mm. Darker silt layers are homogeneous and micaceous and commonly 10 to 15 mm thick. The beds are ubiquitously burrowed so that the pale laminae are disturbed, commonly into subhorizontal and near-vertical pods (burrow fill); in places intensive bioturbation has homogenised the sediments. Weathering in exposures for example [SO 3243 9191] near Snead, produces protrusive harder ‘beds’ 20 to 100 mm thick and softer partings. The effect appears to be due to the relative concentration of silt/sand layers into bundles which are paler and tend to weather proudly.

The Tankerville Flags, Shelve Church Beds and Lord’s Hill Beds are disused names for local litho-variants; the Ladywell and Snailbeach Grits and Flags appear to have been the remnant bulk of the formation (Lapworth, 1916). The Lords’ Hill Beds represent the sandier transition into the formation from the underlying Stiperstones Quartzite; the Tankerville Flags may be considered as an argillaceous transition at the top of the formation into Hope Shales above. The Shelve Church Beds are restricted to the Shelve Hill outcrop and were described by Whittard (1931) as being light grey and micaceous with a peculiar brown-weathering property. He considered the peculiarity to be caused by a high percentage of volcanic dust. Lynas (1985a), however, recorded no abnormality in the rocks around Shelve Church other than that which was attributable to the contact metamorphism of an intrusive dolerite.

The Mytton Flags were deposited on a deepening marine shelf from a transgressive sea. Since the Stiperstones Quartzite represents a very shallow stage in the transgression, the deposition of the Mytton Flags cannot have been in very deep water, even though subsidence of the basin must have been rapid in order to accumulate nearly 1 km of strata (after compaction) in the span of two graptolite biozones.

The deposits were essentially of fine grain size; silt with thin, cleaner, strongly burrowed sandy beds. No massive sand bodies have been identified nor has storm-wave disturbance; channel-fill deposits likewise appear to be absent. The evidence thus indicates a low-energy, yet oxygenated, shelf sea. Atmospheric storms were mild or of small fetch, so that storm wave-base was shallow and storm currents transported only silt and fine sand. The transgressive sea probably flooded a subdued terrain which provided little coarse detritus; the position of the shoreline is not known. There is no evidence by way of variety or slumping in the deposits that the Pontesford–Linley Fault influenced deposition at this time.

Biostratigraphy

The Mytton Flags are referred to the Arenig Series, the lower beds to the Moridunian Stage and the tops beds to the Fennian (Fortey and Owens, 1987, p.98). Fossils are sparse, consisting predominantly of benthic trilobites and brachiopods with a scatter of pelagic graptolites.

The Mytton Flags yield graptolites in two areas, both exposing high parts of the formation. One is on the easterly outcrop at Bergam quarry [SO 3562 9975] (in the ‘Tankerville Flags’), the other is near Shelve Church [SO 3360 9905] (in the Shelve Church Beds). The fauna from the former includes Didymograptus hirundo, D. extensus linearis?, D. cf. goldschmidti (of Fortey and Owens, 1987). These occur in the gibberulus and hirundo biozones in the Lake District and in the rushtoni Biozone of the Fennian of South Wales (Fortey et al., 1990, p.131). In addition, Pseudisograptus? cf. geniculatus has been found in nearby mine waste 120 m east of Tankerville Mine [SO 3564 9946]. The fauna from near Shelve Church (Whittard, 1979) includes Didymograptus cf. deflexus, D. cf. inflexus, D. cf. nitidus, D. sparsus, Eoglyptograptus dentatus, E. shelvensis, and Oelandograptus cf. austrodentatus (group). Strachan (1986) believed that D. cf. deflexus and D. cf. inflexus probably indicate a sub-hirundo position, yet deflexed graptolites also occur higher; Fortey et al. (1990) accordingly attached greater significance to D. sparsus and the biserial graptolites, and assigned the beds near Shelve Church to the hirundo Biozone.

The brachiopod fauna includes 19 articulate and inarticulate taxa, of which Monobolina plumbea, Palaeoglossa myttonensis and Paralenorthis cf. proava are the most common. Mollusca include Redonia anglica, Riberia complanata and Oxydiscus perturbatus. The commonest trilobite is Merlinia major (recorded by Whittard (1964) as Ogygiocaris selwynii), but Fortey and Owens (1987, p.98) gave greater stratigraphical significance to the Neseuretus species, which suggest a Moridunian (early Arenig) age for part of the succession, and to the Whitlandian to early Fennian taxon Cyclopyge grandis grandis. Whittard (1955, 1965) described a variety of trinucleid trilobites — Myttonia, Incaia, Lordshillia and Bergamia — which, though important in the early evolution of the group, are of local occurrence and little value for correlation. The trilobites from the highest parts of the Mytton Formation are of Fennian (late Arenig) age (Fortey and Owens 1987, p.98).

Rushton (1985) referred the trilobites from the Mytton Flags to the Raphiophorid Community of Fortey and Owens (1978) and thus compared them with faunas of outer-shelf origin; low-energy marine circulation might have induced such a fauna to inhabit the shallower water of the middle shelf. The presence of trinucleids in the easterly outcrop, but not in the Shelve anticline, inclined him to suggest that the deeper water lay in the east. A more significant difference may be that between the faunas of the main bulk of the formation and those of the upper parts. The upper beds of the easterly outcrop at Bergam quarry [SO 3562 9975] — the ‘Tankerville Flags’ — have yielded eight trilobites of which only Ampyx cf reyesi (= A. salteri of Whittard) is known to occur in older beds. Shelly faunas from the Shelve anticline near Shelve Church and Wood House [SJ 3387 0016] have nothing in common with those of lower beds and only a few long-ranging species in common with the upper beds of the easterly outcrop. These differences are probably related to differences of depositional facies though it is not clear which is the more distal.

Details

The most southerly exposure [SO 3243 9191] of the Mytton Flags is south of the A489 road, some 40 m south-east of Roveries Lodge. The beds are nearly vertical, younging westward in a dip section 8.5 m long. They consist of thin layers of medium to pale grey arenaceous siltstone and interlaminated, streaky, darker grey, argillaceous siltstone all disturbed by infaunal burrows.

Bioturbated siltstones are exposed in a quarry [SO 3206 9772] near the old school at White Grit. The type specimen of Dictyonema cobboldi (Whittard, 1979, p.22) was obtained here. An even better exposure occurs behind a new building at Foxhill Farm [SO 3223 8847] where there are several metres of streakily bedded, medium to dark grey, argillaceous siltstone with discontinuous and uneven laminae of paler arenaceous siltstone. The rock here has a bedding fissility which imparts the characteristic flaggy appearance.

In thin section (E58872), angular quartz grains, up to about 0.03 mm in size, are accompanied by detrital mica, a little feldspar and indeterminate rock fragments. Some pyrite also occurs. A slightly coarser facies, a dark grey fine-grained sandstone with lenticular laminae, is accessible in a filled shaft west of Gritt Farm [SO 3237 9803] (E58840) where the quartz grains are mostly about 0.1 mm. In the vicinity of Shelve, several dolerites have intruded the flags; concomitant metamorphic effects are usually hardening and bleaching to a buff colour and it is apparently this contact metamorphism which is responsible for the ‘Shelve Church Beds’ facies. There are many good exposures of the flags north-east of the village, particularly in old quarries [SO 3382 9980] and at Shelfield [SO 3412 9968]. In these, the flags are hornfelsed by dolerite intrusions, and at the former site some 10 m of beds are exposed. The beds, bleaching aside, are typical of the area with sandstone laminae mostly less than 10 mm thick. In thin section (E58876), the rocks in the Shelfield quarry contain sandy laminae, up to 15 mm thick, consisting of angular quartz grains less than 0.3 mm in size. These laminae exhibit cross-bedding and are affected by microfaulting, presumably the result of differential compaction.

The contact with the Hyssington Volcanic Member is not exposed; exposures of both formations south of Shelfield indicate a gradual fining upwards and loss of the flaggy bedding. This change, as elsewhere in the area, corresponds with a break-in-slope which is fairly readily mappable.

Transition upwards from the Stiperstones Quartzite is visible at the top of Perkins Beach [SO 3722 9998] and also at Rock House [SO 3510 9639] where fine- to medium-grained, bleached sandstone flags are exposed in the track 30 m west of massive quartzite. Along the higher parts of the south-east side of Perkins Beach, there are exposures of massive feldspathic sandstone, probably equivalent to those at the head of Mytton Dingle. These may represent the ‘Lord’s Hill Beds’ of Lapworth (1916, p.37). The main sector of the outcrop, exposed in Perkins Beach, consists wholly of flaggy, streaky, grey siltstone and fine-grained sandstone. They are visible also in the dingle east of Tankerville, and in a quarry [SO 3587 9935].

There is a number of good roadside exposures and a small quarry [SO 3563 9976] at Tankerville, presumably the ‘Tankerville Flags’ of Lapworth (1916, p.37) and Whittard (1931, p.325). The beds are slightly more argillaceous and hence darker grey than the main sequence (‘Ladywell Grits and Snailbeach Grits (Flags)’, Lapworth, 1916) as seen in Mytton Dingle and Perkins Beach.

The upper contact is taken as the transition between the flags typical of this formation and grey monotonous shales typical of the Hope Shale Formation. The highest exposed flags [SO 3545 9940] are south-west of Tankerville where they are dark grey, poorly laminated, coarse siltstones.

In the quarry [SJ 387 049] at Callow Hill, the formation is typically flaggy, in beds 0.05 to 0.20 m thick. They are medium to dark grey coarse siltstones, laminated and intensely bioturbated. Higher in the succession, in the western part of the quarry, the flags are less well bedded and probably pass gradationally into the overlying Hope Shale which is visible in the older, western quarry.

Mytton Dingle (the type section) [SJ 367 005] provides a nearly continuous section through the entire formation; the interlaminated arenaceous and argillaceous siltstones, which comprise the formation, are abundantly exposed along the north side of the dingle. The lowest beds, exposed below the high-level path at the dingle head, are rather more massive fine-grained sandstones; some 100 to 200 m higher, they pass up into the typical flags which continue without substantial change to Stiperstones village, and are visible [SJ 3634 0061] below the main road.

Hope Shale Formation

The name Hope Shale is derived from Lapworth and Watts (1894 p.316). The unit was redefined as the Hope Group by Lynas (1985a) to incorporate all the volcanic rocks and shales (Figure 6) between the Mytton Flags and the Weston Flags formations (Table 3). This memoir has applied formational status to it as fitting with its homogeneity and mappability. The two main volcanic sequences within the formation are the Hyssington Volcanic Member in the east and the Stapeley Volcanic Member in the west. Whittard considered these members to be geographically separated parts of one and the same unit, the Stapeley Volcanic Member (Whittard 1979, fig. 20 and fig. 25). The shale above he called the Stapeley Shale Member, and the shale below he called the Hope Shale Member.

The Hope Shale occupies by far the largest outcrop of Ordovician rocks in the area (Figure 4), dominating its central area from near Snead and Broadway House in the south to Callow Hill Fort and Lower Grimmer in the north. Estimates of the thickness of the formation are rendered imprecise by the presence within the outcrop of the Shelve Anticline and the Llan (or Ritton Castle) Syncline. There are probably 1500 m of shale, mudstone and volcanic rocks. The shale overlying the Stapeley Volcanic Member is over 200 m thick to the north of Corndon Hill, but dies out to the south.

Gabbroic dolerite on Corndon Hill has hornfelsed the Hope Shale in an aureole 200 to 500 m wide, and these toughened mudstones have been quarried on the west side of the hill as ‘slate’.

The mudstone and shale of this formation are medium to dark grey, homogeneous and commonly silty and micaceous. Rusty weather-stains are common on fracture surfaces. Silt laminae, up to 1 mm thick, occur and are undisturbed; unlaminated mudstone between the laminae, contains small, paler grey, silty blebs which may have resulted from the disruption of laminae. A burrowing infauna might have been responsible for this, but together with a nebulous colour patchiness, it could be the effect of a sparse benthic epifauna present at the time.

Details

Hard and blocky shale occurs in the stream gorge [SO 3095 9987] near Valley Knoll. Shales high in the formation are visible almost continuously along the length of the brook from Grimmer Cottage [SJ 3410 0361] to Moplety Dingle [SJ 3394 0279]. In Ladyhouse Dingle and Cottage Coppice [SJ 3484 0296] the unconformity with overlying Llandovery sandstone is well exposed. The shale is leaden grey with orange to rusty-coloured weathered surfaces. It is also very fossiliferous containing didymograptids and fragments of trilobites. Highly fossiliferous, delicately laminated soft mudstones occur in Whitburn Dingle, west of Pentirvin [SJ 3288 0196] where they yield Ogygiocaris seavilli and other fossils.

Hope Shale is well exposed in Callow Hill quarry [SJ 3843 0489] where it consists of grey, rusty-weathering mudstone with bed-partings up to 1 m apart.

Hyssington Volcanic Member

The Hyssington Volcanic Member is made up principally of acid vitroclastic tuff, crystal and lithic tuff, laminated tuffite, and associated volcaniclastic sedimentary rocks interbedded with mudstones. In contrast, the volcanic rocks of the Stapeley Volcanic Member are mainly of intermediate to basic composition. The volcanogenic beds make prominent, mappable features revealing that individual beds are laterally impersistent. Thus there is a variable sequence of interbedded volcanogenic rocks and mudstones, as can be demonstrated behind Willow Lodge [SO 3129 9406] and because of the lateral variation the top and base of the member are difficult to define. The volcanogenic rocks are generally thickly bedded, and some show a vague foliation possibly due to loading. Normal and reverse grading occurs and, more rarely, cross-bedding. Some beds have been contorted into small ‘slump folds’ or disturbed and broken by thixotropic liquefaction. Others reveal load-casts on subfaces, small listric compactional faults and a ‘neptunian’ dyke, all of which suggest rapid, subaqueous deposition. Plane-parallel-lamination is present in some of the finer grained beds, but bioturbation has not been recorded.

The member crops out on both limbs of the north-east-trending Llan Syncline, but only a distal extremity extends into the western limb of the Shelve Anticline. In the southern part of the syncline it incorporates three separate volcanic units. These can be traced for some distance northwards on both limbs, coalescing in places, but eventually fingering out as the intercalated shales become thicker towards the north. The member also thins westwards and north-westwards, so that on the western limb of the Shelve Anticline no more than two or three tuffitic beds are present around Hope. These are composed of microscopic clasts in a carbonate cement; they are chert-like and are the ‘Chinastone Ashes’ of Watts (1925, p.341). In its area of maximum thickness, around the southern part of the Llan Syncline, the total thickness of the member is 400 to 500 m, within a total thickness of some 600 m of the formation. In the north-west around Gravels [SJ 3350 0010] the thickness is under 100 m, contained in about 200 m of the Hope Shale Formation.

Western limb of the Llan Syncline

The first unit of the Hyssington Volcanic Member, the lowest, occurs on the western limb of the Llan Syncline in the vicinity of Hyssington village. It commences with muddy volcaniclastic beds well exposed behind Pinfold [SO 3139 9447]. The clasts (E58117) are of pumice and a rock with a vitroclastic texture. The crystals are feldspar and corroded quartz set in a deep brown sub-isotropic base. Beds which are slightly higher within the succession are exposed in an old quarry [SO 3119 9404]. These are vitroclastic acid tuffs, over 1 m thick, pale grey in colour, hard and splintery. In thin-section (E58083), they show a well-developed vitroclastic texture, with bimodal shards of spike, rod and bubble forms. Feldspar and hollow quartz crystals are set in this vitric base. Southwards, and stratigraphically higher, an old quarry [SO 3114 9387] affords a fine section (Cave et al., 1985, fig. 2) through tuffs and sedimentary rocks intruded by a sill. A sample from just above the sill contact has (E56084) a well-developed vitroclastic texture and includes hornfelsed siltstone rip-up clasts (due to the proximal intrusion) and ragged clasts of pumice up to 8 mm across.

Slightly higher in the succession 12.5 m of strata are exposed behind Willow Lodge [SO 3129 9406]); above a shale in the first unit, excellent vitroclastic textures and large randomly orientated pumice clasts (E58115) are present in the mid part of the section.

Farther north and separated from the Hyssington exposures by a north-west-trending fault, this lowest unit forms a strong feature at Cefn Bank [SO 317 953]. Exposure is poor, but a thin section (E57548) reveals a vitroclastic tuff with scattered feldspar (albite) and quartz megacrysts. The rock consists largely of pumice volcaniclasts. Shard forms are especially well developed some near-complete bubbles are present, up to 1 mm in size, and sufficiently coarse to be visible in hand sample. The formerly glassy shards have devitrified to a mosaic of quartz-feldspar and a vague foliation is developed, although there is no evidence of welding.

The second unit of the Hyssington Member is well exposed in old quarry sections near the road in Fremes Wood [SO 3165 9395] (Cave et al., 1985, fig. 4) (Figure 7). Bed 1 is a rubbly mudstone. Bed 2 is a massive volcaniclastic sandstone (E58093) consisting of equigranular quartz and feldspar clasts and a few lithic grains including microlithic dolerites or trachytes. Bed 3 is tuffaceous shale. Bed 4 is similar to Bed 2 and has an erosive base. Bed 5 is a graded tuffite which consists (E58094) of an immature wacke-textured sandstone with variable volcaniclasts, largely matrix-supported. The clasts include glass, pumice and trachytic-textured lavas, and abundant plagioclase crystals. Bed 6 consists of laminated tuffaceous siltstones and sandstones. Bed 7 is rather massive, consisting of volcaniclastic sandstones, fine-grained in the lowest 2 m, with 15 m above concealed but forming the crest of Fremes Wood ridge. The top 2 to 3 m consist of streaky crystal-lithic tuffite. Bed 8 is rubbly shale. Bed 9 is another grey massive volcaniclastic sandstone with lithic clasts up to 0.1m long; it is crystal-rich near the base. Bed 10 is a hard grey sandstone, with lithic clasts, passing up into Bed 11, a finely laminated tuffitic siltstone. Bed 12, the highest bed in these quarries is similar to Bed 9. Clasts are concentrated near the base (E50895), and include limestone,trachyte, porphyry and basic lavas with intergranular texture, as well as plagioclase.

The third unit of the Hyssington Volcanic Member is the highest and forms a prominent ridge trending north-east from the Llan to beyond Bank Farm, in the core of the Llan Syncline. A section [SO 3193 9421] at the Llan (Cave et al., 1985, fig. 5) (Figure 8) exposes 27.5 m of mainly crystal lithic tuffs, commonly massive and medium grained, for example beds 7 and 9; Bed 5 is coarse grained and Bed 8 is fine grained. Beds 3, 6 and 11 are tuffites. Mudstone clasts tens of centimetres long are present in a tuffite (Bed 11) and at the top of Bed 9 where they are silicified (E58103); similar clasts near its base (E58102) are smaller and accompanied by volcaniclasts. Volcaniclasts are abundant and of various sizes. In Bed 1 a foliated crystal lithic tuff (E58099) in which the pumice clasts are flattened and, as is normal, original glass is devitrified to a mozaic of quartz and feldspar; Bed 4 is similar. Bed 5 is a coarse crystal lithic tuff of andesitic composition in which volcaniclasts are pink to white, of acid to intermediate composition and up to 100 mm in size. There is a passage onto Bed 6, which is a medium-grained tuffite with few small volcaniclasts. Bed 7 is a massive crystal pumice tuff of andesitic affinity. Top parts contain layers of large green relict pumice clasts showing reverse grading, and chlorite and carbonate alteration products are present (E58100) and (E58101). Ellipsoidal siliceous volcaniclasts are present also. Bed 8 is a very fine-grained crystal tuff, homogeneous hard and splintery, also of an intermediate affinity. Beds 9 and 10 are massive, mainly crystal tuffs with some lithic tuff in Bed 10.

Northwards is Tasker quarry [SO 3251 9565] (Lapworth and Watts, 1894, pl. VIII; Whittard, 1956, pl. 5); this is not the Tasgar (= Tasker) Quarry of Whittard (1979, p.32, Locality 270). It illustrates the diverse nature of the member. In this area, it is more difficult to identify the three volcanic units as here they are further subdivided by beds of shale. However, the volcanic rocks in this quarry are probably equivalent to the top part of the first unit. A composite section (Lynas and Langford, 1985, fig. 3) is shown in (Figure 9), the lower part of which was measured along the road north of the main quarry. The basal bed is a massive crystal-rich wacke. This grades up into mudstone (Bed 2) where only thin layers are crystal rich. Bed 3 is feldspathic sandstone. Bed 4 is a mudstone embracing a few lenticular feldspathic sandstones; it is capped by fine- to medium-grained tuffite (Bed 5), followed by more mudstone (Bed 6). ‘Bed 7’ is composed of medium to coarse-grained crystal lithic tuffites, in beds up to about 0.35 m thick, each with an erosive base. Bed 8 is formed of delicately laminated fine tuffaceous siltstones consisting of angular chips of quartz and feldspar and scattered shards set in an irresolvable matrix. The penultimate layer displays eroded ripples. Bed 9 can be examined on the west side of the quarry, where dip surfaces are prominent and where it has been hardened by contact with a basic sill (10) visible to the north of the quarry. It consists of a crystal-rich wacke with basal load casts and flame structure resting on laminated mudstone with thin layers of tuffite and characteristic circular pale mottles up to 10 mm across; there are scattered remains of trilobites. Above, Beds 11 (except its base, 11a) and 16 are massive feldspathic sandstones. Bed 11a consists of fine-grained tuffaceous sedimentary rock, faintly laminated and contact metamorphosed. Bed 12 is a crystal tuff mostly massive, but bedded in parts, especially near the top where the beds have erosive bases. Crystal tuff also forms thin beds within the tough grey mudstone of Bed 15. Bed 18 is a tuff, pale grey, hard and vitroclastic with colourless feldspars. The sandstones are fine to medium grained although coarser in Bed 17. Lithic clasts, mainly of mudstone up to 10 mm across, are common, and flat ellipsoidal bodies (possibly weathered concretions) of porous sandstone, up to 0.5m long, occur in Bed 16. Bed 18 is hard, pale grey, vitroclastic tuff with colourless feldspars. It is overlain by a dolerite sill (19) (E58815) with an amygdaloidal top and containing darker doleritic xenoliths (E58814)(Chapter seven). Bed 20 is a complex of vitroclastic tuffite with lobes and lenses of coarse crystal tuffite. A thin section across the two component lithologies (E58822) shows the coarser part to be a volcaniclastic sandstone containing feldspar crystals, and clasts of devitrified felsic volcanic rocks and pumice, which commonly exceed 1 mm in diameter. It passes abruptly into Bed 21, an even-grained vitric tuffite with small shards, feldspars and pumice, less than 0.15 mm in size. Two dolerite sills (22 and 23) occur, the upper one has chilled margins, top and bottom, and the upper part is also amygdaloidal. Bed 24 is a massive coarse chaotic wacke with irregular mudstone clasts, considered to be a mudflow breccia. It contains feldspar crystals, large feldspar porphyry volcaniclasts with a devitrified glassy base, pumice and siltstone rip-up clasts set in a matrix filled with pumice and devitrified glass clasts and shards; Bed 26, the top of the section, is closely comparable, but separated from it by grey mudstone of Bed 25, the top surface of which is affected by small normal faults.

Eastern limb of the Llan syncline

The three-fold division can be identified in the south where the first unit is considerably thicker and forms the hump-backed ridge of Cefn Gunthly, but northwards of Brooks Hill [SO 345 970] they are so split into sub-units as to be unidentifiable.

The first unit is exposed near the summit of Pellrhadley Hill [SO 3389 9581] (Lynas and Langford, 1985, fig. 5) (Figure 10) where it consists largely of volcaniclastic sandstone with interbedded tuffaceous mudstone. A thin section from the base of the unsorted volcaniclastic sandstone of Bed 1 shows that the abundant clasts in the base of the bed (E58801) are felsic, mostly feldspar porphyry, and are tightly packed in a dark brown matrix. Bed 2, with a sharp base, is a grey hard tuffaceous mudstone overlain sharply by Bed 3. This is an unsorted volcaniclastic sandstone, with scattered balls of mudstone. Bed 4 is a hard tuffitic mudstone, but from Bed 3 there is a transistion upwards whereby volcaniclastic sandstone, with devitrified pumice clasts, is confined to thin beds which become thinner and sparser. The overlying bed grades up into tuffaceous mudstone through a sequence of interlaminated beds. A thin section from the transitional beds (E58802) shows the rock has a well sorted framework of angular feldspar chips, pumice and devitrified glassy volcaniclasts; most clasts are less than 0.1 mm in diameter. Bed 5 is a grey compact crystal lithic tuff capped by a lenticular bed, Bed 6 which is a much finer grained tuffaceous sediment. Bed 7 is composite of massive beds of pumice and crystal lithic sandstone which have scour and fill bases. Spherical cavities, up to 0.2 m, occur and rafts of fissile fine-grained tuffite, like those of Bed 4, occur in lower parts. A thin section from near the top shows the rock to be unsorted and matrix-supported in places, with angular pumice, feldspar chips and other fine-grained volcaniclasts, commonly 0.25 to 0.2 mm across.

Farther north near Brookshill Farm, a major new excavation [SO 3436 9667] for a rifle range has provided a section through beds at the base of an upper leaf of the unit. The lowest exposures are of iron-stained mudstones with bentonitic or feldspar crystal-rich beds, and thicker tuffaceous beds, some showing penecontemporaneous distortion due to load casting and possible slumping. One such bed contains scoriaceous-looking cobbles which look like volcanic bombs. In thin section (E58829), they are intensely carbonatised, and have the appearance of pumice crystal tuffites. Within the mudstone, beds of soft-weathering laminated clays also occur. Higher up, the mudstones and volcaniclastic beds pass into more massive beds, and these form a distinct slope break to the north. They are largely inaccessible, and consist of pale grey, acid, crystal, lithic tuffs. These rocks die out to the north-east, and are not exposed north of this locality.

Tuffs, in what is probably the uppermost leaf of a divided first unit, are exposed discontinuously at the north end of Cefn Gunthly [SO 3325 9554] to [SO 3328 9566] (Lynas and Langford, 1985, fig.6). Interbedded silty mudstone (E58796) is strongly affected by cleavage. The base of the overlying bed (E58794) consists of an unsorted aggregate of densely packed pumice with a large range of shards and feldspar chips. It is a pumice crystal tuff. The middle of the bed (E58790) is dominated by feldspar crystals, grain size is up to 0.5 mm across; the top of the bed (E58791) is a volcanogenic siltstone with coarser tuffaceous laminae, maximum clast size being 0.1 mm. The topmost bed (E58792), (E58793) consists of a volcaniclastic sandstone with clasts up to 15 mm, formerly cemented by a little carbonate matrix. Clasts are angular and include welded tuff, rhyolite with perlitic cracks (clearly visible in hand samples), pumice, trachytic-textured rocks, flow-foliated felsites (commonest) and feldspar porphyry. The highest part of this bed (E58795) consists of a tightly sutured mosaic of clasts without any matrix; the clasts are dominantly trachytic-textured microporphyry, some of which show clear perlitic cracks, with some feldspar porphyry and felsitic trachyte. The rock appears to have intermediate compositional affinities, but its textural origin is problematical.

The second unit is exposed in Nind Wood, in a quarry [SO 3335 9597] (Loc. 263 of Whittard 1979, p.33). Massive, light grey acid tuffs, in a bed up to about 4 m thick (E58809), form the centre of the section. It contains a scatter of small pumice clasts, a few layers in which they are concentrated and larger lenticular bodies (possibly weathered concretions) of soft sandstone. The bed has a sharp uneven base resting upon 1.1 m of chaotically mixed mudstone, pumice clasts and feldspar chips (E58808). This, in turn, overlies 1.6 m of argillaceous siltstone which contains a bed, 0.25 m thick, of massive acidic crystal tuffite (E58807) near the top. It was this bed that yielded a Calymenid pygidium resembling Neseuretus spp. or Platycalymene tasgarensis. At the base of the section is another feldspathic sandstone and the top 2.7 m consists largely of grey siltstone, with feldspar chips, pumice clasts and shards concentrated in the lower half (E58810).

Volcanic beds and interbedded mudstone of the upper part of the member are well exposed near Ritton Castle where the river flows through a small gorge. The volcanic beds are complex in detail, but continue to be dominated by acid vitroclastic tuffs. They have a cherty appearance. At one exposure [SO 3456 9777], silicified grey vitroclastic tuffs consist of scattered shards, pumice and feldspar crystals set in a brown isotropic matrix (E58833). On the other side of the gorge, tuffites and mudflow deposits are overlain by a massive bed, 0.50 to 0.60 m thick, with a load-cast base and a top which shows scour-and-fill structures. Large porous clasts are common, together with mudstone clasts, both up to 0.20 m across. The base of this bed (E58834) comprises a variety of clasts: feldspars, pumice, scoria, perlitically cracked glassy rhyolite, feldspar porphyry with dark matrix containing tiny microlites, trachytes and siltstone rip-ups. All are angular with sizes up to 3 mm, and they are set in sparse dark matrix. Overlying this tuff bed is another similar, well sorted rock with a few rip-up clasts (E58835). It is succeeded by several metres of flinty commonly vitroclastic acid tuffs.

At the northern end of the Shelve Anticline are the three ‘Chinastone Ashes’ of Watts (1925, p.341) referred to as ‘dust-tuffs’ by Blythe (1938, p.397). The lower two are exposed in the Hope Valley. At one exposure [SJ 3515 0191], the characteristic rock, thickly bedded, cherty acid tuff, banded pale and darker grey is present with some vitroclastic coarser beds. Like counterparts to the west, the beds are lenticular, some with lobes or truncations indicative of wet sediment loading or slumping. In thin section (E59497), the vitroclastic texture is fairly clear; devitrified shards are recognisable throughout, accompanied by pumice clasts.

Details

North of Runnis, by the old ruin of Tasker [SO 3276 9603], coarse feldspathic tuffites are exposed, and illustrate the difficulty of differentiating tuff from tuffites and volcaniclastic sandstones. They form beds 0.10 to 0.20 m thick, with sharp tops and bases which are planar and parallel. They consist of feldspar chips, 1 to 4 mm across, which are commonly euhedral, clasts and matrix. Internal layers, 10 to 30 mm thick, have indistinct, gradational boundaries, and are defined by size and abundance of clasts. No sedimentary structures are visible and the beds are not graded. They may represent ash fall-out into still water, without subsequent reworking.

A small modern quarry on Mucklewick Hill [SO 3309 9704] reveals the nature of some of the generally poorly exposed volcaniclastic beds. The lowest beds are mudstone, with crystal tuff interbedded with laminated fine-grained tuffites; the tuffites show small-scale current bedding and slump structures in places. These pass up into thicker bedded, crystal lithic (mostly mudstone clasts) tuffites which form the east edge of the Mucklewick Hill ridge.

A new quarry [SO 3319 9710] exhibits bedding surfaces, some 40 m long, veneered with thin mudstone beds upon which there are scattered trilobite remains. Branching and gash-like ‘veins’ of the underlying tuffaceous sandstone have been injected into the mudstone by liquefaction and constitute evidence of compactional dewatering. At the north-west end of the quarry, a short dip section through the tuffites reveals load-casting and slumping, and the sandstone can be seen to bulge into the overlying mudstone. These beds are underlain by well-bedded vitroclastic pumice tuffs (E58828) seen also in natural exposures nearby. The freshly exposed bed-surfaces of these tuffs are covered with dark flattened discoidal masses which may be collapsed scoria lapilli; some exceed 10 mm in size, and commonly contain feldspar crystals.

West of Berth House [SO 3333 9731], a 50 m track exposes similar rocks. They include massive medium-grained crystal tuffites with rip-up clasts of mudstone. Thin mudstones show evidence of soft sediment injection. A lithic tuffite contains scattered trilobite fragments, and consists of pink and white clasts making it distinctive. In thin section (E58832), the clasts are seen to be set in a dark shardic matrix and include feldspar crystals, pumice equigranular felsites, feldspar porphyry, foliated felsic rocks with snowflake texture and a more basic rock with feldspar microlites defining a trachytic texture. This last clast type is abundant in part of a section at the north end of Cefn Gunthly [SO 332 955].

North-east of these localities, the tuff members of the formation thin and in some cases disappear completely, and exposures are sparse because of a thick mantle of drift. South of Shelve, on Grit Hill [SO 3393 9809], vitroclastic acid tuffs are visible. The most northerly exposures within the district are afforded by a small quarry north-east of Shelve [SO 3447 9945]. This reveals several metres of variable tuffs and tuffites. In the northern quarry, fine-grained pale grey laminated tuffites are interbedded with volcaniclastic sandstones of basaltic affinity (E58880).

Stapeley Volcanic Member

The member is present only in the western limb of the Shelve Anticline (Figure 4) where it strikes north-north-east and dips to the west. The main part of the member forms the high ground of Todleth Hill, Roundton and Lan Fawr, with lower units cropping out to the east. The outcrop is over a kilometre wide in the south where the member is almost 1000 m thick; it thins northwards to some 50 m at the northern limits of the district.

The Stapeley Volcanic Member is composed of tuffs, tuffites and volcanogenic sandstones, generally of basic composition. Few of the beds were truly pyroclastic; most are interpreted as the deposits of water saturated debris-flow deposits. The base of the member is identified as the lowest easily definable tuff in a sequence, interbedded with mudstone and shale of the upper part of the Hope Shale Formation (Figure 6).

The mudstone interbeds of the Hope Shale Formation are variously medium to dark grey, khaki or pale buff in colour, depending on the proportion of volcanogenic mud which has been incorporated. The paler mudstones are commonly thinly bedded with uneven top and bottom, or they may be lenticular. Some are distinctly graded, coarser at the bottom with sharp bases. The beds appear not to be disturbed by burrowing, yet some dark areas near the tops indicate apatite enrichment so that marine conditions, at least in places, were not anoxic. However, the nature of the substrate and the rapid sedimentary input were perhaps inhospitable to an infauna. Even the mudstones contain much variously sized tephra in places and may have been of density or mass-flow emplacement.

The lowest units of the member are separated from its main part by shales of the Hope Shale Formation. One of these units is exposed in quarries north and south of Hurdley Farm [SO 2955 9416], revealing thickly bedded, massive volcaniclastic sandstone and wackes, with thin interbedded tuffaceous siltstone turbidites which are evenly bedded and laminated (see Details).

The main part of the member is well exposed on Todleth Hill, Roundton and Lan Fawr. In this area, Lynas (1983) distinguished two suites of strata, one centred upon Todleth Hill [SO 2890 9450] the other upon Lan Fawr [SO 297 967]. Exposure on Lan Fawr is comparatively poor, but the sequences are similar so that both can be exemplified by that on Todleth Hill and divided into three units. A lower, tephra unit comprises coarse lapilli tuffs and tuffs with few lapilli, passing up into laminated feldspathic tuffs with lapilli beds. A central unit consists largely of basaltic andesite which is thickest, up to 300 m, on the southern part of Todleth Hill. The rocks are highly vesicular with trachytic textures and are blocky in part. An upper, tephra unit which is a complex of pyroclastic and epiclastic deposits includes bedded volcaniclastic sandstones, tuffs and lapilli tuffs. Interbedded within the lower part of this unit on Lan Fawr is a basaltic lava. South of Lan Fawr, the pyroclastic beds beneath the lava thin rapidly, while the lava thickens and coalesces with the lavas of the central unit. The top part of this unit passes up in places into a matrix-supported boulder bed in which the matrix is a crystal tuffite with feldspar fragments up to 50 mm across. The boulders consist of porphyritic basaltic andesite; they are up to 0.8 m across, abundant and so well rounded as to indicate an origin as river or beach pebbles.

The basaltic andesites of the central unit were originally considered as extrusive by Lapworth and Watts (1894), but later interpreted as intrusive (Watts, 1925; Whittard, 1979). Lynas (1983) supported the earlier view that the rocks are extrusive and notes that they are interbedded with well-bedded lapilli tuffs of the same composition and that there is no metamorphism of incumbent strata. He interprets the block-rich parts as lobe-like flow fronts to the lavas. In addition, clasts in the overlying Weston Flags are identical to the basaltic andesites, and the Weston Flags appear to onlap the lavas in the Todleth Hill–Lan Fawr area, suggesting subsidence of an eroded volcanic edifice. Thus the igneous rocks were either shallow intrusions which were eroded and unroofed rapidly or, more likely, were lava flows and subject to instantaneous erosion.

Details

In the south, 17 m of the lowest beds of the member are exposed in Upper Hurdley quarry [SO 2961 9445] (Hains and Lynas, 1985b, fig. 1) (Figure 11). The underlying mudstone, Bed 1, was seen in an excavation at the northern end of a nearby house. It possesses thin beds of graded tuffaceous sandstone which thicken upwards to become dominant. Bed 2 is a volcaniclastic wacke with large deformed rip-up clasts of mudstone, up to 0.2 m long, near the base. Volcaniclasts, mostly about 0.1 m in size, are both volcanogenic and sedimentary in origin, and small synsedimentary faults occur, throwing down northwards by up to 0.5 m. Bed 3 is a well-bedded tuffite, which weathers to a buff colour, overlain by a massive bed (Bed 4) of volcaniclastic feldspathic wacke with mudstone rip-up clasts at the base. Bed 5 is a composite of thin, evenly bedded, laminated, silicified tuffaceous siltstone turbidite, interbedded with dark grey shale. Each siltstone is about 0.12 m thick and the bases of the lower beds are erosive. Some show load casts and flame structure. Burrow-fills are present too, as are small synsedimentary faults. The top of the section is marked by massive volcaniclastc sandstone (Bed 6) with a few pumice blocks. Exceptionally, volcaniclasts are 0.2 m across.

The lower, tephra, unit of the member is displayed on the east side of Roundton where it forms columnar-jointed crags. The lowest bed [SO 2933 9473], consists of coarse lapilli tuffs interbedded with tuffs containing fewer lapilli. The rock is dominated by prehnite (E56906) which occupies cavities between the closely packed basaltic scoria, and some pumpellyite has also been identified (Bevins and Rowbotham, 1983). Some of the scoriaceous lapilli show both trachytic alignment of feldspar microlites and stretching of vesicles parallel with flow direction. The lapilli also contain clinopyroxene, and show seriate, intersertal and intergranular textures. Vesicles are 0.1 to 0.3 mm in size and in hand specimen give the lapilli a distinctive ‘frothy’ appearance, a characteristic of the lapilli tuffs on Todleth Hill and Lan Fawr. The vesicles are filled with chlorite, carbonate or quartz.

The lapilli tuffs pass up into laminated feldspathic tuffs with lapilli beds, which, higher in the hillside, become wholly massive lapilli tuffs with well-developed columnar jointing. Exfoliating ‘cannon ball’ concretions occur in parts of the section, which may be load casts of the lapilli tuffs into the finer underlying ash beds. The exposures around the summit of Roundton are well-bedded tuffs and lapilli tuffs with distinct lapilli beds 0.5 to 0.10 m thick. At three localities [SO 2930 9482], [SO 2935 9488 and [SO 2938 9507] the lapilli tuffs can be seen passing directly into blocky lavas with a rubbly appearance. Dips in the stratified tephra are variable as might be expected if the tuffs are proximal, and may represent original depositional dips of individual tephra showers on the flanks of a growing volcano. In some sections, they show clear resemblance to modern scoria deposits. The lapilli are invariably scoriaceous and some are pumice-like.

The former Roundton Mine exposes [SO 2927 9473] massive jointed lapilli tuffs at the base. Above, there are roughly bedded lapilli tuffs, about 5 m thick, which in thin section (E56904) can be seen to be basic hyalotuffs. They contain scoriaceous lapilli and numerous glassy fragments. Overlying these are 10 m of massive to faintly bedded, feldspathic tuffs (E56903) which are similar to, but finer than the lapilli hyalotuffs.

The lava unit on Roundton is well exposed on the western slopes forming crags [SO 2927 9482] [SO 2908 9507]. The rock is extremely vesicular, and appears to have been completely autobrecciated. Vesicules are elongate and infilled with calcite. The lava (E56898) has an interstitial to pilotaxitic texture with a large feldspar microlite (less than 4 mm) phase giving a microporphyritic appearance. In contrast, the lavas on Todleth Hill are massive and usually porphyritic with megacrysts of fresh clinopyroxene as well as plagioclase. The lower tephra unit is overlain [SO 2922 9420] by massive, green aphanitic basaltic andesite lava. The rock (E56857) possesses a microlitic groundmass, with an interstitial to intergranular trachytic texture of green sub-isotopic phyllosilicates enclosing plagioclase, possibly after orthopyroxene. Other pseudomorphs show the curved fractures characteristic of olivine. Rounded euhedra of glomero-porphyritic clinopyroxene (augite about 2.5 mm) are common.

The upper tephra unit is exposed on the west side of Roundton at The Pant [SO 2903 9495] as bedded volcaniclastic sandstone (E56912), tuff and lapilli tuff. A small quarry at the summit of Todleth Hill [SO 2884 9453] shows massive vaguely foliated feldspathic tuffs with scattered lapilli. Adjacent are loosely cemented block lapilli tuffs with a steep plane of separation. The blocks exceed 0.5 m [SO 2871 9452] and may be part of a small vent-fill.

On the western slopes [SO 2862 9404] of Todleth Hill, block lapilli tuffs pass up into a boulder bed [SO 2858 9396] where it forms a dip slope. The matrix appears to be a crystal tuff incorporating pebbles, cobbles and boulders, commonly well rounded and some exceed 1 m in diameter. They are composed of highly altered feldspar porphyry with glomero-porphyritic texture and thus similar to the lavas below. A small quarry [SO 2857 9389] exposes the top of the conglomerate overlain by massive feldspathic sandstone of medium grain size which is disposed with a dip (33°W) different from the underlying strata (57°W).

A boulder bed is seen also at the south end of Simon’s Castle quarry [SO 2854 9327] where tuffaceous rocks overlie the lava, a basaltic andesite. The top of the lava contains a thin boulder bed, while rounded and some angular clasts of porphyry, up to 0.40 m across, extend at least 2 m below the bedded material. More boulders are present high in the south-west quarry face.

Resting upon the boulder beds on the western flank of Todleth Hill is an acid vitric tuff exposed [SO 2849 9372] north of Todleth House, similar to one within the topmost Hope Shale farther north. The component, massively bedded, acid, crystal tuffs are exposed [SO 2849 9372] and can be traced north and south for a little over 100 m. The beds are homogeneous and unsorted. The rock (E56876) consists of devitrified and silicified pumice clasts and feldspar crystals, the latter are usually euhedral, about 1 mm in size and exceedingly altered. The volcaniclasts include sparse basic clasts, and are mostly 1 to 3 mm in size. In some beds, the feldspar megacrysts are ill-sorted and of irregular shape. Thin beds with a conchoidal fracture (E56877) possess a vague vitroclastic texture and appear to consist of variably recrystallised glass, crystals and volcaniclasts. Flakes of this siliceous material are present within some of the beds of lithic crystal tuff developed here. In (E56878), small (less than 0.1 mm) shards and broken pumice are clear and, along with feldspars, are set in a dark matrix. Large irregular clasts, over 10 mm across, show slight flattening, while most of the clasts are recrystallised probably from glassy porphyry or aphyric glassy lavas and not true pumice. Scattered, small, rounded, primary quartz phenocrysts are visible.

Individual beds rarely exceed 0.20 m in thickness, and some beds possess a streaky foliation. In thin section (E56879), green porphyritic clasts with cryptocrystalline chlorite are visible probably replacing basic glass. Some of the clasts retain evidence of a pumiceous nature in their centres, but this primary fabric is progressively destroyed by secondary recrystallisation towards the margins which now comprise mosaics of quartz and feldspar. A few shards may be discerned in the brownish matrix. Thin section (E56878) is similar.

Exposures of the member north-east of Lan Fawr are meagre. Clastic material, which was probably redistributed and distal, is visible [SO 3088 9901] to [SO 3140 9978] in the fault-controlled north-west scarp of Stapeley Hill. Adjacent to an old level [SO 3088 9901], the rocks consist of polymict coarse beds, with clasts up to 10 mm across, ranging from felsic porphyry to variable basic rocks and scoria (E57388); the clasts are angular and poorly sorted in sparse high-relief muddy matrix.

The best exposures of the Stapeley Volcanic Member in the north occur between Luckley Hill [SJ 326 018] and Leigh Hall [SJ 3332 0358] with the lowest beds being exposed at Leigh Manor [SJ 3371 0224]. Thin layers of bentonite occur at Blue Barn in a silage pit [SJ 3340 0268]. Tuffs and volcaniclastic sandstones form both the east and west sides of Whitburn Dingle [SJ 3323 0270]. The highest, and most persistent beds in this area are exposed on Luckley Hill. Near Luckley Gate [SJ 3247 0165] coarse- to medium-grained, feldspathic, lithic tuffs are exposed. The rock is unsorted, containing abundant feldspar chips and varied volcaniclasts which include aphyric and porphyritic intermediate rocks, sub-isotopic microphyric lavas and sparse examples of perlitically fractured lavas (E59482).

Laminated feldspathic tuffs are exposed in Whitburn Plantation e.g. [SJ 3313 0307] but at Leigh Wood [SJ 3335 0342] they are massive or crudely bedded. Evenly grained feldspathic sandstone overlies these tuffs and volcaniclastic sandstone and is exposed in quarries north and north-west of Whitburn Cottage [SJ 3316 0307].

A unique conglomerate is exposed [SJ 3365 0400]; [SJ 3343 0372] near Leigh Hall. Clasts range from granules to boulders, 1 m across. They are well rounded and in an exposure several metres thick no grading is discernable. It is clast supported with sparse lithic-feldspathic sandstone matrix. The clasts are all of a monzodiorite, not matched by anything exposed in the district. In thin section (E59496), the rock contains abundant large (up to 1 mm) twinned albite and orthoclase phenocrysts, with conspicuous intergranular quartz making up 5 to 10 per cent of the rock. Lath-like biotite or hornblende pseudomorphs are composed of chlorite, which also forms greenish patches. The relationship of this rock to the rest of the Stapeley Volcanic Member is not known. Whittard (1931) referred to it without comment as ‘volcanic agglomerate’, and Blyth (1938, p.410), also without comment, called it a ‘vent agglomerate’. The monolithological nature of the clasts supports this view, but it seems unlikely that the perfection of rounding shown by the clasts could have been the result of mechanical abrasion in a vent. Furthermore there is a total absence of the distinctive monzodiorite from the rest of the member.

Biostratigraphy

The Hope Shale Formation belongs to the early Llanvirn Didymograptus artus Biozone (Fortey and Owens, 1987). From the mudstones, a mixed graptolite and trilobite fauna has been obtained including the graptolites D. (D.) artus, D. (D.) stabilis, D. (D.) spinulosus (= D. pluto of Strachan, 1986), Acrograptus acutidens and species of Glossograptus and Pseudophyllograptus (Strachan, 1986).

The trilobite faunas of all parts of the formation, of which more than 40 taxa were described by Whittard (1955–1966), are more uniform than those of the Mytton Flags, but they retain similarities to Raphiophorid communities. There is some resemblance to the atheloptic community of Fortey and Owens (1987, p.105), in that most of the trilobite genera are blind, for example species of Ampyx, Cnemidopyge (described by Kennedy, 1988), Stapeleyella, Dionide, Placoparia, Seleneceme, Ectillaenus and some agnostids (Whittard, 1966); some large-eyed cyclopygids are present, but in general they are not common. As a departure from the typical atheloptic community some trilobites had well-developed eyes, for example Platycalymene tasgarensis, Ogygiocaris seavilli, Ogyginus spp. and Barrandia homfrayi.

Apart from Protolloydolithus neintianus and Cnemidopyge pentirvinensis, trilobites from the Stapeley and Hyssington volcanic members are of species found also in the shales and mudstones below them. Like the shales and mudstones, the Hyssington Volcanic Member of the south-east outcrop and the Stapeley Volcanic Member of the north-west outcrop show faunal similarities. Of the 25 species of trilobite known from the Stapeley Volcanic rocks of Whittard (1955–67), but excluding his Stapeley Shales, seven are peculiar to the Stapeley outcrops and seven to the Hyssington; eight, including most of the more abundant species, occur in both Stapeley and Hyssington volcanic members; there are no data for the remaining three. As examples of the differences, Dionide and Placoparia are confined to the Hyssington member and Cnemidopyge and Cornovica to the Stapeley member. Williams (1974) described the brachiopod faunas which is almost wholly of inarticulate taxa. A few mollusca are recorded, including orthoconic nautiloids and hyolithids.

The faunas provide no evidence for there having been depositional facies differences between them, and it is known that both members were deposited within the limited time-span of P. neintianus, although there is indeed no faunal proof for any time differential between them, yet it cannot be excluded. It may be a valid reflection that in comparison with the time spans of ‘mid-shelf to outer shelf’ species of trilobite, these volcanic members are even more fleeting.

The Hope Shale Formation is the biozonal equivalent of the Didymograptus bifidus Shales of the Builth–Llandrindod area in which volcanogenic rocks are sparse or absent, in marked contrast to the Shelve sequence (Table 3). Such a difference supports a belief that there was a separation between the two areas at the time of deposition. Were they contiguous at the time and separated subsequently by major strike-slip faulting, some lateral expression of the Shelve volcanic rocks might be expected. In other respects marine conditions were similar; both formations reflect dysaerobic water and anoxic mud sedimentation with a sparse trilobitic benthos, albeit of different specific aspect.

Weston Flags Formation

The name ‘Weston Group of Grits’ was given to this formation by Lapworth (1887a, b) and later named the ‘Weston Flags and Shales’ by Lapworth and Watts (1910). Although various changes were made in subsequent years, Lynas (1985b) reinstated the term flags and formally defined the unit as a formation; his name is retained as definitive and descriptive, based upon the hamlet of Priest Weston; the quarry there [SO 2918 9729] is the type section.

The formation has an outcrop some 0.8 km wide from Hoarstone in the south [SO 2833 9342] to Leigh Hall where it disappears northward beneath Silurian strata (Figure 4). The thickness of the formation is estimated to vary between 100 m and 600 m.

The formation consists mainly of fine-grained, plane-bedded, flaggy and micaceous sandstone, containing interbeds of siltstone and shale, but within these are several more massive, coarser tuffitic sandstones which occur as individual beds or groups (packets) of beds. These form topographical features and make mappable but unnamed members. For this reason the Weston Flags are considered to be a formation and not a member as Whittard (1979, p.38) preferred. The more massive, tuffitic sandstones attracted quarrying operations for building stone, and such quarries have fostered a biased impression of the nature of the whole formation. As pointed out by Whittard (1967, p.39), the Priest Weston quarry is atypical of the formation as a whole. Bioturbation is common in the formation, particularly Skolithos burrows of up to 10 mm in diameter and bedding-parallel burrow casts.

The massive sandstones are fine to coarse grained, some beds show normal grading and other, thinner, beds within the packets show cross-laminae. They consist of quartz, feldspar and lithic grains, with volcaniclasts such as pumice, trachytic and porphyritic lavas, up to 20 mm across, reflecting an igneous source. Thicknesses of beds of sandstone range between 0.02 and 1 m. Some beds may be composite and many have load-cast pseudonodular bases.

At least two thick packets of beds in the lower half of the formation have been mapped over several kilometres. These are developed best in the central tract of the outcrop, between Priest Weston and Castle Ring where they form high ground including The Rowls and Rorrington Hill. The lower packet reaches a thickness of 100 m, the upper is about 50 m thick. Other beds of coarse sandstone are less persistent and lensoid. Slump disturbance of bedding is present in places, for example [SO 2983 9883] near Middleton Farm.

The tuffites are the only evidence of possible contemporaneous volcanic activity, although it is possible also that the volcaniclastic debris was derived by erosion of the earlier volcanic rocks in adjacent areas. The intensive burrowing seen on the bases of sandstone beds reveals that the marine bottom conditions, poorly oxygenated in Hope Shale times, had become well oxygenated for the Weston Flags deposition. At Builth, conditions remained poorly oxygenated during this time, indicating possibly deeper water. Thus it is unlikely that the Builth area was a source of detritus for the Weston Flags.

Biostratigraphy

The fauna of the Weston Flags is of low diversity and dominantly shelly. The commonest trilobites are Ogyginus corndensis and Platycoryphe vulcani, and of the five species of brachiopod present only the dalmanellid Tissintia prototypa is recorded as common (Williams, 1974). Mollusca are relatively numerous and diverse locally and Iocrinus shelvensis makes its appearance. Didymograptus (D.) murchisoni is the only graptolite recorded and the formation is considered to occupy the lower part of the D. murchisoni Biozone.

The faunas represent a complete change from those of underlying beds. The abundance of Ogyginus and Platycoryphe suggest reversion to a Neseuretus-type trilobite community notwithstanding the presence of Bettonolithus chamberlaini (equivalent to Bettonia frontalis of Whittard). The restricted brachiopod fauna and abundant molluscs are, like the sedimentological characteristics of the formation, typical of fairly high energy, shallow-marine deposits.

The formation correlates (Table 3) with part of the Didymograptus murchisoni Shales, ‘Newmead’ and ‘Builth’ volcanic rocks of the Builth–Llandrindod area to the south (Elles, 1940; Jones and Pugh, 1941). These contain suites of rocks, of great thickness, which include pillowed porphyritic andesites, together with other intermediate lavas and tuffs, agglomerates and acid tuffs. These volcanic rocks are interbedded with graptolitic mudstones indicating the persistence of poorly oxygenated bottom waters, in contrast to the oxygenated shallow conditions which had been established near Shelve. Such differences indicate that the Builth–Llandrindod rocks are unlikely to be a strike-fault displacement of the Shelve sequence.

Details

Behind Hoarstone Farm [SO 2832 9343], small quarries display immature heterolithic tuffites and tuffitic sandstones. Angular clasts, commonly of pale green porphyritic lava, are up to 10 mm in size; rounded quartz grains are of similar size. Feldspar chips and mud pellets are common. A thin section (E56869) reveals clasts of monocrystalline and polycrystalline quartz, feldspar and abundant angular clasts of basic lavas with a trachytic texture. These range in size up to 2 mm to form a non-sorted packstone. Matrix is sparse in the coarser beds. Numerous interbeds of well-laminated siltstone, up to 0.20 m thick, occur.

A quarry [SO 2824 9370] exposes well-bedded, bioturbated, laminated, micaceous siltstones and interbedded sandstones with uneven surfaces. The sandstones range in thickness from 0.01 to 0.5 m. They are polymict, of various grain sizes and clasts include well-rounded quartz, angular feldspars and rock fragments, up to 20 mm across, which include mud pellets, pale grey siliceous igneous rocks, pale green porphyritic lava and other indeterminate igneous rocks. A thin section (E56872) reveals that the angular clasts came from a great variety of lavas, some more felsic than others, and metasiltstones. All are tightly packed and have an unsorted grainstone texture.

Similar rocks are exposed near Old Church Stoke, in particular the bioturbated flags and coarse, speckled sandstone (E56915) exposed in a quarry [SO 2889 9520]. The last reveals rounded, clear quartz grains, some polycrystalline, green basic volcaniclasts and feldspar chips as at Hoarstone Farm.

The highest beds are exposed [SO 2922 9839] in a brook at Little Weston. Flaggy micaceous sandstones and bioturbated laminated siltstones contain Ogyginus corndensis cordensis, Platycoryphe vulcani, Schizocrania? sp., Deceptrix?, Praenucula?, Similodonta?, hyolithid and a crinoid stem (cf. Iocrinus shelvensis).

The feature-forming sandstone packets have been quarried north-north-east of Priest Weston. The lower packet is exposed behind the Miner’s Arms [SO 2933 9725] in Priest Weston and to the north-east at [SO 2960 9753] where flaggy to massive sandstones, tuffites and some vitric tuffs, in beds up to 1 m thick, are exposed. The tuffite (E57348) consists of feldspar (dominant), some rounded quartz in grains (0.5 mm) larger than surrounding clasts (over 0.2 mm), basic volcaniclasts and some brown shards in a green chloritic matrix. Slumping is present in a quarry [SO 2909 9576] to the south.

The upper packet is visible at its southern limit in the type section, where about 3 m of massively bedded, medium-grained, grey-green, quartzo-feldspathic sandstone have been quarried [SO 2918 9729] (Whittard, 1979, p.38). The rest of the beds in the quarry are fine-grained, bioturbated, flaggy, micaceous sandstones and silty sandstones. Some bedding surfaces are uneven, and weathering has produced spheroidal fractures. Branching burrows are present. A greater degree of lamination of the sandstones is present farther north [SO 2964 9804] where small scour-fills are visible.

Similar ‘wash-outs’ together with load casts, rippled bed surfaces and much bioturbation are present in the quarry [SO 2990 9863] near Middleton, also in the upper packet, as are Skolithos burrows (10 mm diameter) and bed-sole ‘fucoid’ marks displayed [SO 3010 9848] in Middleton Beach. A few large round quartz grains (over 1 mm) are present. The majority of clasts are volcanogenic including feldspar chips, scoria and glassy pumice (E57349) and (E57350).

In the north-east, the formation is well exposed along the outcrop, the best being in the sandstone packets, for instance the lower sandstone packet in crags and quarries [SJ 3072 0004] on the south flank of Rorrington Hill (E57376) and (E57377) and the topmost packet around Village Farm [SJ 3205 0225] (E59463) where slumping has occurred.

Betton Shale Formation

Lapworth and Watts (1910) referred to these beds as the Betton or D. murchisoni Shales. Before that, they had been the lower part of the Middleton Group (Lapworth, 1887a, b). Whittard (1931, and following references) consolidated the name Betton as a member within the Middleton Formation, and Lynas (1985b) raised the beds to formational status on the grounds of lithological distinctiveness from the Weston Flags below and the Meadowtown Formation above.

The outcrop extends from near Church Stoke in the south to Betton Dingle in the north (Figure 4). At both ends, the outcrop plunges below Silurian strata of low dips; in the south this unconformity is concealed beneath Quaternary deposits. The outcrop is discontinuous, being broken by at least seven near-strike faults. At Betton Dingle, the thickness of the formation is about 120 m. Farther south it is probably nearly 200 m thick. Exposures are sparse and of limited extent, for the formation has given rise to cultivated low ground. The boundaries with the adjacent formations are poorly documented; Whittard (1979) reports that both the bottom and top the formation show gradational boundaries.

The distinctiveness of the Betton Shale stems from the very dark grey, soft, shaly mudstone, commonly silty and micaceous, which forms the bulk of the formation. It can be seen in Little Weston Brook [SO 2912 9838]–[SO 2922 9839] and in the long strike section of Betton Dingle, north of Lyde [SJ 3176 0149]. It contains much fine pyrite resulting in rusty weather stains. In parts, the formation consists of alternating thin, pale and darker grey beds, each with a sharp, silty base. These mudstone beds have a ‘graded’ appearance and are probably the products of volcanogenic density flows; other parts are vaguely laminar, indicative of undisturbed hemipelagic deposition. At least one bed of disturbed mudstone was encountered, the product of a debris-flow. Thin layers of fine, micaceous sandstone and siltstone occur and these contain volcanic detritus; some are very fossiliferous.

Basaltic tuff and hyaloclastite (Lynas, 1985b), previously thought to be a doleritic intrusion (e.g. Whittard, 1979), forms a pronounced ridge from just south-south-west of New House [SO 289 977] to near Middleton Hall Farm [SO 297 990]. The thin beds of volcanogenic sandstone which cross the stream in Deadman’s Dingle much farther south [SO 2853 9593] are probably at the same horizon. Parts of this rock show a crude bedding foliation in exposures of crystal lapilli tuffs which overlie laminated siltstone. The weathered rock is red, and clasts, up to 40 mm diameter, commonly stand proud on weathered surfaces. They are of basic composition (E57342) and (E57343) and some are scoriaceous and pumice-like, indicating a hyaloclastic origin.

The other main part of the outcrop to the north [SO 2908 9815] reveals a massive hyaloclastite in parts with a pseudo-doleritic texture, but shards and pumice fragments are clearly visible in thin section (E57345). The feature becomes subdued in the north, west of Weston House, where it is formed of crudely bedded medium-grained volcaniclastic sandstone.

Biostratigraphy

The fauna of the Betton Shale contains large pendent Didymograptus referred to D. (D.) murchinsoni and a problematical form akin to Gymnograptus or ‘Lasiograptus retusus’ (Strachan 1986, p.47). These and the trilobites Bettonolithus chamberlaini and Ogyginus corndensis indicate the D. murchisoni Biozone. Trinucleus acutofinalis and early Marrolithus and Whittardolithus make their appearance. The brachiopod fauna is of low diversity, the commonest forms being Palaeoglossa attenuata, Schizocrania salopiensis and Tissintia prototypa. The fauna is generally similar to that of the Weston Flags, but with a lesser resemblance to the Neseuretus community because of the absence of Platycoryphe.

The formation belongs to the upper part of the D. murchisoni Biozone and is thus of late Llanvirn age. It correlates, therefore, with the upper D. murchisoni Shales of the Builth–Llandrindod area (Hughes, 1969) showing that both areas were volcanically dormant at the close of the Llanvirn epoch. It is noteworthy that much of the benthic fauna is preserved in siltstone and fine sandstone and that mudstone contains much laminated hemipelagite indicative of oxygen-deficient benthic conditions.

Meadowtown Formation

Like the Betton Shale Formation the Meadowtown Formation was divided from the Middleton Group of Lapworth (1887a, b) by Lapworth and Watts (1910) who called it the Meadowtown Calcareous Beds or Stage. The name ‘Meadowtown’ was retained by Whittard (1931, and following references), and was raised from member to formational status by Lynas (1985b).

The outcrop trends north-north-east, from near Old Church Stoke to Lower Wood; it is 100 m wide in the south and nearly 800 m wide in the north. It rests on the Betton Shale Formation, but the lower boundary, in particular, has been faulted out along near-strike faults. The formation is about about 80 m thick in the south and about 500 m thick in the north; this is due in part to a high proportion of thick sandstones in the north where there are several mappable members, some up to 50 m thick and one over 100 m thick. In the two places where the base has been observed, the underlying Betton Shale passes gradationally upwards into flaggy sandstone, while at the top the dark grey mudstone of the formation passes gradationally into the soft black shaly mudstone of the Rorrington Formation.

The distinctive appearance of this formation stems from beds of calcareous flaggy sandstone and siltstone, together with limestone, within the dark grey mudstones. The sandstone and siltstone are usually medium grey, with parallel laminae, in beds between 20 to 200 mm thick.

Opinions on the proportions of sandstone and siltstone to mudstones have varied; Whittard (1931; 1960) emphasised the mudstones, recording some limestone, calcareous flags and tuffs, but Lynas (1985b) considered the formation to be ‘broadly arenaceous’. The mudstone is concealed under low cultivated ground, while the sandstone-rich units form ridges with natural exposure and quarries. The calcareous flaggy sandstones are commonly bioclastic, bioturbated in some places, and cemented with calcite; on weathering they decalcify to a brown rottenstone. Bivalves and sparse graptolites are present together with sponge spicules, crinoid columnals, bryozoan or coral fragments, and ostracods. These record the influence at source of shallow water which, at times, clearly extended oxygenated conditions into parts of this Shelve basin to promote a burrowing benthos in an otherwise dysaerobic environment. An abundance of dark-grey, possibly phosphatic nodules 2 to 5 mm in size have been recorded, confined to a few layers up to 76 mm thick.

Biostratigraphy

The fauna of the Meadowtown Formation is diverse. Whittard (1955–1967) described more than 20 trilobite species among which are several trinuclei though much the commonest form is the asaphide Ogygiocarella debuchii. Williams (1974) recorded 17 brachiopod taxa, mostly articulate forms including Dalmanella salopiensis gregaria, Rafinesquina delicatula and Tissintia prototypa. Jones (1986, p.5) recorded unnamed Ostracoda from the Meadowtown Formation and there are a few records of undetermined mollusca and other shelly fossils. The graptolites recorded by Strachan (1986) and Hughes (1989) include Dicellograptus cambriensis, Dicranograptus irregularis, Diplograptus foliaceus and Normalograptus brevis, and is referred to the teretiusculus Biozone.

In comparison with the type area of Llandeilo (Williams, 1953) both the trilobite and graptolite faunas are much more diverse. All but two trilobite genera of the lower Llandeilo at Llandeilo are present in the Meadowtown Formation. Lloydolithus lloydi and Marrolithus inflatus maturus in particular leave little doubt of the early Llandeilo age of the formation while the Llanvirn pendant didymograptids make no appearance at all above the Betton Formation. The presence of Marrolithus favus suggests that the formation includes ‘mid-Llandeilo’ horizons as does perhaps Ogygiocarella debuchii. It should be noted that the subdivisions lower, middle and upper of the Llandeilo Series of Llandeilo by Williams (1953, p.190) exclude the overlying Nemograptus gracilis Shales. Williams (p.194) placed these in the ‘Lower Bala’, that is the Caradoc, but there is little doubt that they are the equivalents of the Rorrington Formation (high Llandeilo) of Shelve (Whittard, 1956, p.66).

Any correlation with the Builth–Llandrindod area must consider the Teretiusculus Shales of Elles (1940) (Table 3) to be equivalent to the Meadowtown Formation. The general aspect of the fauna is similar, as instanced in the high diversity of the trilobites, but there is little specific, or even generic, identity. Both formations are partly calcareous which suggests that the differences are unlikely to be caused by different depositional environments. Geographical separation may be a factor, but one giving little support to the view that Builth and Shelve were separated by movement on a strike-slip fault only after the Llandeilo Epoch.

The inclusion of both formations within the graptolitic biozone of G. teretiusculus depends largely upon the disappearance of pendant didymograptids at the base and the absence of N. gracilis; both are negative criteria and difficult to evaluate. They can be applied to both the Llandeilo area and the Berwyns with equal success but where their absence could be due to hostile sedimentation and preservation. However, Dicranograptus irregularisappears to be predominantly indicative of the teretiusculus Biozone (Hughes, 1989, p.37).

The Berwyn outcrops, to the north-west of Shelve, share none of the lithological similarities which are apparent in the Builth and Shelve areas. In the Berwyns, rocks of the Craig-y-glyn ‘group’ (Wedd et al., 1929) are a sequence of calcareous shales, limestones and ‘ashy’ and sandy shales. They are capped by ‘blue Slates’ (equivalent to the top ‘100 ft’ of Macgregor, 1961 p.177). These lithologies compare much more closely with Williams’ (1953) Llandeilo Series at Llandeilo, especially the middle and upper parts, than they do with the Meadowtown Formation. The same can be claimed of the faunas, for both at Llandeilo and in the Berwyns graptolites have not been recorded, yet they are common at Builth and Shelve. It is as though the shallower and more energetic environments and the thicker sedimentary accumulations lay to the west of Shelve and Builth in pre-N. Gracilis Llandeilo times. Indeed, a problem in any correlation with the Berwyns sequence is the absence of N. gracilis there. Macgregor correlates all but the top 30 m (100 ft) of his sequence (the Craig-y-glyn group of Wedd et al., 1929) with Williams’ upper Llandeilo of the Llandeilo area (Williams, 1953) on the basis of an abundance of M. favus. His concession that mid-Llandeilo horizons may be present is worthy of serious consideration. Nevertheless, these beds clearly correlate with the Meadowtown Formation. The top 30 m (i.e. the ‘blue shales’ in the Garwallt Group of Wedd et al., 1929) would then correlate well with the N. gracilis Shales of Llandeilo and Builth, and the Rorrington Shale Formation of Shelve. Unfortunately N. gracilis has never been recorded from the Berwyns. However, the problem digs deeper and spreads farther than this; it unsettles correlations across the shelly-graptolitic facies divide, for example Addison (in Williams et al., 1972) reported that, in the Hendre Shales near Carmarthen, N. gracilis occurs alongside Lloydolithus lloydi. This places the base of the N. gracilis Biozone no higher than low Llandeilo (Finney and Bergström, 1986) while N. gracilis ranges upwards into both the Coston Beds of south Shropshire, the type strata of the base of the Caradoc Series (Dean, in Williams et al., 1972), and the Spy Wood Sandstone of Shelve, also Caradoc in age (Chapter 4).

Details

In the stream east of Spywood Cottage, hard, calcareous, fine-grained sandstones and siltstones are exposed [SO 2830 9570], forming beds 0.03 to 0.20 m thick. They yield Lloydolithus lloydi, Ogygiocarella debuchii and nearby [SO 2829 9577]Palaeoglossa attentuata was found with them.

In a track [SO 2950 9535] west of Weston Church, flaggy siltstones, which are probably volcanogenic, and grey, calcareous, hard sandstones are exposed. In thin section (E57351), the sandstone is quartzo-feldspathic and bioclastic; fragments include sponge spicules, crinoid columnals, possibly ostracods, polyzoan or coral fragments and sparse graptolites. The sandstone is partially silicified with a carbonate matrix.

In the brook [SJ 3085 0069], between Meadowtown and Rorrington, the gradational base of the formation is exposed. Immediately below the road bridge [SJ 3067 0097], the strata yield very fossiliferous grey mudstone. Beds containing phosphatic nodules (2.5 mm) occur, together with medium-grained calcareous sandstones, up to 0.3 m thick. Lower in the stream, grey shales pass gradationally into the darker shale of the Rorrington Formation, and sandstones are absent.

In the area of Mincop [SJ 3146 0183], sandstones are exposed, for example in the track [SJ 3115 0134]. Thinly bedded (less than 100 mm), medium-grained, faintly banded or laminated, calcareous sandstones are interbedded with shaly siltstone. A thin section (E59459) reveals angular quartz grains (about 0.03 mm) and some feldspar. Calcite makes up 70 to 80 per cent of the rock in the form of shell detritus and matrix. An old quarry [SJ 3123 0137] exposes similar beds, and in a track [SJ 3111 0161] grey flaggy siltstones yield P. attenuata, O. debuchii and Whittardolithus inopinatus.

The most northerly exposure is a quarry [SJ 3148 0241] where several metres of grey siltstones are interlayered with pale grey calcareous sandstones (20 to 100 mm thick). Thin sections (E59467) and (E59468) reveal small isotropic concretions in a few beds which contain sponge spicules and shell chips. Bedding planes are mostly smooth and parallel and some surfaces bear trace fossils. W. inopinatus is common.

Rorrington Shale Formation

Rorrington Group was the name coined by Lapworth for strata currently called the Rorrington Shale Formation. The latter name was established by Lynas (1985b) to replace the Rorrington Member (Whittard, 1979).

The formation crops out from a strike-fault just north-east of Old Church Stoke, in the south, to Lower Wood [SJ 3088 0268] in the north (Figure 4), where it is overlapped by the sandy Venusbank Formation of late Llandovery age. Like the Meadowtown Formation, the outcrop broadens northwards, reflecting a northward increase in thickness of the formation from 200 to 500 m.

The formation is sparsely exposed; it comprises distinctive dark grey to black, soft, shaly mudstone with only slight variety due to micaceous siltiness. Hard siltstone beds occur sparsely, particularly near the base which is ill defined and gradational upward from the Meadowtown Formation. The main difference is the virtual absence of sandstone beds in the Rorrington Shale.

A particularly noteworthy characteristic is the dark brown streak made by the mudstone. It compares closely with those of the N. gracilis Shales near Llandrindod and the Nod Glas Formation of late-Caradoc age, both deposited during periods of marine transgression. The formation is also rich in pyrite and it is interpreted as a concentrated hemipelagite in which the depositional laminae are largely intact. In these respects the formation is unique in the Shelve Ordovician succession.

Apart from a few thin layers of siltstone the only lithological variety in the formation is a more massive mudstone due to a micaceous siltiness for instance near the middle of the formation at the junction of Aldress Dingle with Brynkin Dingle. Here the rocks are slightly colour-banded.

Biostratigraphy

The fauna of the Rorrington Shale is markedly more graptolitic than the Meadowtown Formation and it contains orthocones more abundantly too. Both reflect the large pelagic influence. Dicellograptus salopiensis, D. cambriensis, Leptograptus validus, ‘Glyptograptus’ cf. teretiusculus and Nemagraptus gracilis range through most of the formation. The presence of N. gracilis indicates the gracilis Biozone whilst Dicellograptus geniculatus, which occurs below the range of the N. gracilis, indicates that the lower part of the formation is referable to the upper part of the teretiusculus Biozone (Hughes, 1989). Whittard (1955–1967) described 14 species of trilobite, including Marrolithus cf. anomalis, M. inornatus and Marrolithoides arcuatus, which all occur in the central parts of the formation (Whittard, 1966). Jones (1987) recorded three ostracods species from the Rorrington Formation:Bullaeferum llandeiloense and Laterophores elevatus range from the upper Llanvirn to the top of the Llandeilo of south-west Wales, and Pariconchoprimitia oscillata ranges from the upper Llandeilo to the top of the Costonian. The brachiopod fauna is dominated by inarticulate forms, and though several articulate genera are present, they are represented by scarce juvenile valves.

The only elements of the fauna with comparability in the Llandeilo area are M. anomalis, suggesting correlation with the Middle Llandeilo Flags of Williams (1953) and N. gracilis which occurs in his overlying Nemagraptus gracilis Shales. Williams placed the latter in the ‘Bala Series’, the base of which is considered, on shelly faunal evidence, to be of Costonian (Caradoc) (Williams, 1963 pp.338–339) and thus to be synchronous with the base of the Caradoc Series (as defined in the type area). However, as discussed above, it is difficult to identify this horizon in the graptolitic succession. The Rorrington Shales pass gradationally up into the Spy Wood Sandstone which can be correlated with the Coston Beds of south Shropshire and therefore marks the base of the Caradoc Series at Shelve. Thus the Rorrington Shales must be part of the Llandeilo Series with their top defined by Caradoc strata.

In the Builth–Llandrindod area, the comparable formation is the Nemagraptus gracilis Shales (Elles, 1940). The lithology and fauna of the two formations are very similar but a comparison of thickness (Earp, 1977) indicates that, because of Silurian overstep in the Builth–Llandrindod area, the N. gracilis Shales are representative of only the lower half of the Rorrington Shale. Trilobites are more abundant in parts of the N. gracilis Shales and it is perhaps significant that in those parts the fine hemipelagic laminae are absent, yet are present in the graptolitic parts. This suggests that in the Builth–Llandrindod area the sea bed was periodically more oxygenated than at Shelve and biogenic disturbance of the sediment might have been greater during those periods.

The oldest beds on record from the Tywi Anticline were also contemporary deposits of the Rorrington Shales. Dark grey graptolitic shale with abundant N. gracilis in places (Cave, 1988) are exposed stratigraphically above a borehole sequence which proves some 200 m or more of dark grey mudstone, contemporaneous igneous intrusions and mass flow deposits. Whittardolithus sp., and graptolites of the teretiusculus Biozone occurs low in this sequence indicating a possible equivalence with the Meadowtown Formation (Cave and Rushton, 1996).

The nearest outcrop of comparable age is that of the very dark grey or black Nemagraptus Shales (with N. gracilis) at the base of the succession, north of the Breidden Hills (Watts, 1885; Wedd, 1932; Wedd et al., 1929). At one time, some 240 m were visible beneath the Black Grit, the local equivalent of the Spy Wood Sandstone Formation. This lithological sequence is closely similar to the Shelve equivalents, yet it is not recognisable in the Berwyn equivalents to the north-west (p.29). The separation of the depositional provinces must lie near to, or beyond, the Severn Valley on the west.

In south-west Wales the equivalent beds form the top part of the Hendre Shales and Mydrim Limestone.

Details

The best and almost complete exposure follows the stream bed and banks down Deadman’s Dingle from the base of the formation [SO 2845 9597] to its confluence [SO 2832 9598] with Spy Wood Dingle and thence down that dingle to the base of the Spy Wood Sandstone Formation [SO 2819 9581].

Near [SO 2842 9597] the base of the Rorrington Formation the mudstones are blocky, very soft and very dark grey with a brown streak. Strata in Deadman’s Dingle [SO 2833 9597] to [SO 2843 9597] yield Schizocrania sp., Ogygiocarella? (juv),Primaspis whitei, Climacograptus sp., Dicellograptus sextans, Nemagraptus?, ‘G’. teretiusculus, orthocones, ostracods and crinoid columnals. Nemagraptus gracilis occurs in abundance just above the middle of the formation [SO 2852 9587] where the rock is less dark, slightly silty and finely micaceous. It is also faintly colour-banded [SO 2842 9584]. Higher in the formation, sooty black mudstones again dominate the succession.

Chapter 4 Ordovician: Caradoc

The western part of the Shelve Ordovician outcrop is formed entirely of Caradoc rocks in an unbroken succession some 1.7 km thick. Likewise Caradoc strata make up the whole of the Forden Inlier with its southward extension through Montgomery (Figure 1).

Caradoc rocks are distinctly different from underlying strata, a fact recognised by Lapworth (1887b), when he separated them from lower strata under the collective term Chirbury Series. This he equated with his Caradoc Series east of the Long Mynd, and despite the many subsequent reviews no more accurate correlations, nor datings, have arisen since. The logic of this correlation is so clear that it lends itself as a guide in the classification of mid-Ordovician sequences of other areas, a potential also originally recognised by Lapworth (1887a, p.663). The Chirbury Series remained in use until 1925 (Watts, 1925), and was revised variously as the Chirbury Formation (Whittard, 1979) and the Chirbury Group (BGS, 1991). It is nevertheless superfluous. It is, as Lapworth promoted it, the Shelve representative of the Caradoc Series, albeit incomplete at the top, but in the Montgomery district other Caradoc rocks crop out, in the Forden Inlier. Inclusion of the latter within the Chirbury Group would make the name an ill-defined synonym for the Caradoc rocks of the district; exclusion, however, would leave the grouping as a valueless local encumbrance.

The drab uniformity of the Caradoc mudstones of the district does not allow ready division into distinctive formations. In this respect the Caradoc classification (Table 4) is the least satisfactory of the five Ordovician series present. It uses the intervention of two thick volcanogenic sandstones to divide the Shelve outcrop into five formations, following Lynas (1985b) and BGS (1991), while the Forden Inlier is treated as a simple formation within which there is a volcanogenic mass-flow member (Cave et al., 1988). Correlation between the two outcrops is poorly constrained by sparse shelly and graptolitic faunas.

The base of the Caradoc Series is not exposed in the Forden Inlier, but it is exposed in the Shelve outcrop where a case can be made for placing it at the base of the lowest prominent sandstone, above the Rorrington Shales. Here, sandstones increase in number and thickness upwards through some 5 m of beds, until they form 95 per cent of the sequence. The base of the Spy Wood Sandstone Formation, placed at the bottom of this transition, approximates as closely as can be measured anywhere to the base of the Caradoc Series. It also marks an increase in the diversity of brachiopod faunas which continues throughout the Caradoc Series (Williams, 1974). In both the Forden and Shelve outcrops, the top of the sequence is truncated by either the Silurian overstep or Quaternary deposits so that no horizon higher than Soudleyan is exposed within the district.

There is little doubt that late Caradoc (Longvillian to Onnian stages), early Ashgill and even mid-Ashgill sediments were deposited over the district, but they were removed largely during late Ashgill times, (when there was an inversion of the Shelve–Builth basinal area and a glacioeustatic fall in sea level) and partly during the early Llandovery, before the marine transgression completely re-submerged the area. As the Midland Platform emerged during the late Ashgill times, the shoreline retreated rapidly westward to a position west of the River Severn. Farther south, west of Builth Wells, it retreated almost to the break of slope into the Welsh Basin.

Thus, during the late Ashgill (Hirnantian) times vast quantities of detritus, much of it from young unconsolidated rocks, were removed from the district and redeposited to the west nearer the shelf-edge or on to the basin-margin. On the steeper, possibly fault-controlled, parts of this margin these deposits were repeatedly destabilised, carried away and replenished (Cave and Hains, 1986; Leng, 1990). By the end of the Ordovician most of the district had been stripped of its Ashgill deposits; only in the extreme west did they remain, and probably still do, concealed by Silurian strata (Cave and Price, 1978).

Spy Wood Sandstone Formation (Spy Wood Grit)

‘Spy Wood calcareous grit’ was the name given to the formation by Lapworth (1887a, b), and Whittard (1979) referred to it as the Spy Wood Member of the Chirbury Formation. Lynas (BGS, 1991) renamed it the Spy Wood Sandstone Formation on the supposition that ‘sandstone’ describes the rock more accurately than ‘grit’, and this usage is followed here.

The formation has a very narrow outcrop, trending north-north-east from near Church Stoke church in the south [SO 2795 9433] to some 500 m west-south-west of Lower Wood where it disappears northwards under the Venus Bank Formation of Llandovery age. It marks a strong topographical ridge which is particularly pronounced for 2 km north-north-east of Rorrington. The formation is about 40 m thick, and up to 120 m in the north.

The Spy Wood Sandstone has a distinctive lithology in a sequence of dominantly shaly mudstones below and above. It is a well-bedded, pale to medium grey, medium- to fine-grained calcareous sandstone. Most grains are of quartz and are angular in shape. The beds are planar and range in thickness up to 0.20 m; they thicken upwards in the basal 5 m and become thinner at the top of the formation. Many are parallel laminated micaceous and fissile, splitting into flags of about 20 mm. The sandstones are richly bioclastic, of a shelly provenance and include fragmented crinoids, bryozoans, brachiopods, trilobites and Machaeridia. The appearance is indistinguishable from many of the Caradoc sandstones of south Shropshire, weathering to ochreous brown rottenstones along shell-detrital seams. The type section or ‘stratotype’ is in Spy Wood Dingle [SO 2819 9581] (Whittard, 1960, p.253).

Spy Wood and Aldress Dingle comprise a Site of Special Scientific Interest, information upon which is updated periodically and held by English Nature (see Information sources).

Partings of medium grey, very micaceous, silty mudstone, up to 0.20 m thick, separate the sandstones. In the basal 5 m of the formation, these partings are dark grey, and are indistinguishable from mudstone of the Rorrington Shale. At the stratotype [SO 2819 9581], the base of the formation is chosen at the base of a pale grey sandstone about 20 mm thick, and some 5 m below where sandstone forms 95 per cent of the rock. Some 4 m above the base, the sandstones thicken upwards rapidly, the whole forming a transitional sequence. Lynas (1985b) did not define the base of the formation, but Whittard (1979) drew it at the base of the ‘more massive beds of typical Spywood lithology’. This is a somewhat subjective definition more difficult to replicate and higher than the base defined here.

The Spy Wood Sandstone represents a brief period of very frequent, high-energy influxes of fine- to medium-grained silica sand and bioclasts. The provenance of these was a well-oxygenated, shallow, nearshore sea. This detritus was transported into a low-energy euxinic muddy environment, which then became oxygenated. It is unlikely that the sandstones of the Spy Wood Formation represented a sudden marine shallowing, especially as this was a period of general marine transgression. It is more likely that a rapid marine inundation of the existing basin-margin produced a shallow, marginal sea, vulnerable to the influence of storms and into which bioclastic sand debris prograded. The influences of climatic change at this time could have been considerable too. For instance, had offshore winds been replaced by onshore winds, these conditions would have effected not only marine storm surges, but also the destruction of stratification in the tranquil water of the shallow ‘Shelve Basin’ with the expulsion of euxinic or nearly euxinic conditions.

Inundation of the marginal terrain east of the Long Mynd (Figure 12) took place at exactly this time. The Hoar Edge Grit and Coston Beds rest on an uneven surface of latest Llandeilo and early Caradoc age, consisting of Cambrian and Precambrian rocks. However, it seems unlikely that there was a direct marine link between the shallow sea east of Long Mynd at Church Stretton and that of the Shelve area. Some proof of a Long Mynd land barrier in Costonian times lies in Habberley Brook, on the western side of Long Mynd, where Harnagian rocks immediately east of the Pontesford–Linley Fault rest directly on Precambrian rocks (Dean and Dineley, 1961), and it is logical to assume that the rest of the Long Mynd shared this relationship. Nevertheless, the provenance of the Spy Wood Sandstone must have been in the Long Mynd area, or a little to the north. Further indication of provenance is afforded by the Black Grit of the Breidden inlier. This correlates with the Spy Wood Sandstone and consists of some 6 to 7 m of fine-grained sandstone with a bituminous impregnation (Watts, 1885; Dixon, 1991). Its comparative thinness and lack of bioclasts, other than sponge spicules, implies a greater distance from source than the Spy Wood Sandstone and thus a provenance in the north-west is most unlikely (Figure 12).

Biostratigraphy

Bancroft (1933) gave the age of the formation as Costonian because of its marrolithoid trilobites, and Whittard (1966) confirmed this by his record of Costonia ultima from the upper part of the formation. In south Shropshire, C. ultima characterises the upper part of the Costonian Stage (Dean, 1958) and there seems to be good reason to consider the Spy Wood Sandstone as the exact equivalent of the Hoar Edge Grit and Coston Beds (Whittard, 1966,p.283). A Costonian age is confirmed also by Jones’ (1986, 1987) study of the ostracod fauna. The lower part of the Hoar Edge Grit contains Nemagraptus gracilis and a fragment of N. gracilis was recently (1994) found in silty mudstone near the middle of the Spy Wood Sandstone (Figure 13). In the rest of the formation graptolites are fairly common in calcareous sandstone and these indicate the biozone of D. multidens. The boundary between the N. gracilis and D. multidens biozones is thus within the Caradoc Series at about the middle of the Spy Wood Sandstone.

The fauna of the formation is dominantly shelly. Larger forms are commonly broken and these remains have been transported into place, where they form bedding-parallel concentrations. Williams (1974) recorded 21 taxa of brachiopods, dominantly of the articulate genera Bicuspina, Bystromena, Dalmanella, Glyptorthis and Sowerbyella. These are typically associated with sandy substrates and were presumably derived from a relatively shallow, high-energy environmental setting. Trilobites (Table 5) are rarer and less diverse: Whittard (1966) recorded 12 taxa, all very rare apart from Platycalymene duplicata and species of Marrolithus. Identifiable Asaphacean trilobites, so common below, are unknown from the Caradoc rocks of the Shelve area. Ostracods are common and include Harperopsis bicuneiformis, Histina xanios, Ogmoopsis (Quadridigitalis) siveteri, Piretopsis (Protallinnella) salopiensis and Varilatella (v.) dissita; according to Jones (1986, 1987) the ranges of these species overlap only in the Costonian. The commonest graptolite (Table 5) is Orthograptus uplandicus (Strachan, 1986) which Hughes (1989) considered likely to be identical to O. cf. apiculatus from the same beds. Dicellograptus, Dicranograptus and Nemagraptus are represented only by one or two specimens each. Fragmentary bryozoa and crinoids are abundant and some machaeridian plates (Lepidocoleus sp.) occur.

Details

The main exposure is in Spy Wood Dingle, where the stream exposes the complete formation, although the section is complicated by a fold. An adjacent track on the north side exposes the low and middle parts of the formation. The junction with the Rorrington Formation is placed at a white sandstone and the basal 3 m through the rapids [SO 2819 9581] is gradational into 95 per cent sandstone. In thin section (E56896), these sandstones show angular, well-sorted quartz grains of about 0.03 mm diameter, set in a carbonate cement to give a grainstone texture. Mica, parallel with bedding, is common, but feldspar is sparse. There are slight grain-size differences across the boundaries of bedding laminae, the latter being regular and parallel.

Other exposures are in the dingle [SO 2805 9557] near Brynayn Green, in the dingle [SO 2835 9700] north of Lower Aldress, in a tributary [SO 2839 9745] to Coed Brook, at the northern end [SO 2853 9811] of Coed Brook and in the prominent ridge [SJ 3025 0141] north of Rorrington.

Aldress Shale Formation

The original name for this formation is ‘Aldress Graptolithic shale’ (Lapworth, 1887a, b), but after Whittard’s account (1931) the unit was generally referred to as the Aldress Shales and, later, the Aldress Member (Whittard, 1979). Lynas (1985b) modified the name and redefined the unit as a formation. The type locality is Aldress Dingle some 2.4 km north-north-east of Church Stoke which presents a largely strike section through the outcrop.

The formation forms a narrow outcrop, about 0.5 km wide, from just east of Church Stoke in the south to near Wilmington in the north. To the north-east, it is overstepped by Llandovery strata. Sandstone within the shale forms a low topographical feature in the northern part of the outcrop, north of Kinton. The thickness of the formation increases from about 300 m in the south to about 550 m in the north.

The Aldress Shale consists mainly of soft shaly mudstone, which is usually medium grey or olive-grey-green in colour. It is finely micaceous and, on weathering, gives rise to deep orange-stained joint surfaces. Such mudstones are common throughout the remaining Caradoc formations of the district, which vary mainly in the proportion of interbedded sandstone; the sandstones are generally medium grey and fine grained. Within the Aldress Shale Formation, they are sparse and relatively thin or lenticular.

Thicker sandstones do occur sporadically, generally up to 0.20 m but rarely up to several metres thick. They are usually feldspathic, non-laminate and probably volcanogenic. Some contain abundant allocthonous shell fragments, contrasting strongly with the shales, which are characterised by graptolites. In the north, an unnamed sandstone member can be mapped for about 3 km. It is up to 20 m thick and occurs about one third of the way up the formation. Within the member, individual sandstone beds thin upwards, and grey and olive-green shale partings become thicker and merge.

Biostratigraphy

Brachiopods and graptolites from the formation are recorded by Williams (1974), Strachan (1986) and Hughes (1989).

The presence of such graptolites as Amplexograptus leptotheca (= A. fallax according to Hughes, 1989, but not Strachan 1986), Corynoides cf. curtus, Dicranograptus spinifer and Lasiograptus costatus in strata above the highest Nemagraptus gracilis is taken to indicate the multidens Biozone. In view of the presence of Sowerbyella cf. sericea permixta which occurs also in the Nant Hir Mudstones of Bala, Williams (1974) considered parts of the Aldress Shales to be at least Harnagian in age. Trilobites are sparse; however, the occurrence of Salterolithus caractaci, Broeggerolithus broeggeri, together with Reacalymene limba and Ampyxina wothertonensis in the upper parts of the formation gives an early Soudleyan age, but because there is no evidence of a stratigraphical break above the Spy Wood Sandstone, the lower part of the formation must belong to the Harnagian Stage.

Details

The full thickness of the formation is exposed almost continuously in Spy Wood Dingle [SO 2812 9586] to [SO 2790 9599] thence into Aldress Dingle as far as the base of the Hagley Volcanic Formation [SO 2770 9587]. In addition to the ubiquitous grey and olive, micaceous mudstones, other lithologies include thin siltstones, fine-grained feldspathic sandstones, up to 0.20 m thick, and a massive fine-grained feldspathic sandstone several metres thick [SO 2810 9645]. At about the same horizon [SO 2873 9930] in the dingle south-west of Kinton, 3 m of similar sandstone occurs, in beds up to 0.15 m thick. Geological Survey collections from Spy Wood Dingle and Aldress Dingle contain numerous graptolites: Amplexograptus fallax, Climacograptus cf. antiquus, Cryptograptus sp., Dicranograptus cf. furcatus minimus, Dictyonema fluitans, Diplograptus foliaceus, D. leptotheca, Lasiograptus costatus, Orthograptus cf. amplexicaulis, Pseudoclimacograptus scharenbergi, together with the shelly fossils Orbiculoidea?, Paterula, various bivalves (cf. Concavodonta, Nuculites, Praenucula, Similodonta?), Salpingostoma? hyolithids, orthocones, conulariid fragments, Dionide euglypta quadrataand smooth ostracods.

In the stream [SO 2765 9525] near The Cave, interbedded mid-grey, cross-laminated, calcareous fine sandstones and siltstones are present within grey shales, which also contain beds, up to 0.20 m thick, of rotten tuffitic sandstones. A few metres below, two beds of volcaniclastic feldspathic sandstone occur (E56882) and (E56883). A little lower still, the shales with thin fine-grained sandstones are fossiliferous yielding Dictyonema fluitans and Climacograptus sp., with diplograptids and orthograptids indicating the multidens Biozone.

Hagley Volcanic Formation

The volcanogenic clastic rocks above the Aldress Shales were originally included with the overlying Marrington Group as the ‘Hagley volcanic ashes and shales’ (Lapworth, 1887a, b). Later, Lapworth and Watts (1894) separated these and called the ‘ashes’ the Hagley Ash. Whittard (1931) amended that name to Hagley Volcanic Group, but later (1979) he relegated it to the status of member. There is a clear distinction between the shales above and below the volcanogenic sequence and on account of this and because the ash has several constituent members, it was raised to the status of formation (Lynas, 1985b).

Hagley Quarry [SO 2783 9764] has been listed as the type section (Whittard, 1960, p.145) although it represents but a small part of the formation; indeed, there is no comprehensive section through the formation. The base is exposed in Ox Wood Dingle [SJ 2909 0118] and is transitional from the Aldress Shale. Here, crushed grey Aldress Shale is overlain by a massive, coarse, volcaniclastic sandstone and breccia about 1.3 m thick, followed by a similar but bedded breccia which fines upwards before passing up abruptly into shale of the Aldress Shale type.

The outcrop trends north-north-east from just east of Church Stoke in the south to Big Cuckoo’s Nest [SJ 288 006], where it is widest, about 300 m (Figure 4). The formation forms a prominent ridge, especially between Kingswood [SO 2788 9676] and Big Cuckoo’s Nest where the rocks are coarsest and the formation thickest. It increases in thickness from about 50 m near Church Stoke to over 200 m between Kingswood and Big Cuckoo’s Nest, thinning rapidly north-eastward to 20 m at its northern limit.

The Hagley Volcanic Formation is typified by brindled or mottled, rather pale green to khaki, massive feldspathic sandstone, in beds up to 2 m thick. The colour pattern is caused by darker rings and streaks (Whittard, 1931) and has been described as ‘throstle-breasted’. The formation is a complex sequence of feldspathic sandstones, volcanic conglomerates and breccias, thinner, cross-bedded sandstones and finer vitroclastic tuffites and rhyolitic lapilli tuffs. Also present are beds of rhyolite and andesite which were probably minor lava flows. The massive feldspathic sandstones commonly overlie the breccias and conglomeratic beds, but rapid lateral and vertical changes of lithology are common, suggesting bed lenticularity.

The conglomerates and breccias are ill-sorted, commonly non-graded and some are mud-matrix-supported, sedimentary melanges. In some beds, the clasts are angular, but in others many are well rounded; they range in size up to boulders, some 0.40 m across. Commonly the clasts are of igneous rocks, mainly porphyritic andesite or basalt with trachytic textures, but clasts of sandstone, mudstone and tuff also occur.

Where bed-form is well developed there are sole-structures which include load casts, flute casts and groove casts, indicating transportation from the north-north-east. Flame structure is also common and early compactional micro-faulting occurs in some parts.

Quartz-feldspar porphyry

At the entrance to Aldress Dingle is an exposure [SO 2768 9579] of some 20 m of alternating quartz-feldspar porphyry and breccia (Lynas, 1985b, fig. 3) (Figure 14). At the base is a porphyry (Bed 1) (4.5 m thick) with a planar base and evidence of underlying shale; the top is an uneven surface in contact with overlying breccia. The basal part of the porphyry (E56887) consists of a microgranular felsitic matrix containing feldspar microlites throughout and cumulophyric clusters of feldspar megacrysts up to 2 mm size. There is also a scatter of rounded and deeply embayed quartz megacrysts. One metre above the base, the bed is colour-banded, with a streaky appearance. In thin-section (E56886) it contains amoebae-like, rounded and hollow quartz crystals, 0.5 to 1.5 mm in size. Plagioclase megacrysts commonly contain glassy inclusions and again the matrix has an abundance of feldspar microlites. Similar textures are present 1.5 m above the base (E56885) and also in the centre of the bed (E56891). The feldspar microlites, up to 5 mm long, are flow orientated, but at various levels the matrix shows a relict vitroclastic texture.

The overlying breccia (Bed 2) is some 9 m thick with a 2 m lens of quartz-feldspar-porphyry within the basal 2.3 m. A sample from near the base of the bed (E56888) shows that it consists almost entirely of pyroclasts in various stages of comminution and incipient break-up (pull-apart) from a rock identical with the underlying porphyry. This passes up into a second bed (Bed 3) of colour-banded porphyry (E56889) (2 to 3 m thick) overlain by a further breccia (Bed 4) (E56890).

Interpretation of these rocks is equivocal; they may have originated as flows of rhyolitic lava with associated explosive, pyroclastic eruptions, yet the relict vitroclastic texture leaves the possibility of a pyroclastic (ignimbritic) origin.

Conditions of deposition

As with the Stapeley Volcanic Member of the Hope Formation (Chapter Three), the source of the debris making up the clastic beds of this formation is almost entirely the marine-processed remains of an explosive volcanic centre. The massive nature of sandy beds, the lack of sorting and grading of many of the conglomerates and breccias, the wide range of maturity in the clast shapes and the fact that in some beds the clasts are mud-matrix-supported, suggest that much of the detritus was transported in high-density turbidity flows and mass-flows. They imply locally steep submarine slopes while lateral impersistence of some of the lithologies may indicate channelised transportation.

The presence of graptolites within tuffs (Whittard, 1979) proves submarine deposition, and allocthonous brachiopod debris in the sandstones indicate a shallow marine source. The volcanic edifice may have become subaerial at times, either by aggradation of the core, or during periods of tumescence, but there is little indication of where it lay other than it was nearest to the north-central part of the outcrop. A subaerial phase, involving fluvial or shoreface attrition could account for the highly rounded nature of some of the andesite clasts. The thinner and finer interbedded sandstones may well be low-density turbidites formed during the waning of high energy activity. Cross-bedding, flute casts and groove casts indicate turbidity flow from the north-north-east.

Biostratigraphy

The trilobites B. broeggeri and S. caractaci recorded by Whittard (1966) indicate an early Soudleyan age. Graptolites from the formation are Climacograptus antiquus, C. bicornis [‘peltifer’ form], Diplograptus foliaceus and Pseudoclimacograptus scharenbergi.

The correlative beds in North Wales are the Nant Hir Mudstones at Bala (Bassett et al., 1966; BGS, 1986) and the Llewelyn Volcanic Group of Snowdonia (Howells et al., 1983). These are similar to the mid-Caradoc formations of Shelve in having originated in an oxygenated depositional environment which supported brachiopod, or brachiopod/tritobite faunas, albeit sparse ones. In contrast, non-volcanic black shales dominate the sequences in west and central Wales, for instance the Mesograptus Beds of the Mydrim Shales near Carmarthen and their counterparts in the Sugar Loaf Formation (Mackie, 1993, p.263; Davies, 1928) of the Tywi Anticline.

Details

At The Cave [SO 2753 9527], a fairly complete section is exposed showing massive, feldspathic, medium-grained sandstones with a characteristic and clearly diagenetic mottling on fresh surfaces. It occurs in unlaminated beds 0.30 to 1 m thick, usually with intervening, thin, hard, siltstone partings, some of which are laminated and porcellanous. The mottled characteristic of the sandstone is prominent only in the higher beds.

At the junction of Marrington and Aldress dingles [SO 2760 9580], a fresh roadside section showed over 25 m of mottled, massive, greenish grey sandstones. Within this sequence, sedimentary breccia horizons, comprising siltstone clasts up to 100 mm long, are present within the sandstone beds. Sole markings are common and some beds clearly are loadcast into the underlying sediments. Mottled sandstones can be seen in the steep, wooded slopes north of this locality; at Kingswood [SO 2779 9666], old quarries expose coarse volcaniclastic deposits with clasts up to 0.40 m across. The larger blocks tend to be of sedimentary rocks, whereas the more numerous lapilli-sized clasts are of volcanogenic origin. Some of the clasts are themselves composed of lapilli tuff. Bedding is rare, and the tuffs are overlain by the typical mottled feldspathic sandstones of the formation, although here they are coarser and grain size is more variable.

An isolated knoll [SO 2780 9739] of rubbly pyroclastic breccia shows porphyritic andesite clasts, ranging in size from sub-lapilli to blocks; the blocks (E57335) contain glomeroporphyritic clumps of feldspars set in a seriate intergranular matrix with scattered chloritic pseudomorphs after ferromagnesian minerals.

Farther north, at Ridge and Hagley, the formation forms a pronounced feature and exposures are good. The highest beds are exposed at the type locality of Hagley quarry [SO 2780 9764]. They are massively bedded, medium-grained, buff, feldspathic sandstones, together with more thinly bedded parts and a bed of mudflake breccia. The lower horizons of the formation are exposed in a small quarry, 80 m north-east of the above locality. The beds were clearly water laid and are commonly displaced by microfaults effected by early differential compaction.

On Ridge [SO 2783 9780] to [SO 2794 9813], the highest beds are notably different from those at Hagley Quarry. They consist of a coarse volcaniclastic rock, characterised by massive bedding, and containing feldspar crystals, lapilli-sized volcaniclasts and mud rip-up clasts. Some blocks and boulders are present within the tuff; most are angular, but some are subrounded, and up to 0.5 m in size. The clasts consist of feldspar porphyry (E57341), with plagioclase crystals, some exceeding 5 mm in size. This facies is strikingly similar to the ‘boulder beds’ of the Stapeley Volcanic Member of the Hope Shale Formation (Chapter Three). The lower part of the Hagley Volcanic Formation is exposed [SO 2796 9789] where it forms the top of Ridge. It consists mostly of coarse feldspathic sandstones with irregular intercalations up to 1 m thick of coarse clast-rich sandstones. These beds (E57339) contain volcaniclasts of porphyritic andesite with a trachytic texture. Thin shales and cross-laminated fine-grained tuffites also occur and load-casting is common. These beds pass up into medium-grained clast-rich sandstones (E57340) in which the clasts are of basic lavas with trachytic texture, angular feldspars, some quartz and indeterminate rock fragments; clasts are mostly matrix supported in a wacke base. In a small excavation, a few metres to the south-west of the main locality, two well rounded cobbles composed of light grey aphanitic felsic rock are visible within these sandstones.

Farther north, good exposures [SO 2830 9918] of the lower part of the formation show 2 m of fine- to medium-grained flaggy laminated tuffites, the beds being rather lenticular. This overlies a coarse volcaniclastic unit which forms a steep east-facing bank below this locality. Its highest parts are foliated as it passes into the tuffites and it consists of crystal lapilli tuffs which contain numerous large cobbles and boulders of porphyritic lavas and tuffs up to 0.40 m in diameter, commonly well rounded. The long axes of the clasts lie in the plane of the bedding. The matrix in which the cobbles and blocks are set (E57352) consists entirely of clasts of various sizes of trachytic-textured feldspar porphyry. The cobbles and blocks (E57353), samples of which were used by Rundle (1984 a, b) in an attempt at isotopic age determination (Lynas, Rundle and Sanderson, 1985), show a seriate texture with stumpy altered plagioclase phenocrysts and green mica pseudomorphs after ferromagnesian minerals; they are andesitic to dacitic in composition. Exposures at Cuckoo’s Nest [SJ 2884 0084] show roughly bedded material, parts of which consist of volcaniclasts of varying sizes. A thin section (E57370), reveals a feldspar porphyry with rare quartz megacrysts and a few pseudomorphs possibly replacing amphibole. The feldspar phenocrysts are almost euhedral, show quench textures and some exceed 1 mm in size.

In Ox Wood Dingle [SJ 2909 0118] where the base of the formation is poorly exposed the bedded part at the bottom (E57368) comprises basaltic volcaniclasts of about 2 mm diameter together with pumiceous material, set in a mud matrix. The top of the bed (E57376) contains many mudstone clasts, feldspar and a few quartz grains all less than 0.5 mm across, set in a sparse carbonate matrix.

The section continues a few metres downstream where a 4 m-thick melange is visible. This consists of mudstone with subrounded boulders, cobbles and pebbles of porphyry and tuff. North-east of Ox Wood Dingle, in the vicinity of Wilmington, the rocks in a number of small exposures vary from laminated siltstones and shales to vitroclastic tuffites of intermediate composition (E57375) [SJ 2955 0177]. Shards and pumice fragments are typically less than 0.1 and 0.3 mm across, respectively.

Hagley Shale Formation

The Hagley Shales were separated, as a unit, from the ‘Hagley volcanic ashes and shales’ of Lapworth (1887a, b) by Lapworth and Watts (1894). Whittard (1979, p.53) referred to them as the Hagley Shale Member, and Lynas (1985b) reclassified them as a formation. It is separated from the Whittery Shale by the Whittery Volcanic Formation. The two shale formations are otherwise indistinguishable; indeed, in southern areas where there is no Whittery Volcanic Formation, the separation is notional.

The outcrop of the Hagley Shale is narrow, extending northwards from near Church Stoke along Marrington Dingle as far as Caerbre [SO 276 964] where the formation is about 240 m wide and 200 m thick. Farther north, at Wotherton, the outcrop is at its widest, due to repetition across strike faults and to the introduction of extra members of sandstone and volcanic rocks (see below). The thickness of the formation here is probably about 350 m. Neither the base nor the top of the formation is exposed.

The formation consists dominantly of medium drab-grey to olive-green shaly, finely micaceous, smooth mudstone. It is soft and crumbly with a nodular weathering pattern, and weathers to a pale buff or khaki colour with orange-brown staining on joint surfaces.

Interbedded with the mudstone are thin beds of fine-grained, micaceous sandstone, spaced up to 0.15 m apart, and up to 0.06 m thick. They are medium or pale grey-green in colour, well sorted and predominantly quartzose. Cross-bedding and parallel lamination is common, but many sandstones are non-laminate. Trace fossils, largely trails and bedding-parallel burrows, occur on the bases of many beds.

The normal characteristics of the formation are complicated in the north-central part of the outcrop by the inclusion of several sandstones and thick, but discontinuous, members of volcanogenic breccia of intermediate to acidic composition that were the product probably of renewed, but minor, activity in the Hagley volcanic centre. Such beds are only present along the outcrop immediately east of Rockabank [SO 281 991], where a rhyolite is separated from volcanic formations above and below by shales each with an intermediate bed of sandstone. Some rhyolite was recorded at outcrop [SO 282 994] and [SJ 2845 0500], and correlated across a north-east-trending fault; these outcrops differ from rhyolitic lava [SO 2828 0000] which occurs in the basal leaf of the overlying Whittery Volcanic Formation. (BGS, 1991; in depicting these areas as rhyolite on the 1:25 000 Series map, the proportion of rhyolite has been exaggerated somewhat.) Whittard (1979), however, included the higher outcrop [SO 2828 0000] as part of his Hagley Shale Member.

Biostratigraphy

Broeggerolithus broeggeri, B. soudleyensis and Salterolithus caractaci are the only trilobites which have been recorded (Bancroft, 1933; Whittard, 1966; Rushton, in Lynas, 1985b). They evince an early to mid-Soudleyan age. Graptolites recorded from the formation include Diplograptus leptotheca, Climacograptus antiquus antiquus, Diplograptus foliaceus, Orthograptus amplexicaulis and Pseudoclimacograptus scharenbergi scharenbergi. These establish a position within the D. multidens Biozone.

Williams (1974, p.22 table 10) recorded 15 species of brachiopod from the ‘Hagley Shales’; of these Sericoidea cf. abdita occurs in the micaceous siltstones, whilst all the others are associated with tuffaceous deposits. He identified them as part of an evolving Nicolella association (Williams, 1973 p.242), known to have occupied the Anglo-Welsh Basin sporadically throughout Caradoc times, but they do not assist in a more refined correlation with other areas.

Details

In Church Stoke, a section [SO 2707 9416] (now mostly obscured) revealed micaceous laminated silty shales with some whitish grey, thin, quartzo-feldspathic sandstone beds up to 30 mm, but exceptionally up to 250 mm thick. At the west end of this section, a massive, rotten, dark tuffitic bed less than 1 m thick yields abundant fossils including a dalmanellid, Kiaeromena cf. kjerulfi?, Sulevorthi (= Orthambonites)exopunctata, Rostricellula sp., Cyclonema crebristria, Sinuites pseudocompressus, an indeterminate orthoconic nautiloid, Broeggerolithus broeggeri, B. cf. soudleyensis, Salterolithus caractaci, crinoid columnals and Diplograptus foliaceus. At another exposure nearby [SO 2725 9429] laminated micaceous siltstones and thin feldspathic sandstones are interbedded with decalcified tuffites containing clasts, up to 10 mm in size, of lava and mudstone. The tuffite at this locality is fossiliferous, with coarse-ribbed orthids. In thin section (E56871), it is a highly immature unsorted volcaniclastic wacke, containing clasts of porphyritic lavas set in a carbonate poikilitic base.

Exposures are numerous in Marrington Dingle and the River Camlad south of Church Stoke Vicarage [SO 2715 9391], where beds of medium grey, fine-grained sandstones up to 100 mm thick are common. The grains are angular and mudstone pellets are present.

In the track at Alport [SO 2743 9522], the formation consists of micaceous laminated siltstones and silty shales, all grey-buff in colour. Some beds are tough and vaguely nodular. Similar beds occur in Marrington Dingle [SO 2759 9622] and nearby [SO 2759 9605] an interbedded tuffite yields graptolites in relief and the trinuclids B. broeggeri and S. caractaci. In the small stream [SO 2752 9699] north-west of Kingswood, approximately 3 m of grey, medium-grained, felspathic sandstone occur in beds 0.30 m thick; sandstone is also exposed 170 m to the east.

In the complex northern sector, the underlying Hagley Volcanic Formation appears to pass abruptly up into grey Hagley Shale, but the shale is superseded by a distinctive, fine-grained, cross-laminated, green sandstone. This is exposed better to the north, but forms a positive feature in places [SO 2834 9983] and [SO 2834 9934]. Exposures are few, but they reveal an apparent facies change southwards approaching the fault [SO 2834 9934] the sandstone becoming coarser and more obviously tuffitic.

The sandstone is overlain by another shale with one exposure [SO 2829 9982] and fragmented evidence along its outcrop. This is, in turn, overlain by an intermediate to acid pyroclastic member [SO 282 994]. Also poorly exposed, it produces a strong north-trending ridge, noticeably oblique to the underlying and overlying volcanic formations. Within much of this member the pyroclasts have acid affinities, and thus the outcrop is regarded as unrelated to the underlying Hagley Volcanic Formation. However, the uppermost beds are more andesitic [SO 2814 9960]. In thin section, the rock at this exposure is packed with a variety of volcaniclasts, all feldspar-phyric and varying from acidic to more basic. Some are dioritic, possessing an intergranular to hypidiomorphic granular texture.

No upper or lower contact is exposed, but siltstone fragments [SO 281 992] reveal that the member is overlain by another shale, which coarsens upwards into a sandstone. This is exposed sporadically [SO 2806 9923] to [SO 2812 9975], and is identical with the lower sandstone member described above. The sandstone passes up into a thin, grey shale which underlies the Whittery Volcanic Formation at the Sheds [SO 2801 9945]. Traced to the south, the sandstone becomes finer grained, apparently passing laterally into siltstones and shales.

The intermediate to acid member alluded to above is well exposed [SJ 2845 0066] north-east of Wotherton. It comprises massive volcaniclastic breccias containing blocks up to 0.4 m in diameter, as well as lapilli and ash. The volcaniclasts are very altered, subrounded to angular feldspar porphyry (E57373). The feldspars show remnants of zoning, and there are pseudomorphs of carbonates and indeterminate secondary minerals after mafic minerals. Patchy recrystallisation of the fine-grained groundmass suggests devitrification from a formerly glassy mesostasis.

In the Wotherton area, shale overlies the Hagley Volcanic Formation but gives way rapidly to greenish cross-laminated sandstone [SJ 2848 0032]. It is exposed better in an old quarry [SJ 2856 0087], which shows about 4 m of cross-laminated sandstones and siltstones. The sandstones occur in beds 0.10 to 0.30 m thick, interbedded with similarly thick shales in the lower part of the section, but with the sandstone becoming dominant higher in the quarry. The beds show current-rippled tops and erosive bases with load casts and indeterminate trace fossils. In thin section (E57372), the rock is well sorted with a grain-supported texture; the dominant clasts are angular quartz (about 0.1 mm diameter) with some altered feldspar and detrital mica in a chloritic matrix.

Whittery Volcanic Formation

This part of the sequence was originally included with the ‘Whittery ashes and overlying shales’ of Lapworth (1887a, b), a subdivision of his Chirbury Series, and was later named the Whittery Ash (Lapworth and Watts, 1894). Whittard (1979) introduced the name Whittery Volcanic Member which was subsequently changed to formation status (Lynas, 1985b). Whittery Quarry is considered to be the type section (Whittard, 1979, p.56).

The formation is absent south of Alport [SO 273 953]; northwards, the outcrop is marked by hills in the Rockabank and Wotherton areas and it forms the steep eastern side of Marrington Dingle, just north of Calcot Farm [SO 2736 9592]. In the past, these rocks have been worked from several quarries along the outcrop, on account of the freestone properties of the feldspathic sandstones. No exposure of the base of the formation has been recorded. The top is exposed in the type locality at Whittery Quarry. The maximum thickness of the formation is about 150 m at Rockabank, it thins slightly to the north and dies out to the south near Alport.

The formation consists of a variety of coarse volcaniclastic breccias, fine tuffites, minor rhyolitic lavas and shales overlain generally by massive, brindled, feldspathic sandstones. The rocks are very similar to those of the Hagley Volcanic Formation indicating a renewal of the volcanic activity in the area. The rocks are coarsest, thickest and thus probably nearest to source between Bank Farm [SO 2748 9775] and Rockabank where they form a prominent hill and are exposed in cliff-like exposures as agglomeratic rudite.

Biostratigraphy

The stratigraphical tables of Whittard (1966) do not record trilobites from the Whittery Volcanic Formation. Graptolites have been obtained in a tributary [SO 2770 9699] (Whittard, 1979, p.58), but were not named. Brachiopods too appear not to have been recorded (Williams, 1974), though it seems possible that they have been included with the fauna of the Whittery Shales (Williams, 1974, p.21).

Details

The southernmost exposure [SO 2738 9602] which is unequivocally of the Whittery Volcanic Formation is just north of Calcot Farm. Here, massive feldspathic tuffs are interbedded with shales and include a coarse rubbly volcaniclastic bed. Farther south, a vague feature can be traced for a few hundred metres before being lost around Alport. Northward, an exposure [SO 2741 9628] reveals massive, crystal lapilli tuffs overlying 2 m of wavy-bedded, blocky, fine-grained, feldspathic sandstone which passes down into massive feldspathic sandstone.

South of Bank Farm, 10 m of this variable formation is exposed at the top of Caerbre Quarry [SO 2741 9650] (Figure 15) (Lynas, 1985b, fig.6) as an upward-coarsening sequence of massive sandstone beds, 1 to 2 m thick, and shale partings. Beds 5 and 6 are sandstones with a tightly packed mix of microphyric basic and feldspar porphyry clasts (E56991) and rafts of shale. Bed 7 is a coarser volcaniclastic sandstone which coarsens upwards. At the base are lobes (E56992) containing clasts of altered feldspar porphyry in a matrix of angular feldspar chips of about 0.1 mm size. Bed 8 is a coarse volcaniclastic sandstone which coarsens upwards. The highest bed (Bed 9)(E56993) is a coarse volcaniclastic sandstone; clasts range up to 50 mm in size and include feldspar porphyry and others of a fairly basic composition with trachytic texture. The grains are tightly sutured. Nearby, in the River Camlad [SO 2735 9684], contact with the overlying Whittery Shale is visible. Massive fine-grained, feldspathic, crystal tuff forms a waterfall and on the west side passes, via two thinner tuff beds, into slightly disturbed grey shale. The nature of the contact is unclear.

North of Bank Farm [SO 2748 9775], a rocky knoll exposes a coarse volcaniclastic breccia which underlies the massive feldspathic sandstone facies seen in the nearby quarries (see below). The rock is massive and chaotically mixed with light grey porphyritic clasts up to 1 m, but most clasts are 0.05 to 0.30 m across. Clasts are angular, though some are rounded, and all are packed together in a darker crystal-rich matrix and weather proud of the tuff surface. In thin section (E57337), (E57338), the clasts consist of both trachytic-textured andesitic feldspar porphyry and feldspar pyroxene porphyry of basic to intermediate composition, also with a trachytic texture and with chlorite pseudomorphed after pyroxene. Similar rocks, though not as coarse, are seen 200 m south, below the farmhouse.

The top of the formation is exposed in Whittery Quarry [SO 2747 9810] of Whittard (1979, p.54). Here, about 8 m of massive feldspathic sandstone, in beds 1 to 2 m thick, are overlain by soft grey or greenish grey shale with subordinate, rhythmic, beds of feldspathic sandstone, 0.02 to 0.15 m thick. The massive feldspathic sandstone is brindled, or mottled, and identical in appearance with those from the Hagley Volcanic Formation in Marrington Dingle (Lynas, 1985b, fig.4). They typify the unmappable, but distinct, upper sandstone member of the Whittery Volcanic Formation.

At Bank [SO 2766 9835], feldspathic volcaniclastic breccias are exposed. The clasts range up to 0.20 m in size and consist largely of porphyritic andesite. The rock is packed with altered feldspars and angular feldspar porphyry clasts of all sizes, some with trachytic texture (E57356). In the midst of these occur lenses of fine-grained, well-bedded, brown tuffites, succeeded by coarse tuffs. The latter tuffs are bedded in parts, and pass up into the brindled massive sandstone facies seen at Whittery Quarry.

Near the Sheds [SO 2798 9944], the formation is divided by a mappable leaf of shale. The lower beds are well exposed in a series of knoll-like features trending north towards Rockabank. There is little evidence of faulting and the knolls may simply reflect the nature of the deposits. The exposures by The Sheds are described graphically in Lynas, 1985b, fig. 7. The succession is about 24 m thick with evidence of grey shale above and below. The basal 10 m consist of volcaniclastic feldspathic sandstone. In the centre, with poor exposure above and below, is a bed, 1.4 m of which are visible, in which angular and rounded fragments of volcanic rocks are contained in a volcaniclastic matrix. Clasts consist of strongly banded feldspar-phyric lavas of intermediate composition which contain altered biotite and a little primary quartz (E58106), and of rounded altered feldspar porphyry (E58107). There is some evidence that the clasts become smaller towards the south. At the top are some 4 m of wavy-laminated fine-grained sandstones with a few thicker (0.10 m) more massive beds.

Rock which probably originated as a glassy, acidic lava is exposed on a knoll [SO 2800 9964]. It contains numerous layered or banded acidic pyroclasts up to 50 mm in size, but at the north-west end of the exposures, intermittent banding, 0.2 m thick, suggests either that this is a giant exotic clast, or that it is the flow-banding of a lava in situ. The rock is feldspar-phyric (E57365) and contains a few highly corroded quartz phenocrysts. The matrix is vaguely banded and shows devitrification textures. Some of the clasts in the nearby coarse-grained volcaniclastic breccia may have been derived from this lava. Similar lavas are present just to the north [SJ 2821 0017] and [SJ 2823 0003] in a suite which also includes tuffs and breccias. At the former locality the lowest parts look like massive lapilli tuff but in thin section (E57359) present the appearance of an auto-brecciated acidic lava. The clasts are tightly packed and consist of fine-grained recrystallised quartz-feldspar mosaics with sporadic corroded quartz phenocrysts and flow-oriented feldspar microlites. Some vivid pale green pumiceous clasts occur (E57360). The higher beds seen in these exposures are pale grey to white and siliceous. Large masses of flow-banded lavas are locally distinct (E57361) consisting of strongly recrystallised quartz-feldspar mosaics with distinct banding. The texture of this rock is almost hyaloclastitic in places.

The portion of the formation above the shale ‘leaf’ is exposed in Spring Coppice [SO 2788 9954] where the block breccia is andesitic and identical with that in the Rockabank crags. The highest exposures are of feldspathic tuffites revealing a passage upwards into finer grained and well-foliated rocks. At Rockabank [SO 2801 9978], the entire crag of some 200 to 300 sq.m. is formed of a coarse pyroclastic breccia (Whittard, 1979 fig.37) with clasts ranging in size to over 1 m. The pyroclasts are mid-grey feldspar-phyric andesites and dacites and most exceed 50 mm in size. In thin section (E57358), they exhibit glomeroporphyritic altered plagioclase phenocrysts up to 3 mm in size, and brown pseudomorphs, some of which appear to have been pyroxene. The clasts were isotopically dated (Rundle, 1984a and b; Lynas, Rundle and Sanderson, 1985) to give a Rb–Sr isochron age of 454 ± 4 Ma, and K–Ar ages of 437 ± 10 Ma and 449 ± 10 Ma. Although angular in shape, the pyroclasts have rounded edges and it is clear that the entire rock is composed of lavas preserved at various stages of mechanical brecciation, which was possibly violent. The totally massive nature of this deposit is characteristic of mass-flows as well as of vent agglomerates, but the evidence strongly suggests that the Rockabank crags represents a source vent for part, possibly all, of the Whittery Volcanic Formation.

This facies can be traced northwards [SJ 2808 0010] to Wotherton Mine [SJ 2805 0042], but it is fairly abruptly replaced by crystal tuff. A block-lapilli tuff which is also laterally impersistent is exposed in an old level [SJ 2798 0024], but other exposures in the area are of massive feldspathic sandstone [SJ 2798 0042] as in Whittery Quarry.

Whittery Shale Formation

This part of the sequence was originally included within the ‘Whittery ashes and overlying shales’ of Lapworth (1887a, b). It was named Whittery Shales by Lapworth and Watts (1894) and Whittery Shale Member by Whittard (1979). The sequence contains mappable sandstone members and it was thus given formational status by Lynas (1985b).

The formation caps the Shelve Ordovician sequence and is concealed by an extensive cover of Quaternary deposits in the west (Figure 4). It crops out from Church Stoke in the south to Wotherton in the north. Its broadest part, 2 km, is between Dingle Bridge and Timberth where it is thought to be overlapped by Llandovery strata. In this area dips, though few, are various and opposing, so the total thickness may be no more than 600 m. The relationship with the Ordovician rocks farther west at Montgomery, in the Forden Inlier, is obscure; the two outcrops are separated by Wenlock strata.

The Whittery Shale is very like the Hagley Shale, consisting mainly of olive and grey micaceous shales with subordinate thin, fine-grained, laminated sandstones and a few tuffite layers up to 30 mm thick. In the north, between Whittery Bridge and Wotherton, some tuffitic sandstones are sufficiently thick to form mappable members.

Biostratigraphy

The formation yields a mixed shelly and graptolitic fauna. Trilobites recorded by Whittard (1966) are as follows Amphilichas fryi, Broeggerolithus broeggeri, Flexicalymene cf. acantha, F. planimarginata, Decoroproetus calvus, Stenopareia camladica and Salterolithus caractaci. The association of B. broeggeri and S. caractaci is evidence of a Soudleyan age; the presence of F. planimarginata reinforces the evidence from the three underlying Soudleyan formations that the Whittery Shale is no older than mid-Soudleyan, while S. caractaci precludes the beds from being very late Soudleyan.

Graptolites listed by Hughes (1989) include Amplexograptus leptotheca, Diplograptus foliaceus, Orthograptus amplexicaulis and Pseudoclimacograptus scharenbergi scharenbergi. These indicate the multidens Biozone.

Brachiopods are associated with the sandstones rather than the shales and may have been transported from nearby shallower areas, partly in mass-flows. The list given by Williams (1974, p.23, table 11), incorporates the brachiopods of both the Whittery Shales and the coarser, sandy Whittery Volcanic Formation. Together they comprise a fauna very similar to that of the two Hagley formations (Williams, 1974, p.22, table 10), although the Whittery faunas are both richer and more diverse with 23 genera. Williams regarded the Whittery brachiopods mainly as part of a Nicolella association that typically inhabited lime-rich silty substrates (although the shale of the formation is not notably calcareous).

Details

In the south, grey, micaceous shales with subordinate laminated sandstones and some thicker (up to 30 mm) tuffitic beds are exposed in Marrington Dingle [SO 2725 9610]. The base of the formation is exposed in Whittery Quarry [SO 2735 9685] (p.42) while olive-grey shales higher in the formation are exposed near Whittery Bridge [SO 2706 9820]. Farther north, fine- to medium-grained cross-laminated sandstone members become more important. One sandstone member has been traced from Whittery Bridge north-north-eastwards, and in an old quarry [SO 2735 9888] consists of poorly bedded, rubbly, coarse, feldspathic sandstone (probably volcaniclastic in origin) with thin shale partings; at the north end of the quarry, coarsely ribbed orthid brachiopods occur in massive and coarse volcaniclastic breccias.

In the west bank of the Camlad [SO 2732 9902], another sandstone member is composed of cross-laminated and tuffitic sandstone interbedded with silty shales. The sandstone (E57355) is composed of angular, well-sorted quartz, feldspar, indeterminate chloritic fragments and mica, all up to 0.1 mm in size.

In the extreme north, characteristic shales are exposed on both sides of the River Camlad near Hockleton Farm [SJ 2739 0003] as well as in Rock Coppice quarries.

Forden Mudstones Formation

The formation is based upon the outcrop of Caradoc rocks exposed in the Forden Inlier in the west of the district. Apart from a study by the late Professor D V Ager in the 1950s little attention has been paid to these rocks. Murchison (1839) recorded isolated bodies of volcanic rocks, which he considered to be intrusive, at Nant Cribba (1839, p.287, pl.32, fig. 1), Stallow (Stalloe) and Montgomery Castle (1839, p.288). No description of the great thickness of mudstones within which they are bedded has been given.

The inlier lies mainly within the Welshpool district (Sheet 151). This survey has shown that the inlier is much larger than hitherto believed and extends southwards from Montgomery to near Sarn and westwards across the Severn to Garthmyl (Figure 1). It embraces the two small areas of ‘Greenstone’ shown within the ‘Upper Silurian’ strata on some editions of the one-inch old series geological map Sheet 60 SE at Montgomery and Stalloe. The inlier is fault-bounded and partially concealed by Quaternary deposits.

The recent survey extended some miles north of the district towards Forden, and Ager reconnoitred the whole of the northern part of the inlier but little variation in the rocks was apparent. In large tracts of the outcrop the strata are vertical, or overturned to dip westward. The inlier may be in the form of an eastward-verging anticline and since there is neither top nor bottom by which to calculate the thickness of the formation it can only be estimated that there are well over 1000 m of strata between Garthmyl and Forden, and some 700 m south of Montgomery.

The formation consists dominantly of medium grey and olive-coloured, soft, micaceous, shaly mudstone, thinly interbedded with thin, fine-grained sandstones up to 0.07 m thick. Thicker, buff-weathering sandstones up to 0.33 m thick occur sporadically, or in packets up to 10 m thick, particularly in the extreme south. These thicker sandstones are strongly cross-bedded indicating currents from the east-north-east and some reveal sole marks including groove casts aligned north-north-east to south-south-west and flute casts indicating currents from the north-east (between 025° and 050°). In places, both sandstones and mudstones are bioturbated (Plate 1).

The Forden Mudstone Formation was deposited under conditions similar to those of the Hagley Shale and Whittery Shale formations. These formations were deposits of the same ‘basin’ receiving an abundant supply of mud, silt and fine sand by means of very frequent low energy currents. There is little evidence of wave-interference of the sands, but it is possible the mud was re-suspended on occasions. The sandstones show many of the signs of traction currents, with much cross- and parallel-lamination, but grading and non-laminate bases are present too and simulate turbidites. It is therefore possible to invoke either means of transportation with the currents generated by storm-action on distant shallower areas. Currents in the area south of Montgomery and at Church Stoke came from the east-north-east and north-east, not from the north-north-east and if these reflect the submarine gradient in the area at the time, then the source of the volcanic conglomerate member of the Forden Mudstone Formation may lie in the Worthen area of Shelve rather than in the Breidden Hills.

These conditions remained unchanged throughout early and mid-Soudleyan times, apart from two significant interruptions from nearby volcanic centres which produced ephemeral piles of ejectamenta. Around these, sandstones and conglomerates were distributed to form the Hagley and Whittery Volcanic formations. The small packets of massive though fairly thin sandstones within the shale formations may well have originated from minor activity at the volcanic centres.

The benthic shelly faunas of the three mudstone formations, the Hagley Shale, the Whittery Shale and the Forden Mudstone, are very sparse. The bulky documentation is the result of intensive and protracted collecting mainly by Whittard and this may have been biased towards allochthonous faunas from the volcanogenic sandstones. The substrate was not attractive to these forms of life, either because the water was too deep or because the influx of mud was too rapid. Dark grey, finely laminar hemipelagite is not evident while bioturbation is evident, mainly as burrow-casts on sandstone bases but occasionally within mudstone. Lack of oxygen was not, therefore, the inhibitor.

Volcanic conglomerate and associated strata

Volcanic conglomerate is exposed at Castle Hill, Montgomery, Stalloe, possibly Salt Bridge, Cwm Farm and Forden Castle. It is a pale grey, massive rock, without a bedding foliation. It is composed of rounded clasts of porphyritic andesite within a feldspathic matrix which is so similar in appearance as to make the clasts barely discernable. The clasts are distinct only on weathered surfaces, and in the natural exposures just south of Stalloe they are excellently displayed, and up to boulder size. In general, clast sizes range up to 1 m across. Normally two varieties of andesite are present, a pale grey variety and a darker, greenish grey variety in which the feldspar phenocrysts are zoned. Petrographical descriptions of the rock were published by Sanderson (1975) and Dixon (1988) and it is clear that the clasts are identical with the andesitic rocks in the outcrop between Moel y Golfa and Bausley Hill some 9 miles to the north-north-east of Forden.

The clasts are ill sorted and mainly matrix supported, and the conglomerate has all the characteristics of a mass-flow deposit, or a lahar as suggested by Sanderson. At Montgomery Castle, apophyses of mudstone penetrate several centimetres upwards into the base of the conglomerate upon which a trilobite (S. caractaci) is embedded. It appears to be uncooked and therefore the lahar was probably cold.

The surface is overlain by a thin sandstone sequence with a conglomerate containing rounded clasts of andesite and is probably a local debris flow deposit. The sandstones yield the only diagnostic shelly fauna of the formation.

Each of the volcanic conglomerate outcrops is elliptical in plan, which, therefore, must be the shape of the conglomeratic mass in vertical section. This may imply that they travelled from source as mass-flow trains which eventually bulked up at the bottom into isolate masses, or that the outcrops are the cross-sections of ribbon-like fill of shallow channels. The latter possibility is favoured because all four of the well exposed conglomerates are overlain by a thin sequence of sandstones and conglomerates. This very specific association with the volcanic conglomerate — for such beds do not occur in other parts of the sequence — must signify a genetic relationship. The mass-flow seems to have entrained a series of waning marine tractional currents behind and over it, which continued for some time depositing a thinning and fining sequence of sandstones before the normal marine deposition resumed. Some form of shallow channel, therefore, might well have effected a confining and quickening influence to these flows. There is, however, no evidence of a well-developed turbidite fan.

Biostratigraphy

The Forden Mudstones are almost barren of macrofossils. A few plectambonitid brachiopods occur in mudstones at Garthmyl and a fragmentary climacograptid graptolite was obtained in the extreme south. Professor D V Ager obtained better graptolites from Edderton Dingle, north-west of Forden. These were identified as Climacograptus aff. antiquus (information from Dr I Strachan, 1954). He had noted similar forms in the Hagley Volcanic Formation and the lower part of the Hagley Shale which he considered to indicate the D. multidens Biozone.

The most distinctive fauna, however, is the exotic benthic fauna in sandstone overlying the volcanic conglomerate at Montgomery. It includes acritarchs, chitinozoa, trilobites, ostracods and graptolites (Cave et al., 1988). The trilobites are B. broeggeri and S. caractaci and indicate correlation with the Hagley and Whittery formations of Shelve. These trilobites together with the graptolites of Ager prove a similar link with the middle and upper parts of the sequence of the Breidden Hills (Wedd, 1932). Clearly, the andesitic conglomerates of the Forden Mudstones were magmatically related to the Moel y Golfa Andesite to the north and the andesitic conglomerates of Wedd (the Bulthy Formation of Dixon, 1990). The outcrops of volcanic conglomerate are aligned north–south between Montgomery and Forden, parallel with the regional strike. This may indicate that they all lie at one horizon and are probably the equivalent of the Bulthy Formation of the Breidden area, which had a volcanic centre at Moel y Golfa. A similar centre in the Shelve area might have been at Rockabank on the outcrop of the Whittery Volcanic Formation and others may lie between the two, concealed under Long Mountain. From which of these centres the Montgomery–Forden volcanic conglomerates came is not clear.

The distribution of trinucleid trilobites in these Caradoc sequences raises a perplexing problem. East of the Long Mynd and at Welshpool to the west the abundance of S. caractaci, not occurring in association with B. broeggeri, has been claimed to indicate the Harnagian Stage. In the intervening areas of Habberly Brook (Dean and Dineley, 1961), Shelve and the Breidden Hills this criterion fails; S. caractaci occurs only in association with B. broeggeri, and where these are found together the strata are assigned to the Soudleyan Stage.There are no apparent differences between these Soudleyan S. caractaci and those from supposed Harnagian strata at Welshpool except that examples from Habberley Brook are larger (Cave, 1955). Therefore a sedimentary reason must be sought for their absence from the lowest Caradoc mudstones and shales in the Shelve, Breidden Hills and possibly Forden areas.

Salterolithus caractaci is found most commonly in mudrocks, even the occurrences in the neptunian dykes on Hazler Hill, Church Stretton have a mudstone matrix (Strachan et al., 1948). Thus it is clear that the animal flourished on muddy deposits, yet the muds of the Forden area and the early Caradoc muds of Shelve and the Breidden Hills failed to satisfy its requirements. It appears to have been fastidious; perhaps it did not like deeper water and perhaps during Harnagian times the water in the Shelve and Breidden areas was indeed slightly deeper than that to the east and west. If so, it is significant that the taxon also eschewed the muddy environment of the early stage of the marine transgression of the east-wall of the Pontesford–Linley Fault, where there is no Costonian Sandstone, to appear in abundance a little later in the Soudleyan (Dean and Dineley, 1961, p.369). It suggests that the east-wall platform area foundered initially very rapidly to a depth greater than the tolerance of S. caractaci, then, as elsewhere, the water became shallower during the Soudleyan. This foundering must have arisen by relative movement, down-to-the-east, on the fault, the reverse of earlier movement (p.131).

Micropalaeontology

Mudstone from the bed of the River Severn and exposures to the east yield a diverse assemblage of acritarchs, chitinozoa and sporomorphs. There is evidence (BGS, PD Reports 90/63, 90/61, 87/255 and 87/227) that the acritarch flora is typical of the late Ordovician, and is most likely to be of Caradoc age; Pheoclosterium fuscinulaegerum has been recorded only from the Caradoc (Turner, 1984). Ancyrochitina alaticornis, Spinachitina bulmani and Desmochitina minor cocca are recorded from the Caradoc and Conochitina chydea from the Llanvirn to the Caradoc. Belonechitina robusta may range from the early Caradoc to the middle Llandovery (Grahn and Bergström, 1984), although recent compilations of Ordovician and Silurian chitinozoan ranges (Paris, 1990; Verniers et al., 1995) suggest that B. robusta s.s. may be restricted to the Caradoc. B. capitata and B. micracantha are long-ranging through the Ordovician. The present sparsity of comparative British Caradoc acritarch and chitinozoan data precludes a more precise age determination, although there is undoubted potential for refinement, particularly with Chitinozoa.

The common occurrence of sporomorphs and the nonmarine ‘acritarch’ Moyeria cabottii suggests proximity to a source of terrigenous material. The sporomorph assemblage is significant as many of the taxa have not previously been recorded in strata older than earliest Llandovery.

Details

The southernmost exposures of the formation are in a bank [SO 2178 9264] and [SO 2186 9272] about 2 km north-east of Sarn revealing medium-grey shale and interbeds of grey-green bioturbated sandstone up to 30 mm thick. Similar westward dipping strata are exposed sporadically along a few hundreds of metres of the dingles west of New House [SO 2227 9298] and Cwm Bromley [SO 2265 9343]. In the former, there are bioturbated, medium-grey, finely micaceous mudstones with interbedded thin sandstones up to 0.20 m thick, pale and dark grey, with bioturbated top parts [e.g. [SO 2211 9303]. At this locality, well preserved acritarchs and chitinozoa support a mid-Ordovician age (BGS, PD Report 87/225). Nearby [SO 2209 9303], buff-weathering, silty mudstone yielded a fragment of Climacograptus?.

In Cwm Bromley Dingle, 9 m of buff and olive-grey silty sandstones have been quarried [SO 2215 9373]. The sandstone beds are up to 0.25 m thick and reveal bottom structures which include flute casts, bounce-casts and trace fossils. Palaeocurrents from 020° and 025° are indicated by the flutes, while trough cross-bedding suggests a subsequent flow regime from 050°. Examination failed to reveal palynmorphs.

Some 2 km to the north in the banks of the old coach road [SO 2226 9562], many metres of olive-grey mudstone and interbedded, cross-bedded fine-grained sandstone up to 20 mm thick are exposed. At the south end of the yard [SO 2217 9636] behind the Dragon Hotel in Montgomery, about 12 m of smooth olive-grey mudstones and thin, fine-grained sandstones have been exposed. The sandstones are finely micaceous, grey-buff with wavy argillaceous laminae, up to 70 mm thick and 20 to 200 mm apart. The beds are overturned with a 55° dip west.

Shales and thin sandstones almost in contact with massive volcanic conglomerate are exposed in the banks of the lane [SO 2212 9649] from the town square to the castle; details of the volcanic conglomerate and adjacent sedimentary strata have been published elsewhere (Sanderson, 1975; Cave et al., 1988; Dixon, 1988). The overlying strata have yielded the main shelly fauna so far collected from the formation. On the west side of the castle, under the keep, S. caractaci is preserved, embedded into the base of the volcanic conglomerate which here has a nearly vertical dip eastwards.

At Stalloe, an old quarry [SO 2239 9847] revealed overturned, westward dipping shales and sandstone up to 30 mm thick, with a thin basal conglomerate. This rests on the slightly uneven top of a pale grey, massive volcanic conglomerate, 2.5 m was seen. About 250 m to the south-south-west, a surface of bare rock [SO 2250 9826] shows the coarse, ill-sorted, matrix-supported nature of this conglomerate on a weathered surface which displays clasts of andesite up to boulder size.

The formation is well exposed in the railway cutting [SO 2184 9977] and in a quarry [SO 2058 9974] to the west of Gaer. In the latter, 6 m of olive-brown, smooth, finely micaceous, shaly mudstones are interbedded with thin fine-grained sandstones up to 0.20 m thick, but otherwise similar to beds above. Below and a few metres to the west is a further metre of shaly mudstone with thin sandstones. The overall appearance of these beds is similar to that of the Pwll-y-Glo Formation of Soudleyan age near Welshpool (Cave and Dixon, 1993), but without a macrofauna. Acritarchs, chitinozoa and sporomorphs are, however, prolific (BGS, PD Report 90/61 samples (MPA29311) and (MPA29312) and indicate a probable Caradoc age.

Grey shales with very thin sandstones are exposed in the River Severn nearby [SO 2028 9935] and in the adjacent left bank tributary. Again many acritarchs and chitinozoa are present indicating a probable Caradoc age (BGS, PD Report 90/61 sample (MPA29354)).

Beyond the western edge of the Severn Valley alluvium, the formation is exposed at Garthmyl [SO 1941 9908] where very small plectambonitids and ostracods are present in olive-grey and buff shales. Similar beds are present on the left bank of the canal [SO 1940 9912].

Chapter 5 Silurian: Introduction and Llandovery

Introduction

The Silurian Period in eastern Wales was markedly different from the preceding Ordovician. Change was clearly evident at the end of the Caradoc Epoch and by the onset of the Silurian Period at least four events had contributed to it. First, there was continental or arc-to-continent, impact in late Caradoc times (Campbell, 1984) which ended southward subduction beneath Avalonia and curtailed volcanism. Secondly, crustal compression was relieved along the fault-controlled platform edge by inversion of the Builth–Shelve ‘basin’ (Chapter Three) leaving the deeper, turbidite basin to the west largely undeformed. Thirdly, there was expansion, then collapse, of a polar ice cap and the concomitant rapid fall and rise in sea level. The fourth event was the eustatic rise in the early Silurian.

These events were responsible for radical changes in the style and nature of sedimentation, but a further factor had altered the chemistry of the detritus. Northward continental drift during the Ordovician Period had placed Eastern Avalonia in low southern latitudes. Thus, during Silurian marine inundations, carbonate secreting marine faunas flourished in platform areas and contributed more lime to outer-shelf and basin-margin clastic deposits than is evident in Ordovician rocks.

The effect of the late Caradoc collision was the production of an Ashgill passive-margin sediment repository ‘the Welsh Basin’. This repository was sustained by phases of extension on north–south down-to-basin faults, through to its occlusion at the end of the Silurian Period. Relating the remaining three further events to their respective effects is more difficult and, therefore, so is the dating of those events. Local tectonic activity in the form of the inversion of the Builth–Shelve ‘basin’ and the distant glaciation both contributed to a late Ashgill marine regression from eastern Wales and the Welsh Borderland. Clearly, the subsequent deglaciation partially reversed these effects by returning water to the oceans so producing glacioeustatic rise. However, the district was subjected to a protracted marine transgression of global significance which advanced incrementally throughout the Llandovery (Ziegler et al., 1968) signalling a staged, but progressive, eustatic rise (Johnson et al., 1991, p.147); Vail et al. (1977, p.84) record it as one of their second-order cycles of eustatic change. As such, it is the likely result of geotectonic reduction in the volume of oceanic basins (Hallam, 1971, p.182; Summerhayes, 1986, p.29) as distinct from the volume of oceanic water. Glacioeustatic changes were generally of shorter duration featuring amongst the third-order cycles of Vail et al. (1977, p.93), but to what extent the initial Llandovery glacioeustatic rise preceded the geotectonically driven eustatic rise, or how much they might have overlapped, is not evident here. However, the transgressional pause near the end of the Aeronian might signify some element of relay.

The Montgomery district lay close to or across the shelf edge at the eastern margin of the Silurian Welsh Basin. It was not a static edge, but one which migrated from a position just beyond the western side of the district in Llandovery times to the region of Newtown in Wenlock times, eventually becoming less well defined in Ludlow times, as it was overwhelmed by shelf sediment from the south-east. In common with other Welsh Borderland areas (Allen, 1985; Bassett et al., 1992, p.52) the district was part of a uniformly very shallow sea at the beginning of the Přídólí, and in the Downton Castle Sandstone of Ludlow there is evidence of a transient shore-face (Smith and Ainsworth, 1989). There was then a transition through repeated regression and transgression to an expansive, flat, fluvial to coastal flood plain, suffering semi-arid conditions, possibly with offshore winds. The influence of the transgressive episodes of the generally southward-receding epeiric sea is evident as sporadic occurrences of brackish-water faunas.

All four series are present in an outcrop occupying the greater part of the district (Figure 1) and the lithoformations are devices to express the facies pattern set by this variable system.

Llandovery

The Llandovery Epoch returned marine conditions to the district by an eastward transgression with pauses. After the first flush, probably incorporating much of the glacioeustatic rise, the shoreline established itself along the line of the Severn Valley in late Rhuddanian times ((Figure 16a) and it did not advance farther until late Aeronian times. Thus, Rhuddanian strata are present within the district west of the River Severn, but to the east late Llandovery strata (Telychian and some late Aeronian) rest unconformably upon mid-Caradoc and older rocks. By late Aeronian times, the whole suite of Shelve Ordovician rocks and large tracts of Precambrian rocks had been subjected not only to tectonism, but to prolonged erosion which in places had removed an estimated 10 km of strata. This material contributed to both the basin and shelf sequences from late Ashgill to end-Aeronian times. The Welsh Basin then lay west of the Llangadfan area and west of the Tywi Anticline and its eastern margin was probably fault-controlled. In the north, the basin margin and shoreline were widely separated, probably on two north–south faults ((Figure 16a), while to the south of the district they lay much closer together, with a steep shelf-to-basin slope. Farther east, the Longmynd Scarp Fault and Church Stretton Fault were active, and influenced sedimentation locally from mid-Llandovery times.

Ramsey and Aveline (1848, p.297) with Forbes (1848) were the first to recognise that the late Llandovery rocks (their ‘Caradoc Sandstone’) east of the Severn Valley were the deposits of a major marine transgression onto a landscape of Ordovician (their ‘Llandeilo Flags’) and other rocks (Ramsey, 1853; Salter and Aveline, 1854). The classification of these late Llandovery rocks stems from Ramsay, Aveline and Salter during the survey of the region for the old series one-inch maps of the Geological Survey (1850, 1855). On Sheets 60 NE and SE, around the Shelve Ordovician outcrop, they recognised ‘Upper Llandovery Rock: sandstone and limestone’, but for areas to the east a tripartite division was devised, in ascending order, Coarse grits, Pentamerus limestone and Purple shales. Whittard (1932, p.297) modified the latter scheme dividing the sequence into Pentamerus Beds and Purple Shales. He (Whittard, 1928, 1932) mapped the outcrops in detail, describing the rocks and their faunas. In addition, he interpreted the palaeogeography ((Figure 16b) and thus opened the door to a wider vision of the late Llandovery marine transgression (Ziegler et al., 1968). This survey adopts Whittard’s binary classification, and in detail has substantiated his record, with only minor additions. The Pentamerus Sandstone Formation incorporates the beds assigned by Ziegler et al. (1968) to the Bog Quartzite and, on the north side of the Shelve Ordovician outcrop, the Venusbank Formation.

Llwyn-y-Brain Formation

On the western side of the Severn Valley, rocks of Rhuddanian age discovered just within and beyond the northern margin of the district were described by Cave and Dixon (1993). Their presence contributes to the view that a tectonically controlled barrier existed along the line of the valley and inhibited the marine invasion of the Midland Platform to the east until late Aeronian times. The Llwyn-y-Brain Formation has been named after a farm 90 m west of one of the best sections [SO 1911 9998], which may be considered as the type section, just north of the district, near Garthmyl. It exposes over 10 m of buff-coloured mudstone that is silty, micaceous and bioturbated. Thin beds of fine-grained sandstone, less than 30 mm thick, occur in the top 8 m. There are several exposures in the vicinity of Llwyn-y-Brain, generally with thin beds of sandstone being more abundant. Within the Montgomery district, the formation consists of sandy mudstone which is homogeneous, tough, blocky and olive-grey in colour. It contains sparse quartz pebbles, up to 20 mm across which, in places, are segregated into small nests. Interbedded, darker grey, hard sandstones are up to 70 mm thick; these have retained a primary bedding lamination. In general, the formation is sparsely fossiliferous. A few brachiopods, including a large species of plectambonitid, are present at Fron. The sparse scatter of exotic pebbles can have been derived only from the east, but they appear to have no bed-form and thus have been involved probably in secondary disturbance either via mass-flow, or less likely as bioturbation.

The outcrop follows the base of the scarp-slope which forms the north-west side of the Severn Valley at Fron and passes approximately between Fron Church and the main A483 Trunk Road. Only two exposures of the formation occur within the district and neither bottom nor top of the formation is exposed. Indeed, the lower part of the formation is probably faulted against Wenlock strata along the line of the main road. About 70 m of the topmost part of the formation are present here, but the total thickness of the formation may be as much as 300 m. The outcrop appears to continue northward beyond the district to the type section of the Llwyn-y-Brain Formation, which has not yielded a fauna, and thence to Rose Hill where there is a conglomerate. The conglomerate yields a good Rhuddanian fauna (Cave and Dixon, 1993, p.59) and passes up into strata like those at Llwyn-y-Brain. The formation is thus comparable with the Laundry Mudstones and underlying Powis Castle Conglomerate even farther north.

Details

An excavated face, behind the cottage [SO 1790 9761] at the canal bridge in Fron, exposes 4 m of interbedded structureless sandy mudstones and hard laminated sandstones up to 70 mm thick. The mudstones contain a very small number of pebbles, up to 20 mm across, and a sparse collection of brachiopods. A palynological investigation (sample MPA 29531) revealed a poorly preserved assemblage of acritarchs, Chitinozoa and one indeterminate spore. The assemblage suggests either an Ashgill age or an early Llandovery age. Since the latter is the accepted age of the formation the Ordovician palynomorphs present must have been derived. From this, it is clear that exposed late Ordovician rocks east of the Severn Valley were contributing detritus westward into these deposits. Samples (MPA 29307, 29353 and 31404, PDR 90/63) from about 2 km to the north similarly yield assemblages of Ordovician appearance and presumably had incorporated material from late Ordovician rocks exposed nearby. A lane-side bank [SO 1792 9762], some 20 m north-east of the previous locality, exposes 4 m of strata, which have yielded a few brachiopods both from the structureless sandy mudstones and from the bedded sandstones.

Pentamerus Sandstone Formation

The Pentamerus Sandstone Formation is the modified name for the Pentamerus Beds of Whittard (1932, p.861), in order to apply a formal name to this unit that is sandstone dominated. The outcrop of the formation is lenticular and discontinuous. To the south of the Shelve Ordovician outcrop, the formation is present west of the Roveries [SO 318 922] and east of Linley [SO 359 928]; in the centre of the Ordovician outcrop several small outliers occur north-west of The Bog [SO 357 977], and at the northern edge of the Ordovician outcrop the formation is present around Rea Bridge [SJ 325 025] and Lower Wood [SJ 308 035]. The formation is estimated to be about 60 m thick west of the Roveries, and 40 m thick east of Linley. It was not present in either Church Stretton No. 1 Borehole [SO 3723 8978] or Church Stretton No. 2 borehole [SO 3940 9302] (Appendix 1), just to the east of the district. The formation consists of medium- to fine-grained sandstone with numerous clasts in the form of shells and a scatter of quartz and mudstone pebbles; a pebble conglomerate is common at or near the base. The base rests uncomfortably upon pre-Silurian rocks and Whittard (1932) reported that immediately above the contact in places the formation is incoherent and muddy.

The sandstone is commonly thickly bedded, or massive in the east, but flaggy beds are present particularly towards the west, where the beds pinch and swell. The sandstone is normally very calcareous and grades into a sandy limestone. Lenticular coquinae of shells occur commonly and weather to a brown very porous rock with a framework of shell moulds in places, mainly from pentamerid brachiopods. In polished section, shelly layers are commonly parallel-laminated with shells generally aligned with the laminae, but no other preferred orientation is seen; they may obtrude across several laminae. In thin section, for example (E57369), (E58777) and (E59498), the sandstone is well sorted, formed of subrounded to angular quartz grains generally 0.1 to 0.2 mm in diameter, but may be up to 0.6 mm, with subordinate feldspar and lithic grains. Matrix is sparse, but there is much carbonate cement and some overgrowth on quartz grains. In the outcrop to the east of Linley, massive sandstone makes up the whole of the formation. It is exposed at a number of places between Squire Hall [SO 353 930] and Oaklands [SO 376 923], thinning against Longmyndian rocks. The predominant lithology is a brown-weathering, greenish grey, fine- to medium-grained, calcareous sandstone, usually well bedded, and commonly containing abundant brachiopod moulds and shells.

The massive sandstones, conglomeratic in places, occur again near Roveries, but farther west these pass into finer-grained, thinner sandstones interbedded with silty mudstones as couplets. The sandstones are up to 100 mm thick, with sharp bases and show little bioturbation except as casts on their bases. Each sandstone is probably the result of a single depositional event (Bridges, 1975, p.86. These passages, geographically westward and temporally upwards, represent the migration eastward of storm-sheet sands of the inner shelf and their replacement by mid-shelf ‘turbidites’ of the type envisaged for the Hughley Shales by Benton and Grey (1981). The mudstones are grey-green, silty and bioturbated.

The same passage, from massive sandstone in the east to finer-grained sandstone–silty mudstone couplets in the west, occurs to the north of the Ordovician outcrop and again the same massive facies persists farther westward at the base than in higher parts.

The Bog outliers are no longer exposed. One [SO 351 977] occupies a slightly raised oval-shaped area some 250 m long, but the presence of another, claimed to exist some 150 m to the west (Whittard, 1932, p.879), has not been confirmed. Whittard (1932) recorded these rocks as hard compact flaggy sandstones, commonly coarse and conglomeratic; they are decalcified, and the clasts have been derived from local Ordovician rocks.

Biostratigraphy

The Pentamerus Sandstone of the district yields a prolific shelly fauna in places, and fossil lists have been published by Morton (1869) and Whittard (1932) which include graptolites from Snead. A comprehensive fossil list provided by Dr L R M Cocks, from a quarry near Norbury, is given in the details. Other important records, some reinterpreted by Ziegler et al. (1968), are:

1. Around Norbury, Meifodia subundata, Eocoelia intermedia, Atrypa sp., Pentamerus oblongus and Stricklandia lens.

2. Around Snead, P. oblongus, Calymene sp., Monograptus dextorsus and Stimulograptus halli.

Ziegler et al. (1968, p.746) consider the Norbury fauna to belong to the Pentamerus Community while at Snead the basal faunas are similar, but upwards reveal a deepening of marine conditions by the presence of a ‘Marginal’Clorinda Community. The diverse fauna from the Bog outliers was considered by Ziegler et al. (1968) to belong to a nearshore Cryptothyrella Community with rocky-bottom elements.

The age signified by these faunas is imprecise. Eocoelia intermedia is restricted to the late Aeronian (C3–C4) or halli Biozone (Cocks et al., 1984; Loydell, 1991), while S. halli ranges from latest Aeronian to early Telychian. Monograptus dextorsus has been listed from many places, but none has been confirmed outside Sweden (information from D K Loydell, 1994). Its listing by Whittard (1932) from Snead is therefore discounted for the present.

Details

Norbury area

Three old shafts on the northern side of Shuttock’s Wood were reputedly sunk for barytes. Surface debris from one shaft [SO 3730 9239] suggests it had penetrated the Pentamerus Sandstone into the underlying Longmyndian strata. In spoil from another of the shafts [SO 3730 9242], some malachite is present, and Pentamerus Sandstone debris is veined with barite.

The angular unconformity between the formation and the underlying Longmyndian has recently been exposed [SO 3633 9275] near Hall Farm. The Pentamerus Sandstone, dipping gently southwards, comprises fine-grained, well-bedded, buff-brown sandstone or silty sandstone. It appears to be unfossiliferous and is stained with malachite close to the plane of unconformity.

Close to Squire Hall, sandstone typical of the formation in this area is exposed in a small quarry [SO 3532 9287] and is seen in many other quarries and excavations eastwards towards Alma Cottage [SO 373 922]. The well-bedded calcareous sandstone occurs in beds up to 0.15 m thick; it varies from soft, buff-brown and silty to hard white and almost pure sandstone and contains narrow silty partings. Fossil distribution is uneven, brachiopod (Pentamerus sp.) moulds and shells are abundant in places and a few solitary corals occur. Barytes and malachite occur commonly as cement and secondary replacements. From one quarry [SO 3587 9284] on the north side of the lane, a comprehensive fauna was obtained by Dr L R M Cocks of the Natural History Museum who kindly allows publication of his identifications as follows:

Brachiopods Common: Pentamerus oblongus 36%, Eocoelia intermedia 21% and Brachyprion arenacea 16%. All 1% or less: Dolerorthis psygma, Mendacella lata, Resserella sefinensis, Eoplectodonta penkillensis, Mesopholidostrophia salopiensis, Coolinia applanata, Camerella? sp., Clorinda globosa, Protatrypa sp., Howellella anglica, Eospirifer sp. and Stegerhynchus transversarius. Molluscs (all less than 1%): Platyceras, Holopella, Pagodea, Fusispira? and Cardiola?. Others (all less than 1%, apart from crinoid debris 9%): Favosites, streptelasmatid coral, fenestellid bryozoan, Hallopora sp., serpulid, Tentaculites, Encrinurus and crinoid ossicles.

Snead area

The formation is exposed around Friar’s Rock [SO 3105 9256] and, 600 m to the east, in a stream [SO 3167 9255]. At Friar’s Rock, exposures in a roadside quarry, along a trackway and in the adjacent stream are mainly laminated, calcareous, flaggy sandstone with rather undulating bedding surfaces. The sandstone is interbedded with green-grey, commonly bioturbated, shaly siltstone, which weathers brown. The sandstones are composed dominantly of quartz grains, some being polycrystalline; a few indeterminate volcaniclasts and feldspar grains occur also. Grains are subrounded to angular, moderately well sorted and mostly between 0.2 and 0.6 mm across. Similar rocks are exposed in the stream to the east.

The Bog Outliers (and Hope Valley)

Although the quartzitic sandstones of the Bog Quartzite are no longer exposed, they have in the past yielded an extensive shelly fauna from outliers at Bank (Minsterley), Napp and Bog Mine (Whittard, 1932, p.879; Zeigler et al., 1968). It is dominated by the brachiopods Platystrophia brachynota, Mendacella lata, Leptaena contermina, Leptostrophia compressa, Atrypa sp. and Stricklandia lens: the trilobites Calymene planicurvata and Encrinurus onniensis: and Pitcher (1939) described several gastropods from the same strata.

Hope Quarry [SO 3550 0209] (BGS, 1991) is supervised by the Shropshire Conservation Trust, and provides a good exposure of the formation. The beds dip gently eastward and rest with sharp unconformity on the Hope Shales Formation which dips 35°E; 9 m of Pentamerus Sandstone are exposed. They consist mainly of thickly bedded laminated sandstone, with thinner beds pinching and swelling. Thicker beds are parallel laminated, but cross-bedding and possibly hummocky cross- stratification occurs. Lenticular concentrations of shell debris are also present. Thin partings of silty mudstone separate the sandstones and increase in thickness and abundance upwards. In thin section (E59500), the sandstone is well sorted, comprising subangular to angular quartz grains, about 0.2 mm in diameter. Accessory feldspar is present, and matrix is very sparse.

The basal bed (E59498) is a well sorted mature quartz sandstone with much feldspar. Grains are subangular to subrounded with slight silica overgrowths and their sizes range between 0.1 and 0.2 mm.

From the configuration of the unconformable base of the Pentamerus Sandstone around the Hope Valley, Whittard (1932, pp.893–895) deduced that the sand had been deposited in a valley during the late Aeronian marine invasion.

Purple Shales Formation

The Purple Shales Formation (after Whittard, 1932) rests conformably on the Pentamerus Sandstone Formation, or, where it is absent, it rests unconformably on Precambrian and Ordovician rocks. It is poorly exposed and largely concealed by Quaternary strata. The full thickness of the formation (186.28 m) has been proved in Church Stretton No.1 Borehole [SO 3723 8978]. It was also proved in Church Stretton No. 2 Borehole [SO 3940 9302], just to the east of the district, where it was described by Greig et al. (1968) under the name of Hughley Shales. Westward, the outcrop narrows and Whittard (1932, p.871) clearly believed that the thickness of the formation there was much smaller, quoting about 90 m near Snead. The formation is the shelly facies equivalent of the Tarannon Shales Formation (Ramsay, 1881, p.28) ((Figure 16b) which with its subjective synonym Dolgau Mudstones, is based on a basinal type section (Wood, 1906, p.657).

The formation consists dominantly of mudstone with a characteristic dark maroon or purple colour, and pale greyish green diffuse mottles and thin bands. The mudstone is slightly silty, commonly laminated and finely micaceous. Beds of calcareous ovoid nodules occur in parts, as do scattered thin (20 to 30 mm) limestones which are compact and crystalline. In places, for example near Snead, thin sandy beds have been recorded. Shell detritus is present in parts, concentrated as thin layers and lenses. In Church Stretton No. 1 Borehole, shell detritus is most abundant in the basal metre and in about 8 m of strata near the middle, but it is very sparse in the top third.

The yellow sandstones which occur in the Bog outliers seem to have been considered by Whittard (1932, p.862) to be part of the Purple Shales, perhaps on palaeontological grounds and in the belief that they are coeval (p.880). However they are incompatible on lithological grounds and Ziegler et al. (1968, p.744) indicate that the sandstones are of late Aeronian age like the low parts of the Pentamerus Sandstones nearby.

At the top of the formation near Plowden to the south-east of the district, Whittard (1932, p.869) considered there to be ‘no sharp demarcation in coloration between the Purple Shales and the succeeding Salopian rocks.’ ‘At the base of the latter there is a series of sediments which may be conveniently termed Purple Wenlock Beds’. He believed that the ‘Purple Shales conditions [of deposition] were re-established intermittently in the Salopian, ..... until about 40 feet [12 m] of sediment had been deposited’. This belief, that there are purple beds of Wenlock age, has persisted ever since, along with the opinion that the upper part of the Llandovery series here is missing (Grieg et al., 1968, pp.151 and 170; Cocks and Rickards, 1969, p.217) but no supporting evidence has been found. Evans (1957) logged four sections [SO 3811 8767], [SO 3812 8774], [SO 3936 8698], [SO 4319 9085] near Plowden which reveal that the upward change from purple to grey extends through 0.76 m of beds at most. At Buttington (Long Mountain), Palmer (1970) recorded the presence of the basal Wenlock centrifugus Biozone in grey calcareous beds several metres above strata equivalent to the Purple Shales Formation (Cave and Dixon, 1993, p.71). Beds of bentonite, present in all these sections, have been sampled and geochemical analyses may provide definitive correlations between graptolitic and non-graptolitic sections (Loydell and Cave, 1993, pp.93–94).

Biostratigraphy and palaeogeography

To the east and north of the district, the Purple (or Hughley) Shales have yielded a diverse fauna of brachiopods (Whittard and Barker, 1950), gastropods (Pitcher, 1939), trilobites (Whittard, 1938), ostracods (Harper, 1940) and other forms.

The brachiopods Hindella? cf. furcata, H.sp. and ostracod Beyrichia kloedeni were found by Whittard (1932) from a poor, roadside exposure, south-west of the chapel in Norbury, but otherwise nothing distinctive of age has been recorded at outcrop. Evidence from boreholes (Cocks and Rickards, 1969) and from surrounding districts indicates that the formation is of Telychian age, C3 to C6 divisions inclusive on the shelly fauna scale (Jones, 1925; Williams, 1951, p.128), and topmost part of the turriculatus Biozone to just below the top of the lapworthi Biozone in graptolitic terms.

The boreholes, Church Stretton No. 1 and No. 2, yield shelly fossils indicative of a Clorinda Community, overlain by a Marginal Clorinda Community which Ziegler et al. (1968) suggest implies continual deepening of marine conditions during deposition. The dearth of shell detritus in the top part of the formation points to the same conclusion. Graptolites from both boreholes indicate the presence of a high M. turriculatus Biozone or the M. crispus Biozone at the base to the O. spiralis Biozone at 28.8 m (depth 352 feet 4 inches) below the top of the formation (information from Dr D Loydell, 1990). The shelly fossils have received preliminary identifications and the distribution of selected genera is shown in Cocks and Rickards (1969).

A stratigraphical range similar to that of the Purple Shales is indicated by the purple and green mudstones of the Tarannon Shales, some 18 km to the north-west at Buttington; high turriculatus or crispus biozones are recorded at the base and the spiralis Biozone is recorded about 5.3 m below the top. Nearby, at Meifod, a comparable sequence with more extensive evidence places the top of the formation in the top part of the lapworthi Biozone (Loydell and Cave, 1996). Uniform marine conditions thus prevailed across this part of the Midland Platform at this time, and where Purple Shales rest directly on pre-Silurian rocks, land must have persisted until late turriculatus/crispus time at the earliest. Clearly it was a much diminished terrain, the only coarse sediment coming from remanié detritus of the Longmynd and coastal shell debris suggesting that the climate had become drier and calm. On the other hand, Benton and Gray (1981) indicate that the eqivalent beds, with more abundant and thicker sandstones, 20 to 30 km to the north-east (Hughley Shales)are mid-shelf turbidites, the product of storms acting on a shoreface even farther eastward.

Two faults seemed to have influenced the late Llandovery palaeogeography. A narrow area near Craven Arms, immediately east of the Church Stretton Fault, was not drowned until Wenlock times, while the area immediately west of the fault received marine sediment continuously from mid-Llandovery times (Aeronian, convolutus Biozone) (Cocks and Rickards, 1969). Likewise, the sequence to the east of the Longmynd Scarp Fault thins over its footwall and, although on its west side Aeronian strata are absent (Figure 17), the Purple Shales (Telychian) are thick. Thus it appears that differential movement occurred on both of these faults in late Aeronian and Telychian times, being more pronounced on the Church Stretton Fault.

Llandovery red (and green) beds are a facies of intercontinental distribution across parts of northern Europe and North America (Ziegler and McKerrow, 1975). In basinal areas of Scotland (White et al., 1992), the Lake District (Rickards, 1978) and Wales (Wood, 1906) they occupy the topmost Llandovery, traditionally considered to be the crenulata Biozone. This is true of the type Tarannon Shales section in the Afon Trannon, just west of the Montgomery district, where it was influenced by turbidites of the Welsh Basin. Thus the evidence from within the Montgomery district, and nearby to the north, reveals that the base of the Tarannon Shales/Purple Shales rises sharply westward across the line of the eastern margin of the Welsh Basin. It is suggested that a marked change either in the layering of the basinal water, or perhaps in its depth, occurred at the end of griestoniensis times which extended the chemistry already established in shallower easterly areas to areas farther west. Uranium–lead isotope ratios derived from a K-bentonite from near the top of the equivalent Tarannon Shales Formation at Welshpool, nearby, have revealed an age of 430.1 ± 2.4 Ma (Tucker and McKerrow, 1995, p.376).

Details

Lower Snead Farm

Continuous exposure in the stream [SO 3045 9275] to [SO 3067 9300] which crosses the main road reveals soft, shaly siltstone, mainly purple, with green bands and spots. Fine-grained, hard siltstone, in beds up to 0.10 m thick, occurs sparsely, together with scattered ovoid concretions. No fossils were found. Dips are low and strikes various.

Chapter 6 Silurian: Wenlock, Ludlow and Přídolí

Although the Welsh Basin was more truly a single entity during the epochs Ashgill to Přídólí inclusive than it had been before, strong differences mark the Ashgill Series from the Llandovery Series and the Llandovery from the Wenlock. There is no such demarcation between the Wenlock and the Ludlow. For this reason and in recognition of Shropshire (Salopia) as classical ground for Wenlock and Ludlow rocks, Lapworth (1880, p.48) introduced the term Salopian to embrace the two series. Unfortunately he associated it with the Downtonian too, but the term has been widely used in the geological literature of the region ever since.

The Wenlock–Ludlow sequence is the product of basin margin deposition. Implicit in such a model is a submarine slope and therefore formations which are either diachronous, or pass laterally, one to another. A refined biostratigraphical framework, provided by graptolites, reveals examples of both, but the narrowness of the diachroneity and the diffuse nature of the lateral passage are evidence that the slope was gentle in this district.

Early stratigraphical research into the Wenlock and Ludlow rocks of the district was undertaken by Earp (1938; 1940), Evans (1957) and Allender (1958). More recently, there have been thematic studies of particular formations in the Wenlock (Dimberline, 1987; Dimberline and Woodcock, 1987) and in the Ludlow (Tyler, 1987; Tyler and Woodcock, 1987). Cummins’ studies (1957, 1959a, b), which defined the fundamental lithofacies of these rocks in the basinal area immediately to the west, were the first of this kind, and Warren et al. (1984) used similar lithological definitions for the Denbigh district. Their approach is adopted here, on the lines of the pilot account by Cave et al. (1993). The biozonation used is that adopted by the Geological Society of London (Cocks et al., 1992) (Table 6) and the lithostratigraphy is summarised also in (Figure 18).

By Wenlock times the transgression had effectively flooded the Midland Platform (Figure 19) while in the west the Welsh Basin persisted as the Montgomery Trough (Holland, 1992, pp.40 and 43). After the multifarious facies of the Aeronian and early Telychian stages, Wenlock basinal stratigraphy is strikingly uniform, from Builth in the south to Corwen in the north. Formations are the Nant-ysgollon Shales, the Denbigh Grits (with their local representative the Penstrowed Grits) and Nantglyn Flags, and all crop out within the district. A thick shelf sequence accumulated in the east of the district and beyond, over the Midland Platform. Lateral passage from the basinal sequence into the shelf equivalent is illustrated in (Figure 18).

Ludlow ‘basinal facies’

It is traditional to view the Ludlow Series of eastern Wales and the Welsh Borderland (like the Wenlock) as having a ‘shelf facies’ and a ‘basin facies’, yet without defining what is meant by ‘basin’ in a Ludlow context for clearly it differed from the basin of Wenlock times in position and the nature of infill. The view, well expressed by Lawson (1956) and Lawson and Straw (1956), is based upon the clear and fundamental differences between the rocks of two tracts. On the east, in the Welsh Borderland, the rocks are predominantly calcareous mudstones, with limestones in places, and distinctive shelly faunas, all indicative of shallow (‘shelf’) seas. On the west, embracing the Montgomery district, the rocks are more sandy or silty, less calcareous and less fossiliferous with thicknesses five times greater (Holland and Lawson, 1963). The formations, Oakeley Mynd, Bailey Hill, Knucklas Castle, Cefn Einion (and even Clun Forest), were all viewed as ‘basinal’.

(Figure 20)b illustrates the nature of ‘the basin’ at the level of the tumescens Biozone. At its south-western end there was a subaerial to shallow marine delta (m1 of (Figure 20)b). Current indicators have revealed north-eastward flow and in this direction the deposits were progressively finer and more distal in aspect (n2 and o14). Concomittantly, the waters were deeper, so that from just south of Builth and all the way to the North Wales coast, bottomconditions were anoxic ((Figure 20)b, o14, o15, and m16). The trend to distality, however, was interrupted in the area of Radnor Forest where event couplets, with obtrusive silt–fine-sand bases, occur (t7, s6 — the Bailey Hill Formation). They are incongruous with the general, north-eastward directed, sedimentary ‘gradient’ of the regime described above and must represent an invasion, from the stable platform to the east, of the detritus of another system, the flow of which turned northwards also, thus carrying the Bailey Hill Formation across the Montgomery district.

The divide between the variable ‘basin facies’ and the uniform ‘shelf-facies’ is sharp and coincides approximately with the eastern side of the Church Stretton Fault–Brecon Anticline (Figure 20) (Straw, in Kirk, 1951b, p.73) and it seems inescapable that the fault was the cause of the differences. The ‘basin’ appears to have been a strip of the Wenlock Midland Platform which foundered on the western side of the fault during Ludlow times, like a trap-door hinged at the south-western end. Much detritus was fed into the accommodation so produced from the land-mass at this south-western end, but other detritus seems to have come from a point near Presteigne, or Knill, on its south-eastern margin, during Gorstian times — a pattern remarkably similar to that in the Sheinwoodian (Figure 19).

Although ‘trap-door’ subsidence continued during Ludfordian times (Holland and Lawson, 1963, fig. 14), by Prídólí times it was waning and sedimentation gradually eliminated the difference in water depth between the tract west of the Church Stretton Fault and the Midland Platform.

Within the district, the Prídólí Series is represented by the Clun Forest Formation. The lowest part of the formation (Platyschisma Beds or first facies association) shows a gradual change from fully marine conditions in the underlying Cefn Einion Formation to brackish water conditions. This may be compared with the approximately equivalent deposits, which include the Downton Castle Sandstone, of Ludford Corner, Ludlow, described and interpreted by Smith and Ainsworth (1989) as storm-influenced shoreface deposits. The overlying beds, containing decalcified calcretes, mark the gradual onset of true terrestrial conditions. These sediments were laid down on a broad, distal fluvial plain or coastal flat, and were subjected to pedogenic processes which were enhanced by a hot climate and low seasonal rainfall. Most of the formation (third and fourth facies associations) was deposited on a broad alluvial plain, with sediment supply from the north-west.

Lithology and facies

The two fundamental processes of sedimentation operating in the marine waters of the district in Wenlock and Ludlow times were:

1. Continuous vertical rain of particulate matter, namely the remains of pelagic organisms together with fine inorganic detritus placed in suspension from the atmosphere, from rivers, and by storm agitation of coastal deposits. These produce the background, primary deposits known as hemipelagite.

2. Detritus-laden bottom-currents, provoked by specific events such as submarine slope-failures, or atmospheric storms acting on the basin margin or upon shoreface areas. Thus detritus was recycled and re-deposited either as sand–mud turbidites or as storm-sheet sand–mud couplets, temporarily interrupting the background sedimentation.

Two other factors which affected deposits were oxygenated bottom conditions and shallow water conditions. Both promoted the disturbance or destruction of primary bedding fabrics, by benthic animals (bioturbation) and by wave agitation respectively, the latter imposing a new fabric distinctive of itself.

Hemipelagite

The term was used first by Natland (1876) for sediments formed by slow accumulation on the sea floor of biogenic and fine terrigenous silt particles. Its lithified appearance depends upon the conditions under which it was deposited, and here only marine sediments are considered.

Oxygen-depleted conditions

Hemipelagite characteristic of oxygen-depleted tranquil bottom conditions was described by Cummins (1959) and discussed widely since. A pervasive bedding-parallel lamination comprises laminae of dark grey or brown carbonaceous matter, separating millimetre-thick pale grey silt laminae.

In thin section, the silt layers are composed of discrete ovoid segregatious, flattened in the plane of bedding (Cummins, 1959). The silt layers are not continuous and, to a degree, the carbonaceous layers are anastomosing, a quality most evident where silt is dominant and the carbonaceous content reduced to mere films. The reason for the silt having segregated into subspherical bodies is unclear. They have been explained as faecal pellets (Llewellyn, 1965), but a physical process seems more likely to have been responsible, such as post-depositional disruption of a silt lamina on a soft gelatinous carbonaceous mat, or even pre-depositional flocculation. Faint and narrow curvilinear grazing trails are visible in this facies in several parts of the sequence. The relative abundance of silt and carbonaceous remains is highly variable. It reflects two influences, the availability of nutrients in the water column and the proximity to a source of silt.

This facies is a constituent of all formations up to the basal few metres of the Knucklas Castle Formation. It is the dominant lithology of the deeper, basinal facies, for example the Nant-ysgollon Shales, the Gyfenni Wood Shales Member of the Nantglyn Flags Formation and the Cwm-yr-hob Member of the Bailey Hill Formation.

Oxygenated conditions

In oxygenated depositional environments, the development of this lamination is inhibited by a burrowing and scavenging benthos which consumes the organic remains and constantly reworks the sediment. Under such conditions, hemipelagite may become indistinguishable from other fine-grained deposits or even be mixed with them, for instance in the burrowed top part of a mud turbidite or ‘storm-sheet’. The product then is commonly a medium grey or grey-green mottled or homogenised mudstone.

Above the Bailey Hill Formation, the depositional environment was oxygenated, and a siltstone laminite facies invaded the district. These siltstone laminites are unique to the Knucklas Castle Formation and are described under that heading. It is suggested that the facies is largely hemipelagite, like the underlying Cwm-yr-hob Member, but with a very low carbonaceous content and pervasive bioturbation, these differences resulting merely from the availability of oxygen.

Bentonite

Beds of bentonite, up to 0.10 m thick, are common within the sequence; they are volcanogenic in origin and a very particular form of hemipelagite. They consist of soft, pale cream-coloured clay and are the product of submarine decomposition of vitric volcanic dust. Illite–smectite mixed-layer clays are the dominant constituents, illite being the more dominant in higher metamorphic grades. Bentonites are the terrigenous extreme of a range of water-column detrital precipitates of which biogenic pelagite is the other extreme.

Nine Wenlock and Ludlow Salopian K-bentonites were sampled within and close to the district. Major and trace-element analyses are given in (Table 7) and Nb/Y has been plotted against Zr/TiO2 on a Winchester and Floyd (1977) diagram (Figure 21). It reveals that several early-Ludlow K-bentonites fall within the rhyodacite/dacite field, one early Wenlock sample falls within the trachyandesite field, and a late Wenlock to early Ludlow sample lies just within the trachyandesite field. Their plate tectonic setting was reviewed by Fortey, Merriman and Huff (1994).

Event couplets of argillaceous siltstone or silty mudstone on fine-grained sandstone

This facies is the product of detritus-laden bottom currents and has been described by Lynas (1987) and Tyler and Woodcock (1987). The couplet has a lower division of what is loosely referred to as fine-grained sandstone but commonly is siltstone and an upper division of argillaceous siltstone or silty mudstone. Sandstones/siltstones range in thickness up to 0.50 m. They are composed of fine-grained, subangular, silica sand, with grain-contact, and a calcitic argillaceous matrix. The sandstones/siltstones are pale to medium grey, but on weathering they become porous and fawn to brown. Their bases are sharp, commonly erosive, with flute casts, prod-marks and groove casts. Shell-detrital lags may be present at the bases of some sandstones; current-orientated graptolites may be present. Commonly the tops are also sharp; they may be wavy, mostly symmetrically, with amplitudes up to 0.05 m and wave-lengths of 0.25 to 0.50 m. Large current ripples are also present. The bedding fabric is most commonly a subparallel wavy lamination; low-angle cross-bedding is also common, but steeper, current-ripple foresets occur, particularly in the ripple-form top parts. Convolute bedding is fairly common, mainly in the middle of the bed and may grade laterally from low-angle lamination. Convolutions may be truncated by wavy subparallel laminae concordant with the waved surface. Within the sandstones, there may be many flat, or wavy, truncation surfaces revealing erosion and current reactivation during deposition.

The fabric of the sandstones is very variable, both laterally and vertically. In many sandstone beds, especially those of finer grade, layers of coarser sand alternate with fine sand or silt. The coarser layers are laterally very impersistent, being replaced en echelon by others. Commonly, they converge with the base of the bed against which they pinch out. However, overall there is an upwards-fining into the argillaceous siltstone of the upper part of the couplet.

The upper division of the couplet is generally homogeneous, grey and argillaceous, although a faint lamination may be present. Thin wisps or layers of finely laminated siltstone may occur also. It is an incohesive rock, quickly weathering to rubble. Thicknesses range from tens of centimetres to mere partings between the sandstones; actual amalgamation of the sandstone beds has not been recorded.

The couplets are not so orderly as those of the rhythmite facies described below and exemplified in the Bouma sequences of the older turbidites of the Welsh Basin; curved lamination, subparallel with bedding, and internal surfaces of discontinuity are much more abundant. Likewise, the argillaceous siltstones display lateral thickness variation and not the evenness of the Bouma Te interval as seen in thinly bedded basinal turbidites. Nevertheless, the couplets indicate that each was the deposit of a waning current even if the exact nature of the flow regime which delivered these beds is open to debate.

Within the district, this facies presents two aspects. One, deposited in an oxygen-starved environment so that the original bed-fabric is not bioturbated and couplets are commonly separated by thin beds of laminated hemipelagite. It is typified by the Bailey Hill Formation, a formation of couplets which were considered by Tyler (1987) to have been delivered by storm-generated marine currents. Others (Cummins, 1959b; Bailey, 1969) viewed them as turbidites, and, in the palaeogeographical setting discussed below, this seems the more likely possibility. In its second aspect this facies is typified by the Cefn Einion Formation. Deposited under oxic conditions, the bed fabric of the upper parts of the couplet have been modified, or destroyed, by a burrowing and scavenging benthos.

Rhythmite of fine-grained sandstone–mudstone (‘ribbon-banded mudstones’)

This facies comprises well-ordered, laterally constant turbiditic couplets, between 20 and 100 mm thick, of fine-grained sandstone or siltstone (referred to here as fine sandstone), and mudstone, first described by Cummins (1959b). Warren et al. (1984) used the term ‘ribbon-banded mudstones’ in which fine sandstone, or siltstone, is generally subordinate to a superior partner of homogeneous mudstone. It conforms to Facies C2.3 of Pickering et al. (1986). Within this district and the adjacent areas, such rhythmites were deposited in a dysaerobic environment so that ideally the couplet is always separated from the superincumbent couplet by a thin laminar hemipelagite. The fine sandstones are pale grey, usually no more than a centimetre or two thick, and have a bedding fabric of parallel laminae and current-ripple cross-laminae. Some contain a coarse, shell-detrital lag at the base, a form of grading, so that they conform closely to the Bouma turbidite sequence Tacd or Tcd. The homogeneous mudstone in the upper part of the couplets thus records the Teintervals of the Bouma sequence so that there is little doubt about these couplets being low density turbidites of a basinal environment. Graptolites are common in the turbiditic fine sandstone, where they are current orientated and usually mono-specific. They also occur in the homogeneous mudstone, where they are normally well preserved. Bottom structures are not common, though flute casts do occur.

The Nantglyn Flags Formation is characterised by this facies, and ribbon-banded mudstones is an appropriate term for the particularly evenly bedded strata lying in the upper part of the formation between the Mottled Mudstone Member and the Gyfenni Wood Shales Member. The equivalent facies of oxygenated bottom conditions is typified in such formations as the Devil’s Bridge in mid-Wales (Cave and Hains, 1986).

Bioturbated calcareous mudstone

Bioturbated calcareous mudstone occurs as massive beds and may be silty. It is pale to medium grey in colour, but it decalcifies on weathering and assumes buff and olive-green colours. Burrows are pervasive; they are orientated subparallel with and at right-angles to the plane of bedding. Layers of fine sandstone or siltstone may be preserved within the mudstone, together with the remains of benthic and pelagic faunas including brachiopods of a comparatively deep water community, trilobites and graptolites, including the retiolilid genus Gothograptus. Mudstone of this type constitutes the Aston Mudstone Formation and the Mottled Mudstone Member of the Nantglyn Flags (Boswell and Double, 1940, p.158; Warren et al., 1984, p.46).

Slumped and disturbed beds

Disturbed beds occur sporadically throughout the Wenlock–Ludlow succession as thin discrete beds, thick mappable units and, in the Clun Forest area, as stratiform alternations of slumped and non-disturbed beds making up a large thickness of the Bailey Hill Formation (Earp, 1938; Woodcock, 1973). The disturbance results from soft-sediment instability, though the cause of the instability is less clear. Seismic shock and the build-up of sedimentary slope to super-critical angles are two likely causes. In the main, slumping is considered to have involved translatory downslope movement perhaps becoming debris-flows distally, but some disturbed beds might have resulted merely from readjustment of reversed density gradients through shock-induced thixotropy, more or less in situ (p.95).

Disturbed beds in the Denbigh and Rhyl districts have been categorised and described by Warren et al. (1984), Eva (1992) and Eva and Maltman (1994); slumps in this district were described and their form and distribution analysed by Cummins (1959, pp.177–178; 1969, p.221), Bailey (1964; 1969) and Woodcock (1976 a, b).

Conglomerate

Conglomerate is a sparse but important constituent of the sequence, occurring in association with several other facies. It forms lenticular beds scattered within the Nantglyn Flags, including the Mottled Mudstone Member and, more commonly, at the base of the Bailey Hill Formation, though no higher. The clasts are variously round to angular, derived from most contemporaneous local rock types including ovoid calcareous concretions. Only the small round clasts of black phosphatic mudstone and shell debris cannot be matched with a rock in the local sequence. The clasts are closely packed, some appearing to be matrix-supported, others in contact, perhaps as a result of elutriation of the matrix. All are anomalously large to have been supported by the currents which introduced the sediments of the host formations. Thus, for aqueous transportation to have been responsible, currents would have to have been exceptionally strong and infrequent, able to introduce shell debris, possibly rip-up hard ground substrates and winnow out calcareous concretions before carrying them onwards. A less likely alternative is small, but widespread slope failures which produced clast-rich debris-flows. The matrices are normally of mudstone, including clots of bentonite, and in some cases they are rich in brachiopod shell debris.

Palaeontology

Both graptolitic and shelly macrofaunas are well represented in the Wenlock–Přídólí sequence (Plate 2), (Plate 3). However, graptolites have not been found above the basal beds of the Knucklas Castle Formation high in the Ludlow Series. This corresponds with the upper stratigraphical limit of graptolite occurrences recorded elsewhere in the Welsh Borderland. Shelly faunas become progressively more prominent, though less diverse, higher in the succession, especially within the Ludlow Series. Brachiopods are the most abundant shells below the base of the Přídólí Series, where there is a sharp transition to mollusc-dominated assemblages.

Graptolites

Graptolites occur most commonly in the dark grey hemipelagites (p.57) and are associated with subordinate numbers of orthoceratids; together these represent the principal components of the local indigenous fauna (Plate 2). Current-orientated graptolites are also common in the turbiditic fine sandstones, and are well preserved in the rhythmites of the fine sandstone–mudstone.

Exceptionally, as in the more calcareous beds at the top of the Wenlock Series, graptolites are preserved in relief. However, in general, compaction has resulted in their preservation as flattened films on bedding planes. Where these films are partially ventral or dorsal, the thecal profile is affected, and this can cause identification difficulties. This problem is especially acute within the stratigraphically important saetograptids, the classification of which is based mainly on the number of spinose thecae and the position of the spine bases relative to the thecal aperture. These criteria may be difficult to assess in the case of flattened specimens.

A detailed graptolite biozonal classification of the Wenlock and Ludlow Series in the Montgomery district is inhibited by difficulties of identification, the urgent need for a modern description of the Ludlow species and the absence of long continuous sections. Lack of exposures is particularly acute in the lower and middle parts of the Wenlock Series, and the beds are generally poorly fossiliferous. In younger beds, the poor exposure has resulted in uncertainty in the stratigraphical relationships of isolated collections. Furthermore the collections lack diversity, many consisting of fewer than five taxa.

Nevertheless, graptolites provide the basis for a biostratigraphical classification which can be related to the biozonal schemes of the type areas for the Wenlock and Ludlow Series. Graptolite occurrences can be grouped into six assemblages in each of the Wenlock and Ludlow Series. Their relationships to the standard sequence of graptolite biozones is shown in (Table 8) and (Table 9).

In the Wenlock Series, there is little evidence available for the presence of the Sheinwoodian biozones, that is below the Cyrtograptus lundgreni Biozone (Table 8). However, there is no reason to suppose that there are significant breaks in deposition; poor exposure and the sparsely fossiliferous nature of the beds are the most likely explanations for the lack of information. Evidence for the presence of pre-C. linnarssoni Biozone strata is limited to a single restricted assemblage. It is based principally upon the occurrence of Monograptus antennularius which, however, has an upper stratigraphical limit extending into the lower part of the C. linnarssoni Biozone. Similarly, single records of M. flexilis cf. flexilis and of Cyrtograptus ellesae? with Pristiograptus cf. meneghini are the only indications of the C. linnarssoni and C. ellesae biozones.

The Homerian Stage biozones are all well established in the district. The C. lundgreni Biozone is proved by records of C. hamatus and C. lundgreni. Assemblages dominated by Gothograptus nassa with rare pristiograptids, succeeded by graptolite faunas in which Monograptus ludensis and Pristiograptus jaegeri are abundant, represent the G. nassa and M. ludensis biozones at the top of the Wenlock Series.

The difficulty of relating the Ludlow graptolite assemblages of the district to the standard biozonal scheme, particularly at the level of the Lobograptus scanicus and Pristiograptus tumescens (or Saetograptus incipiens) biozones, is accentuated by the comparative rarity of the eponymous species and their long stratigraphical ranges (Table 9).

However, the abrupt appearance of several taxa, including the saetograptids, which are generally accepted as marking the base of Ludlow Series, is well seen throughout the district. This assemblage, which includes Monograptus uncinatus orbatus, Neodiversograptus nilssoni, Spinograptus spinosus and the saetograptids Saetograptus (Colonograptus) colonus and S. (C.) varians, is succeeded by one in which Lobograptus progenitor (formerly N. nilssoni s.l. pars) and L. scanicus displace M. uncinatus orbatus, N. nilssoni and S. spinosus. These two assemblages are considered to be lower and upper subdivisions of the N. nilssoni Biozone. A younger assemblage, consisting essentially of Saetograptus (S.) chimaera chimaera, S. (S.) chimaera seimspinosus. S. (Colonograptus) colonus s.l., S. (C.). varians and L. scanicus, is succeeded by a more restricted one, lacking colonograptids and lobograptids, but with S. (S.) aff. incipiens in the highest beds. These two assemblages are respectively assigned to the L. scanicus and P. tumescens biozones. Decreasing faunal diversity continues in the overlying strata, where S. (S.) leintwardinensis leintwardinensis with S. (S.) aff. incipiens of the S. (S.). leintwardinensis Biozone is succeeded by Bohemograptus bohemicus tenuis (at two newly recorded localities, p.104) indicating the presence of the Bohemograptus proliferation Biozone, the youngest graptolite biozone proved in Britain.

Monospecific collections of S. (C.) varians have been found at some localities, as well as a single monospecific collection of S. (S.) clunensis; the latter being the only record of this species in the district. The age of these fossils is uncertain; the S. (C.) varians occurrences are probably within the upper subdivision of the N. nilssoni Biozone and/or L. scanicus Biozone, and the collection with S. (S.) clunensis is likely to be from the L. scanicus or P. tumescens biozones.

Shells

Pelagic shelly macrofossils are represented almost exclusively by orthoceratids. They usually occur in hemipelagites and are closely associated with graptolites. They are generally crushed and fragmented (Plate 3).

Benthic shelly macrofossils are commonly found as detrital lag deposits, preserved at the base of fine turbiditic sandstones in the rhythmites of fine sandstone-mudstone; they are considered to be allochthonous. Shelly detrital lags are also common at the base of the event couplets of calcareous, fine sandstone and argillaceous siltstone. These lags are generally considered to have been deposited from storm-generated or turbiditic currents; the shells are mostly fragmented and have certainly undergone some transport, but it is debatable whether they should be interpreted as indigenous or allochthonous faunas.

Indigenous, benthic, shelly macrofossils are preserved in bioturbated calcareous mudstones near the top of the Wenlock Series and in fine sandstone and siltstone of the upper Ludlow and lower Přídólí Series. Of these life assemblages, the Wenlock assemblage consists mainly of well-preserved brachiopods with subordinate numbers of molluscs and arthropods. The brachiopods are predominantly species of small dimensions, including Visbyella trewerna, and the assemblage represents the Visbyella Community defined by Calef and Hancock (1974). The later assemblages (top Ludlow and base of the Přídolí) are also mainly of brachiopods, representing the Salopina Community (Calef and Hancock, 1974) and the Protochonetes ludloviensis Association (Watkins, 1979). It is succeeded by a molluscan-dominated assemblage in the basal beds of the Přídólí Series.

Of the shelly macrofossils, most brachiopod species have a comparatively long stratigraphical range and consequently are less useful in biostratigraphical classification than the graptolites. Other fossil groups such as the molluscs and trilobites are not sufficiently well represented, although ostracods, particularly the beyrichiids, may eventually prove to have considerable stratigraphical value. In the present account, they are largely unclassified. A study of their potential in the district awaits the publication of a modern taxonomic study by Siveter; only the first part of his monograph (Siveter, 1980) has been published so far. However, the shelly macrofossils have contributed to the correlation of the local succession with those of neighbouring districts and are useful in palaeoecological and palaeogeographical reconstructions.

Faunal correlation

The graptolitic biostratigraphical classification of the Wenlock and Ludlow successions in the district has provided a precise stratigraphical control for local correlation, and thus for the interpretation of the complex lateral facies variations across the district (Table 6). Furthermore, it enables a correlation to be made with the Wenlock and Ludlow type areas, and other parts of Britain.

The indigenous shelly macrofossil assemblage in the bioturbated calcareous mudstones of the Mottled Mudstone Member (of the Nantglyn Flags) and Aston Mudstone Formation includes several forms which were first described from the Wenlock of Bohemia by Barrande (for example 1879) and the top of the Cyrtograptus Shale of the Wenlock of Smedstorp, in Scania, Sweden (Hede, 1915); fossils include the brachiopods Bracteoleptaena bracteola, ‘Clorinda’ dormitzeri and Mezounia bicuspis. A closely comparable fauna is recorded from the Mottled Mudstones of the Denbigh district (Rushton in Warren et al., 1984, p.57). Correlation of the Mottled Mudstones of the two districts by means of graptolites is thus supported by the evidence of the shelly assemblages.

The shelly faunas above the Wenlock–Ludlow boundary are virtually identical with those below it, but are much less abundant. However, the brachiopod Dayia navicula occurs low in the Ludlow Series and is a useful stratigraphical indicator where graptolites are not recorded.

In the Bailey Hill Formation, shell assemblages in which large strophomenids are well represented are broadly comparable with those of the late Gorstian (Lower and Upper Bringewood formations) of the type Ludlow area. Rare derived examples of the stratigraphically restricted brachiopod Kirkidium knightii are recorded from the Bailey Hill Formation, also indicating correlation with the late Gorstian rocks of the Aymestrey area, where the Aymestry Limestone contains high energy Kirkidium-bearing limestones of the shelf edge area (Watkins, 1979).

Overlying the youngest graptolite record of the Bohemograptus proliferation Biozone at the base of the Knucklas Castle Formation, the indigenous shelly benthic assemblages in the Cefn Einion Formation and in the base of the Clun Forest Formation compare closely with the faunal sequence across the Ludlow–Prídólí Series boundary of the stratotype section for the Downton Group, near Ludlow (White and Lawson, 1989, p.139, fig. 103) and a firm correlation can be made. Concentrations of Craniops implicatus are present in transitional strata between the Cefn Einion and Clun Forest formations and also in the basal beds of the Clun Forest Formation. A similar concentration occurs in the highest beds of the Upper Whitcliffe Formation (latest Ludlow Series) at the section near Ludlow, mentioned above. Holland and Williams (1985, p.35) report another occurrence near Kington, Herefordshire, at the same high level in the Ludlow Series and consider these concentrations to have regional significance.

Base of the Wenlock Series

International definition has established the base of the Wenlock Series in a shelly facies at the base of the Buildwas Formation (Bassett, 1989) in Hughley Brook, near Much Wenlock, 10 miles east of the district. It was said to coincide with the base of the graptolite biozone of C. centrifugus (Table 6), but without proof this remains only a possibility (Bassett, 1989, p.61; Loydell, 1993, p.323).

The base of the Buildwas Formation lies in a short (1 m) passage from red and green mudstone of the underlying Hughley Shales (= Purple Shales Formation) into buff, olive-green and grey calcareous mudstone with shell fragments. The overlying olive to dark blue-grey mudstone of the Coalbrookdale Formation lack the abundant shell fragments; instead they are graptolitic, with a fauna significant of the riccartonensis Biozone occurring 3 m above the base. Given that thicknesses differ slightly, this situation is very similar to that at Buttington, in Long Mountain, (Cave and Dixon, 1993) north of the district, and at Plowden (Evans, 1957) on the eastern edge, making it reasonable to conclude that it applies over the whole area. At Buttington, the C. murchisoni Biozone has been proved in the basal centimetres of the Trewern Brook Mudstones Formation (Palmer, 1972, p.27) and it remains possible that it occurs in the basal 3 m of the Coalbrookdale Formation too (Bassett, 1989, p.61). Indeed, Das Gupta found evidence of it at Plowden (Whittard, 1932, p.869).

Within the district, Wenlock strata overlie Llandovery rocks with conformity. Along the western margin of the district, the Llandovery Tarannon Shales, or Dolgau Mudstones pass up into the Nant-ysgollon Shales Formation through a thin passage of burrowed dark and pale mudstone. Nearby, considered to be part of the Nant-ysgollon Shales Formation to the north-west of the district, this transition contains part of the lapworthi (or grandis) Biozone and the insectus Biozone (Loydell and Cave, 1996), and the same is probably true within the district, west of the Severn Valley. To the north of the district, but east of the Severn Valley, there is a slightly different transition up into the Wenlock Series. Soft, grey-green and maroon mudstones of the equivalent Tarannon Shales pass up into 9 m of grey silty mudstone which is massive, calcareous and burrowed, and has a shelly fauna (Cave and Dixon, 1993; Loydell and Cave, 1993) followed by anoxic deposits of the murchisoni Biozone (Palmer, 1970; Cocks and Rickards, 1969, pp.226–227). The base of the Wenlock Series therefore lies within these burrowed transition members (Table 6) just as it does in the Hafod yr ancr Member of the Builth Mudstone Formation (Harris, 1987). In the south-east of the district the base of the series is less clearly defined. In Church Stretton No.1 Borehole Cocks and Rickards (1969, p.215) could find no evidence of biozones lower than the rigidus Biozone, but there are 18.3 m of mudstones which could accommodate the lower biozones.

Nant-Ysgollon Shales Formation

Type area and section of the Nant-ysgollon Shales Formation (also referred to as the Nant-ysgollon Mudstones) are in the Trannon area just to the west of the district. The name Nant-ysgollon Shales derives by mistaken spelling from Nant-ysgolion [Welsh ysgolion ladders] a minor stream [SN 9080 9836] where the beds are well exposed, as they are also in the Trannon river (Wood, 1906). The formation can be traced from near Llandrindod Wells northwards to Corwen and beyond. This main outcrop skirts the western side of the district between Llandinam, Caersws and Bwlch-y-fridd. Only small inliers occur within the district near Llandyssil, about 2 km south-east of Montgomery and in the vicinity of Sarn.

The base of the formation is not exposed within the district but near Meifod [SJ 153 132] and Llandinam [SO 025 885], immediately north-west and west respectively, it is placed at the base of a sequence of dark grey partially bioturbated mudstones which are transitional between the pale, oxic Tarannon Shales and the anoxic laminated hemipelagites above (Loydell and Cave, 1993). In the ‘basinal’ west of the district, the formation passes up into the Penstrowed Grits (Denbigh Grits Formation); eastwards as the grits feather out against the basin margin (Figure 22), the top of the Nant-ysgollon Shales Formation becomes progressively younger so that ultimately the formation replaces the lower part of the Nantglyn Flags (Table 6). Estimates of the thickness of the formation are approximate, due partly to this diachroneity and partly to the lack of good exposure. Just west of the district, near Aberhafesp [SO 070 924], the formation is about 60 m thick; farther south near Llandinam [SO 025 885] it is about 200 m. In the north, at Fron [SO 178 977], some 100 m are present, but the formation there represents a much larger proportion of the Wenlock sequence (Figure 18).

The formation consists predominantly of dark grey, anoxic, laminated hemipelagite. The hemipelagite is slightly calcareous, in the silt fraction, and where weathered the rock is brown, soft and fissile. In consequence, the outcrop commonly occupies the lower slopes of scarp faces and is poorly exposed. Interbedded mudstone occurs as very numerous thin beds, mostly less than a centimetre thick, and is slightly paler grey and homogenous; these beds are probably low-density turbidites. They are commonly 1 to 3 mm thick in the lowest three Wenlock biozones (centrifugus to riccartonensis), but much thicker in the higher development in the east. Sandstones are present in this latter development. They are are thin, fine grained, and usually parallel laminated and occur near the top of the formation in the lundgreni Biozone.

Slumping and sliding has affected the formation in places, particularly near to the lateral feather-edge of Penstrowed Grits (Denbigh Grits Formation), for instance near Llanbadarn Fynydd [SO 098 780] and Aberhafesp [SO 070 925] and even as far north as Glyn Ceiriog.

Large ovoid concretions of limestone are scattered within the formation. In the River Ithon [ 0868 7543] the concretions attain lengths of over 1 m along the plane of bedding, and, within some, bedding is crumpled and folded, indicating that the sediments had slumped prior to the formation of the concretion.

Biostratigraphy

The macrofauna is dominantly graptolitic and some orthocones are present. Although not exposed here it is known from surrounding districts (Harris, 1987; Loydell and Cave, 1993; 1996) that the base of the formation lies in the Llandovery, within the lapworthi Biozone, while at the western edge of the district, near Aberhafesp, where overlain by the Penstrowed Grits (Denbigh Grits Formation), the top part of the formation yields Monograptus cf. riccartonensis. In the east, however, at Fron and Sarn, the top of the formation lies in the lundgreni Biozone.

Details

West of the district where the formation underlies the Penstrowed Grits, it is well exposed in a roadside bank [SO 0665 9429] where the contact with the Tarannon Shales is slumped. About 450 m to the south-west, in Nant Rhyd-ros-lan and a quarry [SO 0634 9399], it is also well exposed and the beds yield M. riccartonensis?.

A lithological log of a well exposed section along a farm track on Fron bank [SO 1795 9735] is shown in (Figure 23).

Lower parts of the formation, in which sandstones are sparse and laminated hemipelagite dominates, are exposed in an excavated bank at Dolforwyn Hall Hotel [SO 1612 9563], along a forestry track [SO 1652 9639] to [SO 1671 9680], beside the main road [SO 1682 9667] and in the banks of Llifior Brook [SO 1845 9866] to [SO 1853 9866]. An excavated bank behind a house [SO 1797 9784] exposes shales in the basal part of the formation which are very fissile, very soft, dark grey laminated hemipelagite.

The most easterly exposure of the formation is in a quarry [SO 2123 9139] at Sarn. Several metres of very fissile, calcareous and graptolitic flaggy sandstones with abundant Monograptus flemingii, slumped mudstones with concretions, and interbedded laminated hemipelagites are exposed, representing high horizons in the formation.

Bromleysmill Shales Formation

This new formational name is designed to embrace the poorly exposed, mainly mudrock sequence, which overlies the Purple Shales Formation in the eastern part of the district. It is a formation consisting of silty and calcareous shales which constitute an eastward passage (Figure 18) between dominantly non-bioturbated hemipelagites, with thin partings of turbiditic mudstones, of the Nant-ysgollon Shales Formation, (a basinal and basin-margin facies in the western part of the district) and homogenised, bioturbated, silty mudstones of the Coalbrookdale and Buildwas formations (parts of a Midland Platform facies which lies to the east of the district) (Bassett et al., 1975). Thus in the west, around Montgomery, where only its lower parts are exposed, the formation is a brown-weathering, fissile, silty and calcareous mudstone, with a high content of interbedded laminated hemipelagite. In the hemipelagite, carbonaceous films anastomose between pale grey-brown laminae of terrigenous silty mudstone, up to 0.3 mm thick, in which silty blebs have amalgamated but are still evident. At Pentrehayling [SO 241 929], the mudstone is laminated, but less distinctly. In some parts the dark grey carbonaceous films are concentrated into bundles persistent over at least 0.12 m and 5 mm or more apart. Within the bundles, carbonaceous films are discontinuous, arranged en echelon. In other parts they appear as isolated filaments, well separated and short and yet others consist of medium grey homogeneous mudstone of terrigenous origin with evenly dispersed black tabloid ‘flecks’ parallel with bedding. This may well be the result of micro-bioturbation of once more continuous laminae, but there is no evidence of macroburrows.

A little to the east, at Walcot [SO 263 994] (Whittard, 1932, p.885) the mudstones are buff-brown (weathered) and rather blocky. Though still laminated, the pelagic, carbonaceous films are singular and sparse. Layers of homogeneous mudstone attain thicknesses of about 4 mm, within which a bleb-form fabric is still evident. The blebs are well packed, some being partially enwrapped by curved filaments of dark grey carbonaceous material. Still farther east, at Bromleysmill Quarry [SO 3314 9145] the mudstone is coarsely fissile to blocky, micaceous and non-laminate, yet not obviously burrowed. Graptolites and scattered graptolitic fragments occur and a silt-bleb fabric is visible in the mudstone. A thin limestone with trilobite remains is present just below the quarry, indicating the presence of oxic intervals in the sequence here.

On the eastern margin of the district, south of the River Onny (Jackson, 1997), the formation consists of grey, silty mudstone which weathers to an olive-brown colour. The mudstone is finely micaceous, calcareous and with large ovoid concretions and is laminated in places. The rocks have a bed-form of even flags, up to 40 mm thick, and are similar to the Coalbrookdale Formation of the adjacent Church Stretton district (Sheet 166). Argillaceous limestones, up to 0.10 m thick, form up to 10 per cent of the formation, but wispy filaments of carbonaceous matter, characteristic of laminated hemipelagite, are absent, so completing the eastward passage referred to previously. This disappearance of carbonaceous laminae and the consistent eastward increase in the mud and silt fraction, both within the hemipelagite and as thin interbeds derived probably from weak bottom currents, must reflect a continuing influence from the Midland Platform albeit largely submerged by this time (Figure 19).

Assessments of the thickness of the formation, made sporadically along its outcrop, are imprecise. In the west around Church Stoke, the top has been removed by erosion. To the north of the district, in Long Mountain, Palmer (in Cocks et al., 1992), estimated the thickness of the equivalent Trewern Brook Mudstones Formation at about 457 m. On the eastern edge of the district, Jackson (1997) estimated the thickness to be about 620 m at least. The degree of possible error in these figures renders their differences of little significance apart from the likelihood of a slight thickening of the formation eastward on to the western flanks of the Longmynd.

Bromleysmill [SO 3326 9181] is about 800 m north-north-west of Lydham, and the formation is exposed discontinuously in stream banks between Bromleysmill and about 150 m south of Bromleysmill Pool [SO 3314 9198]. These exposures constitute the type section. The base of the formation is placed above purple-coloured shales where there is a change to grey and brown graptolitic shales (pp.52–67 and Evans, 1957, p.10). This change may occur slightly below the base of the Wenlock Series (Cave and Dixon, 1993) within the lapworthi (or grandis) Biozone, but it is not exposed within the district (p.67). The top is exposed in an unnamed dingle [SO 3038 9095] (p.85), and is marked by a change to more blocky bioturbated mudstone of the Aston Mudstones.

Church Stretton No.1 Borehole [SO 3723 8978] proved the lowest 70.91 m of the formation (Appendix). Graptolites and orthocones are present in the upper part of the core; trilobite and bivalve remains are common and burrows are present. Below 60.2 m, small brachiopods and crinoidal debris occur. White and cream-coloured seams of clay (bentonites), up to 60 mm thick, occur throughout the formation, but are particularly evident in the east.

Biostratigraphy

The Bromleysmill Shales Formation probably includes at least part of the lapworthi (or grandis) and insectus biozones of the Llandovery Series together with the whole of the Wenlock Series up to and possibly including part of the nassa Biozone. Correlations with the lithostratigraphy in other areas are presented in (Table 6).

The formation is generally poorly fossiliferous, but graptolites and orthocones are present throughout (Table 8). Small shells, especially of brachiopods, occur mainly in higher beds but are present lower, as near Crankwell Farm (see below). The graptolites indicate the presence of the C. linnarssoni (for example at Rhyd-y-groes [SJ 247 009]) and C. lundgreni biozones, but there is sparse graptolitic evidence for lower Wenlock biozones (Cocks and Rickards, 1969), although there is no reason to believe that the sequence is incomplete.

Details

An old quarry [SO 2364 9871] near Crankwell Farm exposes 0.5 m of buff-coloured shaly mudstone overlying 2.5 m grey, fissile, calcareous shale. Thin, wispy, dark grey, carbonaceous laminae are present but there are few graptolites except near the top of the calcareous shale. Some 1.5 to 2.0 m above the base, the shale is tough and blocky, with small shells including Chonetoidea sp. and a bentonite, 5 mm thick. Other fossils include ‘Orthoceras’ aff. tenuistriatum, ‘O’. sp., trilobite fragments, Monoclimacis flumendosae, Monograptus antennularius, Pristiograptus dubius and crinoid columnals. These indicate a biozonal position within a range of high riccartonensis to low linnarssoni, that is equivalent to the Penstrowed Grits.

Several metres of like strata belonging to the riccartonensis Biozone are exposed in an old quarry [SO 2630 9941] at Walcot, Chirbury where beds of bentonite also occur (Elles, 1900, p.36; Das Gupta, 1932, p.327). The topmost few metres of the formation are exposed in the unnamed dingle described below (p.85).

An old railway cutting [SO 3676 8982], east of Bow House, displays 2 m of grey, silty, poorly bedded mudstone with a graptolite fauna including Cyrtograptus cf. lundgreni, Monograptus flemingii elegans, M. flemingii, Pristiograptus dubius P. cf.d. ludlowensis and P. cf. jaegeri, indicative of the lundgreni Zone.

A quarry [SO 3664 8854], north-east of Totterton Farm exposes 6.5 m of grey calcareous silty mudstone with a few calcareous nodules and three beds of white bentonitic clay up to 50 mm thick. This quarry, and a road cutting [SO 3789 8797] to [SO 3793 8793] in similar beds at Plowden, both yield M. flemingii, indicative of an horizon in the lower part of the Wenlock Series. Church Stretton No.1 Borehole [SO 3723 8978] (Appendix) proved a sequence through the lower part of the Bromleysmill Shales Formation.

Mudstone, shaly, brown and soft, is exposed between tree roots [SO 3328 9180] at Bromleysmill. It yields Monograptus flemingii and fragments of Cyrtograptus sp., possibly C. lundgreni. Biostratigraphically these beds seem to lie well above the base of the Wenlock.

Some 350 to 500 m downstream and many metres higher in the succession, mudstones are exposed in the stream banks and a quarry as follows:

Quarry [SO 3314 9147] Thickness m
Mudstone, medium grey, very shaly and friable. Fauna from top, in stream, Cardiola sp., Monograptus flemingii, M. flemingii cf. elegans from bottom in quarry M. flemingii, and Pristiograptus pseudodubius 13.0
Stream banks and bed [SO 3314 9149]
Mudstone, medium grey, very shaly; graptolites and bivalves 3.25
Clay, cream-coloured (bentonite) 0.04
Mudstone, medium grey, shaly at top, passing down to blocky and olive grey, where it is mottled (bioturbated) containing ochreous tubules. Three, possibly four, beds of cream- coloured clay (bentonite) up to 0.06 m thick divide the mudstone. Fauna includes dalmanitinid trilobites 4.15
Clay, pale cream-coloured (bentonite) 0.06
Mudstone, olive-grey to buff, blocky, calcareous mottled and with ochreous tubules. A layer of hard calcareous concretions is present 0.13 m from top. Fauna mainly shelly: Glassia sp., orthocone and bivalve fragments, Dalmanites sp., M. flemingii s.l. and P. pseudodubius 0.9
Clay, pale cream-coloured (bentonite) 0.06
Mudstone, grey to buff, rather shaly. Fauna: ‘Orthoceras’ rectinicinctum, Dalmanites sp., Monograptus priodon and P. pseudodubius 2.0
Clay, pale cream-coloured (bentonite) in track on east side of pond seen
Old quarry [SO 3664 8854], 1.5 km, 305° from Plowden Mudstone, grey, silty, calcareous, not well bedded. Isolated calcareous nodules, up to 150 mm across, at 0.5 and 1 m below top 2.1
Clay, white, soft weathered (bentonite) 0.05
Mudstone, grey, silty (as above) 0.4
Clay, yellowish white, soft, weathered (bentonite) 0.01 to 0.02
Mudstone, grey, silty, a few calcareous nodules 1.9
Clay, white, soft, weathered (bentonite) 0.02 to 0.03
Mudstone, grey, silty with a few calcareous nodules 2.0

Penstrowed Grits (Denbigh Grits) Formation

The distinctiveness of the formation was recognised very early as the Denbighshire Sandstones (Sedgwick, 1844) and the Denbighshire Grits (Murchison, 1859). By 1881, Ramsay and his colleagues of HM Geological Survey had described most of the stratigraphical characteristics of the outcrop between ‘Conway and Builth’, including the passage of sandstone into mudstone in the east (Ramsay, 1881, pp.285 and 289). The shorter name, Denbigh Grits, was probably introduced by Boswell (1926) and this became the name commonly used for the formation (Cummins, 1957). The local name Penstrowed Grits was introduced by the Geological Survey of Great Britain (IGS, 1972) in the Newtown area, for an outcrop of uncertain correlation at that time, but now known to be synonymous with Denbigh Grits. The type section (p.73) is Penstrowed Quarry [SO 0680 9095] to the west of the district. Probably the most detailed account of the formation nearby arose from the mapping of part of the Llangadfan Syncline by Cummins (1963). More recently, Dimberline and Woodcock (1987) gave an account of the formation to the south-west of the district around Bwlch-y-sarnau.

The Penstrowed Grits occur in the north-west of the district as part of the main mid-Wales outcrop of the Denbigh Grits. Farther to the north-west (Cummins, 1957; Khan and Kelling, 1991), the thickness of the formation is about 1000 m. Similar thicknesses occur immediately west of the district at Llandinam [SO 026 884] and southwards in the Wenlock scarp to near Bwlch-y-sarnau [SO 040 743]. Eastwards these thicknesses diminish rapidly (Cummins, 1957; Dimberline and Woodcock, 1987) into the western part of the district. Here the formation dies out in mainly turbiditic mudstone along a line from just east of Llanbadarn Fynydd [SO 097 781] to west of Newtown, to Tregynon [SO 098 988] and thence approximately along the National Grid easting 10 to Bryneglwys, 5.5 km west-north-west of Llangollen (Figure 19) (Wedd et al., 1929, pp.70–71). This marks the position of the lower part of the eastern margin slope of the Montgomery Trough of Holland (1992).

The formation consists largely of medium grey mudstone in which beds of greywacke sandstone are common. Mudstones are mainly homogeneous, medium grey and with conchoidal fracture. Dark grey anoxic hemipelagite is present sporadically, as thin layers separating turbidites. The proportion of sandstone to mudstone is very variable, both vertically and laterally. The sandstones may occur as single beds, usually 0.10 to 0.30 m thick but some exceeding 1.0 m, and also as thick packets of such beds. The latter have been quarried in places, for example at the type section in Penstrowed Quarry and thus the best exposures display aspects of the formation not typical of the whole.

Sandstone beds are commonly graded (Ta-d intervals of Bouma, 1962). They are usually planar, but many amalgamate laterally. Flute casts and groove casts are common (Plate 4). The sandstones are medium to fine grained but with some granule grade. Grains are mainly of quartz, feldspar and lithic fragments (Cummins, 1957; Dimberline and Woodcock, 1987) in an illite-chlorite matrix. Lithic fragments include sedimentary, metamorphic and igneous rocks, keratophyres and spilites; tuffs with volcaniclastic textures are absent according to Cummins (1957, p.440), making it unlikely that local lower to middle Ordovician rocks contributed detritus. Particularly distinctive, especially near the base, are massive beds of homogenous, high-matrix sandstone (or sandy mudstone) which weather to a coffee-brown colour. The sandstone contains black clasts of ‘charcoal’ which are possibly carbonised algal clots and are abundant in places.

Sedimentation

The distribution of the Lower Wenlock, sand-rich turbiditic facies — the Denbigh Grits — is illustrated in (Figure 19). On grounds largely of palaeocurrent flow, but also of petrology, Cummins (1957, pp.447–448) envisaged two tectonically controlled Welsh deep marine troughs which acted both as depocentres and as conduits for turbidity currents. One trough lay south-north with its axis through western Montgomeryshire, including the western part of the district, to Corwen; the other, to the north-west, crossed Denbighshire with an easterly or north-easterly flow. They have been named the Montgomery Trough and the Denbigh Trough respectively (pp.56 and 94 and Holland, 1992, p.43).

Dimberline and Woodcock (1987) in the south and Khan and Kelling (1991) to the north (more distally) detailed the turbidite system within the Montgomery Trough. In the Bwlch-y-sarnau [SO 030 746] area, near the south-west corner of the district, Dimberline and Woodcock (1987) viewed the Denbigh Grits as the overbank deposits on the eastern side of a major channel complex. The ‘grits’ exposed on the western edge of the Montgomery district are similar though more distal. To the north, Khan and Kelling (1991) saw in the axial deposits of the trough an aggradation of vertically stacked distal lobes, at least five in number each 5 to 15 m thick.

One of the main uncertainties in the model is the lack of evidence for a western side to the trough, though Khan and Kelling (1991) believed that there was a gentle south-east facing slope in the north, produced by the Bala Fault (Figure 19). They suggested that this deflected the higher concentration flows to the north-north-east along its base, but encouraged the deposition of thick, lower density muds.

Another uncertainty is the provenance of the sediment. Cummins (1957) and Dimberline and Woodcock (1987) saw ‘Pembrokeshire’ as the likely source. This would have required an abrupt geniculation in the flow direction, for the palaeocurrent indicators point only to a northward flow from just north of Llandrindod Wells all the way to Corwen. Thus the currents appear to have flowed from an area where there was only mud being deposited (the Builth Mudstones) (Figure 19) and transposition of Pembrokeshire on the extant north-east-trending fault system would have required 100 km of post-Wenlock strike-slip. More plausible would be concealed or destroyed canyons across this fault system, from the south-east, comparable with those across the Church Stretton Fault in early Ludlow times (p.94).

Biostratigraphy

Graptolites are not common within the formation, but there is extensive evidence to the west and north-west of the district that the base lies at or just above the base of the riccartonensis Biozone, for instance, just to the west of the district, near Llandinam [SO 0459 8945] and in Sgwylfa Wood [SO 0653 9418] near Aberhafesp, and also farther afield (Dimberline and Woodcock, 1987; Cummins, 1957, p.434). The top of the formation in these western areas is possibly as high as the ellesae Biozone (Cummins, 1957; Dimberline and Woodcock, 1987, p.62). However, near the eastern margin of the formation there are indications of onlap of the lower parts against the basin margin with a complementary off-lap of the higher parts (Figure 22), so that both the bottom and the top of the formation are diachronous (Dimberline and Woodcock, 1987). For instance at Rhydyfelen [SO 0848 9297], west of Newtown, the top of the formation underlies shales of the linnarssoni Biozone, while to the north, near Llanfyllin [SO 142 195] the formation dies out eastwards within the riccartonensis Biozone (Wedd et al., 1929, pp.70–71).

Details

Major exposures lie just beyond the periphery of the district. Roadside sections [SO 0844 7562]; [SO 0876 7538] and a castle moat [SO 0899 7541] near Llanbadarn Fynydd expose over 100 m of beds on which seven vertically separated sets of palaeocurrent indicators (flute casts and groove casts) reveal currents to between 040° and 060°. Two others read 002° and 027°. The base of the formation is exposed here, placed at a layer of septarian nodules which show disturbed bedding within [SO 0868 7542]. Road sections immediately south are described by Dimberline and Woodcock (1987, fig. 3).

The type section in Penstrowed Quarry [SO 0680 9095] west of Newtown exposes nearly 100 m of beds, similar to those described above, where sandstones are mostly under 0.30 m thick. Minor exposures include old quarries [SO 0859 9500]; [SO 0899 9512], near Bettws Hill, which expose 20 m and 7 m, respectively, of thick sandstones. The sandstones are graded, from coarse at base to fine grained, micaceous and flaggy at top.

Nantglyn Flags Formation

McKenny Hughes introduced the name Nantglyn Flags in 1894 as a result of his work in the Denbigh area. (1879, 1885 and 1894). The type section is a quarry [SH 9790 5980] at Nantglyn, where the rocks, of early Ludlow age, were exploited for flagstones and for their well- preserved crinoids. As defined, its position lies above the Denbigh Grits and below grits of Ludlow age, a position maintained in the Montgomery district wherever the Penstrowed Grits are present. The formation is equivalent to the Nantglyn Flags Group of Warren et al. (1984), and the Nantglyn Flags Formation of Campbell (in BGS, 1993b). The name is preferred to the synonym Nantglyn Mudstone Formation used on the 1:50 000 series map. The lithology of the type section is more characteristic of the higher beds of the formation in this district than it is of the lower beds, but over all, lithologies are remarkably consistent along the length of the outcrop in Montgomeryshire and northwards.

The formation crops out in the western part of the district and forms a small inlier in the Mule valley between Kerry and Sarn. It is thickest in the north-west, thinning towards the south and east. In the area west of Bettws Cedewain [SO 122 968] it is over 800 m thick. Along the Severn Valley where the Penstrowed Grits and the lower part of the Nantglyn Flags are absent (the Mottled Mudstone Member rests on the Nant-ysgollon Shales), the formation is about 700 m thick near Fron [SO 178 977] and about 600 m around Newtown. Eastwards, in the inlier near Sarn, the formation is about 200 m, with the Gyfenni Wood Shales Member taking up about half of this thickness. The formation disappears abruptly eastwards and, apart from the member, it passes into the Aston Mudstones.

The Nantglyn Flags are essentially a thinly bedded, mud-dominated, turbiditic rhythmite deposited under anoxic basinal conditions. Most rhythms are from 20 to 100 mm thick with sharp bases and each generally possesses a thin basal sandstone (Tcd), a median structureless mudstone (Te) and a dark grey laminated hemipelagite at the top. These are the ribbon-banded mudstones of Boswell (1926) in which the rhythms maintain their appearance for at least as far as can be observed in exposures.

The sandstones are commonly up to 20 mm thick, cross-laminated and parallel-laminated, usually with sharp tops and invariably with sharp bases. Basal burrow-casts are absent and few beds preserve well-formed flute casts. Grading is present in some beds, where graptolitic and finer, brachiopod detritus occurs in the basal laminae. Thicker sandstones, up to 150 mm occur in places and these exhibit convolute lamination.

The structureless mudstones are the dominant part of each rhythm. They are of medium grey colour, soft, and upon weathering fracture readily along curved surfaces. Graptolites are very sparse, but well preserved.

The thinnest unit of the rhythm is usually the laminated hemipelagite, commonly less than 10 mm thick but may be as much as 60 mm thick. Graptolites are most common in these layers, and are usually flattened. Beds of white, hemipelagic clay (bentonite) replace or augment the laminated hemipelagites, occurring on average about 10 m apart throughout the formation. These bentonites are thin, usually much less than 100 mm, and many possess a basal lamina of pink calcite less than 3 mm thick. In weathered exposures, the bentonites are unobtrusive, having been etched deeply into the rock face.

The formation is slightly calcareous throughout, and hard ovoid calcareous concretions, up to 0.30 m across, are not uncommon. They lie along layers which follow specific horizons and are a largely pre-compaction, diagenic product. This rhythmite facies compares closely with the C2.3 facies of Pickering et al. (1986) and is typical of the ribbon-banded mudstones of Warren et al. (1984). Within the district, it occurs mainly in the upper part of the formation, in the ludensis Biozone and part of the nilssoni Biozone.

Three members have been differentiated within the formation, namely the Gregynog Mudstone Member, the Mottled Mudstone Member and the Gyfenni Wood Shales Member. In addition, two other lithologies form mappable but unnamed members shown on the 1:50 000 Series geological map as ‘disturbed beds’ and ‘sandstone’.

The Mottled Mudstone Member, in effect, divides the formation into an upper and lower part but these two parts are not equivalent to the Upper Nantglyn Flags Group and Lower Nantglyn Flags Group of the Denbigh district. The two groups are based on biostratigraphical not lithostratigraphical criteria (Warren et al., 1984). The formation is much more variable below the Mottled Mudstone than it is above and clearly was deposited during a change in the basinal conditions. Indeed the Mottled Mudstone marks an event which appears to have wrought widespread change in the nature of deposits even in the platform areas to the east where there was a change from terriginous mud to carbonate deposition.

Disturbed Beds

These are slumped beds (p.62), up to about 20 m thick, and occur in the north-west of the district. They have been observed particularly at or above the Mottled Mudstone Member especially in the nilssoni Biozone, but throughout the Gyfenni Wood Shale Member. Some outcrops have been mapped for distances up to 1 km and, although they appear on the map to lie at different levels, due to folding and faulting, it is possible — indeed likely — that several lie on the same level and belong to the same slump-sheet.

The beds consist of disrupted, contorted or completely destratified rhythmites, mottled mudstone, some sandstone and hemipelagic silty shale. The top surfaces of these ‘slump-sheets’ are sharp and slightly uneven. They are overlain by undisturbed strata, the basal member of which is usually a turbiditic sandstone, up to 0.15 m thick, and containing exotic shell and crinoid fragments [for example at 1008 9432], [SO 1631 9833], [SO 1703 9982] and [SO 1903 9153]. In places, where irregularities in the top surface of the slumps have been high, the overlying turbiditic sandstones do not overtop the eminences. This is achieved only by Te mudstones from the high level, lower density, part of the turbidity currents (Figure 26). The slumps may be the product of destabilisation, down-slope translation, disruption and redeposition of nearby basinal rhythmites. The overlying turbidite is considered to be the aftermath of that event, an energetic turbitity current which unburdened itself as it spread across the slump-sheet.

Sandstones

Towards the top of the ludensis Biozone the sandstones are thicker and more obtrusive forming a positive topographical feature which has been mapped in places such as Bryn-y-Cil [SO 1393 9705], where the sandstone beds are up to 0.15 m thick with cross-bedding and convolute-bedding, Newtown Golf Course and Sarn [SO 204 909].

Although these basinal sandstones were coeval with the Much Wenlock Limestone Formation of the Midland Platform to the east, they are no more calcareous than other sandstones in the Nantglyn Flags. Their provenance appears to have been more southerly, perhaps the landmass of Pretannia, rather than the Midland Platform.

Gregynog Mudstone Member

This newly named member overlies the Penstrowed Grits in the north-west of the district but largely replaces them over a short distance eastwards where the member is 190 to 275 m thick. It is absent farther to the east. The type section is in a stream near Gregynog Hall (p.81). The Gregynog Mudstone is composed of a greater variety of lithologies than is typical of Nantglyn Flags. It is softer, more shaly and weathers with brighter orange and rusty tints. Where rhythmically bedded, the rhythms are thinner, less perfect and thus less distinct as ribbon-banded mudstones than higher in the formation (Plate 5) and than those of the Denbigh district (Warren et al., 1984, pl.5).

The lower part of the member is dominated by homogeneous turbiditic mudstone, commonly thinly interbedded with fine-grained sandstone laminae and thin beds, up to 15 mm thick, and 0.06 to 0.20 m apart. In some parts the mudstone is massive and probably of mass-flow origin. Other slumped beds, also mainly of mudstone, contain contorted siltstone/sandstone laminae and isolated rounded clasts and concretions. Impersistent conglomerates containing intra-basinal clasts in a mudstone matrix are also present. The clasts are commonly small, rounded and oblate, although some are still angular, tabloid-shaped rip-up clasts. They are normally grey, but may possess dark alteration rims and be accompanied by shell detritus and black/phosphatic pelloids (Plate 6). The shell debris and the pelloids were presumably carried downslope into the basin from shallower waters to the east.

Thicker sandstones, up to about 0.17 m, occur particularly in the lower part of the member. They occur in packets, 1 to 2 m thick, and are probably part of the lateral passage eastwards from the Penstrowed Grits. Cummins (1957) recorded a similar passage in the north-eastern parts of the Llangadfan Syncline (Figure 34) as did Warren et al. (1984) in the Denbigh district.

The upper part of the member is characterised by rhythmites. Laminated hemipelagite is present and locally abundant, although not obviously so in exposures. Sandstones in these rhythmite mudstones are commonly 3 to 15 mm thick. Ovoid calcareous concretions occur, as in other parts of the formation. Beds of intensively burrowed fawn mudstone, 1 to 2 m thick, are present at more than one horizon. They are similar to the Mottled Mudstone ofnassa and basal ludensis horizons and indicate brief periods of oxygenated basinal deposition in linnarssoni–lundgreni times.

The upward transition through the member, from higher concentration mudstone turbidites with mass-flow deposits and slumps (D1 facies of Pickering et al., 1986) to low concentration turbiditic rhythmites (D2 facies), is as apparent at the south-west of the district, between David’s Well [SO 054 783] and Llanbadarn Fynydd, as it is in the north-west of the district between Gregynog [SO 088 979] and Highgate. It records the change from the rather accentuated early Wenlock Montgomery Trough (Holland, 1992), occupied by the Denbigh Grits to the west (Figure 22), to a broader basin with gentler slopes. This basin then persisted through the ludensis Biozone and the lower parts of the nilssoni Biozone accumulating the typical ‘ribbon-banded mudstones’ of Boswell (1926, p.562).

Mottled Mudstone Member

The name Mottled Mudstone was introduced by Boswell and Double (1940). In the west of the district, the member rests on the Gregynog Mudstone Member, but farther east upon the Nant-ysgollon Shales. The thickness of the member averages between 10 and 20 m, but is very variable without any clear pattern. In the north-west of the district, it is nearly 40 m [SO 1574 9567] at Dolforwyn and at Cefn Bryntalch [SO 1759 9638] across the Severn Valley. At Bryncoch [SO 1126 9602], the total thickness is possibly 50 m, but the member is split into two equal leaves by 5 m of ‘ribbon-banded’ mudstones. Similarly, two leaves are present [SO 1052 9414] near Goron where the member is about 15 m thick. Nearby, at Lower Garth [SO 0994 9530] and Middle Garth [SO 0950 9489], the member is not split and thicknesses are 13 m and 18 to 20 m respectively. Just north of Llanbadarn Fynydd, at the south-western corner of the district, the member is again split into two leaves [SO 091 783]. The upper and lower leaves have a combined thickness of 40.5 m and are separated by 75 m of ribbon-banded mudstone, making a total thickness for the member of 115.5 m. At Esgairdraenllwyn Bridge [SO 0859 8242], 4 km to the north, the member is only 9.5 m thick, assuming that there is no second leaf. To the east around Sarn, no thickness greater than 4 m has been recorded.

The Mottled Mudstone Member is lithologically very distinctive and commonly produces a topographical ridge. The mudstone is usually massive, silty, calcareous and intensively burrowed. It is medium grey to greenish and mottled but weathers to a buff colour. Bedding (0.10 to 0.20 m thick) is present in places. In other places thin beds of fine-grained sandstone are abundant; these are commonly parallel laminated and weather to a brown colour. In common with many other bioturbated, or disturbed, mudstones the fracture is uneven and curved. It is this uneven fracture and massive nature that lend toughness to the rock, and thus to the characteristic topographical ridge. Possibly the most distinctive characteristic is the scatter of thin tubules (possibly burrows) and small lumps formed of brown goethitic material (Plate 7). These have been discussed by Boswell (1949, p.38) who thought that they might have been rootlets; Warren et al. (1984, p.49) attributed them to compaction and sedimentary diagenesis. In parts, there are faint bedding-parallel marks, a few millimetres wide, visible on joint facies. But for their tabular form they would be deceptively like burrow-fills.

In the vicinity of the Severn Valley, sandstones, up to 0.20 m thick, occur in the topmost part of the underlying Nant-ysgollon Shales; such sandstones are also present within the Mottled Mudstone of this area. Lenses or beds of conglomerate, up to 0.20 m thick, contain synbasinal clasts, up to 50 mm long, exotic shell debris and black pelloids, in a mudstone matrix [SO 1564 9579]. The conglomerate is similar to that contained within the Gregynog Mudstone Member apart from being fawn in colour and not dark grey. Clearly the sedimentational regime was little different, only the depositional oxicity had changed. Within the member in some places, thin interbeds containing dark laminar hemipelagite reflect intervals of oxygen starvation locally and may correlate with the ribbon-banded mudstones which in some other places divide the member into two parts. The temporary dominance of oxygenated conditions, under which the member was deposited, was clearly finely balanced, a balance where local unevennesses of bathymetry might have made the difference between an oxygenated and an oxygen-starved bottom.

Biostratigraphy

The Mottled Mudstone Member contains a benthic fauna (Table 8) which is similar to that of the westerly exposures of the Aston Mudstone Formation, though more sparsely distributed. This fauna includes several taxa, records of which in Britain have previously been confined to the Mottled Mudstones of the Denbigh district (Rushton in Warren et al., 1984), such as Bracteoleptaena bracteola, ‘Clorinda’ dormitzeri, Mezounia sp., Strophochonetessp., Hyolithes fabaceus and Raphiophorus cf. parvulus. Graptolites are more common, however, and indicate that the member belongs principally to the G. nassa Biozone, but locally extends up into the M. ludensis Biozone.

The commonest and almost diagnostic fossil of the member is the retiolitid graptolite Gothograptus nassa. It appears to have been preferentially preserved in oxic marine sediments. There are several undescribed variants of the species which occur in the member and which may prove to be of stratigraphical significance (see details), for instance the taxa recorded as G. aff. nassa [SO 1928 9210] possibly from just below the Mottled Mudstone reached 1.8 mm in width, almost twice the width of the normal species and developed a meshwork which became more open distally. This propensity for the peridermal meshwork to be more open distally is shown by specimens from many localities and one [SO 1227 9578] is unique in having a branch from the main rhabdosome (Plate 2). These do not belong to the taxon G. nassa, but probably a form as yet unnamed.

At some exposures, all specimens reveal traces of membraneous lobes, as illustrated by Holm (1890) in the original description of the species. However in another collection none of the several specimens had developed such lobes, but the apertural margins were stiffened instead (compare with Bouc×ek and Munch, 1952). The membraneous form seems to be restricted to the lower parts of the Mottled Mudstone, for example the lower leaf (see details), while the second form is more significant of the higher parts.

Another fossil which is characteristic of the Mottled Mudstone and referred to as ‘Hairy Dick’ (White, 1988), comprises a horse-tail-like bunch of short (about 10 mm) black, carbonaceous filaments, possibly of algal origin.

Correlation and depositional environment

The event that made the Mottled Mudstone such a distinctive deposit had little effect on the shape of the basin, or upon the mechanics underlying its sedimentation; its mud-sand turbidites, slumps and even sporadic mud-matrix conglomerates are similar to those in the Gregynog Mudstone Member. The sand component of the turbidites is greater in places, as it is in the Severn Valley outcrops of the immediately underlying strata. The sediment was perhaps more calcareous too. The main effect of the event was to depress the oxygen-minimum layer so that it lay below the level of most of the basin floor for much of the time, and some of the basin floor all the time.

As well as a vigorously burrowing in-fauna, the event also introduced into the basin a benthic epifauna, for brachiopods and trilobites are characteristics of the mudstone. The former are indicative of the Visbyella Community, the deepest indigenous benthic community of Hancock et al. (1974); this community signifies distal shelf, or shelf-slope, environments. The Montgomery and Denbigh troughs of the Welsh Basin must have been occupied by water considerably deeper than that under which the Aston Mudstones and the Edgton Limestone accumulated farther east, yet the brachiopods of these latter two deposits are considered also to represent the Visbyella Community, so that the depth tolerance of this Visbyella Community must have been large.

The suggestion (Warren et al., 1984, p.49) that the event which produced these effects was a general lowering of sea level may be valid, but it must be considered carefully in the light of the Wenlock Limestone being cited as a coeval product of the same event (Warren et al., 1984, p.56) for this is not so. The Mottled Mudstone Member belongs mostly to the nassa Biozone and, in places, to the lowest ludensis Biozone too, whereas the Wenlock Limestone is of post-nassa age. However, a lowering of sea level with its implied marine regression would account for the increased energy of turbidity currents as recorded along the basin margin in the topmost Nant-ysgollon Shales and Mottled Mudstone (topmost lundgreni and nassa biozones). Such an event is reflected also in the shallowing sequences of the Ton Siltstone and Trostrey Limestone of Usk (Barclay et al., 1989), the Farley member of Wenlock (Bassett et al., 1975) and of course the lower parts of the Edgton Limestone.

The fall in sea level is unlikely to have been many tens of metres, for such would have dried out much of the Midland Platform of the time (Holland, 1992). Jeppsson (1990) has shown that small falls in sea level can be attributed to oceanic water contraction in periods (‘P-episodes’) of climatically controlled high latitude cooling. Cool, dense water oxygenates the deep bottom and coincidentally, there is a small regression. Such a situation accords closely with the observations made above on the Mottled Mudstone and its co-relative deposits of the nassa and low ludensis biozones.

The question then arises: did the return of anoxic conditions, as reflected in the ribbon-banded mudstones of later ludensis times, signify a rise in sea level and thus marine transgression? This was the time when carbonates of the Wenlock Limestone were accumulating, and thus any sea level rise would have been no more than the reef builders could match. Jeppsson (1990) points to these later ludensis events as one of his ‘S-episodes’. This would have reversed the ‘Mottled Mudstone event’ by a warming and expansion of oceanic water thus allowing salinity-dense, anoxic water to replace the oxygenated cold polar water in the basin. Sea levels rose and basinal deposits became anoxic again.

Gyfenni Wood Shales Member

This term was introduced by Cave et al. (1993). The member crops out at the top of the Nantglyn Flags, and it is lithologically and biostratigraphically similar to, but laterally separate from, the Oakeley Mynd Formation in the east of the district. The thickness varies from 40 to 50 m in the north-west near Bettws Cedewain [SO 122 968] to over 100 m around and to the south of Montgomery. East of Hopton [SO 2265 9100] the member disappears abruptly. The member consists of thinly interlaminated dark grey hemipelagite and paler grey homogeneous turbiditic silty mudstone. The rock weathers to a very friable, brown shale. It differs from the Nant-ysgollon Shales in being more silty and having more abundant thin siltstone laminae, which produces a biscuity appearance at exposure. Ovoid calcareous concretions which lie preferentially along some bedding planes are fairly common and weather to a soft ochreous ‘gingerbread’.

Thin intervals of disturbed (slumped) strata occur especially in the extreme east of the outcrop along the Caebitra valley, where the member and the formation die out eastward.

Biostratigraphy

The fauna is almost wholly graptolitic, with some orthocones and bivalves. In the east, the member is, perhaps wholly, restricted to the lowest Ludlow nilssoni Biozone, while in the west it extends up into the scanicus Biozone. Thus deposition of the member probably continued longer in the west, where the Wenlock Montgomery Trough had existed, than in the east, just east of Sarn. This diachroneity is matched by a converse one, at the base of the overlying Bailey Hill Formation, so that to the east around Aston, the nilssoni Biozone lies partly within the Bailey Hill Formation and partly within a few centimetres of strata below, attributable to the Gyfenni Wood Shales (Figure 28).

In the extreme east of the district, hemipelagites again dominate the nilssoni Biozone in the Oakeley Mynd Formation. This formation is lithologically very similar to the Gyfenni Wood Shales Member of the Nantglyn Flags Formation, but it is set within a very different sequence of non-turbiditic, largely shallow marine mudstones. The formation is not in continuity with the Gyfenni Wood Shales although both would have been termed the Lower Ludlow Graptolitic Shales by Earp (1938 and 1940) and Holland (1959).

Details

Gregynog Mudstone Member and Mottled Mudstone Member

The Mottled Mudstone Member together with parts of the formation just above and below (Figure 24) are well exposed in a roadside section in the south-west corner of the district. Over 50 m of the Gregynog Mudstone Member is exposed and compared with the member farther west at David’s Well and around Gregynog (Figure 25) it exhibits a lower-energy facies. Thin multilayer bedding is pervasive (rhythmite), but not as obvious as it is in the ribbon-banded mudstones above the Mottled Mudstone. Laminated hemipelagite forms a higher proportion of the rock, forming layers that differ widely in thickness. These characteristics probably reflect the easterly, basin-margin position of the section, where the Penstrowed Grits Formation is very thin or absent.

Graptolites at [SO 0881 7708] include M. cf. flemingii, M. cf. f. compactus, Pristiograptus sp.: at [SO 0879 7872] M. cf. flemingii s.l., cf. Cyrtograptus hamatus and at [SO 0875 7875] M. cf. flemingii.

Gothograptus nassa was found only in the Mottled Mudstone Member. Those from the lower leaf [SO 0860 7893] all possess membraneous lobes (Holm, 1890) and occur with P. pseudodubius and ‘Hairy Dick’. G. nassa from the base of the upper leaf [SO 0852 7933] does not possess the membraneous lobe, but the apertural margins are stiffened and the peridermal meshwork is looser and less regular (Bouček and Munch, 1952). Strata higher in this lower leaf [SO 0868 7906] yield M. ludensis indicating that upper horizons of the Mottled Mudstone Member here belong to the ludensis Biozone.

There are several exposures of the Gregynog Mudstone Member in the Tregynon [SO 097 988] area. A stream section [SO 0910 9794] to [SO 0890 9790] 600 m north-east of Gregynog Hall exposes the basal part of the member (Figure 25).

Some 1 km west of Highgate, about 210 m of the formation are exposed sporadically in a stream and lane banks. Lithologies are very similar to those of the Llanbadarn Fynydd road section and, commencing in the lane side [SO 0960 9536], the lower beds yield Cyrtograptus sp. The Mottled Mudstone Member occurs at the top, in the river banks at Lower Garth [SO 0997 9531]. A similar section occurs at Ieithon Brook; from the core of an anticline [SO 1002 9600], the stream cuts through mudstone and sandstone, probably of the topmost Penstrowed Grits, passing south-eastwards through thinly bedded hemipelagites and thin turbiditic mudstones with some sandstones or ‘grits’ [SO 1031 9586] and an intraformational limestone conglomerate. The sequence passes up into the Mottled Mudstone [SO 1047 9577] with Monograptus ludensis, followed by ribbon-banded flaggy mudstones, which are rather fissile at the base, in the centre of a syncline [SO 1053 9573]. On the south-east limb of the syncline, an abundant fauna of G. nassa and M. ludensis occurs in flaggy fissile sandstones and mudstones [SO 1060 9562] which mark the top of the Mottled Mudstone. In this limb, the Mottled Mudstone is either repeated by strike faulting, or consists of two leaves as at Llanbadarn Fynydd to the south-west. Other river sections occur between the Sewage Works [SO 1254 9611] (Mottled Mudstone) and Bettws Cedewain, between Lower Cwm Harry [SO 1218 9822] and the river bank [SO 1144 9773] (yields M. flemingii) and south-south-east of the Community Centre [SO 1415 9957].

A quarry [SO 1241 9855] in the Gregynog Mudstone Member yields M. flemingii. It exposes 7.5 m of mudstones with thin sandstones (to 15 mm thick), homogeneous (turbiditic) mudstones (up to 0.06 m) and hemipelagites (up to 0.02 m) but becoming dominant downwards. Three parallel-laminated sandstones occur in the top 3 m, and an intraformational mudstone conglomerate, up to 0.08 m thick, enveloped by concretionary limestone is present near the base. In the extreme west in another quarry [SO 0836 9940] probably many metres of medium to dark grey mudstone are exposed. There is no regular bedding fabric, only sparse contorted sandy layers, isolated calcareous concretions, scattered pebbles of rotten sandstone and sparse pockets of greenish clay (bentonite). Such slumped mudstone is reminiscent of low levels in the Gregynog Mudstone Member at David’s Well (Sheet 179) where they are underlain by Penstrowed Grits.

Homogeneous mudstone with thin beds of brown weathered fine sandstone is exposed in a track [SO 1227 9578] east of Penbryn and cf. ‘Clorinda’ dormitzeri and Gothograptus nassa, typical of Mottled Mudstone, are abundant. The latter exhibits membranous lobes with some specimens having rather open peridermal meshwork distally. Similar membranes are present in G. nassa from the lower leaf of the Mottled Mudstone Member near Llanbadarn Fynydd, to the south-west, and it is likely that the Penbryn exposure is at a similar low level. One specimen of Gothograptus reveals a secondary branch from the main rhabdosome (Plate 2iii). It does not exhibit the finely reticulate peridermal meshwork of G. nassa and may be an un-named form.

The most easterly exposures of mudstones considered to belong to the Mottled Mudstone Member are near Sarn, in a bank and ridge [SO 1931 9222], a quarry [SO 2060 9184], a small excavation [SO 2083 9102] and behind a pond [SO 2215 9077]. However, these have not been shown on the 1:50 000 Series map.

Ribbon-banded mudstone, typical of the rhythmites within and above the Mottled Mudstone Member, are well exposed at a number of localities which fall within two biozones, the ludensis Biozone (Wenlock) and the nilssoni Biozone (Ludlow).

Nantglyn Flags Formation (undifferentiated)

ludensis Biozone

A railway cutting [SO 0795 9066] exposes some 90 m of thinly interbedded mudstone and sandstone. The mudstones are calcareous and up to 70 mm thick; the sandstones are fine grained, cross-bedded and in beds up to 30 mm thick. A partially disturbed bed, 0.5 m thick, occurs about 10 m from the base, with a 0.05 m-thick blue-grey, pyritous, shell-fragmental limestone about 3 m higher. flute casts are present indicating palaeocurrents from the south-east.

A quarry [SO 2066 9097] behind Sarn Church exposes about 50 m rhythmites, mostly with inverted dips. The thin rhythms consist of basal fine calcareous sandstone (some with shell debris at the base and others with convolute beds up to 50 mm thick), buff homogeneous mudstone and thin silty hemipelagite tops.

The top of scarp [SO 1995 9196], 400 m east of Great Cefnyberin, exposes 4.1 m of thinly bedded calcareous mudstone with much hemipelagite. Thin, fine-grained sandstones up to 50 mm thick are common; cross-bedding indicates palaeocurrents from the south-south-west.

A cutting [SO 1166 9224] by the Old Toll Gate exposes 40 m of thinly bedded mudstone and sandstone, as for the railway cutting above. Turbiditic rhythms are commonly about 35 mm thick, but in some the sandstones are up to 45 mm.

A stream, 1 km north-north-west of Newtown, exposes [SO 1020 9297] a bed of blue-grey conglomeratic limestone 0.20 m thick. It contains rounded clasts up to 0.15 m long of fine-grained limestone in a matrix of darker, pyritic, shell-fragmental limestone (Plate 8).

A roadside quarry [SO 0858 9330] near Fachwen Pool exposes 9 m of ribbon-banded mudstone; rhythms are up to 0.20 m thick, and consist dominantly of nodular-weathering, grey, homogeneous mudstone with thin dark grey hemipelagites.

A quarry [SO 1130 9573], 400 m north-north-west of Highgate exposes 2 m of grey, tough, silty homogeneous mudstone in beds up to 0.06 m thick, with some plane-parallel-laminated, fine-grained sandstone up to 0.03 m. This rests on calcareous siltstone (0.10 m thick) with fine laminations rolled into concentric ellipsoids — pseudonodules. Orthocones, brachiopods and graptolites are strewn over surface of the underlying 5 m-thick ribbon-banded mudstones. This consists of rhythms of homogeneous mudstone (up to 0.05 m thick), hemipelagite (up to 0.02 m) and thin, basal fine-grained sandstone.

A new quarry [SO 1581 9617], 350 m west-north-west of Lower Bryn, exposes about 10 m of ribbon-banded mudstones with numerous thin sandstones in the top 2.5 m; some are lenticular or have basal shell debris. A 0.03 m-thick, very soft, calcareous mudstone lies 4 m from top of section. Two bentonites, both 0.06 m thick, occur, one at the base of the section and the other about 2.5 m above it. The latter is probably disturbed and rests on 0.20 m of slumped mudstone.

A quarry [SO 1345 9870] near Lower Ucheldre exposes 9 m of ribbon-banded mudstones with up to 40 sandstones occurring per 1 m of strata. Soft, homogeneous, pale grey mudstone, up to 0.10 m thick, occurs 4 m from the top, underlain by a prominent sandstone. Convolute-laminated sandstone, 0.03 to 0.09 m thick occurs 1.52 m from the base.

nilssoni–scanicus biozones

A roadside (A483T) cutting [SO 086 830], 250 m long, exposes some 175 m of fissile, to very fissile, Nantglyn Flags in which the fine sandstones are very thin, but common. There is a high proportion of laminated hemipelagite. Some 200 m farther north, another section [SO 0856 8348] reveals similar beds becoming even more fissile upwards and passing up into, the lateral equivalent of the Gyfenni Wood Shales Member.

In a quarry [SO 0820 8463], 100 m south of Cwmyrhiwdre, 14 m of ribbon-banded mudstones are exposed, in which there is much laminated hemipelagite, in beds up to 10 mm thick; homogeneous mudstone occurs in beds up to 20 mm thick.

A lane-side cutting [SO 1971 9057] 200 m east of Red Gwenthrew, in the core of a syncline [SO 1950 9050], exposes 26.93 m of ribbon-banded mudstones. Several sandstones are thicker than normal, up to 0.20 m with convolute lamination. A bed of disturbed mudstone, 1.5 m thick, rests on a soft white bentonite, 20 to 200 mm thick, about 9.5 m from the top. The precise position of these beds has not been established, but it is near the ludensis–nilssoni biozonal boundary. Other exposures [SO 1915 9032], [SO 1933 9010] occur nearby.

In a quarry [SO 0885 9100] near Bryn-hyfryd, 3 m of strata are exposed; about 70 per cent consist of homogeneous silty mudstone in beds up to 70 mm thick which weather spheroidally; thin, fine sandstone makes up about 20 per cent. Laminated hemipelagite is very subordinate.

The dingle [SO 0891 9120] to [SO 0890 9132], near St David’s, exposes many metres of similar, dominantly homogeneous mudstones.

An accessible river cliff [SO 1050 9157], in Newtown, where there is now a suspension footbridge, exposes medium grey silty mudstone (which weathers spheroidally) with fine sandstones or siltstones, usually less than 10 mm thick. Layers of ovoid calcareous concretions occur sporadically. Dips are in the order of 3° or 4° in various directions, so that during low water the beds are exposed as flats and shelves in the river bed.

A bank [SO 2119 9237] excavated behind Perthybu exposes about 22 m of beds. At the base is a 30 mm-thick, pale grey clay (bentonite) overlain by about 6 m of contorted slumped mudstones and sandstones, including pods of bentonite, on top of which is a non-laminated lenticular sandstone. This trio of lithologies has been observed in many places and the association is almost certainly a genetic one. The overlying sequence is evenly bedded mudstone with thin, wispy, sandstones in the lower 8 m, passing up into homogeneous mudstone (up to 40 mm thick) with much thinner, graptolitic hemipelagites.

A lane bank [SO 2024 9252] near Cwmberllan exposes a similar sequence. The basal 5.23 m of ribbon-banded mudstones contains two thin bentonites mainly evident by the thin laminae of pink calcite always associated with such clays. These are overlain by a low-angle dislocation surface, followed by some 1.2 m of contorted and disrupted beds (slumped). Capping the slumped bed and infilling irregularities on its surface is a sandstone, up to 30 mm thick. This is overlain by 7 m of flaggy calcareous ribbon-banded mudstones and thin sandstones. A quarry [SO 2023 9270] nearby, which lies tens of metres higher in the succession, exposes over 4 m of ribbon-banded mudstones. These beds are more fissile and hemipelagite makes up to 50 per cent of the rock; thus they are similar to the Gyfenni Wood Shale.

Cwm Badarn Dingle [SO 2004 9386] to [SO 1996 9435] exposes many metres of ribbon-banded mudstones in rhythms commonly 20 to 30 mm thick, with thin basal sandstones. The beds are disturbed by slumping in several places. At the south end bedding persists, but dips are very variable. In the centre a quarry [SO 1998 9431] exposes 5 m of disturbed (slumped) mudstones which include a bentonite. This slump-sheet is capped by 70 mm of dark grey, fine sandstone, which is cross-bedded and coarsest in the basal 10 mm and provides another example of the close association between bentonite, slumping and a capping sandstone layer. At the north end a waterfall also marks the top of slumped beds, possibly the same ‘sheet’ along strike. Other bentonites are present in the section: one [SO 1998 9421] at the base of the east bank has been stripped out by the stream which flows over a long expanse of the bared surface of the underlying mudstone.

A lane-side quarry [SO 1377 9568] at Pentre exposes 4 m of ribbon-banded mudstones, mainly buff-grey homogeneous mudstones which weather spheroidally and possess a pencil-like cleavage. Rhythms are up to 0.2 m thick with sandstones sparse, thin (30 mm at most) or lenticular and hemipelagites up to 30 mm thick. Nearby a new quarry [SO 1275 9617] reveals about 6 m of similar mudstone, but with proportionately more sandstones in rhythms only 30 to 70 mm thick. The sandstones show cross-stratification indicating palaeocurrents from south to south-east and grade up into homogeneous mudstones with hemipelagite tops.

In a cutting [SO 1495 9771] at Pen-y-coed Farm yard, a slumped mudstone bed contains small rafts of thinly bedded, ribbon-banded mudstones and a scatter of ovoid calcareous concretions. The overlying thinly bedded strata yield a graptolite fauna indicative of low nilssoni Biozone.

In a gully through Llifior Wood [SO 1614 9796], strata indicating the nilssoni Biozone are well exposed. Sandstones are very subordinate or absent, and the rhythms tend to be thick and dominated by spheroidally weathering homogeneous mudstone up to 60 mm thick, with hemipelagites up to 40 mm. In a quarry [SO 1601 9808], at the north end of the gully, 5 m of mudstone consist of 80 per cent spheroidally weathered homogeneous mudstone in beds up to 0.5 m thick, hemipelagites, up to 20 mm and no sandstone. In a roadside quarry [SO 1552 9871], south-west of Black Wood, 9 m of similar strata occur, but with sandstone up to 5 mm thick.

In the Lower Llifior area, a thick slump-sheet (or debris-flow deposit) has been mapped and its top surface is exposed in the north bank of a stream [SO 1631 9833]. The lower part of the deposit (‘a’ of (Figure 26)) is buff-coloured silty mudstone, with curled and streaked colour and textural differences; up to 3 m is exposed. The top part of the deposit ‘b’ differs from ‘a’ in having a random scatter of calcareous concretions as clasts, and is up to 0.25 m thick. Its top surface is uneven and it is overlain by a fine sandstone ‘c’ except on the peaks of the underlying slump surface. The sandstone shows cross-bedding and wavy, parallel lamination and is up to 0.15 m thick. It appears to have arrived in the wake of the slumping activity as the dense fraction of a turbidite, which failed to overtop the eminences of the slump. It was followed by the less-dense mud fraction ‘d’ which did overtop the eminences, thus forming the cap-rock of those parts. Hemipelagite ‘e’ was deposited during the subsequent interturbidity interval. The overlying turbidite rhythm is also atypical of the formation; both the basal sandstone ‘f’ and overlying turbidite mudstone ‘g’ are thick, suggesting that they may have been a more distal product of a repeat slumping or mass-flow event.

Almost 2 km to the north-north-east another outcrop of slumped mudstone is exposed (8 m) in a quarry [SO 1704 9987] and it seems likely that it belongs to the same slump sheet.

Gyfenni Wood Shales Member

The member is exposed along the forestry road [SO 2171 9057] and the following provides a description of the section from top down (iv to i in (Figure 27)).

Thickness m
Bailey Hill Formation
iv. Sandstones, fine-grained in beds up to 0.12 m with plane-parallel, wavy and convolute lamination; calcareous (blue-hearted), weathering pale brown; basal layer, or lenses of current-transported shell detritus present in some beds. Flute casts,palaeocurrents from 210°, 220° and 225°, and cross-sets with palaeocurrents possibly from north- east and east. A few thin beds of bentonite are present, one possibly having acted as a glide plane to the slumped bed (Db) Pristiograptus minor tumescens present in southerly quarry 35
iii. see (Figure 27)b
Nantglyn Flags, Gyfenni Wood Shales Member
ii. Mudstone, silty, homogeneous, fissile (thinly flaggy) due to interbedded silty laminated hemipelagite and siltstone partings. Yields C. interrupta, ‘Orthoceras’ argus, ‘O’. elongatocinctum, ‘O’. cf. recticinctum, B. bohemicus L. cf. simplex, cf. N. nilssoni (near top), S. (Colonograptus) colonus colonus, S. varians, S. cf. varians. c.20
i. Mudstone, very fissile, dominantly silty laminated hemipelagite. A bed of destratified, rather massive grey-buff mudstone, 3–5 m thick (Db) — a slump or mass-flow deposit — occurs c.15 m from the top. C. interrupta, ‘O’. cf recticinctum, Parakionoceras originale; ?L. critinus, L. progenitor (common), Mcl. micropoma, M. cf. uncinatus orbatus, cf. N. nilssoni, P. macilentus, S. chimaera sl., S. chimaera salweyi, S. chimaera semispinosus, S. colonus, S. varians (common), Spinograptus spinosus seen about 25

In Hopton Dingle, a section [SO 2298 9059] to [SO 2298 9049] exposes about 36 m of similar fissile shales with graptolitic hemipelagite. Within this sequence disturbed strata (slumps), with sharp folds and inconsistent dips, occur as ill-defined packets. One packet occurs near the base and one near the top; massive, homogeneous, buff-grey silty mudstone resulted probably from mass-flow. Graptolites include Pristiograptus dubius aff. ludlowensis, Saetograptus chimaera and S. chimaera semispinosus. Some fine-grained sandstones occur, up to 20 mm thick, and a packet of these, 0.2 m thick, creates a small waterfall [SO 2298 9053]. This succession is overlain by 3.4 m of mudstone in which flaggy to fissile sandstones 20 to 40 mm thick make an appearance. Some 1.4 m below the top S. cf. colonus s.l., P. cf. dubius dubius, Bohemograptus bohemicus and ?Neodiversograptus nilssoni are present, recording probably the nilssoni Biozone. These beds are overlain by a synbasinal conglomerate 0.40 m thick, which is exposed in the cliff on the eastern side of a waterfall [SO 2298 9049]. This marks the base of the Bailey Hill Formation and the beds above are typical of that formation, with abundant sandstones in couplets with silty mudstone. At 0.25 to 0.30 m above the conglomerate they yield Saetograptus sp. and Pristiograptus sp. The saetograptid has at least six spinose thecae. These spines appear to arise from the top part of the thecal aperture thus indicating that the graptolite belongs to the chimaera or leintwardinensis group with a position above the lower half of the nilssoni Biozone. A small fault appears to separate these beds from the overlying beds exposed in the face of the waterfall. The latter consist of 4 m of flaggy sandstones with silty mudstone partings in couplets and at the base of the waterfall they yield the brachiopods Jonesea grayi, D. navicula and Isorthis sp. and the graptolites B. bohemicus, L. scanicus and P. dubius.

An old quarry [SO 2350 9083], at Highfield Cottage, exposes 11 m of very fissile silty graptolitic hemipelagites and minor thin homogeneous mudstones. Graptolites including M. cf. uncinatus (common), cf. N. nilssoni and P. dubius dubius are probably of the nilssoni Biozone. Although one specimen of the M. cf. uncinatus is close to M. uncinatus orbatus, others differ in having ‘stronger’ thecal hooks.

Similar beds in a roadside quarry [SO 2340 9309] about 2 km to the north are rich in graptolites including M. uncinatus orbatus, N. nilssoni (both common) and S. spinosus, low in the nilssoni Biozone. This horizon is almost identical with that at Cwm Badarn [SO 1997 9416] which yields the same graptolites within a ribbon-banded mudstone rhythmite facies, illustrating the westward increase in turbidites.

In other stream sections the member is about 120 m thick [SO 1613 9174] and 40 to 50 m thick [SO 1799 9279] to [SO 1813 9290]. Its thickness near Maenllwyd [SO 1489 9233] to [SO 1488 9240] is about 67 m and there is a passage up from ribbon-banded mudstones into alternations of dark grey parallel-laminated hemipelagite and homogeneous mudstone, each up to 20 mm thick, with sporadic ovoid, calcareous concretions. Sandstone flags (up to 30 mm) appear in the top 15 m and pass up into 5 m of ‘passage beds’ into the overlying Dingle Mudstone Member of the Bailey Hill Formation. The Brynrorin Conglomerate occurs much higher [SO 1485 9208].

Aston Mudstone Formation

The name is derived from the Aston Beds of the Bishop’s Castle area (Allender, 1958). In the Church Stretton area to the east of the district, the equivalent beds were included within the ‘Ludlow Series’ (IGS, 1967), for when that district was mapped (1945–1959) the beds were considered to belong to the ‘vulgaris zone’of Ludlow age (now the Monograptus ludensis Biozone). To the west, the formations have a lateral equivalent in the Nantglyn Flags (Table 6), its western limit being in the region of Cwm [SO 260 910], Church Stoke and Dudston. The type section is in Aston Dingle (pp.87 and 88).

The main outcrop lies along the base of the north-facing scarp of Silurian rocks, from near Cwm [SO 260 910] to The Heblands, just north of Bishop’s Castle. This outcrop exposes the base of the formation, in an unnamed dingle [SO 3037 9091] just east of Aston Dingle. Nowhere else is the base exposed and it cannot be traced accurately westward of Aston Hall where there is a thick cover of drift. East of Bishop’s Castle, the outcrop can be traced from Oakeley Farm to the edge of the district. The formation is about 120 m thick along its western outcrop. East of Bishop’s Castle, it thickens from 235 to 250 m, of which 70 and 150 m, respectively, are Edgton Limestone. Here, the base of the formation has been mapped in a position a few metres below the Edgton Limestone, at the base of a thin transition from the Bromleysmill Shale Formation, probably like the sequence in the unnamed dingle to the west. Again this corresponds approximately to the base of the nassa Biozone.

The formation consists of silty calcareous mudstone, blocky and rather tough. Bedding fabric is poorly preserved, probably because of the high degree of bioturbation. The mudstone is medium grey, weathering to a buff colour in natural exposures where it is decalcified and contains narrow irregular brown goethite specks and tubules (possibly burrows). The rock is similar in appearance to the Mottled Mudstone Member of the Nantglyn Flags and yields a brachiopod, trilobite and graptolite fauna. In the eastern outcrop, beds in the lower part of the formation are very calcareous and they form an impure limestone, the Edgton Limestone Member. There are two minor, but mappable beds of limestone below which, although excluded from the member, serve to emphasise the transitional nature of its base (Greig et al., 1968, p.151). In the adjacent ground farther east, Greig et al. (1968) treated the limestone as a formation and its surface as marking the base of the Ludlow Series.

The formation was deposited between the Midland Platform east of the Church Stretton Fault, where shallow water carbonates were accumulating (Wenlock Limestone), and the turbidite basin west of Sarn (Nantglyn Flags). Terrigenous detritus was scant over the Midland Platform and shorelines were distant, to the south and south-west of Usk (Walmsley, 1959; Holland, 1992, fig. S3b).

Storm-generated layers of bioclastic, or siliciclastic, detritus have not been observed in the Aston Mudstone Formation and while it is possible that some bedding fabrics were obliterated by intensive bioturbation, it seems unlikely that all traces of sand beds would have disappeared in that fashion. Therefore it is concluded that the formation was deposited under quiet, distal shelf conditions, compatible with the presence of a brachiopod fauna indicative of the comparatively deep water Visbyella Community of Hancock et al. (1974).

The presence of the Edgton Limestone Member in the east reveals that carbonate deposition did extend westward of the Church Stretton Fault, but it was highly contaminated with terriginous silt and ended within ludensis times. The lateral facies change is focused on the line of the fault and it has been suggested by Greig et al. (1968, p.7) that penecontemporaneous movement with downthrow to the west promoted this change. In addition, it can be argued that if the Mottled Mudstone of the basin reflects a sea-level low-stand in nassa biozonal times, then the rhythmites which succeed it represent a slight sea-level rise in ludensis times. This eustatic rise would have contributed to the effects of the fault in maintaining deep water west of the Church Stretton Fault. Yet the rise was slow enough for the reef-building organisms on the Midland Platform to have kept pace with it.

Edgton Limestone Member

The term Edgton Limestone was introduced by Greig et al. (1968) for use on adjacent ground to the east. In the Montgomery district, it forms a member of the Aston Mudstone Formation in its easterly outcrop. Greig et al. (1968, p.150) pointed out that the Edgton Limestone is unlike the Wenlock Limestone of Wenlock Edge, but is similar lithologically to the Tickwood Beds (equivalent to the Farley Member; (Table 6)) beneath the Wenlock Limestone and the survey of the Montgomery district has shown that there is almost certainly a correlation between these two rock units. It should be noted that Bassett (1989, p.61) erected an Edgton Member of the Much Wenlock Limestone Formation to denote two facies at the top of that formation. That Edgton Member should be renamed to avoid confusion with the Edgton Limestone of Edgton.

The outcrop follows a distinctive ridge north-westward, from near the village of Edgton, disappearing beneath morainic drift near Lodge Farm [SO 3404 8981]. The member occupies a position in the basal part of the formation east of Bishop’s Castle, but is absent to the west by passage into a mudstone facies. There is a thin, lower leaf of limestone separated by mudstone from the main thickness. Thicknesses are discussed above (p.85).

The limestone is fine grained, silty and of uniform texture. It is medium grey and weathers to olive-green or tan colours. Beds are 0.04 to 0.12 m thick, commonly having undulating bases but flat tops. In unweathered sections, interbedded calcareous siltstone is difficult to distinguish from beds of limestone, but it forms about 50 per cent of the rock. In thin section, the limestone consists mainly of recrystallised carbonate, with about 5 per cent angular quartz. There are very minor amounts of mica, biotite, chlorite, feldspar and sphene.

Biostratigraphy

In the west, as in Aston Dingle (Cave et al., 1993), the lower part of the formation has yielded graptolites, belonging to the nassa Biozone, but has not yielded any brachiopods. The overlying ludensis Biozone yields a sparse benthic, shelly fauna, of low diversity, representing the Visbyella Community. This is true also of other western exposures, such as that at Dudston [SO 2432 9754].

In the east, calcareous beds below the Edgton Limestone yield a sparse shelly fauna which includes Visbyella trewerna and graptolites including G. nassa, M. ludensis and P. jaegeri. Thus the nassa–ludensis boundary may well lie within the passage up into the Edgton Limestone, if not in the limestone itself. Within the Edgton Limestone fossils are sparse. They include simple and compound corals and Atrypa reticularis, Leangella?, Harpidella (H.)cf. aitholix and a phacopid pygidium. Graptolites include cf. M. ludensis and P. cf. jaegeri which, together with the Harpidella (H.) aitholix, are suggestive of the ludensis Biozone, possibly the lower part. Above the Edgton Limestone, the formation contains a diverse, indigenous, benthic shelly macrofauna. Small brachiopods are particularly prominent and include Dicoelosia biloba, Jonesea [Aegiria] grayi, Skenidioides lewisii, Strophochonetes cf. cingulatus and Visbyella trewerna which represent the Visbyella Community (Hancock et al., 1974). Trilobites are also well represented and include species of Ananaspis, Dalmanites, Harpidella and Raphiophorus. The affinities of this shelly fauna lie with the Wenlock, but it possesses nothing diagnostic of either Wenlock or Ludlow. However, the only graptolite species collected, cf. M. ludensis, suggests the whole assemblage is best regarded as Wenlock.

The impression given by these faunas confirms the view that the marine environment was shallowest in the east, east of Bishop’s Castle.

Details

West of Bishop’s Castle

The most westerly exposure of the formation is at West Dudston where a few centimetres of crumbly, buff, bioturbated mudstone occurs in the south-east corner of a slurry pit [SO 2434 9753]. It yields a fauna characteristic of the ludensis Biozone, including Jonesea grayi (common), ‘Clorinda’ dormitzeri, Visbyella trewerna, M. ludensis (of Wood), Pristiograptus? and crinoid columnals, an orthocone, a solitary coral and Beyrichia s.l.

Beds which are transitional between Aston Mudstone to the east and the basinal Nantglyn Flags to the west are exposed in the lane side [SO 2806 9158] beneath the ancient ‘Fort’ at Pentre, Some 1.2 km west of Aston Dingle. Lithologically, they are atypical of Aston Mudstone though they have been included in that formation and belong to the ludensis Biozone. Blocky, buff, bioturbated mudstone is probably subordinate to interbeds of grey-brown graptolitic shale which incorporates much laminated hemipelagite. The section shows:

Thickness m
Shale, grey-brown graptolitic. Thin beds of grey laminated hemipelagite alternate with thin grey- buff homogeneous mudstone. Cardiola interrupta, P. cf. jaegeri seen 4.5
Mudstone, buff to greenish, very blocky in rough beds up to 80 mm thick. It is bioturbated and shows a scatter of brown blotches and ‘pipes’ of goethite (burrow-fill). In the top few centimetres, thin graded beds with silty bases are present. Slava? 0.9
Mudstone, buff, shaly to thinly flaggy and silty due to thin, parallel, silt laminae and thin intervals of laminated hemipelagite. Gothograptus sp. (broad form) seen 1.1

The type-exposures provide an almost complete succession through the formation. The base of the formation is exposed in an unnamed dingle, east of Aston Dingle, about 5 m below the foot of a small waterfall [SO 3038 9094]. The section here reveals:

Thickness m
Aston Mudstone Formation
Small quarry and dingle south of main road [SO 3037 9091]
Mudstone, blocky, calcareous weathered brown, yields decalcified rugose coral and ‘Clorinda’ dormitzeri? c.20.0
North of the main road, mainly unexposed except near waterfall:
Mudstone, tough, blocky calcareous weathered brown. Jonesea?, Glassia sp., Strophochonetes cf. cingulatus, Visbyella trewerna, orthocone fragments, Platyceras sp., Dalmanites sp.,beyrichiacean, crinoid columnals and P. pseudodubius c.20.0
Face of waterfall except base:
Mudstone as above; beds from middle of waterfall yield Leangella segmentum, Ananaspis aff. communis Gothograptus nassa, P. cf. dubius ludlowensis 2.5
At base of waterfall:
Mudstone, grey, calcareous weathering buff, more fissile than above. Jonesea?, S. cf. cingulatus, G. nassa, P. cf. dubius ludlowensis, P. jaegeri seen c.1.0
Downstream of waterfall:
Mudstone, poorly exposed. Jonesea grayi, G. nassa, P. dubius, P. dubius ludlowensis c.5.0
Bromleysmill Shale Formation
Mudstone, slightly silty, grey, very fissile in parts. J. grayi, Ludfordina pixis, Nuculites? and graptolite fragments c.2.0
Unexposed
c.8.0 Mudstone as above; G. nassa, P. dubius, P. pseudodubius? c.2.0
Unexposed c.2.0
Mudstone, grey, shaly. ‘Orthoceras’ argus?, ‘O’. cf. recticinctum, M. flemingii elegans, P. cf. pseudodubius seen

The rest of succession is exposed in Aston Dingle (Figure 28) as buff weathering, blocky, bioturbated mudstone. It is calcareous and grey where fresh, and yields mixed shelly and graptolitic faunas of the G. nassa and Monograptus ludensisbiozones. At locality d (Figure 28)b there is a marked change from rather shaly mudstone of the nassa Biozone below, to more massive and more obviously bioturbated mudstone with a shelly benthos of the ludensis Biozone above. This change appears to be too high in the formation to relate to its basal transition seen at the previous locality. However, the base of the Edgton Limestone to the east occurs at about this level and the fauna above it is also dominantly shelly. The highest graptolites from the formation in Aston Dingle are M. ludensis and P. cf. jaegeri found about 3 m below the top of the formation identified by Dr R B Rickards during a visit by the Ludlow Research Group.

At Upper Heblands, near Bishop’s Castle, a quarry [SO 3252 9029] exposes poorly bedded grey, silty mudstone with a rich shelly fauna, and is host to a microgranite. It is strongly burrow-mottled and weathers to a buff colour revealing brown goethitic blebs and wisps, typical of the Aston Mudstone. The graptolites M. cf. deubeli, M. ludensis? and P. jaegeri prove that the mudstone belongs to the ludensis Biozone (Sanderson and Cave, 1980).

East of Bishop’s Castle

Below the Edgton Limestone Member

An old pit [SO 3753 8647] 250 m south-west of Plowden Hall, exposes 2 m of nodular limestones, in beds up to 70 mm thick and forming 60 per cent of the rock, interbedded with grey-buff-weathering, calcareous, silty mudstone.

The faunas from other exposures [SO 3658 8655], [SO 3717 8775], [SO 3719 8653] and [SO 3758 8651] in this area, very close to the base of the formation comprise Amphistrophia funiculata, J. grayi, L. segmentum, Strophochonetes sp., V. trewerna?, Ananaspis aff. communis, G. nassa, Mcl. flumendosae kingi?, Mcl. flumendosae? s.l., P. jaegeri and P. lodenicensis.

An old quarry [SO 3605 8825] about 230 m north of Totterton Farm exposes:

Thickness m
Mudstone, flaggy and calcareous (graptolitic) c.2.3
Siltstone, calcareous, massive; limestone locally at top 0.3
Mudstone, weathering olive-brown, silty 0.4
Mudstone, grey, silty with lenticular argillaceous limestone beds to 0.15 m thick (much bioturbation) c. 1.0

Farther north, a downfaulted outlier near Lea provides two more quarries; one [SO 3571 8933] 400 m 050° from Lea exposes 4.3 m of grey, calcareous siltstone, finely micaceous in flags 20 to 30 mm thick. Limestones, up to 100 mm thick, form about 40 per cent of the rock, in the lower part and 15 per cent in the upper. The graptolite P. jaegeri indicates the ludensis or nassa Biozone. The other quarry [SO 3549 8939] occurs in similar mudstones, siltstones and interbedded limestones.

Also probably below the Edgton Limestone are about 3.5 m of grey to olive-brown silty mudstone, exposed in a quarry [SO 3595 8934]. The beds are poorly bedded at the top but the basal 2 m are graptolitic and well bedded, with M. ludensis and P. cf. dubius ludlowensis.

Edgton Limestone Member

A steep bank [SO 3519 8666] exposes 4 m of interbedded grey limestone and calcareous fine-grained calcareous siltstone in beds 40 to 120 mm thick. A sparse, mainly brachopod fauna is present.

An old roadside quarry [SO 3621 8629] to [SO 3630 8627] exposes 5.0 m of grey, silty limestone in beds 20 to 100 mm thick, interbedded with grey, calcareous parallel-laminated siltstone. The limestone is nodular in places and constitutes 60 per cent of the rock. Fossils include Bracteoleptaena?, J. grayi, Lingula sp.,Sphaerirhynchia wilsoni, V. trewerna; Dalmanites sp., Harpidella sp., Raphiophorus cf. parvulus; Beyrichia s.l.; cf. M. ludensis, P. cf. jaegeri and P. cf. lodenicensis, indicative of the ludensis Biozone.

From an overgrown bank [SO 3515 8648] A. reticularis, Leangella? and Harpidella (H.) cf. aitholix were found near the top of the member.

Above the Edgton Limestone Member

Immediately above the Edgton Limestone, 1.5 m of strata are exposed along a track [SO 3593 8790] to [SO 3593 8788]. They consist of olive-brown siltstone with isolated nodules of limestone, up to 0.2 m in diameter. A gap, equivalent to about 4 m of strata, separates this from a lower exposure of similar rock, 4.5 m thick. This and nearby exposures [SO 3650 8757], [SO 3661 8757] and [SO 3692 8754], also just above the Edgton Limestone, yield Atrypa reticularis, Dalejina hybrida, Dicoelosia biloba, Eoplectodonta?, Gypidula galeata, Howellella elegans, Isorthis?, Leangella segmentum, Leptaena cf. depressa, Lissatrypa sp.,M. obtusa, Microsphaeridiorhynchus?, Orbiculoidea sp., Pholidostrophia (Mesopholidostrophia)sp., Protochonetes minimus, Resserella cf. sabrinae, Shagamella ludloviensis, S. minor, Skenidioides lewisii, Sphaerirhynchia?, Striispirifer?, V. trewerna; Cornulites?, Cypricardinia, Mytilarca mitilimeris?; ‘Orthoceras’ cf. filosum; Dalmanites sp., phacopid;Beyrichia s.l., crinoid columnals.

A roadside exposure [SO 3457 8625] and [SO 3596 8768] to [SO 3597 8763] in the topmost few metres of the formation yields a solitary coral, decalcified bryozoan, and the following fauna. Brachiopoda: Craniops implicatus, Dalejina hybrida, Dicoelosia biloba (common), Eospirifer radiatus, Gypidula?, Howellella elegans, Isorthis clivosa, cf. I orbicularis, Jonesea [Aegiria] grayi, Leangella segmentum, Leptaena holcrofti, Lingula sp., Lyssatrypid, Mesopholidostrophia?, Salopina conservatrix, Shagamella?, Skenidioides lewisii, Spiriferid? Strophochonetes?, Visbyella trewerna, Whitfieldella?; Cephalopoda: Dawsonoceras annulatum, Kionoceras angulatum, Trilobita: Calymene sp., Dalmanites sp. Harpidella (H.) cf. aitholix, Acanthopyge? phacopid fragment; Ostracoda: Beyrichia s.l. and crinoid columnals. Such a fauna could be of Ludlow age or Wenlock age but V. trewerna favours Wenlock. Beds some 30 m higher contain S. (C.) varians indicative of nilssoni to scanicus biozones.

A quarry [SO 3446 8834], east-north-east of Oakeley Farm, exposures 4 m of greenish grey, finely micaceous and calcareous siltstone from near the top of the formation. This yields solitary rugose corals, Amphistrophia funiculata, A. reticularis, D. hybrida, Howellella sp., Pholidostrophia (Mesopholidostraphia)?, Resserella cf. sabrinae, Visbyella?; Dalmanites sp., odontopleurid fragment; Beyrichia s.l. crinoid columnals and cf. M. ludensis. The record of cf. M. ludensis, so near the top of the formation, must be of significance, in beds where the shelly fauna is rather equivocal of age.

An excavation [SO 3525 8817] yielded Isorthis sp., L. segmentum and R. cf. sabrinae.

Base of the Ludlow Series

The type area of the Ludlow Series lies immediately south-east of the district in the Welsh Borderland. Here the stratotype base of the series has been established at Pitch Coppice Quarry, near Ludlow (Lawson and White, 1989, p.80), fixed to the base of a transition from the Wenlock Limestone into the Lower Elton Formation, a shallow marine facies of soft argillaceous siltstones. An alternative biostratigraphical definition of the base of the Ludlow series (Holland et al., 1969; Cocks et al., 1971, p.105) made it coincide with the base of the nilssoni Biozone. This, it is said, also coincides with the base as defined at the stratotype (White, 1981; Cocks et al., 1992, p.2) making the two definitions complementary. The latter definition can be applied fairly easily where graptolites are common in the Wenlock and lower Ludlow sequences. Thus, in the western part of the district the base of the Ludlow Series is found to lie within the ribbon-banded mudstone rhythmites of the upper part of the Nantglyn Flags (Table 10). Farther east, in Aston Dingle, the same biozonal base lies within a metre of strata below the Bailey Hill Formation (Figure 28) at a sharp change in lithology, which was accompanied by a change from oxic to anoxic conditions of deposition. No part of the Aston Mudstone Formation in the Aston Dingle area is of Ludlow age.

Farther east, up to the boundary with the Church Stretton district (Sheet 166), the base of the Ludlow Series remains at the base of the graptolitic shaly facies (Oakeley Mynd Formation) which rests abruptly on Aston Mudstone, thus it maintains the link with the change from oxic to anoxic deposition. However there is a new factor in this area, the development of the Edgton Limestone, some 150 m below the top of the Aston Mudstone and thus also 150 m below the change from oxic to anoxic conditions. It was this lithological change, marking the top of the limestone, that the surveyors of Sheet 166 chose as the base of the Ludlow Series. They were following the tradition of fixing the base of the Ludlow Series to the top of the Wenlock Limestone with which the Edgton Limestone was in part correlated (Greig et al., 1968, p.150). In any case their faunas collected from the beds immediately above were not sufficiently diagnostic to be certain that any belonged to the Wenlock. Consequently they designated all the beds above the Edgton Limestone ‘Ludlow Series (undivided)’ and thus there is a mismatch across the sheet boundary between the top 150 m of the Aston Mudstone Formation and the bottom 150 m of the Ludlow Series (undivided) (Figure 29).

This mismatch is part of the long-standing problem that no biostratigraphical signature has been found which identifies the base of the Ludlow Series within the shelly borderland sequences, where graptolites are sparse and ineffectual for purposes of correlation. The traditional use of the lithofacies change at the top of the Wenlock Limestone relied on the hope that it was synchronous everywhere, but Hurst (1975) maintained that such a view belies the facts. He posed evidence of a ‘quick transgression’ [eustatic rise], and thus a synchronous datum, which lay close enough to the stratotype boundary for it to serve as the base of the Ludlow Series. Matched against this, the top of the Wenlock Limestone is diachronous, of Wenlock age in the north, but of Ludlow age in the south. Even if valid, exposure in the Montgomery district is too poor to facilitate a search for Hurst’s datum, but there seems little doubt that the top of the Edgton Limestone is transgressive, ageing westwards until it meets the base. The use in the Church Stretton district of such a diachronous surface to mark the base of the Ludlow Series may be the sole reason for mismatch with the base of the series as defined in the Montgomery district.

However, it should be noted that the base of the Ludlow in the Montgomery district is identified not just with the base of the nilssoni Biozone, but also by the coincidence of this base in Aston Dingle with an oxic to anoxic revolution. When traced to the eastern boundary of the district the latter, and seemingly the former too, lie some 150 m above the Edgton Limestone (Figure 1). Farther east, beyond the Church Stretton Fault the only comparable oxicity change is that at the base of the Middle Elton Formation, which introduced a dominantly graptolitic fauna (Holland, et al., 1963, p.107) including N. nilssoni. This circumstance is a reminder of the case made by Shergold and Bassett (1970, p.136) for the base of the Middle Elton Formation being coincident with the base of the nilssoni Biozone and thus the Ludlow Series.

Oakeley Mynd Formation

Oakeley Mynd was first used as the formational name by Jackson (1985, 1997) who adopted it from the hill of that name just east of Bishop’s Castle. The type section is in the quarry [SO 3385 8800] at Oakeley Farm. Evans (1957) has mapped the formation under the name Graptolitic Grey Mudstones. Allender (1958) placed the strata in the Eastern Facies of his Aston Beds.

The thickness of the formation is between 140 to 180 m on the northern slopes of Oakeley Mynd but it thins to the north-west, disappearing under glacial drift [SO 332 890] east of Bishop’s Castle. Beyond this, the Bailey Hill Formation appears to rest directly upon Aston Mudstone, for instance in Aston Dingle, and since there is no obvious unconformity the formation appears to pass laterally westwards into the basal part of the Bailey Hill Formation. It is at about this level that an intra-basinal conglomerate, the Brynrorin Conglomerate, is sporadically present to the west, so that the possibility of a depositional hiatus or sediment-scavenging currents has to be considered.

The Oakeley Mynd Formation consists of uniform silty mudstone, grey to brownish grey in colour, micaceous and slightly calcareous in places. It is fissile, splitting readily into flags about 10 mm thick. It is a hemipelagite intercalated with up to 80 per cent terrigenous mudstone. Carbonaceous filaments are short and wispy, up to 3 mm long, and vaguely concentrated into layers in places. Lamination is thus present but poorly defined. It is also partially destroyed by a scatter of largely bedding-parallel, burrow fills and possibly other smaller scale bioturbation. In general, cross-laminated silt and sand are absent and graptolite rhabdosomes and orthocones show little evidence of current orientation.

The lithology of the Oakeley Mynd Formation compares more closely with easterly occurrences of Bromleysmill Shales than it does with the contemporary Gyfenni Wood Shales to the west with their numerous thin siltstone layers and more acutely orientated graptolites.

The formation contrasts sharply with the underlying blocky and shelly Aston Mudstone though exposures of the base are few. A track section [SO 3588 8751] to [SO 3592 8754], near Totterton Hall, reveals that the change occurs as a rapid passage, in which blocky mudstone alternates with fissile graptolitic mudstone. The formation is exposed around Oakeley Mynd and has been quarried for use on lanes and tracks. In ploughed land, it produces a flaggy brash commonly revealing graptolites.

Biostratigraphy

The Oakeley Mynd Formation yields a pelagic fauna with abundant graptolites indicating the N. nilssoni Biozone with evidence of the L. scanicus Biozone from the highest beds. Orthoceratids are well represented, and specimens of Entomozoa have been recorded; so too has the earliest occurrence in the district of the brachiopod Dayia navicula.

The formation correlates with the Lower Ludlow Graptolitic Shales at Knighton (Holland, 1959) and the Gyfenni Wood Shales east of Sarn. However, the presence of D. navicula and the incidence of burrowing suggest that the Oakeley Mynd Formation sediments were slightly less starved of oxygen. The general view of borderland correlations is that these formations, representing the basal Ludlow Series, equate with the Lower Elton Formation (Cocks et al., 1992) on the basis that the Lower Elton Formation rests upon the Wenlock Limestone the surface of which traditionally marks the base of the Ludlow Series. Difficulties endemic in this view were discussed above (p.89) and evidence in its favour lies in a few specimens of Neodiversograptus nilssoni. An alternative view is sustainable on other grounds as well as graptolitic faunas, that these formations correlate largely with the graptolitic Middle Elton Formation. If the traditional correlation were to be upheld then a sharp bathymetric gradient would have had to be present across the Church Stretton Fault.

Details

The type section in an old quarry [SO 3385 8800], consists of silty, micaceous, very fissile shale, in beds up to 10 mm. Buff homogeneous mudstone is interbedded with brown laminated hemipelagite. Graptolites are abundant, and some are orientated into a 350° trend. Apart from a single Monoclimasis micropoma they are monospecifically Saetograptus (Colonograptus) varians. Other fossils include Cardiola?, cf. Maminka sp., ‘Orthoceras’. mocktreense, ‘O’ reticinctum. In the entrance to this quarry S. (C.) colonus compactus was found.

In a nearby lane-bank exposure [SO 3363 8790], the proportion of laminated hemipelagite is higher. The beds yield C. interrupta, ‘O’. recticinctum, Lobograptus scanicus, Mcl. micropoma, Pristiograptus. cf. dubius, P. cf. minor tumescens and S. (C.) varians.

A temporary section in a trench on the north of Oakeley Mynd [SO 3494 8816] to [SO 3483 8805] exposed typical brown-grey, fissile, silty mudstone through the bottom third of the formation. Debris from the trench yielded a rich fauna (Table 11).

A road bank [SO 3573 8750] to [SO 3573 8753] exposes fissile silty mudstone, which is grey and weathers to an olive-brown colour. The beds are abundantly graptolitic; fauna includes ‘O’. argus, ‘O’. recticinctum, L. progenitor, Mcl. micropoma, Pristiograptus? and S. (C.) varians.

Shale debris from a pond [SO 3655 8485] provided fauna ranging from low to high nilssoni Biozone with conocardiacian?, cf. Slava?, ‘O’. ibex, ‘O’. cf. recticinctum, ‘O’. aff. recticinctum, S. (C.) colonus compactus, S. (C.) cf. colonus, S.(C.) varians, L. progenitor, Mcl. micropoma, M. uncinatus orbatus, N. nilssoni, ?P. vicinus and S. cf. incipiens.

Bailey Hill Formation

The name is adopted from the Bailey Hill Beds (Holland, 1959) of the Knighton area, south of the district; Tyler and Woodcock (1987) substituted ‘Formation’ for ‘Beds’. Allender (1958) used the name Bishop’s Moat Beds for most of the formation and set up a stratigraphical classification for the district (Table 10), but the bulk of his results was never published. A disadvantage with Holland’s definition lies in the fact that base of the formation does not crop out in the type area (Holland, 1959, p.452).

One of the earliest accounts of the stratigraphy of the local Ludlow Series was given by Stamp (1919) who had mapped the Bucknell area to the south-east. He misinterpreted the age of that sequence (Holland, 1959, p.471). Earp (1938, p.130 and 1940) worked on the western part of the outcrop and identified these beds mostly as the Wilsonia wilsoni Grits. This name was in common use for many years but has now been abandoned as it is a biological derivative. Both Allender and Earp recognised a thin group of more fissile graptolite rocks above the Bishop’s Moat Beds and Wilsonia wilsoni Grits, namely the Cwm Mawr Beds (Allender, 1958, p.13) or Monograptus leintwardinensis Shales (Earp, 1938). These beds are merely the end expression of a thinning-upwards sequence in which sandstone/mudstone couplets (p.61) have died out leaving laminated hemipelagite as the more dominant facies. Thus, in this account it forms the top part of the Bailey Hill Formation, as the Cwm-yr-hob Member (Table 10); (Figure 18).

The formation has one of the most extensive outcrops in the district forming some of the major slopes around the edge of the Clun Forest high ground and also extending northward from Kerry to Montgomery. A small outlier occurs about 2 km east of Bettws Cedewain [SO 122 968]. The formation may be as much as 1300 m thick in the south-west of the district, near the headwaters of the River Teme, over 1100 m in the centre around Clun Forest, decreasing eastwards to about 600 m around Bishop‘s Castle. The base of the formation is well exposed in several places, in Aston Dingle [SO 2946 9117] (Figure 28), in Hopton Dingle [SO 2298 9049] and in the Gyfenni Wood forestry track [SO 2137 9016] (Figure 27).

The Bailey Hill Formation is distinguished from the formations above and below it by the dominance of relatively thick and regular couplets of fine sandstone and argillaceous siltstone (p.61). It is further distinguished from the formation above by the absence of bioturbation. Two members are recognised within the formation, the Dingle Mudstone at the bottom and the Cwm-yr-hob Member at the top. In addition, the Brynrorin Conglomerate occurs at, or near, the base of the formation, but it is not a mappable unit.

In general, sandstones are thickest near the base of the formation, up to 0.50 m; higher in the sequence they are no more than 0.10 m thick. In places thicker sandstone beds occur at the base of the formation. In the east Allender (1958) called them Lower Bishop’s Moat Beds and in the west, just east of Newtown, Tyler and Woodcock (1987) named them the Brimmon Wood Member. The defining characteristics of these members are vague and the terms have been abandoned. However, in many places the introduction of the sandstone/silty mudstone couplets, marking the base of the formation, is abrupt and plain to map. The outcrops around and to the east of Bishop’s Castle were excluded from the formation by Bailey (1969) and Tyler and Woodcock (1987) (see also Cummins, 1969, p.229), yet Allender (1958) and this survey found that they differ little from the formation elsewhere. Sandstones account for about 30 per cent of the formation and range in thickness from 0.03 to 0.30 m. However, a tendency for them to occur in impersistent packets is recorded by the symbol ‘S’ on the map.

Concentrations of coarse shell-detritus are present, usually as lenses at the base of sandstones. These formed the basis of Earp’s name, the Wilsonia wilsoni Grits. They weather, by dissolution of the carbonate shell debris, into dark brown porous rottenstone. These traction-carpet deposits are not confined to the bases of sandstones. Some occur within the body of the sandstone as lags above surfaces which may descend, by overlapping or truncating lower layers, to become the base of the bed elsewhere (Plate 9). It is presumed that these were produced by local resurgence of the current during the one event and not by currents of a separate, successive event.

A subordinate component of the formation is silty laminated hemipelagite. It occurs as partings between the couplets and is thin and sparse in lower parts of the formation, increasing in thickness and abundance upwards, so giving rise to the Cwm-yr-hob Member at the top.

The couplets of this formation were considered to be turbidites (Cummins, 1959b; Bailey, 1969; Woodcock, 1976b), deposited in a trough of bathyal depths (‘Montgomery Trough’). Subsequently, Tyler and Woodcock (1987) interpreted them as the deposits of storm impelled currents on a shelf below wave-base, a regime which was terminated by a mid-Ludlow tectonic event which brought about a modification of the palaeogeography and was marked by submarine sliding of sediment. The event couplets of the Bailey Hill Formation do indeed differ in appearance from turbidites characteristic of the Ordovician–Silurian basin of mid-Wales to the west. For instance, the silty mudstone component of the couplet, which would be Bouma’s (1962) Te interval of a turbidite, is more silty and less sharply demarcated from the underlying sandstone. Secondly, apart from the bioclasts of basal lags the sandstones are uniformly fine-grained. However, some of these unusual characteristics, described and discussed in detail by Tyler and Woodcock (1987, pp.80–81), could stem from factors imposed by the grain-size and petrography of the source material and it does not seem that the bedform of the couplets excludes the possibility of transportation by turbidity currents.

Had storm currents been the means of transporting the sediments which comprise the Bailey Hill Formation, then the parent storms would have been felt intensively on the shallow marine Midland Platform, barely 15 km to the east of the outcrops. Under that circumstance palaeocurrent indicators in the formation could be expected to reflect the linear nature of such a source more clearly than they do (Figure 20)a and b. In any case, storms would have left ample evidence in the Gorstian platform sequences themselves, waning rapidly into the Ludfordian, but Watkins (1979, pp.185–193) and Lawson (1992, p.42) demonstrate that they do not. It seems most likely, therefore, that the Bailey Hill Formation was deposited as a ‘fan’ from unconfined density-impelled flows generated in the area of Knill, near Old Radnor, by mass sliding of sediment (Kirk, 1951b; Bailey, 1969, pp.287 and 297; Lawson, 1973, p.270) and by nearby submarine channels from the Midland Platform (Whitaker, 1962), to form a fan in deeper, but not ‘bathyal’, water on a foundering tract of the outer shelf west of the Church Stretton Fault. In its most westerly outcrop, south of Dolfor, the lower part of the formation is dominated by silty mudstone, in beds up to 3 m thick near the base; sandstones are thin. This facies is atypical of the outcrop to the east, and even of the Dingle Mudstone Member to the north and may reflect rapid deposition of silt and mud at the western fringe of the ‘fan’.

The very large quantity of relatively fine-grained sediment which constituted the Bailey Hill Formation seems too great to have been derived directly from the small emergent area of the Midland Platform to the north-east and the promontory of Pretannia to the south-east (Lawson, 1992, p.45), yet there is no known terrestrial alternative. Furthermore there is evidence, from several earlier periods (Kirk, 1951b, p.56; Cave, 1992, p.43 and this account p.130), of shallow marine or attenuated sequences along the eastern (footwall) side of the Church Stretton Fault; Lawson (1973, p.270) and Kirk (1951a, p.72) indicated that the positive nature of this tract continued into the Ludlow, at which time Lawson envisaged it as a ‘shelf-edge ridge’. Such a ridge would have impeded a direct transfer of platform detritus westward into the deeper water except by way of breaches in it. Whitaker (1962) indicated that such breaches did exist, possibly induced by small east-north-east-trending faults. The parallel, but much larger, Leinthall Earls Fault nearby may have been the focus of another breach (Figure 20)b. Thus, it would have been feasible for the input of sediment to have been localised along these, while fault movements may well have concentrated slumping in areas around and west of Knill (Kirk, 1948; compare with Bailey, 1969, p.297).

Evidence of slope ‘wasting’ together with channel (canyon) activity in the region of Wigmore Rolls has been recorded by Whitaker (1962; 1994) with the inference that they were largely the product of a post-Bringwoodian catastrophe (1994, p.39). Clearly the activity ceased during the early Leintwardinian, coincidentally with the demise of the Bailey Hill Formation event couplets (when there might have been a minor rise in sea- level), but on the evidence there seems to be no good reason why the slope wasting had not commenced much earlier, in early Gorstian times, that is coinciding with the start of deposition of the Bailey Hill Formation with its Brynrorin Conglomerate farther west and north-west. Such coincidences of timing underline further the connection between activities on the submarine slope caused by the Church Stretton Fault and deposition of the Bailey Hill Formation north of a line westward from Knill.

Montgomery Trough

This account has not embraced the concept of a Montgomery Trough of Ludlow age which accommodated the Bailey Hill Formation between north-westward and south-eastward facing slopes, and which passed, respectively, through the areas of Builth Wells and Newtown. The concept originated with Cummins (1959b, pp.177–178) who envisaged a remnant of the Welsh Basin lying east of Newtown. He based it on evidence in the Bailey Hill Formation of contemporaneous westward-directed slumping near Builth Wells and in the Aymestry to Ludlow area, of south to north palaeocurrents and of slumps near Newtown (Earp, 1938; 1940). The direction of movement of the Newtown slumps, though crucial to the concept, had not been established, but they lie well to the north-west of the Builth Wells, Aymestry and Ludlow slumps, so it was argued that they must have been generated on an opposing slope and therefore have moved south-eastward.

The trough thus envisaged was called the Montgomery Trough (Cummins, 1959b, p.174) and considered to be the same one that had accommodated the Montgomeryshire Denbigh Grits of Wenlock age (Cummins, 1959b, p.171; 1962, pp.52–53). It was argued that the western margin of the Montgomery Trough (Wenlock) was the Derwen Ridge (or Merioneth Dome), extending south-westwards from near Corwen (Cummins, 1957, pp.147–149; Khan and Kelling, 1992) and it was considered to have remained as the Derwen Ridge into early Ludlow times (Cummins, 1959b, pp.171 and 176; Bailey, 1964, p.17). The Derwen Ridge lay some 40 to 50 km to the north-west of course, yet the main premise behind the concept of a Montgomery Trough of Ludlow age is the relatively steep (Cummins, 1959b; p.177) south-eastward facing slope that produced slumps near Newtown (Woodcock, 1976b, p.173). Bailey (1969, p.303), cast some doubt on the role of the Derwen Ridge in Ludlow times, and indeed on the Montgomery Trough of Ludlow age, when he suggested that the slumping near Newtown may be due to more rapid subsidence in the Clun Forest area and not a persistent Derwen Ridge. Further doubt can be added in that the two formations supposed to have been accommodated in the trough, the Denbigh Grits and the Bailey Hill, show major differences in composition (Cummins, 1959b; and Smith, in Bailey, 1969, p.303), have different provenances and are widely separated spacially (Figure 19), (Figure 20)b). Thus there is little reason to relate the Bailey Hill Formation to the Montgomery Trough of Wenlock origin, but the notion of the formation having been the deposit of a narrow trough confined within a north-west margin near Newtown and a south-east margin through Builth seems equally implausible (Bailey, 1964, p.10). Its axis, as defined by the slumps on either side, would have to curve sharply to the north-east in the region of Llandrindod Wells, yet the flute-cast palaeocurrent indicators do not follow such a path. Secondly, near Newtown and northward, they appear not to have been influenced in any way consistent with having to cross the postulated south-eastwardly facing slope on their way northward and north-westward ((Figure 20)b). It is therefore suggested that a Montgomery Trough, an elongated depression acting as a conduit for turbidity currents, did not exist in Ludlow times (Figure 19), (Figure 20).

Disturbed and slumped beds

Large parts of the formation, especially in the south-west of the district, suffered penecontemporaneous disturbance. This is revealed as linear crenulations forming reticulate patterns on bedding surfaces, contorted bedding (slumps) and resedimented debris-flow deposits. Units of disturbed strata no more than 2 m thick are widespread, but only some thicker units south of Newtown have been mapped. Commonly, there are associated bentonites, particularly at the slump bases, but they may also occur, disrupted, within the slumped mass. This coincidence of slump and bentonite seems too common to be fortuitous, and the bentonites are seen as facilitating movement which seismic shock might have instigated. It is sufficient here merely to list the two main types of penecontemporaneous disturbance of the formation.

One type, forming a bed of disturbed strata, a few metres above the Dingle Mudstone Member, can be mapped in places from Bronyvastre [SO 1202 9010] to the boundary of the district and beyond, but it has not been observed with certainty elsewhere. The maximum thickness of the bed is probably about 25 m in the west near Dolfor dying out eastwards near Bronyvastre. It is composed largely of structureless silty mudstone enclosing balls and contortions of fine-grained sandstone or siltstone. Little primary bedding remains; it is assumed that the sediment suffered reconstitution during flowage, but in-situ reconstitution while it was in a thixotropic state cannot be dismissed. It occupies a position low in the scanicus Biozone.

The second type occurs in the upper part of the formation, in the area east of Dolfor. Packets of disturbed strata are repetitiously interbedded with packets of normal strata. Most, if not all, of these lie within the leintwardinensis Biozone. The thicknesses of the packets are commonly between 0.5 and 12 m changing laterally, over many tens of metres, mainly by small amounts, but possibly enough for some packets to merge.

The degree of disturbance of strata within these packets varies greatly, from mere discordance of dips to contorted folds. It seems unlikely that much lateral ‘downslope’ movement was involved during the disturbance of these beds. Even beds which apparently are undisturbed may display on their surfaces a reticulate pattern of intersecting linear crenulations. Three sets are present at Ring Hole (p.100) where they intersect at angles of 60°, 55° and 65°. These bedding surface crenulations are in fact penetrative microfolds which form steeply dipping ‘kink bands’; they may be akin to the shear zones of Woodcock (1976a, fig. 10b). Such crenulations are characteristic of the beds immediately underlying a ‘disturbed’ packet thus occurring within the transition between disturbed and undisturbed beds. The tops of the disturbed units are usually sharp.

Dingle Mudstone Member

This member forms the lowest part of the formation in the area around and to the north of Newtown. It is thickest around Newtown, where it approaches 150 m, thinning rapidly to disappear east of Kerry and south of Dolfor. The base, as seen in the stream [SO 1488 9240], appears to be a rapid transition up from the Gyfenni Wood Shale Member [of the Nantglyn Flags] which is dominated by laminated hemipelagite. The top of the member is defined by the abrupt appearance of thick sandstones, which elsewhere mark the base of the Bailey Hill Formation and it is noteworthy that the Brynrorin Conglomerate remains associated with the base of those thick sandstones.

The Dingle Mudstone Member consists of grey homogenous silty mudstone, in beds 20 to 30 mm thick, and beds and lenses of fine-grained sandstone or siltstone, 20 to 80 mm thick (Cave et al., 1993, fig. 9). The sandstone shows plane parallel and cross-lamination; some thick beds (up to about 0.20 m) show convolute lamination. Laminated hemipelagite is sparse. The lithology is similar to the sandstone/silty mudstone couplets of the Bailey Hill Formation, but relative proportions differ and thicknesses are smaller.

The type section is in The Dingle below the waterfall which marks the base [SO 1076 8974], of Earp’s Wilsonia wilsoni Grits. The presence of the Brynrorin Conglomerate is indicated only by loose blocks, but there is a 2 m-thick slump sheet at about the same horizon. Other exposures occur in the stream bed [SO 1485 9208]; in the track, [SO 1411 9539] near Brynrorin and the forest road [SO 1642 9295].

Brynrorin Conglomerate

Although not a mappable unit, the Brynrorin Conglomerate is sporadic but of wide distribution in eastern mid-Wales at the sharp base of the thicker sandstones in the Bailey Hill Formation (Bailey, 1964, p.11). It constitutes something of an enigma for the clasts are wholly synbasinal, making the conglomerate comparable with those present in the turbiditic Wenlock sequence nearby. There is no record of a younger Ludlovian conglomerate other than intraformational, fragmented hard-grounds of early Ludfordian platform sequences (Cave and White, 1971; Cherns, 1980).

The conglomerate consists of well-rounded, ovoid clasts of fine-grained limestone, many identifiable as septarian concretions common in the underlying Nantglyn Flags. Other clasts are rounded to subangular and include calcareous fine-grained sandstone and dark grey laminar hemipelagite (Cave et al., 1993, fig. 11). Clasts are set in a calcareous mudstone matrix containing abundant small, round, black, phosphatised pellets or rounded clasts and coarse shell-detritus.

A section [SO 1434 9552] to [SO 1407 9532] near Brynrorin is shown in (Figure 30).

Thickness m
Bailey Hill Formation
Sandstone, calcareous, fine-grained, flaggy; parallel, wavy and convolute laminated, interbedded with buff-coloured silty mudstone seen 4–5
Brynrorin Conglomerate
Conglomerate, clasts up to 200 mm, mostly ovoid concreteous of limestone. Matrix consists of brown silty mudstone with calcite-lined vugs of white clay (bentonite), some shell detritus 0.5
Siltstone and silty mudstone, partially enveloping clasts which protrude up from the bed below 0.14 Conglomerate, as above, clasts up to 100 mm 0.25
Mudstone and siltstone interlaminated; brown weathered calcareous, fissile laminae present, possibly laminated hemipelagite 0.30
Conglomerate, clasts up to 180 mm, include argillaceous limestone concretions. Some argillaceous shell-detrital limestone with black phosphatic granules adhers to their surfaces suggesting that they have been reworked from the basal bed of the member. Matrix consists of silty mudstone with pockets of greenish yellow bentonite c.0.55
Silty mudstone, rather flaggy in an uneven bed. Clasts from underlying bed protrude into the base 0.1–0.2
Conglomerate as below, but clasts (limestone concretions) more abundant, up to 250 mm, and less shell-detrital matrix with phosphatic granules 1.2
Limestone, mainly shell-detrital with argillaceous micritic matrix which also encloses black phosphatic granules and flat clasts of mudstone, up to 80 mm. Oblate ovoid limestone concretions up to 200 mm, are present as clasts near the base, some partially enclosed by yellow-brown bentonitic clay. Shell detritus includes coral, leptaenid rhynchonellid, orthocone and trilobite fragments 0.55
Dingle Mudstone Member
Mudstone, homogeneous, buff to olive grey. Sparse laminated hemipelagite. Interbedded siltstones, as laminae or lenses, up to 30 mm. Bottom structures include slender trails and burrows; prod and bounce marks indicate palaeocurrents from the south and south-east respectively. Fine shell detritus is concentrated at the base of a few beds. Partings of pale grey clay (bentonite) up to 3 mm thick occur below many of the siltstones; a thicker bed 6 m from the top is divided by 0.2 m of soft brown mudstone 20.0
Mudstone, grey, silty and homogeneous, interbedded with mainly thin siltstone, up to 0.06 m, a few up to 0.12 m, some incorporate fine shell debris. Much fissile laminated hemipelagite 10.0
Nantglyn Flags, Gyfenni Wood Shales Member
Hemipelagite, laminated, very silty, fissile and brown weathered, thinly interbedded with grey, homogeneous, silty mudstone and a few thin siltstone layers seen 8.0

A noteworthy aspect of the section is the numerous thin films of bentonite in the Dingle Mudstone Member. These presumably occur everywhere but are not visible in weathered sections. The inclusion of irregular pods and wisps of bentonite within the Brynrorin Conglomerate must imply a measure of transportation by mass-flow. Being composed of several separated beds, mass-flow must have recurred several times in a relatively short period. The conglomerate also recorded the first major incursion of concentrated shallow-marine shelly debris and thus a brief but energetic event.

Cwm-yr-hob Member

The Cwm-yr-hob Member occupies the top part of the formation in its western outcrop. Named after the old farmhouse where the type section, including the basal transition, (p.101) is well exposed in a track [SO 1547 7991] to [SO 1550 8002], it is essentially the same as Earp’s (1938) Monograptus leintwardinensis Shales and Allender’s (1958) Cwm Mawr Beds. Transitional boundaries at both bottom and top make mapping subjective and thickness calculations imprecise. The member is thickest in the outcrop south of Kerry, up to 220 m thick; it thins eastward and southwards. In the west the member was much affected by penecontemporaneous movements including slumping (Earp, 1938; 1940). It comprises very thinly bedded alternations (couplets) of fine sandstone, or siltstone, and mudstone, commonly concentrated into thin bundles a centimetre or so thick interbedded with laminated (anoxic) hemipelagite (p.57). The thin sandstones are planar to lenticular, with low-angle ripples in places and the rock is shaly and grey-brown in colour. In the upper part of the member intervals of bioturbated strata form the transition to the overlying Cefn Einion Formation.

Details

South and east of Dolfor–Kerry–Sarn

The formation is exposed in quarries and cuttings along Gyfenni Wood Forestry road [SO 2138 9015] (Figure 27).

The base of the formation is exposed in Aston Dingle, resting upon Aston Mudstone (Figure 28). Westwards to Great Argoed [SO 254 909], the lowest part of the formation comprises grey flaggy and mainly parallel-laminated siltstone interbedded with calcareous sandstone up to 0.30 m thick. The sandstones normally show a sharp base and top, and there is little evidence of erosion at the base. The thicker beds commonly show convolute lamination, some with parallel lamination in the upper and lower parts of the bed. There is local small-scale slumping. The most complete section is in Cwm Cae [SO 2789 9133] to [SO 2800 9111] where about 125 m is exposed in and adjacent to the stream.

In Hopton Dingle, the Brynrorin Conglomerate is exposed in the right bank below the waterfall [SO 2298 9048] (p.85). It is up to 0.40 m and is composed of ovoid clasts, less than 0.20 m in diameter. The clasts include calcareous concretions and anoxic laminated hemipelagite. The latter can only have been derived from the Oakeley Mynd Formation some miles to the east, or from the underlying Gyfenni Wood Shales.

In a section at the side of a farm track the Brynrorin Conglomerate again marks the base of the formation [SO 1820 8971]. It is at least 0.50 m thick in places, passing laterally from parallel-laminated sandstone into coarse conglomerate with rounded clasts of siltstone and larger (100 mm) calcareous concretions. They are arranged randomly in a sandy, shell-detrital limestone and pebbly matrix. Boulders of an identical rock occur in the stream [SO 1948 8975] at Cwm Earl.

Around Kempton, there are a number of quarries which yield diagnostic faunas. One quarry [SO 3618 8324], 550 m north-north-east of Kempton Farm, exposes massively bedded bioturbated and faintly laminated siltstones with well-developed slump folds. There are a few beds of limestone nodules 0.20 to 0.50 m across; some are richly fossiliferous. The fauna includes brachiopods Sphaerirhynchia wilsoni, Striispirifer plicatellus, Ptychopteria sp. and graptolites Saetograptus (Colonograptus) varians indicative of the nilssoni or scanicus zones, at the base of the formation.

An old quarry [SO 3537 8280], 520 m west of Kempton Farm, shows flaggy bedded siltstone with some more massive beds, up to 0.40 m thick, with scattered bivalves. There are some thin beds of ochreous-weathering, laminated, fine-grained, sandy siltstone with lenses of winnowed fossils at their tops, and some beds of fine-grained siltstone with graptolites. The fauna includes Dayia navicula, Leptaena depressa, Microsphaeridiorhynchus nucula, Goniophora cymbaeformis, Ptychoptera sp., Kionoceras cf. angulatum, Encrinurus rosensteinae and Saetograptus (S.) leintwardinensis leintwardinensis.

A quarry [SO 3547 8193], 75 m south-east of New House Farm, exposes sandy siltstone, siltstone and silty mudstone with coquinas of brachiopods in beds over 50 mm thick. A slumped horizon in these beds is overlain by thinly bedded (less than 10 mm) siltstone with graptolites and comminuted shell fragments. The fauna includes Jonesea grayi, D. navicula, M. nucula, Orbiculoidea rugata, Shagamella minor, cf. S. wilsoni, ‘Lituites’ ibex, ‘Orthoceras’ sp. and S. (S.) l. leintwardinensis. The fauna of both these quarries indicates the leintwardinensis Zone near the top of the formation, possibly equivalent to the Cwm-yr-hob Member in the west of the district.

Between Aston Hill [SO 299 901] and Pantglas [SO 247 896], a number of sections expose the upper part of the formation. These show grey, flaggy, mainly parallel-laminated mudstone and siltstone with some interbedded calcareous sandstones up to 0.10 m thick. The sandstones are fine grained and are structureless or show parallel lamination; a few exhibit convolute lamination and there is small-scale slumping in places. The best sections are in a series of small quarries alongside a track [SO 2739 9054] to [SO 2766 9032], east-south-east of Warbury. The largest of these quarries [SO 2754 9049] shows 14 m of beds, and in the other quarries [SO 2760 9041], and [SO 2815 9018] at Upper Rolva the sandstones pinch and swell along their length, but are otherwise structureless.

Sandstones up to 0.10 m thick, constitute about 50 per cent of the rock exposed in a silage pit [SO 2512 8862] near Lower Dolfawr and along a track [SO 2586 8761] to [SO 2577 8773] in Steven’s Dingle. At the latter locality, shell-detrital lags include S. (S.) chimaera semispinosus and S. (S.) leintwardinensis incipiens of the incipiens, or tumescens Biozone. The sandstones are typical of the formation; they are well jointed, calcareous with soft porous brown-weathered exteriors and hard blue-grey cores. They have planar tops and bases and show wavy, cross-and convolute lamination. Some laminae are emphasised by opaque minerals (E60138). Quartz grains are angular, but well sorted and less than 0.07 mm in diameter. The matrix is ‘dirty’ and chlorite–illite rich, in which white mica is aligned parallel with bedding.

A structurally complex section is exposed along 600 m of forestry road [SO 2553 8694] to [SO 2500 8665]. Lithologies are similar and fossils include D. navicula, M. nucula, S. wilsoni with north–south current-aligned saetograptids.

The narrow north–south outcrop of the Bailey Hill Formation between the Clun and Bettws-y-crwyn outliers lies within the Clun Forest Disturbance (Figure 1). The best section (Campbell, 1989) in this part of the outcrop is along a recently excavated track-side [SO 2591 8307] to [SO 2584 8302], north-east of Wellfield. Because of variable tight to open folding, and steep normal and reverse faults of uncertain magnitude, the thickness of the exposed sequence can only be estimated, at about 15 m. It comprises flaggy grey to yellowish brown siltstone and sandy siltstone with subordinate fine-grained sandstone, in beds 10 to 50 mm thick. Near the base, there is a slumped bed up to 4 m thick. At two horizons, about 5 m above the base, abundant ‘Orthoceras’ sp and rare brachiopods are present in bedding planes. At about the same horizon some carbonate concretions are developed subparallel with bedding. Beds are 10 to 300 mm thick and bedding structures include climbing ripple sets. About 10 m above the base of the section, the strata are very evenly flaggy bedded, with some beds showing load structures. Some interbedded, darker grey, micaceous siltstones yield abundant Saetograptus leintwardinensis aff. incipiens and S. (S.) cf. l. leintwardinensis, indicative of the leintwardinensis Biozone of the upper part of the formation.

In the southern part of Hopton Dingle, a quarry [SO 2289 9005] exposes 10 m of fine-grained sandstone interbedded with non-laminated rubbly silty mudstone, typical of the lower part of the formation. The sandstones are 0.30 m thick and reveal convolute and parallel laminations; some contain lenticular, basal, shell-detrical lags, up to 30 mm thick. Flute casts (Plate 9), (Plate 10) indicate that current flow was from 160°, 235°, 195°, 175°, 185° and 198°. Fossils include decalcified bryozoa, chonetids, D. navicula, Isorthis sp., Leptaena sp., Lingula sp. and S. incipiens. At about the same horizon nearby [SO 2262 9024] P. minor tumescens and P. cf. minor minor are present, and a little lower [SO 2298 9025] Aegiria sp., Isorthis sp.,Kirkidium sp., P. cf. minor tumescens and P. cf. minor minor occur.

Similar beds are well exposed in quarries near City [SO 2067 8985], [SO 2001 8945] where currents from the south-east are indicated and a bentonite, a few centimetres thick, is present. Bentonite is also exposed in another quarry nearby [SO 2129 9000].

A quarry [SO 1737 8874], beside the forestry road in Fronderw Wood, exposes many metres of beds composed of equal proportions of poorly laminated to non-laminated silty mudstone and well-laminated, fine-grained, calcareous sandstone. Sandstone beds are mostly 40 to 100 mm thick, but some are considerably thinner (10 mm) and are no more than starved ripples. Some beds pinch and swell irregularly and erosion surfaces in wavy and parallel lamination are common. Palaeocurrent directions are consistently from the south-east. In another quarry [SO 1753 8866], 200 m to the south-east, sandstones form about 30 per cent of the rock. flute casts and groove casts indicate currents from the south-east, consistent with the direction of derivation indicated by cross-lamination.

Many exposures occur in Drefor Dingle. The lowest, near the base of the formation, is beside the brook above Drefor Farm where slumped beds occur [SO 1697 8896]. Beds of slumped, homogenised, silty mudstone, up to 5 m thick, are common in the Dolfor area. They usually have sharp contacts with evenly bedded strata above and below. One, 2 m-thick bed can be seen in a roadside quarry [SO 1241 8690] and another is exposed in a small gorge [SO 1197 8691] in the River Mule.

A roadside quarry [SO 1070 8550], 1.5 km south of Dolfor, exposes very flaggy, calcareous, mostly micaceous, parallel-laminated sandstone up to 80 mm thick. Some are cross-bedded near their bases and many contain current orientated graptolites (found loose); others contain shell debris too. Some thin fine-grained sandstones show a ripple-like surface trending 065° to 072°, and steep to the north-west. These coincide with loading structures at the base and small-scale slump disturbance, directed towards 332° on some beds. Several sandstones contain current-orientated orthocones with apices pointing to 280°, possibly indicating currents from 100°. An interbedded slump bed, 0.30 m thick, consists of homogenised silty mudstone with inclusions of sharply folded fine-grained sandstone. Shell debris including Dayia sp., Microsphaeridiorhynchus sp.,Cardiola sp., orthocones and graptolites are concentrated near the top of the bed. Like many such slumps, this bed is overlain by a sandstone which ‘smooths out’ the uneven surface of the slump. The fauna at this locality includes A. reticularis, D. navicula, M. nucula, Ptychopteria?, ‘Orthoceras’ sp., P. cf. minor tumescensand S. (S.) leintwardinensis incipiens of the scanicus or more likely tumescens Biozone.

A roadside quarry [SO 0933 8774], about 1.5 km north-west of Dolfor, exposes 3.5 m of slumped splintery silty mudstone and some disturbed bedding. The top is not seen and the base is non-erosive. It rests on 4.2 m of flaggy sandstones, which include two thicker beds of sandstone, 0.25 m and 0.30 m thick and thin interbeds of laminated hemipelagite.

Beds near the base of the formation are preserved in the core of a syncline which can be extrapolated southward into a series of sharp folds through Glôg [SO 090 852]. The lowest beds are exposed in a roadside (A483T) cutting [SO 0918 8585], 650 m north-east of Glôg Farm [SO 0873 8542]. These consist of 76 m of graptolitic shaly mudstone, which is dominantly laminated hemipelagite, and siltstone; fine-grained sandstones are sparse, in beds up to 30 mm. The fauna includes Monoclimacis micropoma (or Pristiograptus sp.), P. cf. vicinus, S. (S.) cf. chimaera and S. (C.) colonus?, indicative of high nilssoni Biozone. These beds are considered to belong to the top of the Gyfenni Wood Shales. At the west end of the cutting [SO 0913 8576], they are overlain by 2 m of rather massive, grey, silty mudstone in beds up to 0.35 m thick, and with partings of laminated hemipelagite but with no sandstones. They pass up into 15 m of homogeneous silty mudstone, which is not typical of either the main bulk of the Bailey Hill Formation or its Dingle Mudstone Member.

Nearby, a hillside track traverses a westward verging sharp anticline. The base of the formation is exposed in both limbs [SO 0978 8581] and [SO 0964 8574] and the Gyfenni Wood Shales, yielding an upper nilssoni fauna, in the core. Here too the basal strata are dominantly homogenous silty mudstone, with thin ‘ribbons’ of fine sandstone. They are overlain by over 20 m of beds, more typical of the Bailey Hill Formation, consisting of silty mudstone with sandstones up to 60 mm thick, but also beds of homogeneous mudstone up to 3 m thick.

Near the base of the formation, in a nearby quarry [SO 0889 8574], at Bwlch Gate and to the east [SO 0905 8574], 20 m of silty mudstone, is exposed, in beds up to 120 mm thick, and contain silty sandstone and laminated hemipelagites, up to 4 mm thick.

Slightly higher, a quarry [SO 0913 8617] exposes a fault disrupted sequence 53 m thick. Couplets of silty mudstone and thin, fine-grained sandstone present an aspect between that of the main part of the formation and that of the Dingle Mudstone Member. A fauna of nilssoni–scanicus age includes P. cf. vicinus, S. chimaera chimaera, S. chimaera cf. salweyi, Isorthis cf. clivosa and Nanospira sp.

The Brynrorin Conglomerate appears not to be present in this area, nor to the south where the main road crosses the base of the formation on both limbs [SO 0885 8473] and [SO 0836 0489] of a large north–south-trending syncline just west of the district. A large quarry [SO 0876 8477], situated on the axis of the syncline, exposes 36 m of evenly bedded, grey, homogeneous silty mudstone, in beds up to 80 mm thick, and fine-grained sandstones. The latter show wavy and convolute lamination and are up to 100 mm thick. A 300 mm-thick bed of homogeneous silty mudstone which lies about 16 m below the top of the succession contains a layer of pseudo-nodules of fine-grained sandstone each showing concentrically arranged laminae and caused by reversed density and viscosity differences in rapidly accumulating deposits.

Down the succession along the roadside south-eastwards towards the Devil’s Elbow bend, the beds pass fairly abruptly into shaly mudstone with much silty laminated hemipelagite of Gyfenni Wood Shales aspect and high-nilssoni to low-scanicus biozonal (Ludlow) graptolites including: Mcl. micropoma, P. cf. dubius, P. cf. vicinus, S. colonus colonus, S. incipiens and also C. interrupta. These beds can be assigned to the Gyfenni Wood Shales. Descending the succession along the roadside northwards, in the opposite limb of the syncline, the Bailey Hill Formation passes into similar fissile silty mudstones with much hemipelagite, weathering with rather rusty stains and containing several thin layers of bentonite. However, after repeated collecting they yielded a ludensis biozonal (Wenlock) fauna of M. ludensis, ?P. dubius, P. jaegeri and C. interrupta. This biostratigraphical imbalance in the synclinal limbs is a puzzle, but equally puzzling is the similar ludensis fauna derived from shales in a landslip [SO 0911 8465] on the south side of Devil’s Elbow and along the roadside [SO 0890 8453], although a fault is invoked to explain the latter. Barely 400 m to the south-west a highnilssoni–scanicus fauna occurs in 14 m of typical ribbon-banded Nantglyn Flags [SO 0819 8463]. Although the area is undoubtedly structurally complex there appear to be facies changes too. The area lies on the Ystwyth Fault lineament.

Basal parts of the formation, containing sandstones up to 130mm thick in some places, rest fairly abruptly upon fissile largely laminated hemipelagic shales in several streams [SO 0997 8468], [SO 0955 8331], [SO 0928 8219], [SO 0966 8059] and [SO 0933 7963]. The Brynrorin Conglomerate was not seen.

In the outcrop along the western side of the Bettws-y-crwyn outlier, there are a number of sections in the formation. A quarry [SO 1147 8069] north of Gwenlas Farm exposes strata in the lower part of the formation, comprising grey sandstones interbedded with about 30 per cent of turbidite ‘sandstone’ (very fine-grained sandstone/siltstone), in beds up to 100 mm thick. Hemipelagite beds, up to 20 mm thick, are also common. The best section in the upper part of the formation is at the Ring Hole (see below) in the headwaters of the River Teme. Farther south, a track-side [SO 1560 7901] to [SO 1570 7913] north-east of Lower House Farm shows grey silty mudstone in beds up to 0.6 m thick. This is massive with a conchoidal fracture, or locally with faint parallel or convolute lamination, and is interbedded with a thinner bedded lithology with diffuse sandstone/mudstone interlaminae and diffuse hemipelagite.

Slump facies

Disturbed beds (slumps) have been detailed in several sections, but the following are illustrative of the two types (p.95).

Type 1. A cutting [SO 1163 9071] south of Middle Brimmon, through a ridge, exposes 10 m of homogenised silty mudstone with inclusions of contorted sandstone. The slump sheet is at least 15 m thick here, but it can be mapped to near Pen-y-banc where some 25 m are exposed in a quarry [SO 0949 8954] dipping steeply westward.

Type 2. This type of slump is common south-east of Dolfor, for example in the stream gorge [SO 1050 8441] and quarry [SO 1096 8250], but one of the best exposures is at the bottom of the Ring Hole [SO 1205 8375] on the north side of the B4355 road. The strata lie within the upper part of the Bailey Hill Formation and the ‘sandstones’ are thin and are probably of silt grade (Earp, 1938; Woodcock, 1973). Packets of evenly bedded strata make up about 25 per cent of the sequence (Woodcock, 1973) and they are well exposed on the northern side of the gorge where dip barely exceeds 10°. The interbedded packets of disturbed strata are generally thicker ranging from 1 to 12 m. In some, the beds are only very gently disturbed and slumping is revealed as a mere discordance with undisturbed beds. A packet consisting of 4 m of disturbed beds overlain by 2 m of evenly bedded strata can be traced for many metres in the side of the gorge, but ultimately it is possible that slumped strata pass laterally into disturbed strata. At stream level, three sets of narrow kink bands can be seen in some of the evenly bedded strata. These reflect stresses generated during the slumping. Such crenulations occur widely, associated with slumped beds, for example in the east of the district in quarried beds [SO 3409 8615] high in the formation near Walcot Pool.

Cwm-yr-hob Member

Transition upwards from the main body of the Bailey Hill Formation into the Cwm-yr-hob Member is exposed in a quarry [SO 1783 8804] near Cwm-golog; the beds dip 35°S. Fissile, buff-grey, silty mudstone and interbedded dark brown graptolitic laminated hemipelagite (more than 50 per cent) form the basal 2 m of the member, and include some rotten, brown sandstone up to 30 mm thick. These beds overlie a 0.17 m-thick sandstone, typical of the main part of the formation below, which rests on about 2 m of flaggy silty mudstone with thin laminated hemipelagic partings.

A stream nearby [SO 1824 8799] exposes several metres of fissile, hemipelagic shales with scattered sandstones, up to 40 mm thick, close to the top of the member. A few metres higher, at the bend in the farm track [SO 1821 8798], some 8 m of very burrowed fissile siltstone, with thin beds of fine-grained sandstone, are exposed. This is the basal part of the Knucklas Castle Formation although there is probably little sedimentary difference between it and beds of the Cym-yr-hob Member in the stream below, other than its bioturbation and higher carbonate content resulting from an oxic depositional environment.

These new sections reveal that the Cwm-yr-hob Member extends farther east than is indicated on the 1:50 000 Series sheet.

The type section for the member is near the southern end of its outcrop in the track [SO 1547 7991] to [SO 1550 8002] north-east of Cwm-yr-hob farmhouse. Here the member is about 100 m thick. The underlying Bailey Hill Formation is dominated by homogeneous silty mudstone in beds 0.10 to 0.20 m thick. There is a rapid upward transition into the Cwm-yr-hob Member which comprises interlaminated fine-grained sandstone, siltstone and mudstone with many beds of laminated (anoxic) hemipelagite up to 10 mm thick and a few bioturbated intervals. In the upper part of the member the frequency of the bioturbated beds increases upwards. There is a gradational transition into the overlying Knucklas Castle Formation and the boundary is taken to be where the bioturbation becomes dominant. The section yields graptolites Saetograptus (S.) cf. leintwardinensis leintwardinensis and S. leintwardinensis s.l. at [SO 1552 7994].

At Lyrchyn quarry [SO 1213 8480], sporadic beds of fine-grained, parallel laminated sandstone comprise about 5 per cent of the sequence. In a stream section [SO 1314 8238] south-west of Cwm-gwyn Hall, there is a transition into the Knucklas Castle Formation.

North and west of Dolfor–Kerry–Sarn

Dingle Mudstone Member

Over 100 m of the Dingle Mudstone Member crops out in The Dingle although the base is missing. The lowest exposures occur in the lane bank [SO 1064 9051], [SO 1067 9050]. About 30 m higher, behind Cil-haul [SO 1073 9021], there are 26 m of siltstone (or silty mudstone) which is homogeneous, mid-grey and in beds 10 to 80 mm thick. They contain a scatter of black carbonaceous and graptolitic fragments and are interbedded pervasively with fissile, pale grey, fine-grained, very micaceous sandstone, in beds up to 20 mm thick and showing wavy parallel lamination. A 0.20 m-thick homogeneous silty mudstone with scattered siltstone clasts, 8 m from the base of the section, is probably a mass-flow deposit. There is a single ovoid concretion about 0.08 m from the base.

Some 35 m higher, 20 to 23 m of similar beds are exposed [SO 1008 9075] to [SO 1001 9676] where sandstones, up to 30 mm thick, are common. The fissility of the mudstone is caused by thin silty laminae, not hemipelagites, which are sparse and thin. These beds, as in other exposures of the member, are not regular rhythmites like those characteristic of the Nantglyn Flags. Some 10 m higher, a lane-side section [SO 1077 8997] exposes another 14 m of the beds including a thin decalcified sandstone 3 m from the top which yields shell detritus, and S. (C.) colonus colonus and S. (S.) chimaera cf. chimaera indicating topmost nilssoni Biozone or the base of the scanicus Biozone. The graptolites have a preferred orientation between 090° and 100°. An exposure [SO 1046 8979] nearby, at about the same horizon, yields the scanicus Biozone fossils: cf. Cardiola sp., ‘Orthoceras’ sp.,Lobograptus scanicus and Pristiograptus. cf. vicinus.

From the farm [SO 1071 8996] to near [SO 1076 8976] the base of the waterfall about 24 m of thinly bedded, dominantly homogeneous, silty, micaceous mudstone, with a faint parallel-lamination and shaly fissility are exposed; interbedded thin, fine-grained sandstone/siltstone beds, 30 to 40 mm thick, are exposed in the stream and banks of the dingle. Several, probably many, thin (up to 5 mm), pale grey bentonites occur and in some places thin, laminated hemipelagites are fairly common. Towards the top, the mudstone becomes more blocky and there is a bed of shelly limestone in the stream which, together with loose blocks of the Brynrorin Conglomerate from the steep side of the dingle, mark the base of the more thickly bedded, fine-grained, sandstone sequence seen in the waterfall. A quarry [SO 1081 8978], on the east side of the dingle, yields S. (C.) colonus colonus and P. cf. vicinus from the upper part of The Dingle sequence.

A quarry [SO 1372 9040], 1 km west of Kerry, exposes 4 m of medium grey homogeneous shaly mudstone, well bedded with silty laminae. Thin, very fine-grained sandstones (or siltstones) are present. Some silty mudstones show cross-bedding; weathering opens up a fissility which follows the silty laminae and the parallel laminae of some very dilute (inflated with non-pelagic mud), but common, hemipelagites. The lithology accords with its position at the eastern extremity of the Dingle Mudstone Member. The fauna indicates a position high in the nilssoni Biozone, or more probably, the scanicus Biozone and includes Cardiola sp., Orthoceras sp., Lobograptus crinitus, P. vicinus (common) and a retiolitid. A new quarry [SO 1308 9062] nearby, almost devoid of hemipelagites and a little higher within the member, exposes 8 m of silty mudstone, commonly with thin sandstones at their bases. Most sandstones are very fine grained, and about 90 mm in thickness, with cross-bedding and wavy lamination; one contains a shell-detrital base several centimetres thick. These sandstones serve to reaffirm the close genetic relationship of the member with the beds above. Flute casts are present on several beds revealing palaeocurrents from 130°, 155° and 163° [SO 1262 9037].

Good exposure occurs beside the forestry track [SO 1652 9288] in the Mule gorge (Figure 31). The presence of the Brynrorin Conglomerate is indicated by debris found at the base of the thick sandstone sequence above the Dingle Mudstone Member. An elongate small inlier, not shown on the 1:50 000 Series sheet, lies across the Mule Gorge [SO 164 937] to [SO 169 941].

In the bank of a track at Brynrorin [SO 1434 9552] to [SO 1407 9532], the complete thickness (about 30 m) of the Dingle Mudstone Member is exposed (Figure 30). The section is noteworthy for the number of thin (up to 5 mm) bentonite beds, detectable when the section was fresh.

Above the Dingle Mudstone Member

The Dingle, 1.8 km south-east of Newtown, exposes the lowest 30 m or so of this part of the formation in two waterfalls. The larger waterfall [SO 1077 8972] shows the following sequence:

Thickness m
In waterfall
Sandstone, calcareous, parallel and convolute laminated, beds up to 0.25 m thick; subordinate, interbedded, homogeneous, silty mudstone 15–20
Below waterfall
Mudstone, silty, slumped 1
Sandstone and silty mudstone as in waterfall 3
Mudstone, silty slumped 1–2
Sandstone and silty mudstone as in waterfall 4
Mudstone, shaly, homogeneous, with sparse sandstones up to 80 mm thick c.2
Dingle Mudstone Member
Mudstone, shaly, homogenous with thin, fissile, fine-grained sandstone showing wavy lamination; some thin laminated hemipelagite

The smaller waterfall [SO 1080 8975], which descends the north side of the gorge downstream of the larger waterfall, provides a good exposure of about 20 m of sandstone and silty mudstone, partly equivalent and partly higher than the section exposed in the larger waterfall. Sandstone, 3 to 4 m thick, forms a vertical face about two-thirds the way down the section; this sandstone also forms the top of the larger waterfall. The sandstone occurs in beds up to 0.40 m thick and shows convolute bedding. The base rests on a pale fawn, soft mudstone (bentonite) 20 to 30 mm thick. A feature-forming sequence with sandstones overlies the Dingle Mudstone [SO 1433 9077] east of Weston. The sandstones, up to 0.30 m thick, show parallel, wavy and convolute lamination, and north of here to Lower Bryn-mawr [SO 1481 9157] there are many stream exposures of this part of the formation with sandstones up to 0.20 m. The base of the sandstone sequence, with the Brynrorin Conglomerate, is exposed in the stream [SO 1483 9207].

Further exposures occur to the east near Cefn-caled [SO 1522 9156], [SO 1519 9166], [SO 1533 9180] and near Maen Hwyd [SO 1556 9178]. The base of the sandstone sequence, with sandstones up to 0.20 m, thick rests upon the Brynrorin Conglomerate in the stream [SO 1609 9183].

The formation is exposed unusually well in the Mule Gorge. The base of the sandstone sequence occurs at the roadside [SO 1637 9293] and in the forestry road section [SO 1643 9296]. Higher strata, with thinner sandstones occur along this road to the north [SO 1675 9352], in places dipping less shallowly than the slope beneath. The consequences of exposing the strata in the road and bank and loading the top of the slope with rock debris has been landslipping rendering the road useless.

A 40 m-cliff section [SO 1673 9371] with sandstones up to 0.17 m and silty mudstone interbeds gives indication of palaeocurrents from the east and south-east (cross-bedding) and from 150° and 160° (flute casts). On the opposite side of the gorge, exposed in the core of an anticline 23 m of sandstone rest on Dingle Mudstone [SO 1650 9373] (p.101); nearby flute casts indicate palaeocurrents from 134° and 110°. The contact is also exposed farther east in the stream [SO 1823 9292] near Lower Cwmgwnen.

A rock face behind The Den [SO 2173 9301] yields a fauna including S. (S.) chimaera salweyi (common) and D. navicula. An old quarry [SO 2133 9308] exposes 15 m of calcareous sandstones in beds up to 0.24 m thick, many having shell detritus at the base. Cross-bedding indicates currents were from the south-east. A temporary quarry [SO 2174 9332] exposed 11 m of argillaceous siltstone with some sandstones which are parallel laminated, thin, fine grained and with shallow scour-fill bases. Silty laminated hemipelagites are common and a bentonite (50 mm thick) is present. Fossils include S. (S.) leintwardinensis incipiens and S. varians.

Near Montgomery, a dingle exposes easterly dipping beds including sandstones up to 0.20 m thick. The sequence rises to a quarry [SO 2208 9549] where a sheet, at least 3 m thick, of slumped silty mudstone occurs. It contains rafts and internally contorted pillows of fine-grained sandstone up to 0.55 m across. The slump sheet rests sharply and concordantly upon evenly bedded strata typical of the formation with sandstones. Fossils include B. bohemicus (common), S. (C.) cf. varians, Shagamella minor and C. interrupta.

Typical flaggy sandstones and silty mudstones and a slump [SO 2208 9643] are exposed along the lane leading up to the castle in Montgomery. These exposures confine the position of the Montgomery Fault to a position just east of the lane. Four metres of sandstone, in beds up to 0.15 m thick, are exposed [SO 2140 9655] dipping 62° ESE and passing north-west across Ffridd Faldwyn to the main road [SO 2199 9734] where they are vertical.

South of Llandyssil, a quarry [SO 1953 9494] exposes 20 m of sandstones in beds mostly about 0.10 m thick, some up to 0.30 m thick and one 0.93 m thick, 5 m from the base. In the basal 5 m, sandstones are thinner, up to 30 mm. Flute casts indicate currents from 110° and about 245°. Another quarry [SO 1824 9556] exposes 10 m of flaggy, fine-grained sandstones less than 100 mm thick, but more commonly about 50 mm, with silty mudstone interbeds. Flute and groove casts on several bases indicate currents from 140° to 150°.

The remaining outcrops comprise two outliers to the west of the River Severn. In one, at Brynrorin (Figure 30) the basal few metres of sandstones are exposed above the Brynrorin Conglomerate [SO 1415 9540]. In the other outlier, some 68 m of beds are exposed in a stream [SO 1470 9657] (base) to [SO 1468 9665]. The basal 10 m, overlying the Dingle Mudstone Member, form a packet of couplets in which sandstones are up to 0.15 m thick, each grading into overlying silty mudstone. Some basal lags of shell detritus, up to 0.10 m thick, are present. Above, the beds are similar but with a higher content of silty micaceous mudstone. Flute casts on different beds show current derivation from 135°, 155°, 152° and 143°. There is no Brynrorin Conglomerate here, but it is exposed e.g. [SO 1345 9624] in stream sections nearby. One metre of conglomerate composed of chaotic, rounded clasts, up to 0.30 m diameter, of fine-grained sandstone and calcareous concretions in a sand, granule and shell-detrital matrix rests on the Dingle Mudstone Member. To the north-east [SO 1505 9715], flaggy sandstone, up to 0.15 m thick, rests with uneven contact upon at least 4.5 m of similar conglomerate with very fine sandstone (40 to 60 mm thick) near the bottom. The base of the conglomerate is not visible.

Other sections of the sandstone sequence occur around The Hill [SO 1503 9697] and to the north-east [SO 1524 9719] (base).

Knucklas Castle Formation

The Knucklas Castle Formation was designed by Holland (1959) to accommodate beds which differ from his Bailey Hill Beds, largely by the absence of graptolites (Holland, 1959, p.495). Lithological changes both at the bottom and top of the formation are transitional through many metres of strata. Holland’s definition has been amended slightly, in order to match it to lithology rather than to the absence of graptolites. The base is now taken at the appearance of dominantly burrowed (oxic) strata. Thus, the thin transitional sequence, where oxic and anoxic strata are interbedded, is incorporated largely within the Knucklas Castle Formation. For reasons not understood graptolites do not persist above the basal few metres of this transition despite the presence of some beds of laminated hemipelagite suitable for their preservation.

Previous work in the west of the district by Earp (1938, 1940) established a bio-formation, the Dayia navicula Beds; the lower two-thirds of this formation is equivalent to the Knucklas Castle Formation; other nomenclature is shown on (Table 10). Evans (1957) working in the east of the district included these strata in his Upper Ludlovian Cviia beds.

The formation crops out on the major scarp slopes around the high ground of Clun Forest. In the north of the outcrop, around Cwm-golog and Pont-y-folog [SO 187 887], there is a marked step in the scarp formed by a unit (6 to 7 m thick) of closely spaced sandstones, in beds up to 20 mm thick.

The thickness of the formation shows a regular increase across the district from about 110 m in the east to 500 m in the south-west (Figure 1). The thickening affects the whole of the formation, so that even the transitional sequence at the base increases from a few metres in the east to between 100 m and 200 m in the west. Such a large difference in thickness over a short distance, and without major differences of facies must indicate greater subsidence in the west. There is no obvious diachroneity of the base, but there may be a little at the top.

The basal part of the formation is transitional from the shaly top part of the Cwm-yr-hob Member of the Bailey Hill Formation. The sandstone/silty mudstone couplets common in that formation, thin upwards and become sparser through the transition. In addition, anoxic hemipelagite, which dominates the Cwm-yr-hob Member is confined to discrete packets interbedded with burrowed siltstone laminite. Within the transitional beds, these packets become thinner and more widely spaced upwards. In the west, they are absent at 100 to 200 m above the base; in the east, they die out just above the base. Near the top, the formation is less fissile and more flaggy, due to the appearance of sandstone/silty mudstone event-couplets. These become dominant at the contact with the overlying Cefn Einion Formation.

The formation is characterised by a siltstone laminite facies (p.60) but there are sporadic, thin, fine-grained, sandstone/mudstone couplets, mainly near the base and the top, in which sandstones are up to 60 mm thick. Unweathered, the rock has a massive appearance; it weathers rapidly to become fissile and the closely spaced thin laminae, up to 3 mm thick, then become obvious. Above the transitional beds, all parts of the formation are burrowed. Bioturbation varies in intensity; in parts the laminae remain largely continuous and are clearly visible. Intercalated, intensively bioturbated horizons increase in thickness westward from 0.60 m to a few metres; the overall intensity of bioturbation increases slightly upwards.

Ovoid calcareous concretions are present sporadically through the formation, and there are examples near the top which are 1 m long and 0.20 m thick. In the type area of Castle Hill, Knucklas (Knighton, Sheet 181), 20 mm of purple beds were noted in the railway cutting [SO 248 744] (compare with Earp, 1938, p.134).

In the west, thin parallel-laminated fine-grained sandstones are more abundant (up to 30 mm thick), and are obtrusive, forming packets up to 5 m thick. This greater abundance of sandstones must bear a direct relationship with the westward increase in thickness of the formation.

Marine conditions were tranquil and similar to those which prevailed during the deposition of the upper Bailey Hill Formation, but were oxic. Carbonaceous sediment was not preserved, and was presumably consumed by the infauna which disrupted the silty lamination. The sparsity of epifaunal remains is not satisfactorily explained. Certainly, bottom currents were too weak to have transported shell debris from the platform to the east and it seems that either marine bottom-waters were so charged with suspended silt that they were inimical to filter feeders, or the silt substrate was alien to their life-style.

Biostratigraphy

Macrofossils are very sparse in the bulk of the formation, being slightly more common at the bottom and top where sandstone/silty mudstone couplets are more abundant and brachiopods and trilobites are present. Elsewhere, there are scattered occurrences of orthocones, bivalves, myodocopes including Bolbozoe cf. anomala and small black carbonaceous ‘scabs’ perhaps of algal origin. Indigenous brachiopods are almost absent; Dayia navicula is fairly common at a few horizons within the formation; Microsphaeridiorhynchus nucula and Salopina lunata, more characteristic of the overlying Cefn Einion Formation, are present near the top.

Graptolites are absent except in beds transitional from the underlying Bailey Hill Formation. One locality at the base [SO 1811 8806] has yielded Saetograptus (S.) leintwardinensis lentwardinensis. Another locality [SO 1558 8014] yields Bohemograptus bohemicus tenuis from anoxic strata within a few metres of the base of the formation, and represents the Bohemograptus proliferation Biozone. Although anoxic hemipelagite occurs higher, it appears to be barren of graptolites, so that it can be assumed that they were excluded from the local marine environment before conditions alien to the preservation of their remains prevailed.

Details

South-eastern corner of the district

Exposure is poor because the beds lie generally at the base of the scarp formed by the overlying Cefn Einion Formation. A roadside cutting [SO 3695 8124] and quarry [SO 3701 8119], south-east of Little Broughton, both show interbedded flaggy and thinly bedded siltstones. The lower part of the cutting shows recumbent slump folds and there is a major slump fold in the eastern part of the quarry.

Periphery of the Clun outlier

The best section through the formation in the district follows the forestry road on the southern side of Sunnyhill, westward from the car park [SO 3335 8396]. The latter was a small quarry where olive-coloured silty mudstone is exposed. Above this to the west, the basal part of the Knucklas Castle Formation is exposed sporadically showing very fissile olive siltstone which yields [SO 3308 8376] Leiopteria?, ‘Orthoceras’ sp., a ceratiocarid?, and Bolbozoe aff. divisa. A small quarry [SO 3300 8365] exposes about 3.5 m of very fissile and micaceous, burrowed, olive-coloured, silty mudstone and thin, parallel-laminated siltstone, in beds up to 20 mm thick, which weather brown. Along the next 450 m of track exposure of the upper part of the formation is better. The silty mudstone is more blocky and probably more burrowed in the approach to the top of the formation and the quarry [SO 3263 8339] which is in the overlying Cefn Einion Formation.

Periphery of Bettws-y-crwyn outlier

In the eastern part of this outcrop, the formation is very poorly exposed. An old quarry [SO 2305 7877], 750 m east of Cwm, displays about 6 m of parallel-laminated, flaggy siltstone, with a few sandstones up to 0.10 m thick. The sandstones are either structureless or show poor convolute-lamination. Dayia navicula is present.

A quarry [SO 2005 8737], north-west of Shadwell Hill, exposes beds near the top of the formation. They comprise laminated sandy siltstone with some fine-grained sandstone up to a few centimetres thick and sporadic trains of starved ripples. Decalcified concretions up to 1 m long by 0.02 m thick occur. Orthocones and serpulids are present.

North-east of Two Crosses, in a deep glacial channel [SO 2445 8726], beds of similar lithology occur towards the base of the formation and yield ‘Orthoceras’ sp., Anticalyptrea (Spirorbis) sp. and possible eurypterid remains.

A good, though discontinuous, section through the entire formation is present in Trefoel Brook [SO 1326 8009] to [SO 1476 8049]. Grey ‘siltstones’ near the top of the formation [SO 1460 8045] to [SO 1470 8047] vary from thin and flaggy (up to 30 mm) to thick bedded (30 to 100 mm); the bed thickness reflects the degree of bioturbation. The thinner beds are dominated by millimetre-scale sandstone/mudstone interlaminae, whereas in the thicker beds the lamination has largely been destroyed by bioturbation. Lower in the sequence [SO 1413 8030] to [SO 1451 8040], the lithologies are similar but there is an alternation, on a scale of one to several metres, of strata in which intensively bioturbated beds dominate and those in which laminated beds dominate. Near the base of the formation [SO 1339 8012] to [SO 1379 8021], the thinly bedded laminated facies becomes steadily dominant downwards and a few beds of hemipelagite occur.

The lower part of the formation can also be seen in a stream section [SO 1314 8237] to [SO 1356 8256], south-east of Cwm-gwyn Hill, where there are two one-metre-thick slumped beds at its eastern end.

A good section at Cwm Quarry [SO 1658 8712] exposes several metres of thinly bedded, flaggy, dull grey, calcareous, coarse siltstone and fine-grained sandstone from near the top of the formation. The beds are generally laminated throughout; some bed surfaces show bioturbation, including feeding trails.

A quarry [SO 1811 8806], 550 m east-north-east of Cwm-golog, shows beds transitional to the underlying Bailey Hill Formation. These comprise fissile parallel-laminated siltstone with interbedded, coarse, non-laminated, calcareous siltstone, and yield a fauna including Saetograptus (S.) leintwardinensis leintwardinesis, Jonesea grayi, Microsphaeridiorhynchus sp. and cf. Furcaster leptosoma. Nearby [SO 1791 8785] to [SO 1798 8791] good exposures rising up the formation southwards occur in a farm track. Bioturbated calcareous sandstones up to 50 mm occur and cross-bedding reveals that palaeocurrents came from the south. Sandstones here are markedly thicker and more abundant than they are in the east at Sunnyhill.

Cefn Einion Formation

The recent survey of the district adopted the name chosen by Allender (1958) (Table 10), and selected as the type section the exposure along the forest track on the south flank of Sunnyhill; the section is continuous through from the Knucklas Castle Formation (p.104). The formation corresponds largely with the Dalmanella lunata Beds of Earp (1938) and combines Holland’s (1959) Wern Quarry Beds and Llan-wen Hill Beds. It has a gradational base [SO 3272 8342], well exposed in the type section, where it is placed at the point where thin beds of fine-grained sandstone attain 10 per cent of the sequence.

The formation has been quarried locally for hardcore and other purposes. It forms many of the hills and much of the high ground in the southern part of the district, around Clun Forest, where it has a thickness of about 300 m. In the west the thickness is less than 200 m. It is possible to explain this as due largely to diachronism of the base of the formation, for it is logical that the siltstone laminite facies of the Knucklas Castle Formation would have persisted longer in the distal, western part of the district, while the influence of shore-face, storm-generated, bottom currents took time to prograde out.

Fine sandstone/silty mudstone couplets (p.61) are the essential characteristic of the formation. Although they are probably event couplets they are unlike those of the Bailey Hill Formation in having been deposited in an oxic and shallower environment. Bioturbation is prevalent, especially in the silty mudstones, and thus the differentiation between the sandstone part of the couplets and the silty mudstone portion is not always sharp. This is particularly true in the lower part of the formation, well illustrated in the forestry road section [SO 3270 8341] to [SO 3221 8349] and [SO 3214 8340] to [SO 3220 8328] on the south-west side of Sunnyhill. Indeed, in the basal 15 to 20 m, the arenaceous part of the couplet is very fine grained, and is visible on joint faces only as paler layers about 10 mm thick, in couplets which are barely 30 mm thick.

In the low and middle parts of the formation, these sandstones reach a thickness of about 0.30 m in a thickening upward sequence. A few sandstones at the top of this sequence attain thicknesses of 1 m. In general, the fine-grained sandstones are grey to buff. Most sandstone beds are of even thickness, with sharp bases and distinct tops, but some pinch and swell. In the upper part of the formation, the beds are again thin. This part is the Llan-wen Hill Beds of Holland (1959). The sandstones are not graded, except for a coarse shell-detrital lag present in some beds. Commonly, they have a bed-fabric of subparallel, flat to wavy lamination, rising into ripple cross-lamination in places. Convolute lamination has also been recorded. In some beds, parallel laminae become wavy upwards by thickening under convexities in the wavy surface. This is a characteristic of hummock cross-stratification and may have the same origin (Brenchley, 1985; Duke, 1985).

The sandstones are composed mainly of quartz (up to 70 per cent), with subordinate, detrital, white mica and biotite and accessory rutile; matrix consists of sericite/ chlorite with some carbonate cement.

Silty mudstone is generally bioturbated and rubbly; cylindrical burrow-fills are common. Where burrowing is less intensive, laminae are visible, and are reminiscent of the siltstone laminite of the Knucklas Castle Formation. Thus the silty mudstone is probably a bioturbated mix of the upper parts of sandstone/mudstone couplet and a siltstone laminite which may be largely hemipelagic. Towards the top, where the sandstones thin, fine detrital mica becomes abundant. Earp (1938) described these beds as ‘grey green micaceous shales’ which have retained their primary lamination and show small-scale rippling (compare with the Llan-wen Hill Beds of Holland, 1959).

Disturbed (slumped) strata occur sparsely (Earp, 1940, p.6). The remains of brachiopods, orthocones and, in the higher beds, ostracods and serpulitids are much more common than in the Knucklas Castle Formation.

The Cefn Einion Formation represents the returning influence of strong bottom currents, and wave activity is indicated by the bed-form of some sandstones. These factors suggest that the sea was shallower in late Ludlow times and that the sediments were deposited just below fair-weather wave-base. Storms were probably more influential than during the deposition of the Bailey Hill Formation (Watkins, 1979), at least until the late stages when marine conditions were very shallow.

The formation is the final product of the post-nilssoni shallowing, open-marine, platform margin. It had been an unstable margin having accommodated up to 2000 m of compacted sediment where slumping was also a prominent constituent. The changing balance between deposition from bottom currents and that of subsidence was one factor which divided the sequence into three lithoformations (Figure 18). The second factor was the fundamental change in oxygenation of marine bottom-water from anaerobic to aerobic just after the Bohemograptus proliferation Biozone.

Biostratigraphy

The formation possesses a shallow marine fauna of low diversity dominated by brachiopods of the Protochonetes ludloviensis Association of Watkins (1979). This is the equivalent of the Salopina Community of Calef and Hancock (1974). Dayia navicula is common in the beds transitional from the Knucklas Castle Formation below.

Details

South-eastern corner of district

A good section at Oaker quarry [SO 378 815] displays about 27 m of well bedded, massive to flaggy siltstones. Laminated siltstone alternates with bioturbated siltstone in beds 0.05 to 0.30 m thick. Some beds contain siliceous nodules which overprint the lamination, and rotten calcareous nodules also occur. The quarry is the type locality for the brachiopod Salopina lunata and there are numerous other Ludfordian brachiopods at this locality.

Periphery of Clun outlier

The formational type section is along the forest road on the south side of Sunnyhill where the lower part of the formation is exposed. Adjacent to the track, a quarry [SO 3263 8339] exposes 5 m of blocky, grey to fawn, argillaceous siltstone unlike the underlying Knucklas Castle Formation which is thinly fissile. On weathered joint faces, pale bands about 10 mm broad and little more than 10 mm apart represent fine-grained sandstones which appear within the succession at the base of the formation. The joint face banding is a characteristic of this formation and occurs to the south at least as far as Builth. The basal beds yield a typical Ludfordian fauna, largely concentrated in the arenaceous layers which includes Microsphaeridiorhynchus nucula, Protochonetes ludloviensis, Fuchsella amygdalina, Ptychoparia?, ‘Orthoceras’ cf. trachaele, crinoid columnals, Calcaribeyrichia torosa and Serpulites?

Along the road to the west and north-west, and up the sequence, the arenaceous layers become thicker, 30 to 60 mm and more rarely up to 150 mm thick, more widely separated and slightly coarser. Fossils become plentiful and bioturbation pervasive, except in the lower parts of the sandstones where parallel and wavy lamination is preserved.

From a bend in the road [SO 3223 8360] south-westward for some 190 m and then down the track [SO 3207 8349] to [SO 3214 8341], about 34 m of beds characteristic of the formation are exposed. Sandstones are commonly 20 to 100 mm thick, one lenticular bed is 0.40 m. The silty upper part of the couplets are up to 0.30 m thick, very bioturbated, and are commonly very fossiliferous contrasting with the sandstones. The latter contain transported shell detritus; the former appear to preserve a more indigenous fauna. The fauna includes Aulopora sp., M. nucula, Orbiculoidea rugata, P. ludloviensis, Salopina lunata, F. amygalina, Goniophora cymbaeformis, Leurocycloceras whitcliffense and ‘Serpulites’ longissimus.

Farther to the west, in Red Wood, near the boundary with the overlying Clun Forest Formation, fossils become less common, laminations more common [SO 3072 8362] and the beds are more micaceous [SO 3111 8036].

To the north, the formation is well displayed in a number of quarries [SO 3200 8552]; [SO 3146 8514]; [SO 3175 8555] on the isolated hill of Acton Bank.

In the northern part of this outcrop, a quarry section [SO 2716 8629] west of New House Farm shows the transition into the Clun Forest Formation. Some 20 m of beds are seen, comprising blocky, calcareous, massive, coarse siltstone in beds up to 0.20 m thick. The beds pinch and swell and are interbedded with fissile siltstones which are both unlaminated and wavily to poorly laminated with irregular calcareous horizons. They yield a typical Ludfordian fauna which includes Microsphaeridiorhynchus nucula, Protochonetes ludloviensis and Salopina lunata. However, the uppermost beds contain the Přídolí assemblage Turbocheilus [Platyschisma] helicites, Modiolopsis complanatus and fish fragments. The faunal change is abrupt but there appears to be little change in the lithofacies.

To the south-west of the outlier, there are a number of sections on and around Rock Hill [SO 280 793]. An old quarry [SO 2857 7959], 350 m west-south-west of Cockford Hall, exposes 4 m of grey to khaki, wavy-bedded, finely micaceous siltstone with an abundant Ludfordian fauna including M. nucula, P. ludloviensis, S. lunata, bivalves and ‘Orthoceras’ sp. A track-side [SO 2762 7892] on the south side of the hill showed 5 m of thickly bedded siltstone, heavily bioturbated but with a few thin laminar beds. The fauna includes Fuchsella amygdalina and ‘Serpulites’ longissimus. Similar beds were seen at intervals along the track for about 300 m on either side of this section.

Periphery of Bettws-y-crwyn outlier

An old quarry [SO 2222 7871] at Cwm shows about 5 m of extensively bioturbated, flaggy, bedded siltstone with a few laminar beds. On the opposite side of the valley, a quarry [SO 2251 7890] exposes 10 m of bioturbated siltstone with laminated beds of siltstone and fine-grained siltstone up to 80 mm thick. Lamination is not widely developed in the upper part of the section. Farther north, a quarry [SO 2755 8005], 500 m south of Hill House, near the base of the formation, shows 5 m of siltstone and sandstone which is almost entirely bioturbated with some well developed burrows. These beds contain F. amygdalina and ‘S.’longissimus. North of the Clun Valley, there are numerous sections along a track on Caldy Bank [between 2371 8307] and [SO 2359 8328] in flaggy bioturbated siltstone with some fine-grained sandstone beds up to 50 mm thick. A section farther north, north-west of Caer-din Ring, yields an extensive late Ludfordian fauna including M. nucula, Orbculoidea rugata, P. ludloviensis, S. lunata, T. helicites?, ‘Orthoceras’ sp. and Polygrammoceras bullatum.

The basal beds are well seen around Cwm Moch [SO 2159 8700] where, for example, a quarry at the back of the house shows 3 m of mainly laminated, greenish grey, coarse-grained siltstone and fine-grained sandstone. Some of the sandstones are up to 0.30 m thick and are mainly massive apart from partially disrupted laminae in the top few centimetres. A new quarry [SO 1800 8737], also at the base of the formation, exposes 8 m of blocky siltstone; bioturbation has destroyed most of the original sedimentary structures but some laminated, wavy, and lenticular horizons are preserved. Grazing ‘trails’ and burrows occur on the surfaces of mudstone partings and fragments of the brachiopod Dayia are present.

Around Kerry Pole, the formation is not well exposed. To the south, however, where the outcrop is repeated by faulting there are a number of good sections. The best of these is in Crochan Dingle [SO 1420 8161] to [SO 1472 8149], where there is an almost complete section through the formation. At the top (eastern end) of the sequence, the dominant lithology is fine-grained, pale grey sandstone, with beds up to 0.20 m thick, wavy to lenticular bedding and local ripple cross-lamination; scattered sandstone beds up to 1 m thick are present with faint parallel lamination and little or no grading. The thicker sandstones die out gradually downstream. The sandstones are interbedded and intergraded with subordinate siltstones, which have wavy partings, up to about 5 mm thick. Down sequence west of [SO 1453 8152], the maximum thickness of the sandstone gradually decreases to about 0.10 m, as the proportion of interbedded ‘siltstones’ increases (interlaminated sandstone/mudstone lithology, commonly bioturbated). Near the base of the formation, there is a slump bed, 1 m thick, underlain by a few sandstones up to 0.50 m thick.

An excellent new section [SO 1536 8220] to [SO 1541 8220] at Hendre, at the base of the formation, shows about 8 m of beds of interlaminated mudstone (dominant) and sandstone, with sporadic bioturbation which rarely destroys the lamination. These beds are interbedded with about 30 per cent of fine-grained sandstone, up to 0.30 m thick, with planar or slightly wavy lamination, commonly giving way upwards to ripple lamination. The sandstones have sharp tops and bases and are planar but with some lenticular bedding. Carbonaceous debris is present on some bedding planes.

Clun Forest Formation

This formation was designated by Lynas (1987) with a type section (p.109), defining its base, at Bryn [SO 2951 8535]. The formation occurs in two outliers and therefore has no defined top. It is about 660 m thick in this district and includes all the strata in the district formerly grouped in the Downton (now Přídolí) facies of the Old Red Sandstone including the Platyschisma Beds, Green Downtonian and Red Downtonian (Earp, 1938, 1940; Allender, 1958; Holland, 1959). Its relationship with the Downton Castle Sandstone, Temeside Shales and Ledbury Group in the Church Stretton (Greig et al., 1968) and Ludlow (Holland et al., 1963) districts to the east is not precisely known, hence the reason for establishing the new formation.

The outcrop of the formation is limited to two large outliers in the southern half of the district. The western area has been called the Bettws-y-crwyn Outlier and the eastern area the Clun Outlier by Stamp (1919, p.232).

The Clun Forest Formation is a lithologically heterogenous deposit in which a sequence of four poorly defined facies associations can be recognised (Zalasiewicz, 1991).

The first and lowest of these facies associations (Table 12) is the Platyschisma Beds of Earp (1938, 1940) and other authors. In many aspects it is lithologically similar to the uppermost beds of the Cefn Einion Formation and comprises thick planar and lenticular beds of fine-grained sandstone with thinner beds of sandstone, siltstone and mudstone. Locally, the beds differ from the typical Cefn Einion rocks in containing abundant mica and/or carbonaceous debris (including recognisable plant and eurypterid fragments on bedding planes) and in the dark grey colour of some of the interbedded mudstones. Most characteristically the deposits include the gastropod Turbocheilus [Platyschisma] helicites, associated with Lingula, and the bivalve Modiolopsis complanata, together with fish fragments (thelodont scales and acanthodian spines). These beds are up to 10 m thick in the west of the outcrop and reach 20 m in the eastern outlier.

The second facies association overlies the ‘Platyschisma Beds’. It consists of up to 20 m of grey, green and brown, variably bedded, fine- to medium-grained (locally coarse-grained), micaceous and feldspathic sandstone, interbedded with siltstones and mudstone. The deposits are commonly massive or irregularly bedded, with irregular blotchy variegations of colour and texture, including green and red veining. The sandstone commonly contains intraclasts, together with friable patches up to 20 mm in diameter which probably represent decalcified carbonate nodules. These irregular, locally nodular to rubbly weathering beds are interbedded with subordinate, thinner bedded, wavy to ripple laminated sandstones. These beds form the lower part of the ‘Green Downtonian’ of Earp (1938, 1940).

The third facies association includes the greater part of the Clun Forest Formation and comprises a succession of micaceous silty mudstone which includes fine-grained micaceous sandstone. The mudstone shows a blocky to indistinct horizontal bedding; the sandstones occur in bedded units up to several metres thick. The deposits are green to grey near the base of the sequence, becoming red-brown to purple-brown upwards. The change in colour is irregular and gradational; strata near the base of this association comprise grey-green sandstone interbedded with purple-red mudstone. Individual sandstone beds are rarely over a metre thick, planar or lenticular, with sharp tops and bases, ungraded, and internally structureless or with faint horizontal to subhorizontal lamination. The sandstone units are lenticular; north of Beguildy for example around [SO 1910 8155] they give rise to low topographical ridges which can be traceable for 200 to 500 m. There appears to be an increase in the proportion of sandstone to mudstone upwards. The lower, predominantly green, part of this facies association forms the upper part of Earp’s ‘Green Downtonian’: it ranges in thickness from about 50 m in the west to 100 m in the east. The overlying red beds are the ‘Red Downtonian’ of Earp, and about 580 m are preserved. A notable feature of the ‘Red Downtonian’ in this district is the paucity of pedogenic carbonate in comparison with areas to the east (e.g. Downton Group of Ludlow district, Allen, 1974, p.111) and south-east (e.g. Raglan Mudstone of Hereford district, Brandon, 1989).

The fourth facies association comprises a number of sandstone units, up to about 7 m thick, which are intercalated within the highest beds of association three (Table 12). These units are medium-grained sandstones made up of a number of individual beds of high-angle, planar cross-sets each 0.2 to 0.8 m thick. Current directions are consistently from the western quadrant.

The Clun Forest Formation has not been subdivided on the map. Earp (1938, 1940) subdivided his ‘Downtonian’ on the western side of the Bettws-y-crwyn Outlier and Stamp (1919) attempted to extend the subdivisions of the Ludlow area (Elles and Slater, 1906) into the southern part of the Clun Outlier. However, over most of the outcrop the paucity of exposure and the gradational nature of the boundaries would render the mapping of such subdivisions highly subjective.

The first facies association (Platyschisma Beds) shows a gradual change from the fully marine conditions of the underlying Cefn Einion Formation to brackish water conditions. The second facies, which includes decalcified calcareous nodules (calcretes), marks a transition towards terrestrial conditions. These sediments were laid down on a distal fluvial plain or coastal flat, subject to pedogenic processes which were enhanced by a hot climate and low seasonal rainfall (see Allen, 1985; Brandon, 1989). The third facies was laid down on an alluvial plain; mudstone represents distal overbank deposits, and fine-grained sandstone represents the more proximal overbank and crevasse-splay deposits. The medium-grained sandstones of the fourth facies association are fluvial in origin, deposited in channels, and indicate sediment supply from the north-west, thus confirming that uplift and erosion had clearly commenced in the far north-west during the Přídolí. The relative sparsity of pedogenic deposits, apart from those in the basal part of the green beds of the second facies association, may be due to the district being slightly closer to the source of sediment supply than the greater part of the Anglo-Welsh basin to the south and east.

Biostratigraphy

The base of the Přídólí Series is defined on graptolite faunas (Martinsson et al., 1981); in their absence in Britain it has been taken at the base of the Ludlow Bone Bed Member at Ludford Lane, Ludlow (Siveter et al., 1989), where most of the marine species characteristic of the underlying Ludlow Series abruptly die out. The equivalent horizon in this district is not as clearly defined, but there is nevertheless a rapid transition from marine to brackish conditions.

The first facies association of the Clun Forest Formation contains a few marine fossils such as the brachiopod, Salopina lunata, but the fauna is dominated by gastropods and bivalves notably Turbocheilus helicites, Palaeoneilosp., and Modiolopsis complanata, with some ostracods, thelodont scales and acanthodian spines. Bedding planes covered with an abundance of the brachiopod, Craniops implicatus, are present in places. Above this horizon fossils are sparse. Earp (1938, 1940) has recorded eurypterids, bivalves, ostracods and Lingula sp. from the overlying green beds and also from green beds within the ‘Red Downtonian’ (second and third facies associations). Sandstones yield Lingula sp. and fish remains. He estimated (1938, p.140) that the highest faunas are at least 385 m from the base of his Downtonian (i.e. Clun Forest Formation).

The position of the upper limit of the Přídólí Series (the Silurian–Devonian boundary) cannot yet be determined with any accuracy within the terrestrial (Old Red Sandstone) deposits of the Anglo-Welsh area. Allen and Williams (1981) argue that the boundary must lie in the middle or upper part of the Ledbury Formation of the Clee Hills area to the south-east, below the lowest calcrete concentration of the ‘Psammosteus Limestone’. Since definition on strict faunal grounds is as yet impossible, they suggest that the Townsend Tuff Bed, some 60 to 100 m below the ‘Psammosteus Limestone’, be used as a second-best measure. If this is accepted, then the thickness of Přídólí beds in the Clee Hills is about 430 to 510 m. The total thickness of the Clun Forest Formation is estimated at about 660 m, but there is no evidence of the existence of the Townsend Tuff Bed or any calcrete concentration in its upper part. As both the Tuff Bed and the calcretes are widespread throughout the Anglo-Welsh region it would seem likely that their absence in this district is due to thickening of the underlying deposits, possibly due to greater proximity to their source. Consequently, it is considered likely that the whole of the Clun Forest Formation lies within the Přídolí.

Details

Bettws-y-crwyn outlier

An old quarry [SO 1682 8558] west of Anchor Bridge, displays beds in the higher part of the formation. These comprise impersistent, lenticular, greenish brown, micaceous, medium-grained sandstones separated by 1 to 2 m of brownish red siltstone. The siltstones have irregular green mottlings and there are a few bright green bands, up to about 50 mm thick. The lenticular nature of the sandstones suggests channel-fills; planar cross-bedding at the base of one sandstone dips at 3° towards 115°.

In a quarry [SO 2342 8500] 500 m west of Caer-din Ring, the basal beds of the formation are well exposed (Lynas, 1986, p.6). The main and lower part of the quarry exposes blocky, mostly non-laminated, green-grey siltstone, in beds up to 0.15 m thick, interbedded with fissile siltstone. A few brachiopods occur at the bases of some beds. Several beds at the western end of the quarry are packed with Turbocheilus helicites and poorly preserved Modiolopsis sp. These first facies association sediments are about 1 m thick and consist of bioclastic limestones in beds up to 0.10 m thick. They pass up into well-laminated, thin-bedded, green, micaceous siltstone.

An excellent section [SO 1472 8481] to [SO 1472 8485], adjacent to the Nant Medwaledd, exposes the lower part of the formation (Zalasiewicz, 1991). This comprises some 15 m of thick- and thin-bedded sandstones and siltstones. The thick sandstones, which make up about 50 per cent of the sequence, are up to 0.50 m thick, mainly fine grained, commonly lenticular with sharp, channelled bases and sharp or gradational tops. They show faint parallel lamination, horizontal to low-angle wavy lamination, and low-angle ripple cross-lamination. The intervening strata are either thin-bedded (10–30 mm) fine-grained sandstones, or siltstones with slightly irregular parallel lamination. At the southern end of the section, two sandstones about 0.30 m apart grade laterally into lenses with intraclasts of dark grey mudstone, up to 20 mm long and 2 mm thick. A thin sandstone within these beds yields a fauna including T. helicites?, Palaeoneilo sp., Modiolopsis complanata and cf. Lingula minima; fish debris is present at the base of a sandstone just beneath the mudstone lenses. Slightly higher beds (in facies association 2) are seen in Crochan Dingle [SO 1479 8152] to [SO 1485 8150]; the north-east end of the section shows greyish fine-grained sandstones, of both massive and thin-bedded types, which pass downwards to thick-bedded green rubbly structureless silty mudstones with small friable decalcified carbonate nodules.

Beds within the upper part of the formation are well exposed at a number of locations in the long dingle [SO 1877 8071] to [SO 1851 8245] north of The Moat. A typical section [SO 1881 8088] displays about 4 m of blocky purple-red siltstones with, in the upper part of the sequence, two units 0.5 to 0.6 m thick of interbedded, planar-bedded, green-grey, fine-grained sandstones and siltstones, generally with weakly defined bedding. The lower unit is markedly lenticular passing laterally, within 0.15 m, into grey mudstones and silty mudstones.

The fluvial sandstones (facies association 4) in the uppermost part of the formation are exposed near Bettws-y-crwyn [SO 206 812] and Curney Farm [SO 207 845]. Near Bettws-y-crwyn, a quarry [SO 1997 8137] shows 3 m of greyish purple, medium-grained, cross-bedded, feldspathic sandstone. It comprises about eight sets of planar and high-angle cross-sets varying in thickness from 0.2 to 0.8 m. Foreset laminae indicate current directions mainly from between west and south-west (dips 18° to 28°, directions between 20° and 110°). Around Curney Farm and Hall of the Forest [SO 209 837], two thick sandstones and several thinner ones are marked by parallel topographical features, and the thicker sandstones have been mapped. They are similar in lithology to that at Bettws-y-crwyn, again with current flow mainly from the west, and they have been extensively worked for local building stone.

Clun outlier

The transition from the Cefn Einion Formation into the overlying Clun Forest Formation can be seen in the vicinity of Bryn [SO 294 853]. Lynas (1987, p.7) has used these sections as his type sections to define the base of the Clun Forest Formation. The uppermost beds of the Cefn Einion Formation, exposed in a track [SO 2957 8527], are calcareous bioturbated siltstones interbedded with greenish grey fossiliferous blocky siltstones up to 0.3 m thick and parallel to wavy-laminated siltstones. To the north-west, another track [SO 2951 8575] shows these beds overlain by greenish grey, blocky, vaguely and well laminated siltstone with a fauna of thelodont scales and Londinia sp. These basal beds of the Clun Forest Formation are also seen in an adjacent quarry [SO 2947 8531] where deep brown decalcified fossil beds yield T. helicites, horny brachiopod fragments and thelodont scales. The fossil beds are also exposed in a quarry [SO 2716 8629] near New House Farm.

Around Maesyrhaime [SO 265 830] and Bicton [SO 289 826], the first facies association is sufficiently well exposed for it to be locally mappable (Campbell, 1989). A section [SO 2651 8322], north of Maesyrhaime, shows 9.2 m of well-bedded muddy siltstone, siltstone, sandy siltstone and sandstone. Variably massive to ripple cross-laminated, bluish grey to brown, calcareous, bioturbated, sandy siltstone and sandstone (in units of 0.02 to 0.20 m) make up 50 per cent of the section; the remainder are finer grained lithologies. Shelly lags, up to 0.05 m thick, at the bases of the coarser units yield common Turbocheilus helicites, withLoxonema gregarium, Pharetrolites murchisoni, Tritonophon trilobatus, Deceptrix? subaequalis, Lingula sp., and cf. Salopina lunata. This facies is overlain by 2 m of brown siltstone, with flaggy sandy siltstone and fine-grained sandstone.

Near Bicton, a quarry [SO 2897 8296] in the overlying second facies association shows massive green mudstone with calcareous nodules, and silty mudstone, up to 2 m thick. The calcareous nodules are up to 20 mm in diameter and occur in bedding-parallel swarms and banks up to 0.50 m thick. There are also some diffuse beds of muddy sandstone.

The upper part of the formation crops out north of Whitcott Keysett and Bicton and around Weston Hill [SO 270 810], and a number of sandstone beds have been mapped in these areas. Most quarries exploit the sandstones and so are not representative of the whole of the sequence. One such quarry [SO 2710 8051], on the southern side of Weston Hill, exposes:

Thickness m
Sandstone, yellowish brown and greenish grey, fissile 1.0
Siltstone, greenish grey, sandy, predominantly parallel-laminated and low-angle cross-laminated 0.75
Siltstone, greenish grey and yellowish grey, fissile, laminated and low-angle cross-laminated, with subordinate sandy siltstones 0.45
Sandstone, greenish brown, fine-grained, coarsely micaceous, feldspathic, relatively massive to cross-bedded 0.5
Siltstone, relatively fissile, grey; laminated, sandy siltstone and subordinate greenish grey, fine- grained sandstone 0.6
Sandstone, greenish grey, very fine-grained, micaceous, relatively massive, poorly laminated 0.4 Siltstone, yellowish brown, fine-grained, sandy, fissile, low-angle cross-laminated 0.75
Sandstone and sandy siltstone, yellowish brown, fine-grained, micaceous, laminated 0.75

Some finer-grained units are visible in a pit [SO 2651 8018] near Burfield where the section comprises:

Thickness m
Siltstone and sandstone. Siltstone is strongly weathered, brown and reddish brown, yellow and greenish grey, sandy, laminated and cross-laminated; sandstone is fine grained, feldspathic, micaceous 0.6
Sandstone, greenish grey, fine-grained, feldspathic, weakly laminated 0.3
Mudstone and siltstone, reddish brown, conchoidally fractured 0.4

Beds in the lower part of the third facies association are seen in a number of sections around Pen-y-wern [SO 307 791] and in this area it has been possible to map a number of thin sandstones within these beds. For example an old quarry [SO 3100 7909] exposes:

Thickness m
Sandstone, greenish grey to greyish green, fine- to very fine-grained, with thin interbeds of greyish green silty mudstone and siltstone. The sandstone shows sphaeroidal weathering c.2.5
Mudstone, greyish green to bluish green, silty in places, blocky weathering, poorly bedded, some small calcareous nodules c.1.5

Chapter 7 Intrusive igneous rocks

Intrusions into Longmyndian rocks

There are several quartz-dolerite dykes intruded into the Longmyndian. They vary in width from about 2 m up to 60 m, but commonly are only 2 to 4 m wide. The fresher dolerites are massive, dark bluish grey rocks with clear feldspar and with small amounts of visible pyrite. Sericitication, epidotisation and chloritisation are all developed to varying degrees although some of the rocks appear to be relatively unaltered. In thin section, they typically show good ophitic texture and are composed of interlocking subhedral prisms of plagioclase feldspar which may be wholly or partly enclosed by anhedral plates of augite. In some cases quartz occurs as a relatively abundant accessory, in anhedral crystals interstitial to the feldspar. The petrography of this doleritic suite is fully described in Greig et al. (1968, pp.53–58).

Most of the dykes trend approximately west-north-west and are probably close to vertical. This trend is roughly perpendicular to the major syncline in the Longmyndian.

A composite dolerite dyke, 9 to 15 m wide and trending at 110°, is well exposed between Norbury and Upper Gravenor [SO 3700 9355]. It is intruded into massive sandstones, mudstones and siltstones which are thermally metamorphosed at the contact. The dyke splits into at least two parts, and was also seen nearby to the west [SO 3683 9360] and [SO 3658 9368].

A 60 m-wide dyke is exposed at Lower Gravenor [SO 377 941] and in a quarry [SO 3750 9425] near Upper Gravenor. The dolerite is coarse grained and roughly columnar. A number of dykes are exposed in the deeply incised valley of the River East Onny; one [SO 3788 9452] varies in thickness from 1.5 to 3.5 m over a distance of 200 m and has a sill-like branch.

Intrusions into Ordovician rocks

The intrusive rocks of the Shelve area (Figure 32) are mainly dolerites, with the exceptions of a picrite dyke at Cwm Mawr and an andesite dyke at Lower Wood [SJ 3085 0255]. The intrusions take the form mainly of sills, but a small dyke swarm occurs in the north between Estell and Lower Wood. The largest intrusion in the district is that at Corndon Hill. No intrusive rocks have been mapped in the Forden inlier in the west. The intrusive rocks of Shelve have been described by Blyth (1944), and Lynas et al. (1985) published an account of the pyroxene chemistry which includes radiometric age projections.

The major basic intrusions of the Breidden Hills and Shelve possess dual magnetic polarity (Piper, 1995), having cooled during a reversal of the field and before folding occurred. This supports a Caradoc age for the intrusive activity (Lynas et al., 1985) and is compatible with Late Ashgill folding.

Dolerite

Dolerite intrusions range in composition from alkali-olivine to tholeitic and in texture from subophitic to gabbroic. They are considered to have been co-magmatic with the Ordovician lavas of the district (Lynas et al., 1985). The rocks are composed largely of plagioclase (up to 65 per cent), which is calcic (up to the composition of labradorite), clinopyroxene (25 to 30 per cent) and ilmenite. The texture shows a largely devitrified mesostasis indicative, possibly, of intrusion of a largely crystalline magma to a high level where rapid cooling of residual magmatic liquids occurred. The more finely grained margins of the intrusions are less altered than the coarser gabbroic parts which cooled more slowly and were more susceptible to hydration.

Less commonly, the dolerite has intermediate affinities, oligoclase feldspar occurring as microlites in a groundmass associated with pale coloured, isotopic phyllosilicate (probably replacing pyroxene). A few patches of secondary quartz are present. This type of dolerite may be gradational into the more common pyroxene dolerite. It is usually vesicular and slightly porphyritic; the amygdales, up to 10 mm across, are occupied by calcite, some pyrite and quartz. It is exposed in a quarry [SO 298 934] at New House (E56863), where phenocrysts of oligoclase feldspar reach 2.0 mm across, and also occurs in a sill (E58085) exposed in a quarry [SO 3114 9387] south of Hyssington and at Kinton Farm [SO 2894 9946].

Details

Mytton Flags to top Stapeley Volcanic Member

More Quarry [SO 325 934] (Lynas et al., 1985; ‘Squilver’ of Blyth, 1944) was formerly worked for roadstone. It exposes green gabbroic dolerite consisting of intensely saussuritised plagioclase, ophitically enclosed by pale chlorite and other phyllosilicates (E58782). Clinopyroxene may exceed 6 mm in length; it is fresh, commonly twinned and in places encloses pseudomorphs, probably after olivine. A few cognate xenoliths are present. Iron oxides are abundant as accessory minerals and apatite is common. Common secondary minerals are epidote, prehnite and phyllosilicates in aggregates. Near-vertical flow-banding is present and in places a more leucocratic variety of dolerite is present. Magnetic susceptibility measurements revealed large differences across these bands over a distance of a few centimetres. A finer grained marginal facies contains more plagioclase (about 75%) with less pyroxene and opaque minerals (Blyth 1944 p.175) and narrow veins of prehnite and epidote occur within the dolerite near its upper contact in the old quarry [SO 3229 9336]. Blyth considered these to be the product of late-stage soda-bearing solutions filling fractures in the upper parts of the cooling intrusion and the adjacent country rock.

Several other sills and minor dykes are associated with the Hyssington Volcanic Member. Three rather dark brownish green doleritic sills (beds 19, 22 and 23, (Figure 9)) intruding the member are exposed in Tasker Quarry [SO 3251 9565]. Bed 19 (E58815) possesses two generations of plagioclase feldspar microlites and vesicles which are occupied by quartz and calcite. Also present are slightly darker coloured xenoliths, several centimetres across, of a coarser dolerite (E58814). This is somewhat altered, with no fresh ferromagnesian minerals; vesicles contain secondary quartz. Thin sections (E58817) and (E58819) are from the centres of beds 22 and 23, respectively. The former reveals a rock identical with that in (E58815) (Bed 19) and the latter with the xenoliths (E58814) of Bed 19. The base of Bed 23 (E58818) is vesicular, but here infillings are of quartz and calcite and there is a pronounced trachytic texture, like that of many clasts in the volcaniclastic sandstones. The top parts of both sills show faint flow foliation and are variously vesicular, chlorite being the infill mineral. The sections (E58816) and (E58820) also show that the tops are more finely grained than their centre parts, indicative of marginal chilling.

The largest of these is intruded into the Mytton Flags Formation and is exposed [SO 3385 9908] north-east of Shelve, where it has a nodular appearance. The nodules (E58875) comprise highly altered plagioclase with large poikilocrysts of clinopyroxene, up to 10 mm across. Its hornfelsed aureole is exposed in the places [SO 3381 9981], [SO 3412 9968] and shows the usual retrograde assemblage of secondary phyllosilicates and diffuse spots (E58876).

To the north, a small Ordovician inlier is exposed within the lower Silurian rocks. Near Estell, an old quarry [SJ 3602 0394] exposes an unusual dolerite. The rock is grey with strong parallel planar joints and sporadic trains of pyrite concretions, up to 10 mm thick. It is feldspar-phyric (E59501), locally cumulophyric with rather slender phenocrysts, up to 1 mm across. The groundmass is granular with no other primary minerals, but there are many patches of calcite and chlorite. The body is evidently a sill; hornfelsed siltstone is present at the eastern end of the quarry and an intrusion breccia at the eastern entrance. The same intrusion, but with more mafic constituents (E59502), is exposed in a quarry 100 m to the north-east.

Corndon Hill

This is the largest of the intrusions and forms a prominent hill in the north of the district. Watts (1886) considered it to be a laccolith, but Blyth (1944) argued that it is a concordant intrusion emplaced in the crest of a fold. This survey failed to reveal any evidence for the geometry of the base, but a magnetic and gravity survey in 1983 indicated that the igneous body is at least 200 m thick, which contradicts Blyth’s opinion ‘that it need not be more than some 200 ft’.

Adjacent to contact, the dolerite is fine grained (E58105) with feldspar megacrysts set in a fine-grained intersertal base which includes some quartz. The main part of the intrusion is gabbroic dolerite with plagioclase feldspar and clinopyroxene; small patches of primary quartz subophitically enclose the clinopyroxene in parts (E58104) and there are large crystals of skeletal ilmenite (E58098).

A similar rock type occurs in a number of smaller intrusions to the south and east of Corndon Hill. One of these contains abundant biotite [SO 311 963] (E58120). Hornfelsed country rock is associated with these minor intrusions. It is exposed in an old quarry [SO 3047 9566] near Corndon Farm, where 2 m of hornfelsed shale separates two sills of dolerite, and farther east at ‘Contact Crag’ (Blyth, 1944) [SO 3121 9600]. Thin sections of the hornfelsed shales (E57546), (E57547) show a fine-grained spotted aggregate of phyllosilicate, carbonate, quartz and feldspar.

Stapeley Volcanic Member to Aldress Shale Formation

East of Roundwood, the gabbroic facies is exposed in a quarry [SO 2970 9400] where it is markedly pyritous. A thin section (E56985) shows chlorite, plagioclase and patches of high relief with subophitic plagioclase. A finer grained dolerite to the north [SO 2974 9427] contains fresh plagioclase and clinopyroxene and shows a good intersertal to subophitic texture.

The Lan Fawr dolerite [SO 297 968] is massive, with vaguely columnar joints and lenticular quartz veins. In thin section (E56985), it is a fresh rock with a well-developed subophitic texture.

At Stapeley Hill, a sill-like intrusion is exposed on the summit and in an old level [SO 3088 9901]. The dolerite is fresh and shows a poikilitic texture with plates of clinopyroxene, up to 6 mm across enclosing cloudy, altered plagioclase (E57387). The gabbroic central parts of the intrusion are exposed along the summit, where there is distinct layering [SO 3123 9895]. Weathering by exfoliation is a characteristic which has produced many oval boulders, a common feature of the Shelve gabbroic intrusions.

A number of dolerites intrude the lower part of the Aldress Shale between Rorrington and Wotherton [SJ 2935 0039].

Dyke swarm

Dykes of ophitic dolerite occur between Estell [SJ 357 041] and south of Lower Wood [SJ 309 024] trending 080° in the west veering to 060° eastwards They are similar to the dolerites elsewhere in the Shelve area but are darker in colour and fresher (Blyth, 1944, p.191). One of these dykes has been quarried [SJ 3481 0378] in Stanage Coppice. It is nearly 100 m wide in places and consists of 46 per cent plagioclase (labradorite), 25 per cent pyroxene, 8 per cent opaque minerals and about 21 per cent secondary minerals, mainly chlorite and calcite.

Picrite (olivine pyroxenite)

The picrite of Cwm Mawr [SO 3026 9470](Figure 32)(Watts, 1887; Blyth, 1944,) is poorly exposed. In general, the rock is strongly decomposed. No contact with the country rocks has been observed and a magnetometer survey failed clarify the margin of the intrusion.

North of Cubbulch [SO 3067 9500] (said to be the site of the bronze age stone axe factory) fresh rock is exposed; it is dark coloured, hard and jointed. Large plates of pale pink clinopyroxene poikilitically enclose serpentinised olivine and possibly orthopyroxene. Slightly altered labradorite is a relatively minor constituent and there are scattered phlogopite laths (E57551). Elsewhere [SO 3053 9519], poikilocrysts of clinopyroxene preserve fresh olivine (E57549). The rock in its more common, largely altered, condition occurs in a new (1984) section 0 to 85 m west of Cwm Mawr Farm house (E59524) to (E59531). Parts look fresh and are blue grey. Patches of high-relief, glassy mesostasis comprise a mixture of calcite and chlorite. Plagioclase is mostly altered to carbonate and phyllosilicate, but a little remains and this fairly calcic. Pale green serpentinous chlorite is abundant and apatite is a common accessory mineral.

Andesite

At Lower Wood [SJ 3089 0264], massive, porphyritic, trachytic andesite forms a small hill (Blyth, 1944, p.193) and has been quarried for roadstone [SJ 3085 0251] and [SJ 3087 0258]. The rock is pale greyish blue with phenocrysts of plagioclase, up to about 3 mm long. Strong joints occur in the top part of the intrusion on the west side of the quarry where alteration is correspondingly greater. Across the quarry (E59469), (E59472), (E59472), (E59472), (E59472), the rock exhibits good trachytic textures in intensely saussuritised plagioclase, and interstitial quartz is common. The plagioclase phenocrysts are almost wholly replaced by secondary minerals and indeterminate chlorite is probably pseudomorphed after mafic minerals. Patches of calcite and small aggregates of pyrite are abundant. The geometry of the intrusion remains uncertain.

South-west of Lower Wood, a small intrusion [SJ 3061 0211], 150 m long, forms a knoll. Texturally, it is unlike the andesite described above, but is similar in composition. The central portion is sparsely porphyritic and amygdaloidal. Thin section (E59473) reveals a highly altered, granular groundmass with carbonate-filled irregular amygdales, small secondary quartz spherulites and a scatter of acicular plagioclase. Exposures near the margins are brecciated (E59474). Fragments show a scatter of small plagioclase phenocrysts set in a felted microlitic, nearly isotropic groundmass which originally might have been glass. The breccia is bound by a carbonate cement. The north-east edge of the knoll displays isolated pods of intrusive rock which appear to have detached from the main mass and been enveloped in wet sediment.

Intrusions into Silurian rocks

Microgranite

West of Owlbury, a small quarry [SO 306 919] exposes microgranite intruded into Wenlock strata. Similar, but smaller outcrops of microgranite occur to the south-east, near Upper Heblands [SO 3259 9028], [SO 3265 9053], [SO 3269 9061] and [SO 3279 9061] (Murchison, 1839; Ramsay, 1858; Blyth, 1944; Sanderson and Cave, 1980). The microgranite is obviously younger than the Middle Wenlock rocks which it intrudes, but it is unlike nearby Carboniferous igneous rocks; the intrusions may be associated with late Wenlock volcanism.

Near Owlbury, the rock (E40657) is a quartz-porphyry (Table 13), exhibiting some fluxion orientation of the feldspar laths. Scattered prismatic phenocrysts of albite-oligoclase also occur.

At Upper Helbands farm [SO 3254 9026], the microgranite (E40646), (E40651) has a fine-grained hypidiomorphic-granular texture in which tabular (up to 0.23 mm) and lath-shaped (up to 0.2 mm) crystals of turbid albite-oligoclase are set in a matrix of quartz and potassium feldspar (Table 13). Quartz occurs also as partly absorbed xenocrysts and aggregates, up to 0.3 mm diameter. Chloritised biotite is the only ferromagnesian component while accessory minerals include apatite, ilmenite and pyrite, with some secondary leucoxene and epidote. The microgranite intrudes into Aston Mudstone of Wenlock age (Sanderson and Cave, 1980). In the east and south sides of the quarry, the contact is very uneven and the sedimentary rocks are hornfelsed and brecciated. The breccia is invaded by anastomosing veins of intrusive rock which, near its margins, includes bodies of metasomatised mudstone. To the north, in a shallower quarry [SO 3265 9053], microgranite and hornfelsed mudstone are in faulted contact.

Chapter 8 Quaternary

Extensive Quaternary deposits of glacial, periglacial and postglacial origin cover the district but there is evidence for only one period of glaciation. This is generally accepted as being that of Devensian age, though there is no positive dating of any deposits. Traces of previous glacial periods were obliterated by the Devensian ice, and indeed all extant deposits appear to relate mainly to the retreat stages of the glaciation and the subsequent periglacial and postglacial regimes.

During the Devensian, ice from two sources affected the northern part of the Welsh Borderland. One ice sheet extended eastwards from the Welsh mountains; somewhat later, another extended southwards from the Irish Sea into North Wales and the Shropshire–-Cheshire plain. Only Welsh ice directly affected the district, but the presence of Irish Sea ice to the north and north-east during the retreat phases had an indirect effect, particularly in relation to the development of several proglacial lakes. At its maximum, the Welsh ice extended round the southern end of the Long Mynd into the Church Stretton Valley, but there is no evidence that it overrode the whole of the Shelve area or the Long Mynd itself (Greig et al., 1968, fig. 21). It also affected all of the rest of the district, though ice on the high ground of Clun Forest and the adjoining areas was probably limited in thickness. During the retreat phase, the ice to the south of the Long Mynd (the ‘Plowden Glacier’ of Greig et al., 1968) retreated sufficiently to allow water from the East and West Onny rivers to be diverted around the southern end of the Long Mynd into the present channel while the probable former channels south of Lydham and Plowden were still blocked by ice. As a result, the broad valley from Bishop’s Castle to Purslow is now occupied by the misfit River Kemp while the broad hollow [SO 375 870] south-west of Plowden is dry. Other changes in drainage are considered in the section on Lacustrine deposits.

Dwerryhouse and Miller (1930) published a wide-ranging review of the glaciation of much of the Welsh Borderland. Rowlands (1966) studied the Church Stoke area, and also the Marton and Church Stretton valleys to the north and east of the district, and Brown (1971) covers almost the whole of the district.

Morainic deposits

These deposits were previously known as ‘Morainic Drift’ (IGS, 1972; Cave and Hains, 1986). They are characterised by fairly level but uneven topography little modified since the retreat of the ice. They are a heterolithic assemblage, mainly of ill-sorted gravel generally up to cobble size, but with a scatter of boulders and with a high content of fines. Parts have been sorted and deposited by meltwaters to create beds of sand and gravel; other parts are very clay-rich (matrix-supported) non-layered and till-like. Clasts are commonly well rounded and derived from fairly local feldspathic greywackes, fine siliceous sandstones, mudstones and vein quartz. Some field brash [SO 3457 9017] in the region of Lydham includes weathered igneous rocks and purple porphyry, possibly from the Shelve area.

Morainic deposits are confined to the low ground of the main valleys, the glaciated Severn Valley in the west and the ground just west of Sarn; in the east they occur between Bishop’s Castle and More and around Clun. The surface of the deposits is uneven to almost flat and commonly it remains unmodified except by civil agency.

The deposits are considered to be the product of the final retreat of major valley glaciers and thus they were subjected to some reworking by meltwater. Initially, they also entombed bodies of stagnant ice, or were deposited upon such ice, which, upon melting, caused collapses as witnessed by oversteepened dips in stratified deposits and peat-filled kettle-holes (p.115). Apart from those in the area of Lydham, which were the deposits from unconfined ice at the end of a valley glacier fed from the south, morainic deposits are probably broad kame terraces. The best examples occur between Scafell and Aberbechan, upstream and downstream of Newtown respectively.

Working and reworking of glacial detritus by meltwater tended to level the deposits and probably aggrade them, while stagnant ice below and alongside wasted, for example in the area of Newtown cemetery [SO 1185 9190]. Thus, there is a tendency for thick morainic deposits to include less stratified material in their lower parts where they may be indistinguishable from till, as in the base of the cliff section [SO 140 926] near Penarth. The retreat of the ice towards mid-Wales was followed by the erosive activity of the proglacial river and this was effective for a longer period in the lower reaches of the valley than the upper. As a result, most morainic deposits downstream of Abermule have been recycled into fan deposits, river terraces and finally alluvium. In places, the early meltwaters erosively benched the morainic deposits [SO 141 928] and such benches are hard to distinguish topographically from river terraces which differ in being aggradational. However, their surface levels are haphazard, without an even downstream profile (or thalweg).

Details

Severn Valley

The difference in height between the surface of the morainic deposits and the present-day river floodplain increases downstream slightly, from about 9 m [near 089 908] to 11 m at Penarth Weir [SO 1400 9267], and the boundary is marked by a steep bank. Immediately west of Vaynor, this bank exposes up to 3 m of ill-sorted gravel, mainly up to cobble grade, but with a few boulders; it rests on Nantglyn Flags. It is silty with streaks and lenses of fine-grained sand and gravel. A similar section occurs in the railway cutting [SO 0942 9068].

The hillock of morainic deposits [SO 0916 9044] was excavated for industrial building and exposed a face, 15 m long and 4 m high, mainly in vaguely stratified, ill-sorted gravel up to cobble size (230 mm). Clasts are subrounded to subangular, of local Welsh sandstone, greywacke and mudstone including Nantglyn Flags. The lowest metre is very silty and unstructured, with a surface covered by a thin layer of gritty and pebbly clay; in the top 1.5 m is a lens, upto 0.3 m thick, of gritty, pebbly cross-bedded sand.

Construction work exposed a north-west-trending rock face [SO 1170 9190], on the south-west side of Newtown cemetery, where up to 2 m of well-stratified gravel is banked against a rising subglacial surface of Nantglyn Flags. This surface is, in fact, the north-west side of the preglacial Severn Valley; the morainic deposits, from the cemetery and golf course on the north-west, to the railway and beyond on the south-east, span the entire width of this valley. The postglacial Severn has not returned to this valley, but has cut a gorge, some 20 m deep, into the Nantglyn Flags on the north-west side of the preglacial valley. It did so, initially, as meltwater flowing along the north-west, and thus warmer, side of a wasting valley glacier. Once the incision was established the river maintained this course, so preserving the morainic deposits that fill the preglacial valley to the south-south-east from postglacial river erosion. Farther east, downstream, the River Severn has occupied its original preglacial valley. Between Glan-Hafren [SO 1313 9230] and the weir [SO 1400 9265] up to 11 m of gravelly silt and clay, including boulders, are well exposed and are at present being eroded in the south bank. Morainic deposits occur just to the north of the district [SO 200 979] in the abandoned valley north of Llandyssil and they probably underlie the fan deposits and lacustrine deposits of the valley too.

Mule valley

Rather flat areas of gravelly ground occur around Sybwll [SO 1882 9030] and Llwyncowrid [SO 1981 9118]. They form broad shoulders in a minor tributary valley to The Mule and prior to the incision of this tributary they were probably part of a continuous expanse of eastwardly inclined ground. The valley-ice from which they were deposited, therefore, probably retreated westward or south-westward. On its surface three kettle-holes were left, each about 50 m in diameter [SO 1863 9041], [SO 1877 9050] and [SO 1903 9050]. Exposure is sparse and composition is expressed only as ill-sorted gravel in fields and hedges.

Lydham–Bishop’s Castle area

A large triangular area of uneven ground extends north-north-east from an apex at Bishop’s Castle and represents the end moraine of a glacier that emerged from the valleys to the south.

Unlike the Severn Valley ice, no major river accompanied or succeeded the wastage of this ice. Thus stratification of the morainic deposits seems to be less significant; silt is a dominant constituent and the surface is more hummocky. Kettle-holes are common, for example south of Lydham [SO 3355 9010] to [SO 3395 9075] where there is a line of four up to 200 m across. A thick deposit of peat occurs in the bottom of each (p.122).

In banks [SO 3377 9071] near Heath Farm an excavated section 2 m high and 25 m long revealed gravel with a silty matrix and rounded pebbles (up to 60 mm across) of grey mudstone, siltstone and greywacke sandstone, all probably Silurian; many pebbles are strongly iron-stained. Westward dipping sand and granule lenses show cross-bedding with a dip about 16°W; at the west end of the section the clasts are imbricate with dips of 50 to 60°W. Such steeply inclined imbrication is most likely to have been due to post-depositional collapse. Igneous and other Ordovician rocks are not represented, although to the east [SO 3458 9017] clasts of weathered dolerite and purple porphyry were observed in field brash. However, these could have been carried in fluvial input from the north, for there is no compelling evidence to suggest that ice came from that direction.

Near Newton, the 4 m bank [SO 3499 9060] of an overflow channel, probably leading from Lake Camlad–Caebitra, was cut into morainic deposits and here 1.6 m of non-laminated silt was exposed.

Clun area

In this area mounds, up to 15 m high, of reddish brown silt, sand and cobble- to boulder-grade gravel occur, typically along valley sides and at the mouths of valleys. The deposits are usually clearly differentiated from the alluvium by an erosional bank. They are most extensively developed in the vicinity of the confluence of the rivers Unk and Clun, and also extend over the col between Rock Hill [SO 281 794] and Weston Hill [SO 264 810] to the valley north-west of Brook House [SO 2712 7943]. An exposure [SO 3005 8135], behind workshops about 300 m north of Clun, showed 5 m of brown silt with rounded boulders to 300 mm across. There are some small, peaty, clay-filled kettle-holes [for example [SO 2963 8062]; [SO 2989 8122]; [SO 3083 8122] on the surface of the deposits.

Till (boulder clay)

The till of the district incorporates lodgement till, ablation till and flow till from the Devensian ice sheet. It was spread liberally over the hilly terrain north-west of a line from Church Stoke to Kerry and Llanbadarn Fynydd, forming ill-drained uneven ground. Eastward of this line, till is sparser. The surface of till, especially on slopes, has been modified by solifluction and landslipping.

It consists of brown, yellow and grey, silty clay and silt with a variable content of ill-sorted, rounded to subangular clasts mainly of rocks derived from central Wales, including greywacke, quartzose sandstone, siltstone, mudstone and vein quartz. The igneous rocks of Shelve do not feature as erratics beyond their source area, and the existence of an ice cap there, with glaciers radiating from it, during the last glaciation is doubtful. Indeed, much of the till in the area south and east of Corndon Hill might have been soliflucted during freeze–thaw conditions.

The thickness of the till differs greatly from place to place. On the high ground of Clun Forest, it is a thin veneer, but in places west of the River Severn it is thick, though largely unquantified. Drumlins almost certainly present the thickest accumulations, for instance on Reservoir Hill in Newtown a borehole [SO 1007 9110] proved just over 7 m, with probably a further 30 to 35 m below. The 7 m consist of brown, silty clay with clasts up to large boulder size; the top 5 m is recorded as ‘firm’ and below as ‘hard’. The Montgomery Drumlin, likewise, was penetrated to a considerable, though unrecorded, depth in the well at The Old Rectory [SO 2234 9461]. The total thickness is probably over 20 m.

Drumlins are common in the north-western part of the district, many exhibiting the classical up-turned boat shape. Others are less perfect due usually to instability of the steep flanks in the immediate postglacial period. They range in size from up to 1 km long and 50 m high [SO 0885 9464], [SO 0980 9561], [SO 1109 9777] and [SO 1202 9740] to mere ‘poncynau’ a few metres long and high [SO 1115 9080].

The long axes of most drumlins are aligned south-west to north-east, parallel with the trend of the Severn Valley, but a marked deviation to west–east is apparent east of Newtown [SO 1641 9107], [SO 1728 9107], [SO 1669 9157], [SO 1704 9193], [SO 1710 9215] and [SO 1753 9250]. This is indicative of a major ‘spill-over’ of ice from the Severn catchment into the valley of The Mule.

In the south-eastern part of the district there is very little mappable till. However, there is a number of exposures of till in stream banks where the valley sides and floors are covered by later glacial deposits, soliflucted deposits (head) or fluviatile deposits. Examples of such are [SO 3022 8454] near Acton and [SO 3296 8213] near Steppleknoll, where grey clay with angular to rounded pebbles is exposed. In these cases, the valley sides and bottom are covered by mappable thick head. These sections and other similar ones suggest that till may be present under more recent deposits in many of the valleys in this part of the district (see also Valley-bottom drift).

Valley-bottom drift (concealed)

The classification of valley-bottom drift is dependent on boreholes, and these require careful interpretation; the drilling process commonly destroys the fabric of non-lithified sediment and might even alter the grain-size distribution. The thickness of the deposits too is unpredictable because the underlying rockhead is notoriously uneven. In places, this unevenness may be a preglacial inheritance, but in others it has been generated subglacially or perhaps proglacially. Thus boreholes provide reliable evidence of thickness at the point of drilling, but they have little predictive potential in a glaciated valley like that of the River Severn.

Drilling at Newtown indicates the presence of two linear depressions in the rockhead, one across Broad Street, the other under New Road. These appear to deepen eastward to join near the bend of the River Severn [SO 1106 9158]. There, over 13 m of cobbly gravel were proved, yet barely 100 m away [SO 1098 9161] and [SO 1114 9155] Nantglyn Flags are exposed in the river banks and streets. The depression, or channel, continues eastward where boreholes [SO 1127 9163] penetrated 13 m of deposits. These prove silt and gravel at the top passing down into gravel and boulders (at least 1 m across) of Nantglyn Flags which rest on Nantglyn Flags. The formation is exposed in the river bed and banks, only 100 m away from the boreholes. This infilled channel appears to lie to the north of the river, joining the marginal channel (p.115) near the cemetery. It seems unlikely that this deposit can be of glacial origin, rather it must have resulted from proglacial to present-day aggradation in the overdeepened marginal channel. Thus it is reasonable to suppose that the infill in the linear depressions to the south-west is of similar origin, though the scouring of the channels may, in part, have occurred subglacially.

Downstream, near Abermule, two boreholes (150 m west-north-west and 250 m north-north-west of The Court) penetrated 12.19 m of superficial deposits, mainly of sand and gravel without reaching solid rock. Another borehole [SO 1599 1466] proved a boulder bed to a depth of about 6 m, and this is probably a part of the Mule Fan Deposit. Farther downstream near Fron, three boreholes [SO 1755 9703], [SO 1757 9705] and [SO 1759 9706] on the left bank of the River Severn, did prove solid rock at depths of 55.5 m, about 53.0 m, and about 34.0 m (from west-south-west to east-north-east respectively). Since the distance between the last two holes is only 30 m, they reveal a steep rockhead gradient of about 33°, suggestive of aggressive scour similar to that which occurred at Newtown. The drift deposits here consist of silt, thin layers of clay, sand and gravel with boulders. They too are similar to the channel-fill at Newtown and probably originated as proglacial detritus, aggrading into recent river alluvium.

There are no boreholes in the upper Camlad valley to indicate the depth of Quaternary deposits. Geophysical reconnaissance surveys between Church Stoke and Lydham to determine depths to rockhead beneath the valley floor (Figure 38) were carried out by Professor Brooks (University of Wales, Cardiff) and co-workers, in cooperation with BGS. Gravity and seismic refraction methods were employed along five north to south profiles across the valley. A single subdrift channel was identified, in places asymmetrical to the present valley margins, with rockhead lying at depths of up to 80 to 100 m below the surface. A unidirectional fall in the channel was not discerned in the geophysical results. Such a deeply buried channel would be only about 50 m above OD and thus could not have had gravitational egress via the valley of the River Onny to the east or indeed to the south via Bishops Castle, leaving subglacial overdeepening as the only explanation.

Sand and gravel of uncertain age and origin

Small areas of sand and gravel of uncertain age and origin have been mapped in the south-western part of the district near the source of the River Ithon [SO 103 848], in the Teme valley [SO 197 791] south of Beguildy and adjacent to the Gwenlas Brook [SO 104 792]. They may include deposits of both fluvioglacial and solifluction origin (Zalasiewicz, 1991).

Glaciofluvial sand and gravel

These deposits were laid down by glacial meltwater towards the end of the glacial period. They are thin and occur on the lower slopes and floors of a number of valleys and may also be present as valley-fill, obscured by later periglacial and fluvial deposits. The main areas of occurrence are in the Teme valley near Beguildy, the Caebitra valley between Sarn and Brompton Hall [SO 250 934], and in the eastern part of the district in the valleys of the Onny, Kemp and Clun.

Details

Teme valley

Broad sloping terraces, commonly hummocky, of glaciofluvial sand and gravel are present along much of the lower valley sides between The Moat [SO 188 806] and the southern margin of the district. A good section [SO 1968 7965] on the river bank adjacent to Beguildy School showed 4 m of beds, the lower 2 m being structureless, very coarse, poorly sorted sand and gravel and the upper 2 m coarsely bedded with lenses of fine-grained gravel and sand. A roadside [SO 2170 7884] south of Redgate showed 1 m of pebble to cobble gravel with a buff-brown silty clay matrix. The rounded to subrounded clasts consist of green-grey siltstone and mudstone and grey to yellowish grey fine-grained sandstone.

Caebitra valley

The origins of two gravelly deposits [SO 220 920] and [SO 240 925] lying from about 1 km north-east of Sarn to Lynwood [SO 2496 9283] are still unclear. They are depicted as Glaciofluvial Sand and Gravel on the 1:50 000 Series map. The surface of the gravel is nearly flat, showing a steady decline to the east and east-north-east down the broad Caebitra valley. Fan deposits of the Cwmberllan valley abut against them in the west at a slightly (about 2 m) lower level, thus ruling out any association with the immediately postglacial melt-waters and suggesting that it is an older deposit. Initially, they were mapped as a morainic deposit and indeed a direct glacial association is still the most likely, relating the deposit, like a kame, to the ablation of a mass of valley ice while ice persisted longer on high ground to the north. There are no sections through the deposit, but field brash shows subangular and subrounded clasts of very local sandstone, up to 150 mm across.

Onny, Kemp and Clun valleys

Glaciofluvial sand and gravel is present on the south side of the River Onny at Eaton [SO 375 895] and is widespread in the upper reaches of the Kemp and its tributaries between Bishop’s Castle, Acton and Lydbury North. In these areas, it gives rise to broadly undulose terrain and a pebbly loam soil. Downstream from Lydbury North to Little Brampton, sloping terrace-like features are present on either side of the valley, and similarly west of Clunton in the Clun valley. Exposure is very poor, but everywhere this sand and gravel appears to consist of rounded pebbles and cobbles in a matrix of silt to fine-grained sand.

Glacial sand and gravel

Deposits mapped as Glacial sand and gravel are largely confined to the south-eastern corner of the district, along the valley of the River Kemp and near its confluence with the Clun. They blanket slopes up to 70 m above the river and are truncated downslope by later glaciofluvial sand and gravel and by alluvium. Pits [SO 3661 8118] near Little Brampton Farm show up to 5 m of pebble to cobble gravel with rounded clasts of purple sandstone, quartz and siltstone. Some horizons have a silt matrix; elsewhere the gravel is mainly grain supported with a low percentage of fines, mainly silt.

Lacustrine deposits and glaciolacustrine deposits

Lacustrine deposits, including glaciolacustrine deposits, consist largely of gravel-free grey silt and silty clay. Some are thinly laminated and others appear to be homogeneous. Exposure is poor and sparse, and no lithological distinction is made between proglacial lake deposits and postglacial lake deposits; indeed the former pass upwards imperceptibly into the latter. It has not been possible to date any of the deposits; repeated pollen and spore analyses (in 1988) of a laminated clay in the south bank of The Mule [SO 1782 9042] at Penygelly by Drs Bonny and Harland yielded nothing of interest, suggesting deposition in glacial conditions. The clays are liable to become plastic and, in low places, erupt into ‘mud-lumps’, for example in the Camlad valley [SO 3113 9149] and [SO 3157 9137] near Owlbury. Where such eruptions occur along a stream [SO 3105 9210], [SO 3176 9189], they cause minor diversions of flow and are liable to fluidise, becoming quick and treacherous [SO 3181 9188] to [SO 3187 9179].

Within this district, three lakes formed as Welsh ice decayed during the Devensian, and they persisted into Holocene times. These have been named Lake Camlad–Caebitra (Figure 33), Kerry Lake and Lake Llandyssil. The continual downward erosion of overflow channels and build-up of sediment eventually obliterated these lakes, and their sites have become the floodplains of minor streams. In addition, the sites of two other glacial lakes, Lake Crankwell and Lake Kinton, have been identified near to and in the Shelve area (Figure 33).

Camlad–Caebitra Lake

Camlad–Caebitra Lake was formed when a glacier, the Bishop’s Castle glacier, probably dammed the valley at Lydham. At its maximum extent, the surface level of this lake was controlled by the highest point along the line of Marrington Dingle prior to the erosion of this northward overflow channel. With this level at about 162 m above OD, the lake would have extended to Sarn in the west and to Lower Heblands in the east. The earliest overflow may well have occurred around an ice margin at a level nearer 180 m above OD, slightly east of Marrington Dingle, because the initial level of the lake was at this height. Proof of this lies to the east in the superb overflow channel, up to 100 m wide, which commences at Poolshead [SO 3386 9185], at a height of about 179 m above OD, and describes a reverse S-shape around the western side of the More Motte and Bailey. It exits into a sharp elbow in the channel of the West Onny at a point [SO 3425 9123] 200 m south-south-west of More Church. Clearly, as the Bishop’s Castle Glacier retreated from here, the level of the Camlad–Caebitra Lake dropped. Had the egress of the lake across the moraine persisted a little longer the River Camlad would probably have regained its eastward flow. However, the ice margin against the western side of the Shelve high ground was retreating down-slope faster than the More overflow channel was being eroded into the moraine and so Marrington Dingle usurped its role.

A rock ‘threshold’ across the valley at Sarn, at about 180 m above OD, blocked extension of this lake westward towards The Mule valley, except possibly in its very early stages. However, it is probable that some ice occupied ground there, for till occurs at a minimum height of about 147 m above OD, just lower than the rock threshold, and this might have acted as a barrier too. So from a maximum of about 180 m above OD during the Devensian, the level of the lake fell to 125 m above OD during the Holocene, when the bogland area developed between Brickyard Bridge [SO 2674 9403] on the Caebitra and the Camlad [SO 2925 9245]. The rate of the decline is uncertain and probably varied; the only certainty is that by the time Late Devensian ice vacated the ground between Montgomery and Marrington Dingle the level of the lake had dropped below 147 m above OD [SO 2644 9505].

Details

Laminated, grey, silty clay deposited from the lake is visible in a ditch [SO 2663 9425] just west of Church Stoke, and the plethora of elongate hollows in fields between the old brickyard [SO 2675 9400] and The Ditches [SO 2485 9395] is probably the aftermath of a local brick industry based on lacustrine clays.

An old brick pit [SO 3058 9155] exposes laminated grey clay. The bricks made from this clay and from a similar clay [SO 3262 9176] near Roveries can be identified by their content of unground fragments of laminated clay. The banks of a ditch [SO 3190 9078] to [SO 3184 9207], at 155 to 152 m above OD, expose about 2 m of brown silty clay with rootlets and scattered pebbles, possibly head of postglacial age, overlying about 1 m of cryoturbated gravel which channels into clay with some lamination. Laminated clay is visible nearby [SO 3196 9113] and 0.67 m was well exposed in a bank [SO 3308 9092], 400 m west-south-west of Lydham Motte and Bailey at about 164 m above OD. Heavy clay ground occurs in fields here, passing eastwards and up-slope into gravelly ground at about 175 m above OD, which is either a lake-shore deposit or part of the moraine which dammed the lake here.

Since the completion of this account, Gurney and Worsley (1996) have described a group of relict periglacial cryogenic features near Owlbury. The features, approximately circular, form ponds and marshy ground within lacustrine deposits, and Gurney and Worsley consider them to be relict palsas. Their formation may well have been contemporaneous with the cryoturbation in a bed of gravel which overlies, or is within, the laminated clay nearby (see above).

To facilitate the formation of these features the level of Lake Camlad–Caebitra must have fallen below their level (about 142 m above OD) and thus the ice had already retreated from the site of Marrington Dingle. It is possible that they formed in the cold period of the Loch Lomond Readvance, around 11 000 ka.

Kerry Lake

The rock threshold at Sarn about 180 m above OD separates the catchment areas of The Mule and Caebitra rivers. It is possible that a single glacial lake existed when the Severn Valley ice-front lay across the site of the Lower Mule Gorge. As the ice-front dropped below 180 m above OD, Kerry Lake and Camlad–Caebitra Lake must have become separate entities. The emptying of Kerry Lake was accomplished partly by the erosion of a south to north channel through the sand around Old Court. Most water escaped via the channel still used by The Mule (Plate 11). This flows eastward, bending sharply to the east of Penygelli where it has cut a rock gorge, the Upper Mule Gorge, between about 160 to 140 m above OD. The main, Lower Mule Gorge, has a depth of nearly 90 m, and is eroded mainly into solid rock, and partially into drumlins. The volume of water required to produce this gorge must have been enormous.

Details

Lacustrine deposits of the former Kerry Lake are exposed in river banks [SO 1782 9040] near Trefeen Bridge; 7 m of soft, grey, laminated clay occur at a height of about 160 m above OD and are overlain by a 0.10 m of sand and grit. This passes up into 5 m of ill-sorted subangular gravel (clasts up to about 150 mm) of the Miheli Fan Deposit [SO 1609 9005].

Sand, with a flat surface at 180 m above OD, fills the north–south-aligned preglacial valley at The Old Court [SO 1692 9111]. The northern limit of the sand is formed by a steep slope with a gradient of 1:5, dropping by about 12 m, and this is considered to represent the thickness of the body. The slope could be either a feature of erosion, which is unlikely, or depositional. If it is depositional, the slope is too steep to be a delta front, leaving the possibility that the sand accumulated against the ice barrier. The railway cutting [SO 1696 9105] affords a section up to 7 m high in which rounded, small gravel at the top overlies clean sand. The same constituents are evident around the fields as far south-west as Cwmderw [SO 1634 9054] at 180 m above OD.

Lake Llandyssil

Glaciolacustrine deposits west of Llandyssil lie within a large, almost dry valley which runs in a semicircle from the Severn Valley just north of Abermule to Llandyssil and rejoins the main valley 4 km downstream at Montgomery Station. The highest point in this valley is just north-east of Llandyssil, on the Llandyssil Fan Deposit. Without the late glacial and postglacial deposits, the valley would probably fall continuously from south-west to north-east and is probably an abandoned section of an earlier Severn Valley.

Details

Glaciolacustrine deposits form the flat area west of Llandyssil and probably extend eastward some distance beneath and perhaps into the silt and gravel of the Llandyssil Fan Deposit. Grey and orange, silty clay is exposed [SO 1843 9574] in the west; in the east, just west of Llandyssil, two boreholes [SO 1949 9542] and [SO 1921 9566] proved 14.3 m and about 18 m, respectively, of silt and clay with gravelly layers.

The minor stream which flows westward across this alluvial flat enters a small rock gorge west of Lower Bryntalch [SO 1797 9090], through which it descends steeply into the Severn Valley. The gorge probably marks the position of the ice-front which dammed Lake Llandyssil at one end. Presumably the ice-front dammed the opposite end of the dry valley for a while too.

Lake Kinton

Lake Kinton was a minor lake dammed on the western slopes of the Shelve area. Its presence is recorded by an oval area of lacustrine alluvium, some 0.5km long, occupying the northern end of the valley immediately south of Kinton [SO 289 994]. Sparse exposure of the deposit reveals orange and brown mottled clay, which is grey where fresh. Parts of the outcrop are not perfectly level, probably due to an accumulation of material soliflucted from adjacent slopes.

At a height nearly 200 m above OD, this deposit is higher than the lacustrine deposits of the Camlad to Caebitra valley and of the edges of Marrington Dingle just to the west on the same slopes. Thus, it seems that this lake was dammed by the main ice that fronted the western slopes of the Shelve area when it stood at over 200 m above OD, thus blocking the valley south of Kinton. A small gorge on its north-west side [SO 2880 9926] may have arisen as a marginal channel overflow.

Lake Crankwell

Lake Crankwell lay along the north side of the district and occupied the lower Camlad valley between Woodmoor [SJ 276 005], at the exit from Marrington Dingle, and Thornbury [SO 208 996]. At its eastern end, the lake deposits rise to just over 90 m above OD. To the west the highest ground between the site of the lake and the Severn Valley is only 81 m above OD, near Thornbury Roman Fort [SO 208 989]. Here morainic deposits separate lacustrine deposits from River Severn alluvium, forming an eminence clearly inadequate as a dam for the lake. An ice barrier, probably the Severn Valley Glacier responsible also for the morainic deposits, must have performed this function, and as it receded meltwater escaped northwards along its margin, or around its snout, so cutting the shallow gorge, the Forden overflow channel, between the railway [SO 217 993] and the old-ox-bow lake of the River Severn [SJ 210 003].

Exposure of the lacustrine deposits is sparse, restricted to ditch and river banks where some of the deposit must consist of recent river alluvium, for example 2 m of stoneless grey clay with brown mottles [SO 2165 9930] and 4m of brown clayey silt with some wood (possibly oak) [SO 2235 9924] and [SO 245 993]. Proximal to the exit of Marrington Dingle, the lacustrine deposits are likely to be gravelly. The total thickness of the deposit has not been established.

River terraces

These deposits are largely confined to the higher reaches of the River Clun, and the River East Onny. Small areas of undifferentiated river terrace are associated with the Severn and the Teme.

Three terraces have been recognised on the Clun between Badger Moor [SO 200 838] and Duffryn [SO 226 822], the highest being about 15 to 20 m above the alluvium. They are all composed of pebble gravel, mainly sandstone, with a sand matrix. Downstream between Newcastle and Clun and along the tributary Folly Brook north-west of Newcastle, two terraces are developed. The lower terrace is some 5 to 6 m and the higher one is about 8 to 10 m above the alluvium. They comprise pebbly and cobbly sands and silts.

A First river terrace, some 1 to 2 m above the alluvium is developed in places along the River East Onny, especially south of Hardwick [SO 370 906]. A Second river terrace, some 3 to 5 m above the alluvium, is present on the east side of the East Onny north-east of Hardwick. The terraces are composed of a pebble gravel; subrounded to subangular clasts of quartzite and purple sandstone are set in a purplish red silt matrix which makes up a high proportion of the deposits.

Fan deposits

Gravel

Fan deposits occur in many of the major valleys of north and mid Wales; within the district these are the Severn, the Camlad–Caebitra and Clun valleys. Fan deposits are fluvial in origin, and usually size-graded down slope, from gravel to sand and silt. Some display the shape typical of a symmetrical fan, but most are asymmetrically skewed downstream in the valley of entry. The Dulais Fan [SO 0885 9025], south-south-east of Newtown (IGS, 1972) is a symmetrical fan; to the east the Mule Fan [SO 1614 9556] is an example of one distorted by the flow in the major Severn Valley. They were formed during a short period in the final stage of decay of the last glaciation, when meltwaters were copious and enhanced seasonally or by dam bursts; run-off was rapid and unstable detritus was plentiful, fed by frequent landslips and mud-flows.

The fans were the deposits of energetic lateral tributaries and occur where the steep lateral gradients slacken into the floor of the major valley. This is also where the tributaries were released from confinement and thus able to spread laterally, further reducing rates of flow and load-carrying capacities. The thalweg of these deposits grades imperceptibly to the gentler thalweg of the contemporaneous floodplain of the major, probably braided, river, so that clear separation of the two sets of deposits is impracticable. At the end of the period, when run-off diminished, bed loads decreased and rivers retreated into single meandering channels eroding into their earlier deposits and creating narrower floodplains up to 3 m or so lower. The interval between the two levels differs from place to place.

Thus, beyond the reach of the new flood levels, the gravels, sands and silts of the early floodplain, input longitudinally, became river terrace deposits, but because they merge so smoothly with the laterally input fan deposits, lines on the map purporting to mark boundaries between them would be misleading artifacts. In such cases the fan deposit symbol has been used also on the adjacent river terrace. For instance the Mule Fan [SO 1650 4480], input from the south, and the Dolforwyn Farm Fan [SO 1614 9556], input from the north-west, are spread across and down the Severn Valley as one; likewise are the Aberbechan Fan [SO 1426 9347] and the Cil-gwran Fan [SO 1442 9292].

It is noteworthy that the fan deposits of the Camlad–Caebitra valley entered from the south and west, mainly at its western end, supporting the belief that the Devensian ice had little effect on the Shelve area. The fluvial deposits of the West Onny River, however, do reflect a late-glacial or early postglacial period of copious run-off from the highest parts of the Shelve area in that a fan of fluvial gravel, up to 3 m above the present flood-plain, was spread extensively between its apex at Linley [SO 348 929], Lydham [SO 340 910] and Lydham Heath [SO 354 904].

Silt/clay

Some fan deposits consist of silt and clay rather than sand and gravel. They rest on gentle slopes and occupy small drainage hollows that must have been eroded by local valley glacier meltwater. Their silt was sourced on valley sides as silty elutriant from till and carried down by meltwater inadequate for the removal of the clasts — pebbles, cobbles and boulders. These fans have steeper thalwegs than their gravelly counterparts, but they grade distally to the same levels.

The best fan deposit of this type is probably that which spread north-westwards from the foot of the till slope between Fridd Wood [SO 098 899] and Dingle Cottages [SO 107 901], Newtown. In its higher parts, it was deposited around hummocks and drumlins in the till, for example east of Plantation [SO 1000 9034], whence it fingers downslope between the remnant highs of the eroded surface of the morainic deposits [SO 0955 9038] and [SO 0973 9057] to merge with the Dulais Fan (gravel) in the region of the Llanidloes Road railway bridge [SO 0956 9065]. The source of this deposit of fines must have been the till, largely landslipped, on the slope to the south-east. The deposit was exposed in the sides of the brick pit [SO 1015 9070] as 1 to 2 m of fawn silt with scattered small stones. It was reported, by one who had worked in the pit, that about 1 m of gravel overlies the silt but this was not visible in 1969.

Fan deposits of the Camlad–Caebitra Valley

Although fan deposits are essentially fluvial in origin, in the Camlad–Caebitra valley their distal parts, at least, were probably lacustrine. Thus, the fans grade from ‘Fan Deposit, gravel’ in their proximal and early stage parts, probably largely fluvial, into ‘Fan Deposits, silt/clay’ of the distal and late-stage parts which may have been largely lacustrine, for example the Pentrenant Fan [SO 2394 9113], the Coed Cyfenni Fan [SO 2101 9032] and the Cwm Earl Fan [SO 1948 8984] and the small fans on the northern side of the Cwmberllan valley [SO 202 924]. Here again, lines on the map separating ‘lacustrine deposits’ from ‘fan deposits’ are rather poor artifacts, like the lines separating ‘Fan deposit’ from ‘River terrace’ in the Severn Valley.

It is clear that, during this period, the Camlad–Caebitra valley was supplied more abundantly and more vigorously with detritus from the west than it was from the east. Fan deposits are absent in the eastern, Camlad, half of the area and the surface of the lacustrine deposit there rises by only 10 m in 4.5 km between Church Stoke and Snead, while westward, along the River Caebitra, it rises about 10 m in 2 km. This is further indication that there had been little ice in the Shelve and Bishop’s Moat [SO 292 896] areas. Slope failure leading to solifluction may have been the contemporary processes active there [SO 317 908] — see under Head.

Alluvium

The only alluvial floodplains of any significance are those developed along the River Severn, the Camlad–Caebitra, East and West Onny, the Clun and its tributaries and the Teme. Some other small streams have narrow and discontinuous strips of alluvium associated with them.

Severn

Compared with the overall width of the valley, the floodplain of the River Severn is narrow, widening, to 1.5 km, only north-east of Abermule. The reason for this, downstream as far as Trwst-Llewelyn [SO 190 984], is probably the rock barrier across the valley there which has impeded the egress of floodwaters, so causing a build-up of the floodplain. North-west of Newtown golf course, where the river was marginal to the Severn Valley Glacier and was forced to establish a channel in solid rock, the flooplain is only 90 m wide [SO 1207 9234]. In a similar circumstance, just east of Milford, the floodplain is 70 m wide, being restricted partially by landslip from the north-north-west flank of the Reservoir Hill drumlin [SO 1007 9110].

The top 2 m or so of alluvium consist typically of silt and fine sand, representing the overbank deposits of the floodplain. Below this, and more particularly at and below normal river level, the alluvium is gravelly, representing the deposits of a meandering and constantly shifting channel. The total thickness of the alluvium is not quantified, for below it merges gradually downwards with late glacial or even glacial morainic deposits (see Valley-bottom drift). The alluvial gravel was derived largely from glacial deposits, and consists of conglomerate, rounded clasts of feldspathic sandstone, siltstone, mudstone and vein quartz.

Camlad–Caebitra

The Caebitra has an alluvial plain up to 200 m wide, the deposits being interstratified silt, silty clay, sand and gravel. The Camlad, east of Church Stoke, has a very broad alluvial plain, up to 700 m wide, which merges eastwards and westwards into underlying lacustrine deposits (p.117). Sections in the river bank show up to 2 m of structureless, stoneless, greyish brown clay and silty clay which may be partly of lacustrine origin. In the east, interstratified, thin, gravel beds are more definitely of fluviatile origin. From Church Stoke southwards, in Marrington Dingle, the alluvium is about 50 m wide and ranges from silt to gravel.

East and West Onny

In the upper reaches of the West Onny, the alluvium is up to 150 m wide. It is discontinuous between Linley and Lydham Heath, being interrupted by alluvial fan deposits. Downstream it broadens to 250 m and to 350 m east of the confluence with the East Onny. The deposits comprise light brown silt with thin gravel beds; clasts are derived mainly from the Ordovician rocks of the Shelve outcrop. The East Onny and its tributaries also have a well-developed alluvial plain; the deposits comprise purplish silt with a few thin gravel beds, derived from the purplish brown rocks of the Wentnor Group.

Clun

Upstream from Duffryn [SO 225 822], there is a narrow almost continuous strip of alluvium. This broadens downstream to 100 to 350 m wide, narrowing abruptly to 50 m where the river passes through morainic deposits at Clun. Downstream it widens gradually to about 600 m at Cwmlow. The tributaries of the Folly Brook and Unk also have narrow alluvial strips, while the Kemp has a wide alluvial plain, mostly in brown silt, between Brockton and Walcot Farm, narrowing as the river turns southwards towards its confluence with the Clun. In the upper reaches of the river, the Clun alluvium is a reddish brown silt, with some gravel beds, derived from the Clun Forest Formation. East of Newcastle and along its tributaries, the deposits are light brown and light reddish brown silt, fine-grained silty sand, sand and sandy gravel derived both from the Clun Forest Formation and lower beds within the Silurian. There are many exposures [SO 3584 8058] to [SO 3671 8046] south of Purslow showing brown silt, up to 2 m thick, locally with pebble beds, resting on hard cemented pebble gravel, seen to 1 m.

Teme

There is continuous alluvium from [SO 124 836] south-west of Cilfaesty Hill to the margin of the district [SO 210 786]. Sections up to 1.5 m in the river banks show medium- to coarse-grained, imbricate gravels overlain by 0.5 to 1 m of grey-brown silty clay. The two lithologies interdigitate.

Alluvial cone or alluvial fan deposits

These deposits were mapped as alluvial cones and are depicted as such on the BGS 1:10 000, 1: 10 560 and 1:25000 Series maps. On the 1:50 000 Series map they appear under the new name ‘Alluvial fan deposits’. An alluvial cone is essentially the small deposit of alluvium which is forming where a side stream, with a steep gradient, debouches onto the almost level floodplain of the main valley. It has a thalweg which grades downstream to the present-day level of the floodplain alluvium of that main valley and the two deposits merge. This definition is stressed to avoid confusion, under their new name, with ‘Fan deposits’. The latter were the products of side valley floodwaters which prevailed for a short period at the end of the last glaciation and ensued largely from melting snow and ice. Like river terraces to which the thalwegs of ‘Fan deposits’ commonly grade, these have suffered erosion ever since the melt waters abated, whereas ‘Alluvial fan deposits’ are still aggrading. The larger areas of these deposits occur at Pentre [SO 278 920], Linley [SO 349 927], Colebatch [SO 320 870] and near Clunton [SO 332 813].

Pentre

This area comprises two fans, one issuing from Cwm Cae and the other from the two valleys above Lower Cwm. Northwards the two fans merge, and the combined deposits merge northwards into the alluvium of the Camlad. Exposure is very poor, the best being south [SO 2599 9119] of Lower Cwm where 2 m of pebble-boulder gravel with a clayey silt matrix is seen. The gravel is crudely bedded and the clasts are mainly rounded slabs of siltstone.

Linley

This deposit extends for about 2.5 km southwards from the mouth of the West Onny valley at Linley. It cuts through the older morainic deposits and merges southwards into the West Onny alluvium. This area of alluvial fan deposits comprises uncemented, poorly sorted, pebble to cobble grade gravel with a silty sand matrix, commonly overlain by loamy silt. The gravel occupies a plexus of channels, with the silt as an overbank deposit, and has a fluviatile bedding fabric.

Colebatch

A large area of fluvial fan deposits covering about 1 sq km, extends eastwards from Goathouse Dingle to the alluvium of the River Kemp. There are three smaller alluvial fans in adjacent valleys [SO 320 859], [SO 312 850] and [SO 315 843]. A borehole [SO 3180 8301] at Red House Farm, Colebatch, in the upper part of the main alluvial fan proved 6.7 m of ‘stony marl’ resting, possibly, on solid rock.

Clunton area

There are several fan-shaped splays of silty gravel on the side of the Clun valley near Clunton. They occur south [SO 317 809] of Hurst, where exposures show imbricate blocks and pebbles in a grey silt near [SO 3180 8190] Hurst Mill, north [SO 342 809] of Clunton Coppice, at Meadows [SO 349 807]. Two others, which merge distally, occur at Greenfield [SO 341 814] and Quarry Brutches [SO 344 813].

At most of these sites the fans are associated upstream with a glacial meltwater channel floored by head. The coincidence of comparatively dry channels with fan-shaped areas of gravel, emergent downstream, strongly suggests a common origin from glacial meltwater and therefore the likelihood that the deposits are ‘Fan deposits’ rather than the more recent ‘Alluvial fan deposits’

Peat

Within the district, deposits of peat are limited to kettleholes in morainic drift and a few poorly drained areas, mainly on drift deposits.

Kettleholes in morainic drift to the south of Lydham e.g. [SO 3386 9063]; [SO 3367 9043] contain peat deposits exceeding 1.3 m in thickness. Similar sized kettleholes to the east of Newton e.g. [SO 3562 9118]; [SO 3565 9100], however, are infilled with lacustrine clay, probably because of their larger catchment areas.

The largest area of peat in the district is at Brookshill Marsh [SO 347 972] where it has formed on gently sloping till. There are patches on the moorland on the north side [SO 337 945] of Heath Mynd, on poorly drained gravel at More [SO 342 914] and in hollows in lacustrine clay west e.g. [SO 253 941] and south e.g. [SO 2630 9284] of Church Stoke. Elsewhere, peat has developed in hollows in till, as for example near [SO 186 866]; [SO 188 857] Riddings and between [SO 215 877] Ditch Dingle and Long Plantation.

Head

Head is a very variable deposit. It normally comprises angular rock debris in a clay or silty clay matrix. The rock debris may be derived either from local solid rock or from pre-existing drift deposits such as till. It was formed mainly under periglacial conditions at the end of the glacial period as a result of solifluction, but as mapped it may also include more recent slope deposits, hillwash and some alluvium. It is difficult to distinguish with certainty from till and indeed in some areas may include till remobilised under periglacial conditions. It is widespread, particularly in the south-eastern part of the district, from Clun Forest eastwards.

Much of the head of the district may have resulted during freeze–thaw conditions enhanced during the Younger Dryas cold period (11 000 Ma) which followed the main retreat of ice. The major deposit of head from Cwm Iago and Seven Wells (see below) may be of this period as were the late moraines at Rhiw-gam near Machynlleth (Cave and Hains, 1986) and the cirque moraines in the Brecon Beacons (Lewis, 1970).

Shelve and Norbury area

Thin head (up to 1 m thick) is widespread over areas where the solid rock is not obscured by other drift deposits. It is thick enough to be mapped only in a few restricted areas, as for example on parts of the Stiperstones ridge and the south-western slopes [SO 333 936] of Heath Mynd. On the Stiperstones there is a cover of bright orange silty clay with angular clasts of quartzite and shale. Other evidence of periglacial activity here is demonstrated by the development of stone stripes (Goudie and Piggot 1981), particularly on either side of the ridge between Cranberry Rock [SO 3656 9813] and Devil’s Chair [SO 3686 9917]. Further evidence of periglacial conditions were seen in many shallow excavations as for example [SO 3318 9708] south-west of Berth House and the Hollies [SO 3270 9388]. The Berth House section shows ice- wedges in head and both sections show that the solid rocks, shales and sandstones, have been progressively cambered downslope by mass-movement to lie almost horizontal at the surface. This effect is common on slopes, and shows that structural measurements on small exposures are liable to be misinterpreted.

South of Onny–Camlad–Caebitra valley and east of Grid line 20 N

Head is almost ubiquitous throughout this area; it is generally less than 1 m thick and has not been differentiated on the 1:50 000 Series geological map. The cover of head accounts for the general lack of natural rock exposures on the hill tops and valley sides. Mappable accumulations of head are present mainly in the lower parts of the valley sides and in valley bottoms where stream erosion has been absent or insufficient to remove them. The composition of the head is largely governed by the local bedrock, though in places till has become incorporated within it.

In the northern part of this area, there is a major drape of rock debris [SO 320 900] to [SO 295 920] below the Wenlock/ Ludlow scarp, comprising ill-sorted subangular clasts of sandstone, siltstone and mudstone, up to cobble size, in a clayey silt matrix. Up to 5 m is seen in banks north-west of Plas Madoc [SO 302 910]. Much of this debris appears to be derived from nivation cwms at Cwmago [SO 300 905] and Seven Wells [SO 307 902]. It appears to predate the lacustrine deposits at the base of the slope. Farther west, there are thick deposits of head in the valleys south-west of Cwm [SO 259 909] and south-east of Drewyn [SO 261 905].

On the northern side of Clun Forest, there is an extensive spread of head between Long Plantation [SO 200 879] and Pantglas [SO 247 895]. This blankets much of the Kerry Ridgeway and the slopes to the north and is one of the few areas where the head cover on the hilltops is thick enough to be mapped. Such head was formed in-situ during periglacial conditions and has not been affected subsequently. There are few sections, but in Cwmlladron [SO 238 893] over 3 m of angular siltstone fragments in a brownish grey clay matrix was seen.

Southwards and eastwards, there are head deposits along the lower parts of almost all the valley sides and in many of the smaller valley bottoms. These deposits are particularly notable along the valley of the misfit River Unk where the small stream has removed little of the periglacial debris. An example of head which incorporates glacial deposits is seen [SO 268 882] north-west of Mainstone where it contains various types of igneous rock, possibly derived from the Shelve area to the north during an earlier glaciation. It appears to have originated, in part, from till deposits which have been reworked by periglacial action.

Scree

Uncemented block scree occurs mainly on hills of igneous and volcanic rocks in the Shelve area. It is present on the eastern side of Todleth [SO 293 941] and Roundton [SO 295 949] hills, and the southern and eastern flanks [SO 311 962] of Corndon Hill. Scree of Stiperstones Quartzite Formation is present on the southern side [SO 343 956] of Black Rhadley Hill (erroneously depicted as head on the 1:50 000 Series map). Most of the scree formed under periglacial conditions and little material is being added at the present day.

Landslip

Of the many occurrences of landslip within the district, most of those recorded are in till. This is an unordered deposit, usually of unknown thickness, resting on an uneven surface. These factors combine to render instabilities on till slopes largely unpredictable and make an evaluation of them less precise than it is for those in well bedded formations.

A few landslips have occurred in solid-rock formations and thin beds of bentonitic clay almost certainly acted as slip planes in otherwise strongly lithified rock sequences. Such landslips are bedding-parallel rock-slides and do not possess the listric (concave) basal shear-plane characteristic of rotational landslip.

Landslip occurs predominantly in the north-west part of the district. This may be because till is more widespread and thicker there than it is to the south-east. However it is not implausible that, in some earlier period, repeated earthquakes centred along the Severn Valley Faults System helped to destabilise wet till, albeit via disruption of its internal drainage.

Landslips in till

Many slopes on till are disfigured by small step-like scars and raised lumps. These are symptoms of repeated, probably quite shallow (less than 10 m), rotational movements and serve to identify landslipped ground. The upper edge of such a slope is usually the steepest part representing largely unslipped till or even solid rock stripped of drift. The lower parts of the slope were commonly formed of lobes of flowing or sliding clay and in narrow valleys they have diverted the flow of streams. Slope-failure is not usually catastrophic though when there is human interference it can be. Normally, the process is incremental of small spasmodic movements.

The main period of slope instability occurred postglacially at the thaw of permafrost, during the seasonal melting of snow and perhaps higher precipitation. These processes introduced water into the till, especially along planes of weakness and layers of high porosity, and water is the main ‘lubricant’ of landslips. Subsequent conditions have been drier; a drainage system matured and vegation flourished, so that movements have diminished to current levels where events are usually insignificant and associated with hill-slope springs and periods of heavy rain.

The north-west slopes of Penarth Tack [SO 138 916] and the deep dingle on the north-east side where slipping continues, are typical locations for these landslips. Other examples are the slopes north-west of Fridd Farm [SO 1019 8962] and Highgate Rough [SO 1028 9560]. An active (1969) landslip occurs in a field north-west of Croes Forgan [SO 1028 8920]. Here the arcuate, west-facing, back-scar, some 100 m long, exposed up to 3 m of gravelly clayey silt. Below this, an 80 m-long slope is very uneven with arcuate ridges, troughs and minor scars, with an overall gradient of about 7°. The toe of the landslip is composed of homogenised mobile and cracked wet clay which was a mud-slide in places; it diverted the stream westward by several metres. This landslip is associated with the point where storm water from the lane above has been diverted on to the top of the slope!

The Ring Hole [SO 1191 8371] which has been sculpted from thick till, at the headwaters of the River Teme, is also an active landslip, but instigated in a very different way. Constant removal of till from the base of the slope, by the river, keeps the gradient of the slope above a critical angle of stability. The back-scar has thus retreated to within a short distance of the main road.

Steep linear slopes, preferentially those facing north-west, were the loci of most landslips and thus some drumlins have been left with arête-like summits where there is a preponderance of stones and boulders left after elutriation of the fines. The Reservoir Drumlin [SO 1008 9111] is one such with the reservoir perched upon its summit, above about 30 m of till. There are many other landslipped drumlins [for example 1452 9249], [SO 1568 9372], [SO 1630 9390], [SO 1173 9408] and [SO 0926 9621] depicted on 1:10 000 Series geological maps.

There is also a number of slips in till on valley sides [SO 135 868] and [SO 170 885] south of Kerry, and also at Neuadd-goch Bank [SO 104 836] and south-east of Felindre [SO 173 806].

Landslips in solid rock formations

Devil’s Elbow [SO 0910 8465]

Thinly bedded silty mudstones and sandstones, yielding graptolites probably of Wenlock age, dip northwards into the road and stream on the south side of Devil’s Elbow. The stream and latterly road works have undercut the steep slope removing support for strata above. Cohesion between beds was insufficient to hold them together and they slid. An investigation of the slide for Powys County Council in 1977, which included cored boreholes, revealed that several thin layers of clay are interbedded with the mudstones and sandstones and at least two of these were proved to have 15 per cent of smectite. Others were mainly of illite with some quartz and feldspar. Almost certainly these are the layers upon which the slope foundered and clearly two are bentonites. The others also are probably bentonites with a variable addition of rock-flour or other clay impurities introduced once sliding commenced.

Cefn-faesdre [SO 1306 9114]

Till on the north-facing slope south of Cefn-faesdre has suffered from landslips which have carried it beyond the main road to the bottom of Tower Brimmon glacial spillway. Near the top of this landslip [SO 9088 1290] fissile silty mudstones of the Dingle Mudstone Member are exposed with dips high, variable, and obviously disturbed within ‘moundy’ topography. The solid rocks here appear to have been caught up with a landslip mainly of till.

Lower Mule Gorge [SO 1666 9333]

Well-bedded sandstones and silty mudstones crop out in Fron Ffraith Wood on the south-east side of the gorge and belong to the Bailey Hill Formation. They dip, with the steep slope, to the north-west, and about 200 m of the newly constructed (about 1987) forestry road foundered on bedding-parallel slides. Thin bentonites are common in the rocks here too and there is little doubt that their weakness facilitated the movements. It seems likely that excavations made during the construction of the road, here near the top of the slope, exposed the strata allowing direct ingress of storm waters to weaken the bentonites, while loading with spoil from the excavations loaded the top of the slope.

Caer-din Ring [SO 237 848]

This extensive slip (Lynas, 1986) in the Cefn Einion Formation appears to have developed along a number of westerly-dipping bedding planes, forming a series of terrace features. There is a well-developed north–south-trending back scar with a spring line at its base.

There are several other small landslips in the Cefn Einion and Clun Forest formations [SO 248 852], [SO 278 816], [SO 260 813], [SO 258 793], [SO 272 792], [SO 311 798] all of which appear to have developed as shallow slips along bedding planes rather than deep-seated rotational slides.

Chapter 9 Structure

Although the district is not structurally complex, its three salient elements, stemming from a position on the margins of Welsh Lower Palaeozoic basins, are of regional importance. The first of these is the presence of Precambrian strata which give an insight into the early structural history of the Welsh Borderland and into the possible composition of the basement beneath the Welsh Basin. The second illustrates clearly the effects of Ashgill tectonism. The third is the Welsh Borderland Fault System, a series of faults which was influential in defining the eastern margins of successively subsiding Welsh basins. At least two of these faults, the Pontesford–Linley and the Church Stretton (Figure 34), have Precambrian origins and were present within the Avalonian continental crust at the start of the Palaeozoic Era and the third, the Severn Valley Faults, lies very close to the end-Caledonian fold-front.

The Welsh Lower-Palaeozoic Basin developed as an amalgam of marine depocentres on the thinned crust of Eastern Avalonia, a continental microplate on the south side of the Iapetus Ocean (Soper et al., 1987; Cope et al., 1992). Southward subduction of oceanic crust beneath Eastern Avalonia probably ended in late Caradoc times, possibly because of arc-to-continent collision (Campbell, 1984) and volcanism ceased at the same time (Cave, 1965). This effected a fundamental change so that the Welsh Basin became and remained a more singular depo-system of mainly siliciclastic turbidites. At about the same time the Caradoc marine transgression finally submerged the Midland Platform which lay to the east.

The Ordovician rocks of Shelve along with those of Builth and the Breidden Hills differ in many respects from those of the Welsh Basin. They form a narrow Shelve–Builth basinal strip which bordered the Midland Platform and lies largely within the Borderland Fault System. This basinal strip suffered inversion and folding, together with a marine regression in the late Ashgill, but it is not possible to determine how much the regression was due to tectonic uplift and how much was caused by the nearly contemporaneous Gondwanan glaciation. Within much of the district, the return of marine deposition did notoccur until late Aeronian (sedgwickii Biozone) times (Cocks and Rickards, 1969) and by then the youngest remaining Ordovician strata were of early Caradoc (Soudleyan) age. Such evidence would date the tectonic event imprecisely, but when traced westward into the main part of the Welsh Basin the magnitude of the unconformity decreases progressively, probably into Hirnantian horizons (Cave, 1992, fig.3; King, 1928; Brenchley, 1993). How much of the non-sequence below the unconformity is due to the tectonic event and how much is the result of the glacioeustatic fall is not known, since both events were initiated at the same time. Thus, dated as Hirnantian, this tectonic event is quite separate from the Ashgill (pre-Rawtheyan) deformation that occurred in the Welsh Basin, for instance around Bala (Bancroft, 1928, p.484), and from the earlier biostratigraphical non-sequence in the basal Ashgill (Pusgillian and Cautleyan Stage). Indeed, the end Ashgill tectonic event of Shelve–Builth had no structural expression within the Welsh Basin, but, in conjunction with the glacioeustatic fall, it gave rise to a massive and rapid transfer of sediment from shelf (which by then incorporated the inverted Builth–Shelve basin) to basin (Cave and Hains, 1986).

A difference in the depth of early Llandovery erosion on opposite sides of some of the Borderland faults, and thus to movement on the faults in Ashgill times, was noted by Woodcock and Gibbons (1988, fig.2, traverse a–a8). However, in this respect their figure is misleading; differential erosion across the Severn Valley Faults is insignificant. What little there is, between the nearest two exposures on opposite side of the faults, comes well within the compass of the progressive regional overlap of the transgressive Llandovery strata. The ‘Pen y Garnedd Shale’ and overlying Ashgill beds are in fact overstepped several miles west of the faults, so that Llandovery strata rest on rocks probably of Soudleyan age on both sides of the faults. On the other hand, the comparatively deep erosion between the Church Stretton Fault and the Pontesford–Linley Fault is factual, but attributable largely to periods of Cambrian and pre-Caradoc erosion. Only the residue is attributable to the early Llandovery and/or late Ashgill so that the difference in heights of the Pontesford Shale on the east side of the Pontesford–Linley Fault and the notional position (above ground level) of the Spy Wood Sandstone on the west side is the measure of post-Caradoc, displacement effected by the fault. Furthermore, it seems likely that this was largely dip-slip and not due to a 40 km-dextral strike-slip (Woodcock and Gibbons, 1988).

Further mild deformations, largely fault-related, occurred during the Silurian, particularly in the late Llandovery. They led to localised facies and thickness differences within Silurian strata, as recorded west of Llandrindod (Jones, 1947; BGS, 1993a) to the south and at Garn Prys (Cave et al., 1992; BGS, 1993b) to the north. However, the main deformation of the sedimentary fill of the Lower Palaeozoic Welsh Basin took place during the end-Caledonian, Acadian orogeny, which peaked in early to mid-Devonian times (Jones, 1956; Soper et al., 1987) producing pervasive folds and cleavage, and imposing low-grade metamorphism. In the Montgomery district, these dissipate eastwards across the Severn Valley faults and a line from Llanbadarn Fynydd to Kerry.

A point of note regarding the age of fault movements within the district is that, apart from some faults in the Shelve Ordovician rocks, they lack a stratigraphical ceiling. In those cases, there is the possibility that some or even all of the movement was of post-Caledonian age and the truth of this is illustrated by the Guilsfield Fault (Figure 34). This has a trend identical with that of the faults of the mid-Wales orefield, but north-eastward it is aligned with the Wem Fault which passes from Silurian strata into the Triassic rocks of the Cheshire Basin.

Folds

In Precambrian strata

In the north-east corner of the district, the Longmyndian strata (Linley, Bayston–Oakwood and Bridges formations) occupy the overturned limb of a major east-south-east-verging syncline (Whittard, 1952, p.149). The syncline is isoclinal and thus difficult to demonstrate. It has an axial trace trending north-north-east (Figure 3), cropping out from near Whitcot [SO 3770 9160] in the south to Bridges [SO 393 965] and beyond. Inversion of the westerly limit is demonstrated largely by cross-stratification and grading in thin siltstones of the Linley Formation [SO 3450 9416] and [SO 3443 9417]. In appropriate lithologies, for example the Portway Formation, there is a pronounced cleavage with a strike oblique to that of bedding and bedding/cleavage relationships assist in determining way-up. This is true particularly in the many minor isoclinal folds present (Pauley, 1990, p.346).

The folding of the Longmynd Syncline is of a Precambrian age, the evidence for which comes from just east of the district. There, Cambrian rocks rest with sharp unconformity on red grits probably of the Wentnor Group (Longmyndian) and likewise on Uriconian rocks (Greig et al., 1968, pp.94 and 95). Although such relationships are not exposed west of the Longmynd, both the Uriconian and the Longmyndian are involved in the Longmynd Syncline (Greig et al., 1968, fig.18) so that the folding must have been post-Longmyndian and pre-Cambrian. Pauley (1990, 1991) believed that the fold developed when a Uriconian magmatic arc terrane was juxtaposed with Longmyndian strata by sinistral strike-slip on the Church Stretton and Pontesford–Linley faults, in a late Precambrian to possibly early Cambrian regime of sinistral transpression.

In Ordovician strata

The Ordovician rocks of Shelve lie east of the end-Caledonian ‘front’, and in general dip westward at angles of 40° to 70°. A major fold-pair, the Shelve Anticline and the Llan Syncline (or Ritton Castle Syncline), interrupt the westward dips and it is apparent from the map that a monocline in the westerly dips lies beneath Corndon Hill (BGS, 1994, cross-section 1).

The hinge of the Llan Syncline can be traced between Lower Santly Wood [SJ 3530 0070], to the north of the district, and Maesissa Green [3167 9360], near Hyssington. Acid tuffs of the Hyssington Volcanic Member can be matched across this tight fold (interlimb angle about 80°). The hinge of the Shelve Anticline is much less well defined, because of extensive drift cover, and extends from the Hope Valley [SJ 3382 0053] to near Llanerch [SO 3040 9348], passing about 1 km east of Corndon Hill.

There is no expression of either the Llan Syncline or Shelve Anticline in the Llandovery and younger strata immediately to the south. The abruptness of this termination, like that of similar structures in the Ordovician inlier at Builth, illustrates clearly the unconformity between Silurian and Ordovician strata. The folds are part of the late Ashgill deformation.

If the generally westward dips in Shelve had been imposed regionally, for example upon the older rocks on the east, then the isoclinal Longmynd Syncline would have been recumbent originally, verging eastwards. However, this is implausible and it seems more likely that the Precambrian strata had become resistant to further folding and therefore acted as a buttress.

The Forden Inlier lies to the west of Shelve (Figure 1). The folds within it are tight and eastward-verging, and are very different from those of Shelve. The inlier is poorly exposed and way-up criteria are not common in the fine-grained rocks. At the southern end, west of New House [SO 2228 9298] and Stone House [SO 2237 9365], dips are consistently westward between 18° and 87°. However, steep eastward, to overturned westward, dips are present in a tract from Montgomery through Stalloe [SO 2240 9852] to Cwm Farm [SJ 2343 0053] and an old quarry [SJ 2361 0076]. To the west, in the railway cutting [SO 2183 9986], the strike varies between north-east and north-west with dips of 40° west to overturned 78° north-west, while near Munlyn [SJ 2110 0105] dips and strikes are even less consistent. At Thornbury [SO 2023 9932] and [SO 2057 9974], dips are lower, describing a gentle anticline plunging north-north-eastwards, and across the river to the west dips at Garthmyl are 18° east-north-east.

Relationships between the Ordovician rocks of the Forden Inlier and Llandovery strata are not exposed within or close to the district, but the nearest Silurian strata to the south, the Ludlow rocks west of Montgomery are also strongly folded suggesting that the Forden Inlier has been subjected to end Caledonian (Acadian) folding. This is supported at Buttington, 10 km to the north, where relationships are exposed and Caradoc strata are folded congruously with the overlying Llandovery.

In Silurian strata

Folds in these strata developed during the end-Caledonian deformation with a fundamental, but diffuse, structural divide crossing the district from Llanbadarn Fynydd, Dolfor, and thence north-eastward to Kerry and Montgomery (Figure 35). Regular folds, with north-east-striking axial planes, occur to the north-west of this line and mark the eastern limit of the pervasively folded Welsh Basin. Some persist for several kilometres and appear to be replaced en echelon by other folds, although undetected faults may account for the termination of some. The steepness of limbs varies greatly up to vertical and generally there is slight asymmetry with a south-eastwards vergence.

Illustrative of these folds are two outliers of Gorstian strata, including the Bailey Hill Formation. They have been preserved in two periclinal synclines 1 to 2 km south-east of Bettws Cedewain. The larger is a composite of two synclines arranged in echelon, while in the smaller the axis of the syncline has been skewed sinistrally into a north-east trend, against a curvilinear, possibly reverse, fault on its north-western side.

East of the Llanbadarn Fynydd–Kerry divide (Figure 35), the Silurian strata are mainly gently dipping, with a few sharp monoclinal flexures. The dominant structure here is the north–south Clun Forest Disturbance (Figure 34), a complex of ruptured and tightly folded strata dipping steeply both to the east and west, which bisects the area and is bordered on the east by the Mainstone Fault. West of the disturbance lies the almost bowl-shaped Clun Valley Syncline (Figure 35) and to the east is the smaller and more north-to-south-elongate Cefn Syncline [SO 288 838]. Like the similar Long Mountain Syncline in the Welshpool district to the north, they are occupied by outliers of Přídolí strata. Farther east, near the edge of the district, is the complementary Kemp Valley Anticline [SO 3600 8150], which has a gentle southward plunge.

Cleavage

The main fissility in the rocks of the district is bedding parallel. A slaty cleavage does occur in mudrocks of the Precambrian, and is also present sparsely in the Ordovician strata of Shelve, in parts of the Hope Shales. A spacedfracture cleavage occurs locally in the Mytton Flags; it is approximately vertical and strikes at about 15°. This compares with the 030° axial trend of the Llan Syncline and Shelve Anticline. Cleavage, approximately axial planar, occurs in the Silurian mudstones, north-west of the Dolfor–Kerry structural divide. This is mainly vertical and without a preferential vergence direction.

Major faults

Church Stretton Fault

This is one of the most important faults of the Welsh Borderland and although it lies just outside the district, a few hundred metres beyond the south-east corner, it exerted considerable influence over sedimentation during Silurian times (Figure 34). The fault can be traced from the vicinity of Wellington, where it splays north-eastward into Triassic strata, to just north of Brecon, where it enters the Brecon Anticline. Many geologists have continued the fault farther south-west, linking it with the Carreg Cennen Disturbance (for example Owen, 1974).

The fault appears to have affected Llandovery and early Wenlock deposition near Coston, at the south-east margin of the district [SO 386 802]. On its up-throw side, Wenlock strata rest directly upon Caradoc strata (Grieg et al., 1968), whereas they overlap Llandovery farther north and immediately west of the fault there is a thick Telychian sequence. This implies that the Church Stretton Fault was active in the late Llandovery (Telychian), a period of marine transgression, and that the transgression failed to overtop the footwall, possibly because of footwall uplift.

A similar situation occurs to the south of the district, at Old Radnor [SO 2420 5800], where early Wenlock limestones rest upon the Precambrian inlier separated by barely a metre of Llandovery conglomeratic sandstone (Kirk, 1951a, pp.56 and 57). Later in the Wenlock, the Church Stretton Fault separated calcareous, shallow marine muds and silts in the east, from deeper water muds, within the district. The disposition of these two facies suggests that the fault suffered syndepositional, normal, down-to-basin, movement. There is little evidence of appreciable strike slip, either then or since.

Activity recurred in Ludlow times, when the fault separated thin deposits on the east from a thick, rapidly accommodated, clastic sequence on the west. Indeed the footwall probably experienced submarine wasting (Lawson, 1973; Whitaker, 1962). The products of this may include the slumped strata which occur at the bend in the Church Stretton Fault where it joins the Leinthall Earls Fault, to the south of the district. Again, the facies changes and transport patterns are clearly elements of a single sedimentary system, implying normal fault displacements.

Long Mynd Scarp Fault (James 1956, p.883)

This fault, also known as the West Longmynd Boundary Fault (Whittard, 1932, p.331), enters the district at Lynchgate [SO 3700 8570] and curves southwards as far as Kempton [SO 360 830] where it is asymptotic to the north–south Kemp Valley Anticline. Here it cuts the Bailey Hill Formation, indicating late or post-Silurian displacements. However, there is also evidence of syndepositional control. A thick Llandovery sequence on the west side is without a basal Pentamerus Sandstone facies, while conglomeratic Pentamerus Sandstone does occur in the thinner sequence on the footwall (Whittard, 1932, p.331). The Wenlock does not overstep the Llandovery across the fault, but otherwise the fault behaved in the same way as the nearby Church Stretton Fault in the Silurian.

Pontesford–Linley Fault

The Pontesford–Linley Fault is another of the major north-north-east-trending faults of the Welsh Borderland. It juxtaposes the largely complete Ordovician succession of Shelve, on the west, with Precambrian (Western Uriconian) rocks overlain by a very incomplete Ordovician sequence on the east, where only the Caradoc is present. It is not a single fracture, but comprises a narrow, nearly vertical zone of gently curved, anastamosing fractures which embrace lenticular masses of Uriconian rocks. It has no appreciable affect on the base of the unconformable Silurian at Snead [SO 320 920] proving that movement on the fault was pre-Telychian (BGS, 1994).

The contrasts in the Ordovician sequences on opposite sides of the fault indicate large Ordovician movements of pre-Caradoc age; late Precambrian movements, similar to those on the Church Stretton Fault, also seem likely (Pauley, 1990a, 1990b, 1991). Other proposals have been advanced, but in some measure all are circumstantial or speculative. Grieg et al. (1968), for example, suggested that Longmyndian sediments accumulated on Uriconian strata in a graben between the Church Stretton and Pontesford–Linley faults. Woodcock (1984) considered that the fault was active intermittently from mid-Ordovician to Triassic times and that strike-slip was particularly important during the late Ordovician and early Llandovery in dividing the Shelve Ordovician ‘basin’ and shifting the eastern part south-westward by 40 km relative to the western part. In the proposal, the eastern part subsequently became the Builth Ordovician Inlier (Woodcock, 1984, p.331; Woodcock and Gibbons, 1984, p.917). However, the volcanic episodes of the two parts are of different ages and would be incongruous if juxtaposed. There is evidence of localised post-Llandeilo strike or oblique-slip within the Builth Inlier and the Shelve area, but such movements were probably not large. A simpler interpretation would be that post-Tremadoc uplift in the east, along with downwarping and normal or oblique faulting to the west, probably produced an Arenig shoreline in the vicinity of the western Long Mynd. The uplifted area then remained positive, accumulating little or no sediment, until the Caradoc transgression changed the pattern.

This later event produced an influx of coarse detritus to several areas marginal to the Welsh Basin (Cave and Rushton, 1996) and also oxygenated the Shelve depositional environment. However, neither phenomenon is a common effect of marine transgression.

Severn Valley Faults

Like the Tywi Anticline farther south, this belt of faults behaved as the margin of the Lower Palaeozoic Welsh Basin during end-Caledonian deformation, so that folding and cleavage are widespread to the west, and weak or absent to the east. Several other comparisons with the Tywi Anticline can be drawn, such as the general anticlinal nature of the belt and the sub-Telychian unconformity to the east, but not to the west.

Entering the district from the north, the main displacement on the Severn Valley Faults transfers near Garthmyl [SO 194 988], to the Montgomery Fault and a swarm of small, probably westward-dipping, faults and thrusts; these comprise the Dolforwyn, Glan Hafren and Cwmdockin faults, and the New House Thrust which branch from the point of transfer (Figure 34). The geometry of these faults differs greatly from the rectilinear pattern and near verticality of the rest of the Severn Valley Faults. They may be the products of stresses at the transfer point of the main faults during Acadian movements. South-westwards all these small faults swing away westward from the main Severn Valley, and splay into folded strata of Wenlock and Ludlow age. Satellite (LANDSAT) imagery of this area reveals numerous closely spaced, south-westward-trending ridges and gullies in the Silurian outcrop on the western side of the Severn Valley, but not on the east. Ground examination revealed that these are not stratal dip-and-scarp features. Like the small faults just described, they close northwards, and curve, into the Severn Valley. It is most probable that those features reflect other small faults and their characteristic curvature implies that the Severn Valley Faults extend northwards towards Llanymynech (Figure 34), rather than turning abruptly eastwards into the Wem Fault, as widely believed (Cave, 1995).

The New House Thrust is a small south-eastward-directed thrust, with a dip as little as 30°, which crops out [SO 1859 9775] in the Severn Valley, at the northern edge of the district. Nant-ysgollon Shales are exposed in the core of a tight hanging-wall anticline and the limbs comprise Mottled Mudstone. The thrust seems to separate the folded and mildly cleaved rocks to the west from relatively unfolded, non-cleaved rocks to the east. It probably also lies on the line against which the Rhuddanian marine transgression halted.

Faults on the line of the Severn Valley clearly existed during the Lower Palaeozoic because they constrained the Llandovery marine advance. (Cave, 1965, 1992). Slumped and inverted Rhuddanian (early Llandovery) strata near Berriew (Cave and Dixon, 1993), on the western side of the faults, suggest that they were not merely a passive topographical barrier, but there is no evidence that there was activity on the Severn Valley Faults in Wenlock and Ludlow times.

Despite their more northerly strike, Weston Madoc and Montgomery faults are also included in the Severn Valley Faults. They define the margins of a horst of Ordovician strata. Northwards, this links with the rest of the Forden Inlier, which is 1 to 2 km wide. The precise course of the Weston Madoc Fault is ill-defined, because of extensive drift, but its presence is indicated at the southern end of the inlier by the close proximity of lower Ludlow strata near [SO 2340 9310] Pentreheyling, and Soudleyan strata at Cwm Bromley [SO 2230 9370]. At the northern end of the inlier, less than 1 km outside the district, lower Wenlock strata near Crankwell Farm [SO 2365 9871] occur only 0.5 km east of exposures of Soudleyan mudstones. At both the north and south ends of the inlier, the Silurian strata on the east side of the fault are nearly horizontal.

Both the Weston Madoc and Montgomery faults throw Ludlow strata against Caradoc, and the latter has a downthrow of at least 1.4 km to the west, making it one of the largest faults in the district. Beyond the southern end of the Ordovician inlier at Hopton, where these two faults might be expected to cross the Ludlow scarp, there is but a single fault with small downthrow. The two faults are therefore thought to meet under the alluvium just east of Sarn [SO 2190 9130], effectively cancelling out their respective throws. South-west of Sarn in the Knighton district (Sheet 180) Wenlock and Ludlow rocks crop out on the line of the Severn Valley Faults where their only expression appears to be a group of minor north to south faults east of Cilfaesty Hill [SO 128 841]. These throw down mainly westward and may form a link with the Tywi Anticline and the major Cefncynfal Fault at Crossways [SO 145 741] which throws down eastwards.

North-east-striking minor faults are also widespread in the monotonously folded strata west of the River Severn. The majority are strike-parallel and to the north-east they too curve into the Severn Valley. Most cannot be tracedfar to the south-west and generally they have downthrows, both to the north-west and south-east, of less than 200 m. The exception is the Highgate Fault which extends from Berriew in the north-east to Upper Rhyd-y-felin [SO 0834 9313], with a downthrow of about 300 m juxtaposing Penstrowed Grits and Nantglyn Flags. Nearby parallel faults throw in the opposite direction, forming a minor graben. The Highgate Fault also lies directly in line with the mineralised Castell Fault of the Van Pericline [SN 920 875] of mid-Wales (Figure 34).

Faults of the Precambrian and Shelve Ordovician rocks

Two sets of small faults have been mapped in the Longmyndian Stretton Group with trends north-west and west-south-west respectively. They are very similar to faults in the Stiperstones Quartzite and Mytton Flags, 2 to 3 km to the west. Pauley (1990a, p.346) also described two sets of conjugate faults in the Stretton Group which postdate the Longmynd Syncline, but he recorded their trends as north to south and west-north-west to east-south-east respectively.

In the Ordovician rocks of Shelve there are two main sets of faults. North-north-east-striking faults are the more continuous. Some cut the inlier from end to end. However, there is no evidence, at either north or south end, that they pass into the unconformably overlying Llandovery strata. If they do, then only a small component of movement is post-Ordovician.

The second set of faults strikes north-west in the west, but close to east in the east. Most are short, usually under 1 km in length, and they do not offset folds, thus they lack appreciable strike-slip. Throws are small and down to the north-east and south-west. They are important because they are mineralised with sulphate and sulphide minerals (Chapter Eleven).

All the faults are depicted on the map as normal, dip-slip faults, but there may have been strike-slip on some of them. For instance, a north-east fault causes a small dextral offset of the Shelve Anticline at its north end (BGS, 1991). The Llan Syncline is also offset by a small fault at its southern end [SO 3200 9400] and by the Shelve Pool Fault to the north of the district (BGS, 1991). However, in most cases the evidence is ambiguous and the offsets can be explained by normal faulting.

Lynas (1985a, b, 1988) believed that the pattern of folds, faults and doleritic intrusions in Shelve was caused by late Ordovician dextral transpression. This fits with the views of Woodcock (1984a) who compared Welsh Lower Palaeozoic structure with the strike-slip tectonics of the Southern California Borderland. However, as outlined above, the evidence for strike-slip is tenuous.

Regional metamorphism

The appearance of cleavage west of the Severn Valley Faults and Dolfor–Kerry structural divide is reflected in an increase of low-grade metamorphism, although this increase remains small as far west as the Central Wales Syncline, around Talerddig (Bassett, 1955).

X-ray diffraction (XRD) analysis was used to determine the Kubler indices of white mica (illite) crystallinity for 19 mudstone samples ranging in age from Caradoc to Ludlow (Table 14). Sampling was mainly confined to areas around Montgomery and Newtown and is not necessarily representative of mudstones across the entire district. The Kubler index (KI in³.2U) measures small changes in the half-height width of the mica 10Å XRD peak which occur when authigenic clay/micas recrystallise in response to increasing metamorphic grade. Details of the sample preparation methods, machine and measuring conditions are given by Roberts et al., (1991). More than half of the samples show KI values greater than 0.42, indicating recrystallisation in the late diagenetic zone at temperatures no greater than 200°C. The remainder show values in the range 0.33 to 0.4 1, characteristic of low anchizonal conditions of recrystallisation at a possible temperature range of 200 to 250°C. A plot of KI values against stratigraphical age shows no evidence of a depth-related pattern of metamorphism (Figure 36). The inverted pattern revealed by the figure, with Ordovician mudstones generally at lower grades than those of Silurian age, may reflect the restricted numbers and area of sampling. However, similar relationships exist across the Tywi Lineament of the Rhayader district (BGS, 1993a), where diagenetic zone Ordovician mudstones are overlain by Silurian mudstones with grades ranging from diagenetic to low anchizonal. Low rates of subsidence across the Tywi Lineament probably prevented deep burial of the succession and the development of a depth related pattern of metamorphism (Merriman et al., 1992). A similar explanation may apply to the Lower Palaeozoic rocks of the Montgomery district.

Chapter 10 Geophysics

Gravity data

Gravity stations were measured within the district over a number of years as part of the national survey programme. This coverage was augmented subsequently (1984–1987), nearly doubling the station density to about 1 per 1 km2, to improve the definition of anomalies associated specifically with the Shelve Ordovician outcrop and with faulting in the Lower Palaeozoic sedimentary rocks elsewhere. Gravity values are referred to the National Gravity Reference Net of 1973 (Masson Smith et al., 1974) and the Gravity Reference System of 1967 (Woollard, 1979).

Measurements on samples of Precambrian and Lower Palaeozoic rocks from the Welsh Borderlands show a significant overlap in densities (Cook and Thirlaway, 1955; Powell, 1956; Brooks and Fenning, 1968) within the range2.67–2.78 Mg/m3. Palaeozoic mudstones from the deeper parts of the Welsh Basin have densities higher than those of typical ‘shelf’ deposits but the Montgomery district generally lies within a more transitional part of the sequence. A set of 38 density determinations made on samples (Entwistle, 1984), mainly of volcanic rocks, from the Shelve Ordovician outcrop gave relatively low values overall: calculated grain densities average 2.71 Mg/m3, although individual values range from 2.60 Mg/m3 for acid tuffs to over 2.80 Mg/m3 from the more basic specimens. The density of recent sediments is typically in the range 1.90–2.20 Mg/m3.

The Bouguer gravity anomaly map (Figure 37) was derived using a constant reduction density of 2.7 Mg/m3 and some topographical correlation will occur where this differs from the density of the rocks in situ: a change of 0.1 Mg/m3 will result in a discrepancy of 0.4 mGal over an elevation range of 100 m. Use of a small contour interval (0.25 Mgal) to bring out additional details is justified by the large number of stations and the accuracy of the data.

The highest Bouguer gravity anomaly values (G1 in (Figure 37)) are found over Precambrian rocks, east of the Long Mynd Scarp Fault at the eastern margin of the district. These form part of a positive axis which can be traced to the south-south-west for at least 20 km towards the Precambrian of Old Radnor. A reduction in the anomaly over the western part of the outcrop (G2) is consistent with the more arenaceous beds of the Wentnor Group, being less dense than the rocks of the Stretton Group and its underlying basement.

The zone of relatively high gravity anomaly values extending west from the Long Mynd includes much of the Shelve Ordovician outcrop, and it culminates locally within the district over the dolerite intrusion at Corndon Hill (G3). However, it provides no evidence of a density contrast between the Precambrian and Ordovician formations and no distinctive gravity anomalies are directly associated with the Pontesford–Linley Fault. A persistent south-west–north-east contour alignment from south-west of Kerry, and passing through Corndon Hill, crosses the Precambrian north of The Stiperstones, following a segment of the Shelve Pool Fault (G4). This trend, also seen in the contact between Precambrian and Carboniferous rocks to the north-east, aligns with the Ystwyth Fault in the south-west and is probably fault-controlled overall. Somewhat higher anomaly values to the north (G5) are associated with the northern limit of the Shelve Ordovician outcrop. There are further indications of a major structure within the main outcrop itself, such as the local Bouguer anomaly low over the Shelve anticline, but the relatively low amplitudes of the gravity anomalies generated and the lack of detailed station coverage over much of this open ground limits their definition.

Gravity values decrease west of the Shelve Ordovician outcrop into a well-defined trough (G6) with a graben-like form, bounded to the west by the Severn Valley faults (G7). It is aligned along the extension of the deep gravity low over the Mesozoic sedimentary rocks of the Cheshire Basin to the north-east, and to the south-south-west this zone continues through Beacon Hill to beyond Builth Wells. The thick Silurian sequences found south of Kerry and in the Long Mountain Syncline, just to the north of the district, lie within this zone and probably contribute to the observed gravity anomaly lows. However, the westwards gradient into the gravity anomaly low (G6) commences well within the Ordovician outcrop, above the steeply dipping strata between Corndon Hill and Marrington Dingle. This implies additional structure in the underlying Cambrian and Precambrian rocks. It may reflect the western edge of a block of shallower basement, or a change in basement type, associated with the igneous activity expressed at Corndon Hill and the volcanic centres of Lan Fawr, Todleth and Roundton. The gradients on the western margin of the main gravity anomaly low are partly taken up by a contour trend (G8) which diverges from the Severn Valley fault system and runs southwards for 20 km from Welshpool to link with the mapped faults near Montgomery.

A distinctive east–west gravity anomaly trend (G9) is seen only in the north-west of the district, with values decreasing steadily to the south as far as a line from Newtown to the southern limit of the Shelve Ordovician outcrop. This gradient may arise from variations in the underlying basement or a change in depositional environment within the Lower Palaeozoic. The occurrence of seismic activity (see below) close to its southern limit is consistent with there being crustal discontinuity at depth.

The decrease in gravity values south-eastwards across the Church Stretton fault system (G10) may reflect the transition to shallow-water deposits in the Silurian and a thinner cover on the Midlands microcraton which lies south-east of the fault system. Steep gravity gradients are developed across the fault zone close to Church Stretton itself, although the character of the anomaly pattern shows considerable variation overall, according to the nature of the near-surface geology which includes a variety of intrusions and Uriconian rocks.

The largest density contrasts within the district occur between Quaternary sediments and the bedrock, giving rise to a number of short-wavelength, Bouguer anomaly lows (G11a–d), most conspicuously in the valley of the Camlad (G11a). These lows were poorly defined or missed altogether in the regional gravity survey as they reflect relatively narrow channels. Much closer control on the form of the anomalies was obtained by detailed traversingbetween Church Stoke and Lydham (Figure 38) together with the additional infill.

Other short wavelength gravity lows seen in the Severn Valley are assumed to have a similar origin. The best examples are developed near Marton (G11b) where Aylesford Brook joins the Camlad, and bordering the district, to the east of Caersws (G11c) and south of Welshpool (G11d). These features may in part overlie older structural trends but their location was probably determined more specifically by the topography when they developed and their localised nature suggests that subglacial scour overdeepened the valley in places.

Magnetic data

Aeromagnetic data (Figure 39) were acquired during two separate surveys, flown in 1958 and 1960 at a nominal flying height of 305 m above ground level. The former used east–west flight lines separated by about 2 km, with some infill at 1 km; north–south tie lines at 10 km spacing provided control for the data reduction. The flight pattern, with a line spacing of 2 km, was rotated by 90° for the 1960 survey which covered all except the north-eastern corner of the district. These analogue data were digitised subsequently to allow further processing. A number of detailed ground surveys have also been undertaken to support the geological investigations, both to locate the aeromagnetic anomalies more precisely and to map the extent of Uriconian rocks beneath drift cover.

The Ordovician volcanic rocks of the Shelve outcrop are an important source of palaeomagnetic data; there are few other sites at which the rocks retain the magnetic imprint from this period of large-scale continental movement. The original studies of Piper (1978) were followed up more recently by McCabe and Channell (1990;1991) in an attempt to resolve ambiguities in the palaeogeographical position of southern Britain during the Ordovician. Further discussion provided by Trench and Torsvik (1991a, b) indicated that most of the available data could be reconciled, with the Shelve results representing an early stage in the closure of the Iapetus Ocean when northern and southern Britain were still separated by over 2000 km.

Ordovician volcanic rocks in this area are weakly magnetic in general (Entwisle, 1984) and only a limited number of the basic intrusions show a significant magnetisation. The sedimentary rocks also have a low magnetic susceptibility of about 0.5 3 10-3 SI. Results from the Church Stretton district (Brooks and Fenning, 1968) established that the Uriconian rocks accounted for the stronger magnetic anomalies, although their degree of magnetisation varied considerably.

The Church Stretton fault zone separates the Montgomery district from a region of more intense, short-wavelength magnetic anomalies associated with the shallower basement of the Midlands microcraton to the south-east. The clear expression of the fault here is due largely to a discontinuous belt of intrusions and steeply dipping Uriconian volcanic rocks: the aeromagnetic anomaly gradient (M1) represents the western flank of highs developed over these bodies.

A parallel belt of steep magnetic gradient (M2), some 12km to the north-west of Church Stretton fault zone, delimits the south-east margin of the magnetic low associated with the Cheshire basin. It may link to the Pontesford–Linley Fault although the gradient dies away rapidly to the south. Another axis of high magnetic anomaly lies some 10 km to the south-east of Church Stretton, with the source here buried beneath the cover of Lower Palaeozoic sedimentary rocks at a depth interpreted as 1500 to 2000 m (Brooks and Fenning, 1968). This pattern of parallel, large-scale features is consistent with regional models invoking analogies from North America, of accreted segments of Precambrian crust bounded by major strike-slip faults (Lynas, 1988; Woodcock and Gibbons, 1988).

A broad magnetic high (M3) centred near the south-west corner of the district forms part of a belt of anomalies which follows the margin of the Lower Palaeozoic Welsh basin southwards to beyond Builth Wells, and then westwards to reach the coast near St David’s Head. Strong magnetic anomalies are commonly associated with the more dense, basic rocks but in this case the gravity anomaly values are relatively low. Both the Uriconian and the Pebidian near Pembroke (Cornwell and Cave, 1986) give an analogous response and Precambrian basement of this type at a depth of several kilometres seems a more likely source than, for example, an extensive suite of basic intrusions. The alignments (M4a–c) of the contours flanking this anomaly are also reflected in the gravity data, the clearest example being between M4a and G12, and there is almost certainly a direct structural relation between them although the gravity anomalies probably originate from higher levels. The occurrence of another aeromagnetic high (M5) centred over the similar Silurian sequence in the Long Mountain Syncline supports the suggestion of an association between deep basement structure and Lower Palaeozoic sedimentation. There is no clear expression of the Welshpool–Montgomery and Severn Valley fault systems (M6 and M7) between these magnetic anomaly highs, suggesting that either they lie within a zone of weakly magnetised basement rocks or become less significant with depth.

There is little direct evidence of the Pontesford lineament in the geophysical data of the district, although they have been invoked to support geological ideas of its location (Woodcock, 1984b). The relation between the magnetic anomaly pattern and the Pontesford–Linley Fault depends on the configuration of the magnetic components of the Uriconian: where seen, these are preserved on its eastern side, but they could be displaced at depth depending on the geometry of the fault zone. An anomaly high (M8) with a width of about 1 km and an amplitude of just over 100 nT coincides with Uriconian volcanics lying against the fault to the west of Linley Hall [SO 3340 2930]. This anomaly attenuates rapidly to the north but a subdued ridge (M9) can be traced southwards for 10 to 12 km to Clun. This is perhaps surprising in that further outcrops of Uriconian are seen to the north while there are no inliers within the Silurian sedimentary rocks to the south. In fact, the aeromagnetic flight line records do show small responses of 2–10 nT along the mapped course of the fault (M10) which are lost in the contour presentation. Other large magnetic anomalies (M2), closely associated with a gravity ridge (G13), occur further to the north-east over the Uriconian and intrusive rocks found near Pontesford.

A number of distinctive local aeromagnetic anomalies are associated with the southern margin of the Shelve Ordovician outcrop. A short wavelength magnetic anomaly (M11), 3 km west of Linley, can be tied directly to the dolerite body which is exposed in the Disgwylfa Hill quarry. Ground-traversing showed a large anomaly with a width of 250 m just to the south of the quarry and indicated a lenticular form, terminating within 300 m to the north and tapering out to the south-west over a distance of more than 1 km. The gravity data also show a local anomaly high in this area.

Intrusive dolerite at Corndon Hill, which has been described as being in the form of a phacolith (Earp and Hains, 1971), is associated with an aeromagnetic anomaly (M12) of 125 nT and with a gravity anomaly of nearly 1.5 m Gal ((Figure 38)a). The aeromagnetic anomaly, which is almost certainly underestimated as no flight line crossed directly over the hill, is significantly larger than that of about 50 nT seen over a similar dolerite within the Ordovician outcrop of the Breidden Hills. Susceptibility measurements established that most of the exposed dolerite on Corndon Hill is only moderately magnetic. This may be due partly to alteration and weathering effects but the ground-traverse data also indicate a surprisingly subdued response overall. The latter survey delineated a strip of higher, variable anomaly some 100 m wide, running along the easternmost high ground, and another local anomaly high on the north-western flank of Corndon Hill. Discrete magnetic anomalies were also located near the standing stone by Mitchell’s Fold, 2 km farther north, and over Lan Fawr to the west of Corndon Hill.

Corndon Hill lies on a magnetic high which extends 10 km to the south-south-west, taking in Todleth Hill and Roundton (M13), and has some expression in the contour alignment for a similar distance to the north-north-east. There was evidence of near-surface magnetic rocks over the basaltic lavas and tuffs on Todleth Hill but not on Roundton. Lynas (1983) suggested that the Todleth and Lan Fawr members of the Stapeley Volcanic Formation have a similar, contemporaneous origin as volcanic islands and it seems likely that the more magnetic rocks represent a specific episode of this activity as, in general, the Ordovician volcanics are only weakly magnetic. There may also be an underlying source related to faulting: the continuation of the anomaly southwards across the Camlad does run into a fault mapped within the Silurian near Cwm Cae [SO 3277 2914] and a cross-cutting gravity trend (G4), near Corndon Hill itself, was mentioned above.

Seismic data

The district is notable for a number of felt earthquakes, one of which, in 1990, was of relatively high magnitude for the UK. Results obtained from monitoring of the more recent events are discussed in the section on seismicity below.

The LISPB deep seismic refraction profile (Bamford et al., 1976) passed close to the western margin of the district. An interpretation of the results (Manchester, 1983) shows a horst between the Pontesford lineament and Church Stretton faults in which the Precambrian basement lies within 0.5 km of the surface, contrasting with a depth of about 4 km outside the horst zone. Velocities increased from less than 5.5 km/s for the direct arrivals to 5.8–6.1 km/s at the first (basement) refractor. A mid-crustal refractor, dipping down to the south at an angle of 1.5° (4 km in 150 km), was also identified and this appeared to explain much of the regional gravity gradient across this part of Wales.

The hydrocarbon potential of the Lower Palaeozoic rocks has been largely discounted and no commercial seismic reflection surveys extend into the area. However, BGS commissioned a line which provides coverage from the south-east, across the Church Stretton faults, and terminates just beyond the south-western end of The Long Mynd. The data quality from these dipping, older sedimentary rocks was poor, with few coherent reflections, but an interpretation consistent with the generalised model of Coward and Siddans (1979) was presented (Smith, 1987) implying that the Church Stretton faults are linked to a thrust plane dipping down to the north-west.

Detailed seismic refraction studies (see Chapter Eight, p.116) over Quaternary deposits in the Camlad valley, south of Church Stoke, indicated the presence of channels, 500 to 1000 m wide, in which depths to bedrock typically reach 70 m (maximum of about 100 m). The seismic velocities of the channel fill are very uniform at 1.5–1.6 km/s, consistent with their representing lacustrine clays of the type encountered in the shallow shot holes underlain by typical channel deposits; more variable refractor velocities, of 3.75±0.6 km/s, suggest that the bedrock in which the channel lies is weathered or highly fractured: laboratory measurements of sonic velocities on Ordovician sedimentary rocks gave values close to 5 km/s.

Interpretation

Combined two-dimensional modelling of the magnetic and gravity data along a profile close to the line of Section 1 on the 1:50 000 Series map illustrates schematically the main components needed to match the geophysical responses (Figure 40). A datum shift of -90 nT has been applied to the magnetic data and Polygon 1 is a magnetic basement needed to match the regional variation; a mid-crustal source is included in the model to account for the change in background gravity anomaly values along the profile. The gravity anomaly low seen in the central part of the profile is modelled by Polygon 2, which is assigned a density slightly less than its surroundings. The shape of this body is controlled by the main faults and the margin of the Shelve Ordovician outcrop, and mainly represents the effect of post-Llandeilo rocks east of the Montgomery Fault. Some contribution to the gravity anomaly low is also needed from Quaternary deposits in order to account for the sharp curvatures near the fault zones.

The source of the Corndon Hill anomalies (Polygon 3) can be modelled in a number of ways. The absence of any clear aeromagnetic low around Corndon Hill supports the view that the magnetic component of the intrusion has a significant depth extent, but remanent magnetisation also makes an important contribution to the anomaly pattern. The direction of the remanence vector deduced from the modelling was in good agreement with the values quoted by Piper (1978) for the intrusion: declination of N95°E, inclination 40°. The Polygons 4 and 5 at the eastern end of the profile represent the Stretton and Wentnor groups respectively of the Longmyndian, with a higher density assigned to the latter. It is noticeable that small changes in both the gravity and magnetic profile data occur across the Pontesford–Linley Fault, although these have not been modelled explicitly.

Gravity modelling (Figure 38)b of the Quaternary channel beneath the Camlad shows that a density contrast of about 0.55 Mg/m3 gives the best agreement between gravity and seismic interpretations; this implies a typical clay density of 2.15–2.2 Mg/m3 for the channel fill. A small residual aeromagnetic anomaly which coincides with the gravity low in (Figure 38)b appears to be genuine, although its amplitude is close to the sensitivity limits of the survey; this may represent a slight concentration of magnetite derived from the adjacent volcanic and intrusive rocks.

Seismicity of the country around Montgomery

Contemporary seismicity in the district is dominated by the magnitude 5.1 local magnitude (ML) Bishop’s Castle earthquake of April 1990 which was felt over a wide area of Britain. Damage was minor and the maximum intensity in the epicentral area was 6 EMS (European Macroseismic Intensity Scale; (Table 15)). Slip on the causative fault was by dominant strike-slip motion at the focal depth of 14.1 ± 3.8 km

The magnitude 5.4 ML Lleyn event of July 1984, although with an epicentre 100 km north-west of Montgomery, represents one of the most significant earthquakes to affect the district, and was felt strongly throughout the area at intensity 4 EMS. Fault movement was also strike-slip but with a deep-crustal focus at around 20 km.

Several significant historical earthquakes have also occurred near the Mongomery area, at Hereford and Ludlow. The area in general is thought to have higher levels of seismicity than most of the United Kingdom.

Historical seismicity

Historical seismicity is considered, along with contemporary seismicity, to provide a more complete record of regional earthquake activity in the UK and a more accurate seismic hazard assessment. The global pattern of seismicity correlates closely with the position of known plate boundaries; however, since the UK lies within a plate and the Mid-Atlantic Ridge is some distance away, it does not suffer from devastating inter-plate earthquakes. It does, however, still have moderate levels of seismicity and has the potential for damaging earthquakes. The return periods for these infrequent, larger events are much longer than in high-seismicity regions, hence the need for a complete historical record. For example, the UK could expect an event of magnitude 5.6 ML or larger approximately every 100 years, with a maximum expected magnitude of around 6.5 ML. The macroseismic magnitude of historical earthquakes, or those events which occurred prior to instrumental monitoring, is estimated by modelling magnitude linearly against the logarithm of the isoseismal or felt area. The algorithms used are described in Musson (1993).

The BGS historical database contains data for the general Montgomery area for the period 1700 to 1969, before the start of instrumental monitoring. Prior to 1700, the catalogue is incomplete and only some of the larger events, above magnitude 4.0 ML are well represented (Musson, 1994). For the period 1700 to 1969, the catalogue is thought to be complete for events above magnitude 4.0 ML, and includes many events above magnitude 3.0 ML that were felt on mainland UK.

Two small historical events are known to have occurred in the Montgomery area.

Larger earthquakes which occurred in the general region include the Hereford event of 6 October 1863 which occurred south-west of Hereford in the Golden Valley area (Musson, 1994). The felt area was extensive and covered most of England and Wales, south of the Lake district. The maximum intensity was 6 EMS in the epicentral area and the magnitude was about 5.2 ML.

On 17 December 1896, an event of magnitude 5.2 ML, was felt over most of England, Wales, and the east of Ireland (Musson, 1994). The maximum intensity was 7 EMS in the epicentral area, about 6 km east-south-east of Hereford. A sequence of foreshocks occurred prior to the mainshock which was followed by some weak aftershock activity.

On 15 August 1926, an earthquake with a magnitude of 4.8 ML, occurred close to the villages of Little Hereford and Tenbury Wells, east of Ludlow, and was felt from Plymouth to Hull (Musson, 1994). The maximum intensity was 6 EMS in the epicentral area.

Contemporary seismicity

The BGS earthquake database contains data for all instrumentally recorded seismicity to the present time. It was searched from 1 January 1970 to 1 September 1995 for the district.

The completeness threshold was approximately magnitude 4 ML in 1970, improving to magnitude 2.5 ML in 1990 for mainland UK. This reflects the expansion of BGS monitoring networks in the UK over the years. Elevenearthquakes, ranging in magnitude up to 5.1 ML, were detected within the area: the Bishop’s Castle mainshock, seven aftershocks and three small events representing background seismicity. The earthquakes are discussed in chronological order below, with the Bishop’s Castle sequence treated together.

The ‘Felindre’ earthquake of 15 April 1984, at magnitude 3.3 ML, occurred at a mid-crustal depth of 12.7 ± 8 km, to the south of the town of Felindre in Powys (Musson, 1996). It was felt from north of Newtown to the south of Llandrindod Wells and from Rhayader to Knighton. Following a macroseismic survey, the maximum intensity, locally at the epicentral area, was assigned 5 EMS. Events of magnitude 3.0 ML and above occur, on average, three times a year somewhere in the UK.

The largest earthquake in the Montgomery region, and one of the largest in Britain this century, was the magnitude 5.1 ML Bishop’s Castle earthquake of 2 April 1990. The initial earthquake hypocentral parameters were published in Ritchie et al. (1990); however the earthquake hypocentre, magnitude and focal mechanism have since been revised. More detailed analysis was carried out using digital data rather than the original analogue data. The epicentre of the earthquake was near the village of Clun, 7 km south of Bishop’s Castle [SO298 826 (±1.1km)] and the focal depth was 14.1 ± 3.8 km. The magnitude, 5.1 ± 0.4 ML, is an average of readings from three low-gain, vertical seismometers, as high-gain horizontal instruments had saturated and no low-gain horizontal instruments were available. The magnitude is, therefore, a conservative estimate. The event was felt over a large area, from Ayrshire, in the west of Scotland, to Cornwall in the south and Kent in the east to Dublin in the west. Results from a BGS macroseismic survey carried out at the time revealed the extent of the felt area, 245 000 km2 at intensity 2 EMS, and a maximum intensity of 6 EMS in the epicentral area. Although felt widely, damage was minor and limited to the epicentral area, Wrexham and especially Shrewsbury. Damage consisted of cracks in chimneys, cracks in plaster and falls of small amounts of plaster.

Following the mainshock, a dense network of monitoring stations was installed, during April 1990, to fully examine any possible aftershock sequence and improve the constraint of the locations. Only seven aftershocks were detected in the four months following the mainshock, with one small event on 5 April too small to be located (Table 16). The largest of the aftershocks, at magnitude 1.5 ML, occurred the day after the mainshock. The aftershock of 17 April 1990, with a revised magnitude of 0.5 ML, was located, using the recently installed network in the area [at 3309 2843 ± 0.6 km] and at a focal depth of 15.3 ± 2.1 km. The distribution of the small number of aftershocks, including error circles, shows a roughly north–south alignment and may delineate the causative fault (Ritchie et al., 1990).

The small number of aftershocks following the Bishop’s Castle mainshock suggests that most of the strain energy was released by the mainshock and may reflect a high stress drop. The higher frequency content of the Bishop’s Castle seismogram relative to that of the magnitude 5.4 ML, 1984 Lleyn earthquake agrees with a high stress drop (Ritchie et al., 1990; Turbitt, personal communication). The absence of a marked aftershock sequence contrasts with the majority of events of similar magnitude in intraplate situations. The 19 July 1984 Lleyn earthquake was followed by an extensive aftershock sequence, with 20 aftershocks per week with magnitudes greater than 1.0 ML occurring in the four weeks following the mainshock (Marrow and Walker, 1988), and related activity is still being detected over ten years after the initial earthquake.

Earthquake focal mechanisms are a basic tool in the investigation of both regional and local tectonics, providing information on the nature of the ambient stress regime, local transient stress conditions and the style of fault movement. In the UK, only the infrequent, larger events usually have enough polarity data to produce well-constrained focal mechanisms; however, solutions have also been recently obtained for some smaller events using SV/P amplitude ratios to tighten the constraint of the focal planes (Snoke et al., 1984).

Preliminary focal mechanisms were published for the Bishop’s Castle mainshock and one of the aftershocks (Ritchie et al., 1990). Both mechanisms have, however, been revised using improved analysis techniques. The focal mechanism for the Bishop’s Castle mainshock (Figure 41) was obtained using Kisslinger’s (1980) technique and a computer package by Snoke et al. (1984). Input to the program consisted of 56 P-wave polarities from stations in the UK, plotted on the upper focal hemisphere, with 21 possible solutions obtained, allowing for one polarity error at station YRH on the Lleyn Peninsula. This anomalous P-polarity is thought to represent a source-effect caused by non-double couple motion (Ritchie et al., 1990). The mechanism obtained represents dominant strike-slip faulting with a component of reverse or normal movement on either a north–south or east-north-east-striking fault plane. The two sets of orthogonal nodal planes dip west and south-south-east, respectively. The orientation of the three principal stress axes, with maximum compression acting north-west–south-east, agrees with that commonly observed for the UK. Initial studies of focal mechanisms obtained for UK microseismicity suggest a dominance of strike-slip mechanisms with an axis of maximum compression horizontal and orientated approximately north-west–south-east. This is a result of the pressures acting on Britain as a result of combined forces generated by spreading at the Mid-Atlantic Ridge and the compressional forces resulting from the movement of Africa (Ritchie and Walker, 1991). Recent studies, however, show an increased number of reverse faulting mechanisms with the maximum compressive axis also horizontal and orientated north-west–south-east.

A revised focal mechanism was also obtained for the magnitude 0.5 ML aftershock of 17 April. Input consisted of 8 P-wave polarities and 3 SV/P ratios and one solution was obtained. The mechanism also represents dominantstrike-slip faulting about either a north–south-or east–west-striking fault plane. The two sets of planes have almost vertical dip. The mechanism is similar to that of the mainshock and the principal axis of maximum compression is approximately horizontal and orientated north-west–south-east.

Resolving the ambiguity as to which of the two nodal planes represents the fault plane is not possible by examination of the focal mechanism alone. Surface fault breaks, usually only available for interplate or large, shallow intraplate earthquakes, can be correlated with the focal planes to identify the direction of the fault plane. The epicentral distribution, of the limited Bishop’s Castle aftershocks, trends approximately north–south and is consistent with one of the planes of both the mainshock and aftershock fault-plane solutions, suggesting this could be the orientation of the fault plane (Ritchie et al., 1990). The fault radius for a magnitude 5.1 ML event would, however, only be of the order of 300 m (Burton and Marrow, 1988) and any weakness at the focal depth, around 14 km, could be reactivated without any surface expression.

Further background seismicity in the area includes a small, magnitude 0.1 ML, earthquake which occurred approximately 7.5 km north of Montgomery on 6 April 1990 at a depth of 5.3 km. Another small, magnitude 0.9 ML, earthquake occurred at the southern limit of the district on 1 January 1995, about 8 km north-west of Knighton, Powys and at a depth of 12.6 km. Approximately 140 earthquakes of magnitude 1.0 ML and above occur in the UK every year.

Earthquakes since 1970 affecting the Montgomery region

A large, magnitude 5.4 ML, earthquake occurred on 19 July 1984 on the Lleyn Peninsula at a focal depth of around 20 km (Turbitt et al., 1985). It was one of the largest earthquakes in Britain this century and was felt over a large area, around 250 000 km2 at intensity 3 EMS. The maximum intensity in the epicentral area was 6 EMS and it was felt in the Montgomery area at intensity 4 EMS. The mainshock was followed by an extensive series of aftershocks, concentrated around the mainshock. The aftershock distribution defined a thick planar zone, trending east-south-east and dipping steeply to the north-north-east at between 19 km and 24 km (Marrow and Walker, 1988). The deep-crustal activity is relatively unusual for the UK as most seismicity is concentrated in the upper 15 km of the brittle crust. The focal depth suggests that the base of the seismogenic zone is depressed on the Lleyn Peninsula relative to the rest of the UK and infers either a low geothermal gradient or unusual crustal composition. A focal mechanism for the mainshock represents dominant strike-slip faulting with a component of normal motion about either an east–west- or north–south-striking fault plane. The axis of maximum compression is almost horizontal and orientated north-west–south-east and is in agreement with that obtained for most of the UK. The east-south-east-trending planar zone, delineated by the aftershock foci, correlates well with the orientation of one of the planes on the mainshock mechanism and may represent the fault plane. There is, however, no surface fault or feature on the Lleyn Peninsula which corresponds to this plane.

Two other significant earthquakes occurred just outside the district. On 24 August 1975, a magnitude 3.5 ML event occurred at Hereford, with a maximum intensity of 4 EMS in the epicentral area. It was felt over an area of some 400 km2 at intensity 3 EMS.

A magnitude 3.1 ML earthquake, followed by two small aftershocks, occurred on 17 March 1994 at Newtown, Powys, just west of the district (Ritchie et al., 1995). It was felt over an area of 1100 km2, at intensity 3 EMS, and the maximum intensity was 4 EMS in the epicentral area. The focal depth was 21.6 ± 4.4 km, in the lower crust. This unusually deep focus combined with low heat flow data for the area may indicate a depressed geothermal gradient. Analysis of the body-wave spectra provided more details on the focal parameters: the seismic moment was estimated as 3.1 3 1020 dyne cm, moment magnitude, 3.0 and the fault radius was approximately 155 m. The focal mechanism represents dominant reverse faulting, with a component of strike-slip movement on either a north-north-east- or east-north-east-striking fault plane. The direction of the axis of maximum compression is horizontal and orientated north-west–south-east, in agreement with that usually observed for the UK.

An earthquake with a magnitude of 3.0 ML occured on 20 September 1996 at 04:04 UTC at Llanddewi Ystradenni, approximately 9 km NNE of Llandrindod Wells at 309.4 kmE and 269.7 kmN ± 1.0 km and about 9 km due south of Llanbadarn Fynydd. The event had a focal depth of 14.4 ± 2.7 km and was reported to be felt in Llandrindod Wells, Knighton, Rhayader, Builth Wells and the village of Llanbister. People as far as Knighton, 20 km from the epicentre, reported being awakened from sleep and other observations describe shuddering, house shaking and windows rattling, indicating an intensity of at least 4 EMS.

A focal mechanism was attempted using 15 P-polarities, four compressional and eleven dilational arrivals, together with a single SV/P amplitude ratio from a low gain instrument. Other nearby high-gain instruments had saturated and could not be used in SV/P ratio calculations. A poorly constrained mechanism was obtained with many possible solutions and is, therefore, not presented. The solutions are, however, consistent with a north-west–south-east maximum compressive stress direction generally observed for UK earthquakes.

Chapter 11 Economic geology

At the present time there are no significant commercial mineral workings within the district. Formerly, the West Shropshire (Shelve) mining field was intensively exploited for lead, zinc, barytes and calcite; sandstones, volcanic and intrusive rocks were used for building stone and aggregate; Quaternary clays were used in the manufacture of bricks, tiles and pipes.

Metalliferous minerals

The lead, zinc and barytes ores of the West Shropshire (Shelve) mining field are the only significant metalliferous mineral deposits in the district. These deposits were exploited by the Romans and were actively worked in the12th and 13th centuries. During the early part of the 19th century many mines were active but there are few detailed records apart from those given by Murchison (1839) and Morton (1869). Published records for lead productionstart in 1845 and later for zinc and barytes. Mining activity declined rapidly in the early part of the 20th century, though barytes production continued until the 1940s (Dunham and Dines, 1945). In addition to the descriptions of Murchison and Morton, the mining operations have been described by Hall (1922), Smith and Dewey (1922), Wilson et al. (1922), Dunham and Dines (1945), Dines (1958, 1959), Adams (1962, 1968) and Bailey (1977). Ofthese, Dines (1958) and Adams (1962, 1968) give the most detailed accounts of the mines.

Mineralisation is mainly located along two sets of faults trending approximately north-west and east-north-east, and the main ore bodies are commonly located at fault intersections, as at Tankerville [SO 356 995] and Bog [SO 358 978] mines. There appears to be only one period of mineralisation. Dines (1958, p.5) has pointed out that although the mineral deposits are locally brecciated by subsequent movements on the faults there is no evidence of recementing by later mineralisation. The mineralisation was dated first by Moorbath (1962) and later by Ineson and Mitchell (1975) who presented strong evidence that it was late Devonian in age.

The location of the deposits is clearly determined by the host rock lithology. The mudstones remain largely sealed along the faults while the flags, quartzites, volcanic and intrusive rocks are more brittle and have fractured, with cavity formation. All the ores are fracture fills and occur predominantly in the Mytton Flags Formation, below the thick mudstone of the Hope Shale Formation. It is possible that the hardening of parts of the Mytton Flags Formation, by hornfelsing adjacent to numerous dolerite intrusions, may have allowed generation of more fracturing than if the dolerites had not been intruded (Lynas and Hains, 1985). In places, the dolerites themselves havealso been mineralised. The few small-scale occurrences of ore, mainly barytes, present in strata above the Hope Shale Formation are in the Stapeley Volcanic Member, sandstones in the Weston Flags Formation and in the Hagley Volcanic Formation. There are no workable lead and zinc ores east of the outcrop of the Stiperstones Quartzite Formation. At Shuttock’s Wood, near Norbury, there are three shafts in the Pentamerus Sandstone Formation, reputedly sunk for barytes. From tip material it appears that at least one shaft [SO 3730 9239] penetrated into the underlying Longmyndian rocks. Traces of malachite and barytes are present in spoil. These may be the shafts in the Norbury district which were mentioned by Murchison (1839, p.261) (see also Greig et al., 1968, p.322).

Moseley (1991; 1994) has recorded the occurence of haematite, with traces of malachite, azurite and bornite, in the Stiperstones Quartzite at Manstone Rock [SO 3675 9859] and elsewhere, and haematite in the Linley Formation at The Knolls [SO 3711 9741]. He considers that the iron was derived mainly from an iron-rich sedimentary cover of Upper Carboniferous or Triassic age; a possible subsidiary source is from the oxidation of local pyrite and chalcopyrite and the alteration of dolerites.

Bailey (1977) concluded, from fluid inclusion studies, that the mineralising fluids varied in temperature between 135°C and 160°C and that the deposits showed remarkable similarities to the Mississippi Valley orefield of the USA (p.184). He suggested that late faulting ‘allowed the release of connate brines from deep in the sedimentary basin. These brines migrated up dip until eventually becoming trapped under the unconformity at the base of the Silurian or beneath impermeable shales in the crests of anticlines’. He thought that the heavy petroleum residues commonly found along joints in the Stiperstones Quartzite Formation (Lynas, 1985a, p.5) and associated with the mineralisation (Dines, 1958, p.23) were the residue of petroleum which had ‘migrated along the same channels, eventually being trapped at high levels in the same areas as the hydrothermal solutions’ (Bailey, 1977, p.187). Bituminous residues occur widely in the region, commonly filling voids in fossil brachiopods. Pattrick and Boswell (1991) described the sphalerite stratigraphy of the orefield but could not positively identify the source of the mineralising brines.

In the early 1980s, geochemical work carried out by BGS in the area around Todleth, Lan Fawr and Hagley (Lynas, 1985b, p.22; Hains and Lynas, 1985, p.17) and also around Cefn Gunthly and Pellrhadley Hill (Cave, et al., 1985, p.18) indicated that low-grade, stratabound, sulphide deposits may be associated with the volcanic formations. Geochemical research was also conducted by A P Garnett, University of Leicester, in 1983 to test the area of the presumed southward extension of the Shelve Anticline for vein-type mineralisation which might be associated with a postulated subcrop of Mytton Flags Formation concealed by Quaternary deposits between Corndon Hill and Hyssington. The work was hindered by the drift deposits but some anomalies were found (Hains and Lynas, 1985, pp.17–18).

Brick and tile clay

There are no major deposits of brick clay within the district, though some drift deposits have been utilised locally. The most significant workings were at Goetre Brick Works [SO 174 921], about 3 km south-south-east of Abermule, where until recently the till (Brown, 1971, p.99) was worked for the production of bricks and tiles. Bricks, tiles and pipes were produced from pits [SO 3263 9175] at Roveries and 500 m east-south-east [SO 3358 9149] of Lower Aston in lacustrine silty clays (Cave et al., 1985, p.22). The lacustrine clays were also exploited [SO 267 940] west of Church Stoke (Harmer, 1907).

Rock aggregate, building and walling stone

All rocks of reasonable hardness have been exploited in the past, mainly on a small scale, but there are now no working quarries in the district. Small quarries are still opened from time to time to supply local needs such as aggregate for farm tracks and forestry roads.

Precambrian Massive purple sandstones, locally conglomeratic, of the Bayston–Oakswood Formation have been quarried near [SO 3526 9495] Beach Farm.

Ordovician Most of the volcanic formations have been utilised at some time. A large quarry, in a volcanic conglomerate within the Forden Mudstone Formation, on the north side [SO 2215 9695] of Montgomery Castle Rock, was used to supply local building stone and roadstone. Vitroclastic tuffs in the Hyssington Volcanic Member [SO 3121 9387], near Hyssington, have been exploited, as have volcaniclastic sandstones in the same member at Fremes Wood. There are small old quarries in volcaniclastic sandstones and other lithologies in the Stapeley Volcanic Member north and south of Hurdley [SO 295 940 area], and medium- to fine-grained feldspathic sandstones in the Hagley Volcanic Formation have been quarried [SO 2746 9434] near Church Stoke.

Llandovery There are several extensive quarries in the massive calcareous sandstones of the Pentamerus Sandstone Formation in the Norbury District e.g. [SO 3698 9243].

Wenlock/Ludlow There are a number of small quarries in the Edgton Limestone Member, as at Eyton [SO 3716 8776] and Five Turnings [SO 3625 8629]. These were used for local roadstone and poor-quality building stone, and may also have supplied crushed limestone for agricultural purposes.

The Nantglyn Flags Formation has been quarried around Bettws Cedewain e.g. [SO 1173 9602]; [SO 1327 9714], and elsewhere, mainly for use on farm tracks and roads prior to the introduction of asphalt.

The Bailey Hill Formation has been widely utilised for local building stone and roadstone throughout the district, especially the thicker sandstones commonly present in its lower part. Examples are the quarry at Castlegreen, Bishop’s Castle [SO 3238 8926], that in Cwm Cae [SO 2795 9130], quarries [SO 1265 8769]; [SO 1241 8690] south-west of Penarron and around Drefor Farm [SO 169 891] and Clithriew [SO 158 887].

There are numerous quarries in the hard calcareous siltstones and sandstones of the Cefn Einion Formation. Examples include the old roadstone quarry at Oaker and the very extensive workings at the Rock of Woolbury [SO 3145 7975].

Sandstones in the Clun Forest Formation have been used for local building stone and aggregate. There are a number of old quarries in flaggy to massive, micaceous, medium- to coarse-grained sandstone in the Curney Bank area [SO 203 846] and in a thick sandstone [SO 2000 8137] about 600 m west of Bettws-y-crwyn.

Intrusive igneous rocks The gabbroic dolerite at More Quarry, Disgwlyfa Hill [SO 325 934], was worked for roadstone for many years until bought out in the early 1980s. There are small quarries in many of the quartz-dolerite dykes in the Longmyndian, for example near Upper Gravenor [SO 3750 9425]. Several of the small microgranite intrusions have been quarried, as at Upper Heblands [SO 3252 9029] and Owlbury [SO 3059 9091] (Sanderson and Cave, 1980).

Water supply

There are no major aquifers within the district. The valley-bottom drift deposits of the Severn Valley are used for water supply upstream of the district near Llandinam (Sheet 164) and drilling has recently been carried out in similar deposits in the Newtown area, with a view to possible water extraction. A borehole [SO 3330 9149] about 1 km east-south-east of Bishop’s Castle Church was sunk in 1957, to 30.78 m, in glaciofluvial gravel and till, to supply Bishop’s Castle. It yielded 10 000 gallons per hour. Elsewhere, some farms and residences are supplied from boreholes, wells or springs. Drilling for water in the solid rocks is speculative since obtaining a supply is usually dependent on the borehole intersecting a fracture zone.

Information sources

BGS publications dealing with this and adjoining districts

Books

Maps

1:10 560

The Newtown area. Special Geological Sheet for parts of SO 08 NE, 09 SE, 18 NW, 19 SW, 1972

1:10 000

Geological 1:10 000 scale National Grid maps included wholly, or in part, in the 1:50 000 scale Montgomery Sheet 165 are listed below, together with the initials of the geological surveyors and dates of the survey; in the case of marginal sheets all surveyors are listed.

The first geological survey of the Montgomery district, at the scale of one inch to one mile, was conducted by W T Aveline, H W Bristow and A C Ramsay and published by the Geological Survey of England and Wales in 1850 as Sheet 60SE. The eastern boundary was mapped by D C Greig, B A Hains and J E Wright during 1957 to 1959 as part of the survey of adjacent districts.

The primary 1:10 000 survey of the district commenced during 1980 and was completed in 1991.The surveyors were R Addison, S D G Campbell, R Cave, B A Hains, A A Jackson, R L Langford, B D T Lynas and J A Zalasiewicz.

Copies of the fair-drawn maps have been deposited in the library of the British Geological Survey (BGS) at Keyworth for public reference and may also be inspected in the London Information Office, in the Geological Museum, South Kensington, London. Copies may be purchased directly from BGS as black and white dyeline sheets.

SO 07 NE RC 1989
SO 08 NE RC 1989
SO 08 SE RC 1989
SO 09 NE RC 1988
SO 09 SE RC 1988
SO 17 NW RC, JAZ 1989–1990
SO 17 NE JAZ 1990
SO 18 NW BDTL, RC 1986
SO 18 NE BDTL, RC 1986
SO 18 SW RC, JAZ 1989–1990
SO 18 SE BAH, JAZ 1990
SO 19 NW RC 1988
SO 19 NE RC 1988
SO 19 SW RC 1988–1989
SO 19 SE RC 1986
SO 27 NW BAH 1990
SO 27 NE BAH 1988–1990
SO 28 NW BDTL 1985–1986
SO 28 NE BDTL 1985
SO 28 SW SDGC, BAH 1988, 1990, 1991
SO 28 SE RA, SDGC 1986–1988
SO 29 NW RC 1986
SO 29 NE BDTL 1982
SO 29 SW RC 1986
SO 29 SE BDTL, BAH 1982, 1984
SO 37 NW BAH 1988, 1990
SO 37 NE BAH 1957, 1988, 1990
SO 38 NW AAJ 1985
SO 38 NE BAH, AAJ 1980, 1985
SO 38 SW RA 1985
SO 38 SE RA 1989
SO 39 NW BDTL, RLL 1983
SO 39 NE RLL, BDTL 1983–1984
SO 39 SW RC, RLL, BDTL 1982–1984
SO 39 SE DCG, JEW 1958–59,
SO 39 SE BAH, RLL 1981–83

BGS Reports

SJ 20 SE: Hains, B A, and Lynas, B D T. 1985. Geological notes and local details for 1:10 000 sheet SJ 20 SE (Rorrington and Long Mountain). British Geological Survey Technical Report, WA/DM/85/4.

SJ 30 SW: Lynas, B D T, and Hains, B A. 1985. Geological notes and local details for 1:10 000 sheet SJ 30 SW (Hope and Brockton). British Geological Survey Technical Report, WA/DM/85/5.

SJ 30 SE: Lynas, B D T. 1985. Geological notes and local details for 1: 10 000 sheet SJ 30 SE (Snailbeach). (British Geological Survey Technical Report, WA/DM/85/6.

SO 17 NW, SO 17 NE, SO 18 SW, SO 18 SE: Zalasiewicz, J A. 1991. Geological notes and local details for 1:10 000 sheets SO 18 SW (Cilfaesty Hill), SO 18 SE (part) (Felindre), SO 17 NW (part) (Ddol) and SO 17 NE (part) (Beguildy).British Geological Survey Technical Report, WA/91/29.

SO 28 NE: Lynas, B D T. 1987. Geological notes and local details for 1:10 000 sheet SO 28 NE (Mainstone) British Geological Survey Technical Report, WA/DM/87/29.

SO 29 NE: Lynas, B D T. 1985. Geological notes and local details for 1:10 000 sheet SO 29 NE. (Chirbury and Priest Weston). British Geological Survey Technical Report, WA/DM/85/13.

SO 29 SE: Hains, B A, and Lynas, B D T. 1985. Geological notes and local details for 1:10 000 sheet SO 29 SE (Churchstoke). British Geological Survey Technical Report, WA/DM/85/11.

SO 38 NW: Jackson, A A. 1997. Geological notes and local details for 1: 10 000 sheet SO 38 NW (Bishop’s Castle). British Geological Survey Technical Report, WA/97/47.

SO 39 NW: Lynas, B D T, and Langford, R L. 1985. Geological notes and local details for 1:10 000 sheet SO 39 NW (Shelve and Corndon). British Geological Survey Technical Report, WA/DM/85/8.

SO 39 NE: Langford, R L, and Lynas, B D T. 1985. Geological notes and local details for 1:10 000 sheet SO 39 NE. (Bridges and Stiperstones). British Geological Survey Technical Report, WA/DM/85/7.

SO 39 SW: Cave, R, Lynas, B D T, and Langford, R L. 1985. Geological notes and local details for 1:10 000 sheet SO 39 SW (Hyssington and Lydham). British Geological Survey Technical Report, WA/DM/85/9.

SO 39 SE: Hains, B A, and Langford, RL. 1985. Geological notes and local details for 1:10 000 sheet SO 39 SE. (Norbury). British Geological Survey Technical Report, WA/DM/85/10.

List of main boreholes

Borehole data for the district are catalogued in BGS archives (National Geoscience Records Centre) at Keyworth on individual 1:10 000 scale sheets. Each catalogue consists of a site map at 1:10 560 or 1:10 000 scale and a borehole register together with the individual records. At the time of going to press about 160 borehole records are held for the Montgomery district and 15 for the Shelve area. These range from shallow site-investigation boreholes to deeper holes; 27 are over 20 metres deep. Two boreholes, Church Stretton No. 1 and No. 2, quoted in the text were drilled in 1961 for stratigraphical purposes. Church Stretton No. 1 (Wentnor) lies in the Montgomery district and has been referred to in the past as Eaton Farm borehole (Appendix). Church Stretton No 2. (Robury Ring) lies to the east on Sheet 166.

Bore Name/Location BGS registered number Grid reference Depth m
Abermule Boreholes/1 SO19NE/9 [SO 1759 9706] 31
Abermule Boreholes/2 SO19NE/10 [SO 1757 9705] 31
Abermule Boreholes/3 SO19NE/11 [SO 1755 9703] 56
Abermule By-Pass./BH.4 SO19SE/3 [SO 1600 9466] 6
Abermule By-Pass./BH.5 SO19SE/4 [SO 1576 9462] 12
Abermule By-Pass./BH.22 SO19SE/15 [SO 1577 9443] 12
Bishop’s Castle SO38NW/2 [SO 3330 8811] 30.78
Colebatch SO38NW/5 [SO 3180 8701] 24
Colebatch SO38NW/6 [SO 3198 8710] 12.49
Dolfor Slip/BH.1 SO08SE/1 [SO 0915 8463] 12
Dolfor Slip/BH.2 SO08SE/2 [SO 0915 8468] 20
Dolfor Slip/BH.3 SO08SE/3 [SO 0910 8466] 10
Dolfor Slip/BH.4 SO08SE/4 [SO 0900 8461] 10
Flood Protection Scheme/Newtown/BH 9A SO19SW/37 [SO 1109 9155] 12.19
Llandyssil SO19NE/6 [SO 1950 9542] 26.5
Newton Bridge/BH D SO19SW/6 [SO 1127 9163] 19

Other sources and types of information

Biostratigraphy

Macrofossils and micropalaeontological residues for samples collected from the district are held at BGS Keyworth.

Ordovician

The Ordovician rocks of the Shelve area are famous for their fossils, and many major museums have material from them. Important early collections by Murchison and the Geological Survey are held with the Biostratigraphical Collections of the BGS, along with more recent collections from selected sites. The most comprehensive collections, however, are those of Whittard; his type specimens of trilobites are in BGS, and his collections of brachiopods and graptolites, together with much other important material, are in the Natural History Museum, London. Information on all these collections are available from the curators of the respective museums.

Silurian

The main collections of fossils from the Llandovery rocks are those of Whittard and his co-workers. In his lifetime Whittard presented the type material stemming from their work to the BGS. The Natural History Museum has suites of other material, including Brachiopoda. Enquiries regarding these may be made to the Curators of those institutions. Most of the Wenlock and Ludlow biostratigraphy in the district depends on collections made during this survey, and are now held by BGS; the collections may be examined by arrangement with the Curator of Biostratigraphy Collections.

BGS lexicon of named rock unit definitions

Definitions of the named rock units shown on BGS maps, including those shown on the 1:50 000 Series Montgomery Sheet 165 and adjacent Shrewsbury Sheet 152, are held in the Lexicon database. Information on how to consult the database can be obtained from the Lexicon Manager at BGS Keyworth.

Geophysics data

Gravity and aeromagnetic data are held digitally in the National Gravity Databank and the National Aeromagnetic Databank at BGS Keyworth.

Minerals

United Kingdom Minerals Yearbook

Mineral-related information (mines, quarries, mineral occurrences, planning information etc.) is held in the minerals GIS on-line (MINGOL) system of BGS Minerals Group and will be operational from 1997. The system can provide hard copy and digital products tailored to individual client’s requirements.

Operating mines and quarries. Directory of Mines and Quarries, 1997.

Mineral resource information for development plans. Shropshire: resources and constraints. in prep. BGS Technical Report.

Mineral Exploration

1. Data arising from investigations in the Shelve area, Shropshire Mineral Reconnaissance Programme. Data release No. 6, British Geological Survey.

2. Anon. 1988. North Snailbeach AE157. Open file report Mineral Exploration and Investment Grant Act (MEIGA), 1972. British Geological Survey

Petrology and mineralogy

There are about 661 registered samples, with corresponding thin sections, which have been collected over the years from the Montgomery and Shelve district as a part of the Geological Survey field work programme. The lithologies include epidotised gabbros, olivine and prehinitised dolerites, granophyres, andesites, albite porphyries, chloritised and amygdaloidal basalts, devitrified rhyolites, vesicular lavas, crystal, welded and lapilli tuffs and tuffaceous sandstones, as well as mudstones, sandstones, greywackes and phosphatic nodular beds.

The samples are coded as ‘E’ (England and Wales) Collection, and form part of a larger collection of approximately 250 000 specimens of rocks, minerals and thin sections. All the samples are catalogued and curated as the National Rock and Mineral Collection. The registered data has been electronically preserved in an Oracle database and selective modes will be available in a user-friendly form for research.

Enquiries should be directed to the Curator, Petrology and Mineralogy Group, BGS, Keyworth.

Remote sensing

BGS holds enhanced Landsat Thematic Mapper (TM) satellite imagery covering the whole of the area included under this memoir. The data were acquired during the winter period at which time vegetation cover is least and the low sun illumination emphasises surface topography. The images are thus well suited to a range of geomorphological and environmental purposes: used in conjunction with the geological map, they serve to highlight the relationships between geology and relief, structure and land-use. The data are available to special order either as single band black-and-white, or as false-colour composite, photographic images geo-corrected to the National Grid. Images can be supplied at scales up to 1:50 000. In addition, BGS has produced a digital mosaic of Landsat imagescovering the whole of Wales and the border region. This is available as a bilingual poster (‘Wales from Space’ 1:350 000) and as special order photographic enlargements.

Sites of Special Scientific Interest

Addresses for data sources and summary of main services and products available

References

Most of the references listed below are held in the Libraries of the British Geological Survey at Edinburgh and Keyworth, Nottingham. Copies of the references can be purchased subject to the current copyright legislation.

Adams, D R. 1962. Survey of the south Shropshire lead mining area. Shropshire Mining Club, Account, No. 2.

Adams, D R. 1968. Survey of the south Shropshire lead mining area. Shropshire Mining Club, First Supplement to Account No. 2.

Allen, J R L. 1974. Sedimentology of the Old Red Sandstone (Siluro-Devonian) in the Clee Hills area, Shropshire, England. Sedimentary Geology, Vol. 12, 73–167.

Allen, J R L. 1985. Marine to freshwater: the sedimentology of the interrupted environmental transition (Ludlow–Siegenian) in the Anglo-Welsh region. Philosophical Transactions of the Royal Society of London, Series B, Vol. 309, 85–104.

Allen, J R L, and Williams B P J. 1981. Sedimentology and stratigraphy of the Townsend Tuff Bed (Lower Old Red Sandstone) in South Wales and the Welsh Borders. Journal of the Geological Society of London, Vol. 138, 15–29.

Allender, R. 1958. On the stratigraphy and structure of an area of Ludlovian and Lower Downtonian rocks near Bishop’s Castle, Shropshire. Unpublished PhD thesis, University of Wales.

Anderton, R, Bridges, P H, Leeder, M R, and Sellwood, B W. 1979. Dynamic stratigraphy of the British Isles. (London: George Allen and Unwin Ltd.)

Bailey, J B. 1977. The physico-chemical conditions of galena and sphalerite deposition. Unpublished PhD thesis, University of Manchester.

Bailey, R J. 1964. A Ludlovian facies boundary in south central Wales. Geological Journal, Vol. 4, 1–20.

Bailey, R J. 1969. Ludlovian sedimentation in south central Wales. 283–304 in The Pre-Cambrian and Lower Palaeozoic rocks of Wales. Wood, A (editor). (Cardiff: University of Wales Press.)

Baker, J W. 1973. A marginal late Proterozoic ocean basin in the Welsh region. Geological Magazine, Vol. 110, 447–455.

Bamford, D, Faber, S, Jacob, B, Kaminski, W, Nunn, K, Prodehl, D, Fuchs, K, King, R, and Willmore, P. 1976. A lithosphere seismic profile in Britain — I. Preliminary results. Geophysical Journal of the Royal Astronomical Society, Vol. 44, 145–160.

Bancroft, B B. 1928. On the unconformity at the base of the Ashgillian in the Bala district. Geological Magazine, Vol. 65, 484–493.

Bancroft, B B. 1933. Correlation tables of the stages Costonian–Onnian in England and Wales. (Gloucester: Blakeney; privately printed.)

Barclay, W J, Taylor, K, and Thomas, L P. 1989. Geology of the South Wales Coalfield. Part II. The country around Abergavenny (3rd edition). Memoir of the British Geological Survey, Sheet 232 (England and Wales).

Barrande, J. 1879. Systême silurien du centre de la Bohême. Ière partie. Recherches paléontologiques. 5, Classe des Mollusques. Order des Brachiopodes. Prague and Paris.

Bassett, D A. 1955. The Silurian rocks of the Talerddig district, Montgomeryshire. Quarterly Journal of the Geological Society of London, Vol. 1, 293–264.

Bassett, D A, Whittington, H B, and Williams, A. 1966. The stratigraphy of the Bala district, Merionethshire. Quarterly Journal of the Geological Society of London, Vol. 122, 219–271.

Bassett, M G. 1989. The Wenlock Series in the Wenlock area. 51–73 in A global standard for the Silurian System. Bassett, M G, and Holland, C H (editors). National Museum of Wales Geological Series, No. 9.

Bassett, M G, Bluck, B J, Cave, R, Holland, C H, and Lawson, J D. 1992. Silurian. 37–56 in Atlas of palaeogeography and lithofacies. Cope, J C W, Ingham, J K, and Rawson, P K (editors). Memoir of the Geological Society of London, No. 13.

Bassett, M G, Cocks, L R M, Holland, C H, Rickards, R B, and Warren P T. 1975. The type Wenlock Series. Report of the Institute of Geological Sciences, No. 75/13.

Bates, D E B. 1972. The stratigraphy of the Ordovician rocks of Anglesey. Geological Journal, Vol. 8, 29–58.

Benton, M J, and Gray, D I. 1981. Lower Silurian distal shelf storm induced turbidites in the Welsh Borders: sediments, tool-marks and trace fossils. Journal of the Geological Society of London, Vol. 138, 675–694.

Bevins, R E, and Rowbotham, G. 1983. Low grade metamorphism within the Welsh sector of the paratectonic Caldeonides. Geological Journal, Vol. 18, 141–167.

Blyth, F G H. 1938. Pyroclastic rocks from the Stapeley Volcanic Group at Knotmoor, near Minsterley, Shropshire. Proceedings of the Geologists’ Association, Vol. 49, 392–404.

Blyth, F G H. 1944. Intrusive rocks of the Shelve area, south Shropshire. Quarterly Journal of the Geological Society of London, Vol. 99, 169–204.

Boswell, P G H. 1926. A contribution to the geology of the eastern part of the Denbighshire moors. Quarterly Journal of the Geological Society of London, Vol. 82, 556–585.

Boswell, P G H. 1949. The middle Silurian Rocks of North Wales. (London: Edward Arnold.)

Boswell, P G H, and Double, I S. 1940. The geology of an area of Salopian rocks west of the Conway Valley, in the neighbourhood of Llanrwst, Denbighshire. Proceedings of the Geologists’ Association , Vol. 51, 151–187.

Bouček, B, and Münch, A. 1952. Retioliti str×edoevropského svrchnîho wenlocku a ludlowu. Sborník ústredrúho Ústavu geologického, oddíl paleontologicky´, 19, 1–54 (in Czech), 104–151 (in English).

Bouma, A H. 1962. Sedimentology of some flysch deposits: a graphic approach to facies interpretation. (Amsterdam: Elsevier.)

Brandon, A. 1989. Geology of the country between Hereford and Leominster. Memoir of the British Geological Survey, Sheet 198 (England and Wales).

Brenchley, P J. 1985. Storm influenced sandstone beds. Modern Geology, Vol. 9, 369–396.

Brenchley, P J. 1993. The Ordovician of the South Berwyn Hills. 39–50 in Geological excursions in Powys. Woodcock, N H, and Bassett, M G (editors) (Cardiff: University of Wales Press, National Museum of Wales.)

Bridges, P H. 1975. The transgression of a hard substrate shelf: the Llandovery (lower Silurian) of the Welsh Borderland. Journal of Sedimentary Petrology, Vol. 45, 79–94.

British Geological Survey. 1990. Palaeontology Dept Report, No. 90/61.

British Geological Survey. 1991. The Shelve Ordovician Inlier and adjacent areas. 1:25 000 Classical areas of British Geology Series. Parts of Sheets SO 29, 39, SJ 20 and 30. (Solid and Drift edition.). (Keyworth, Nottingham: British Geological Survey)

British Geological Survey. 1993a. 1:50 000 Series England and Wales Sheet 179 Rhayader. Solid. (Keyworth, Nottingham: British Geological Survey.)

British Geological Survey. 1993b. 1:50 000 Series England and Wales Sheet 120 Corwen. Solid. (Keyworth, Nottingham: British Geological Survey.)

British Geological Survey. 1994. Geological map of Wales (Map Daearegol a Gymru). 1:250 000. (Keyworth, Nottingham: British Geological Survey.)

British Geological Survey. 1995. Wales from space; an enhanced Landsat Image at 1:350 000 scale. (Keyworth, Nottingham: British Geological Survey.)

Brooks, M, and Fenning, P. 1968. Geophysical investigations. 307–321 in Geology of the country around Church Stretton, Craven Arms, Wenlock Edge and Brown Clee. Greig, D C, Wright, J E, Hains, B A, and Mitchell, G H. Memoir of the Geological Survey of Great Britain, Sheet 166 (England and Wales).

Brown, M J F. 1971. The glacial geomorphology of parts of west Shropshire and Montgomeryshire. Unpublished PhD thesis, University of London.

Burton, P W, and Marrow, P C. 1988. Seismic hazard and earthquake source parameters in the North Sea. In Earthquakes at North-Atlantic passive margins: Neotectonics and postglacial rebound. Gregerson, S, and Basham, P (editors). NATO ASI series, Vol. 266. (Kluwer Academic Publishers.)

Calef, C E, and Hancock, J J. 1974. Wenlock and Ludlow marine communities in Wales and the Welsh Borderland. Palaeontology, Vol. 17, 779–810.

Callaway, C. 1877. On a new area of Upper Cambrian rocks in south Shropshire, with a description of a new fauna. Quarterly Journal of the Geological Society of London, Vol. 33, 652–672.

Campbell, S D G. 1989. Geological notes and local details for 1:10 000 sheet SO 28 SE (Whitcott Keysett). (Keyworth, Nottingham: British Geological Survey.)

Campbell, S D G. 1984. Aspects of dynamic stratigraphy (Caradoc–Ashgill) in the northern part of the Welsh marginal basin. 390–391 in Abstracts of current research in Wales, Geological Magazine, Vol. 95.

Cave, R. 1955. The stratigraphy of the Welshpool area (Montgomeryshire). Unpublished PhD thesis, University of Cambridge.

Cave, R. 1965. The Nod Glas sediments of Caradoc age in North Wales. Geological Journal, Vol. 4, 279–298.

Cave, R. 1979. Sedimentary environments of the basinal Llandovery of mid-Wales. 517–526 in The Caledonides of the British Isles — reviewed. Harris, A L, Holland, C H, Leake, B E (editors). Special Publication of the Geological Society of London, No. 8.

Cave, R. 1988. In BGS boreholes 1984–1986. Report of the British Geological Survey, Vol. 19, No. 1, 6–7.

Cave, R. 1992. The Llandovery Series. 37–51 in Atlas of palaeogeography and lithofacies. Cope, J C W, Ingham, J K, and Rawson, P K (editors). Memoir of the Geological Society of London, No. 13.

Cave, R. 1995. A review: Wales from space; an enhanced Landsat image at 1:30 000 scale. British Geological Survey 1994. Teaching Earth Sciences, Vol. 20, 116.

Cave, R, Dean, W T, and Hains, B A. 1988. Age of the Ordovician andesite conglomerate of Castle Hill, Montgomery, Powys, Wales. Geological Journal, Vol. 23, 205–210.

Cave, R, and Dixon, R J. 1993. The Ordovician and Silurian of the Welshpool area. 51–84 in Geological excursions in Powys, Central Wales. Woodcock, N H, and Bassett, M G (editors). (Cardiff: University of Wales Press, National Museum of Wales.)

Cave, R, Evans, J A, and Campbell, S D G. 1992. Garn Prys: a mid-Silurian canyon, feeder to the Denbigh Grits of North Wales. Geological Journal, Vol. 27, 301–315.

Cave, R, and Hains, B A. 1986. Geology of the country between Aberystwyth and Machynlleth. Memoir of the British Geological Survey, Sheet 163 (England and Wales).

Cave, R, Hains, B A, and White, D E. 1993. The Wenlock and Ludlow of the Newtown area. 85–112 in Geological excursions in Powys, Central Wales. Woodcock, N H, and Bassett, M G (editors). (Cardiff: University of Wales Press, National Museum of Wales.)

Cave, R, Lynas, B D T, and Langford, R L. 1985. Geological notes and local details for 1:10 000 sheet SO 39 SW (Hyssington and Lydham). British Geological Survey Report, WA/DM/85/9

Cave, R, and Price, D. 1978. The Ashgill Series near Welshpool, North Wales. Geological Magazine, Vol. 115, 183–194.

Cave, R, and Rushton, A W A. 1996. The Llandeilo Series in the core of the Tywi Anticline, Llanwrtyd, Powys. Geological Journal, Vol. 31, 47–60.

Cave, R, and White, D E. 1971. The exposures of Ludlow rock and associated beds at Tites Point and near Newnham, Gloucestershire. Geological Journal, Vol. 7, 239–254.

Cherns, L. 1980. Hardgrounds in the Lower Leintwardine Beds (Silurian) of the Welsh Borderland. Geological Magazine, Vol. 117, 311–326.

Cocks, L R M. 1968. Some strophomenacean brachiopods from the British Lower Silurian. Bulletin of the British Museum (Natural History): Geology, Vol. 15, Pt 6, 283–324.

Cocks, L R M, Holland, C H, Rickards, R B, and Strachan, I. 1971. A correlation of Silurian rocks in the British Isles. Journal of the Geological Society of London, Vol. 127, 103–136.

Cocks, L R M, Holland, C H, and Rickards, R B. 1992. A revised correlation of Silurian rocks in the British Isles. Special Publication of the Geological Society of London, No. 21

Cocks, L R M, Holland, C H, Rickards, R B, and Warren, P T. 1975. The type Wenlock Series. Report of the Institute of Geological Sciences, No. 75/13.

Cocks, L R M, and Rickards, R B. 1969. Five boreholes in Shropshire and the relationships of shelly and graptolitic facies in the Lower Silurian. Quarterly Journal of the Geological Society of London, Vol. 124, 213–238.

Cocks, L R M, Woodcock, N H, Rickards, R B, Temple, J T, and Lane, P D. 1984. The Llandovery Series of the type area. Bulletin of the British Museum (Natural History), Geology, Vol. 38, 131–182.

Cook, A H, and Thirlaway, H I S. 1955. The geological results of measurements of gravity in the Welsh Borders. Quarterly Journal of the Geological Society of London, Vol. 111, 47–70.

Cope, J C W, and Gibbons, W. 1987. New evidence for the relative age of the Ercall Granophyre and its bearing on the Precambrian–Cambrian boundary in southern Britain. Geological Journal, Vol. 22, 53–60.

Cope, J C W, Ingham, J K, and Rawson, P K (editors). 1992. Atlas of palaeogeography and lithofacies. Memoir of the Geological Society of London, No. 13

Cornwell, J D, and Cave, R. 1986. An airborne geophysical survey of part of west Dyfed, South Wales, and some related ground surveys. Report of the Mineral Reconnaissance Programme, British Geological Survey, No. 84.

Coward, M P, and Siddans, A W B. 1979. The tectonic evolution of the Welsh Caledonides. 187–198 in The Caledonides of the British Isles — reviewed. Harris, A L, Holland, C H, and Leake, B E (editors). Special Publication of the Geological Society of London, No. 8.

Cummins, W A. 1957. The Denbigh Grits: Wenlock greywackes in Wales. Geological Magazine, Vol. 94, 433–451.

Cummins, W A . 1959a. The Nantglyn Flags: Mid-Salopian basin facies in Wales. Liverpool and Manchester Geological Journal, Vol. 2, 159–167.

Cummins, W A. 1959b. The Lower Ludlow Grits in Wales. Liverpool and Manchester Geological Journal, Vol. 2, 168–179.

Cummins, W A. 1963. The geology of the northern part of the Llangadfan Syncline, Montgomeryshire. Unpublished PhD thesis, University of Liverpool.

Cummins, W A. 1969. Patterns of sedimentation in the Silurian rocks of Wales. 219–237 in The Precambrian and Lower Palaeozoic rocks of Wales. Wood, A (editor). (Cardiff.)

Das Gupta, T. 1932. The Salopian Graptolite Shales of the Long Mountain and similar rocks of Wenlock Edge. Proceedings of the Geologists’ Association, Vol. 43, 325–363.

Davies, K A. 1928. The geology of the country between Rhayader (Radnorshire) and Abergwesyn (Breconshire). Proceedings of the Geologists’ Association, Vol. 39, 157–168.

Dean, W T. 1958. The geology of the Ordovician and adjacent strata in the southern Caradoc district of Shropshire. Bulletin of the British Museum (Natural History), (Geology), Vol. 9, 257–296.

Dean, W T, and Dineley, D L. 1961. The Ordovician and associated Precambrian rocks of the Pontesford district, Shropshire. Geological Magazine, Vol. 98, 367–376.

Dimberline, A J. 1987. The sedimentology and diagenesis of the Wenlock turbidite system. Unpublished PhD thesis, University of Cambridge.

Dimberline, A J, and Woodcock, N H. 1987. The southeast margin of the Wenlock turbidite system, mid-Wales. Geological Journal, Vol. 22, 61–71.

Dines, H G. 1958. The West Shropshire mining region. Bulletin of the Geological Survey of Great Britain. No. 14, 1–43.

Dines, H G. 1959. The West Shropshire mining field in Future of Non-Ferrous mining in Great Britian and Ireland. Institution of Mining and Metallurgy, London,295–303.

Dixon, R J. 1988. The Ordovician (Caradoc) volcanic rocks of Montgomery, Powys, N Wales. Geological Journal, Vol. 23, 149–156.

Dixon, R J. 1990. The Moel-y-Golfa Andesite: and Ordovician (Caradoc) intrusion into unconsolidated conglomerate sediments, Breidden Hills Inlier, Welsh Borderland. Geological Journal, Vol. 25, 35–46.

Dixon, R J. 1991. The Ordovician (Caradoc) igneous and sedimentary rocks of the Breidden Hills, Shelve and Forden Inliers, Welsh Borderlands: rift-related volcaniclastic sedimentation in a back-arc tectonic setting. Unpublished PhD thesis, University of Wales.

Duke, W L. 1985. Hummocky cross-stratification, tropical hurricanes, and intense winter storms. Sedimentology, Vol. 32, 167–194.

Dunning, F W. 1975. Precambrian craton of central England and the Welsh Borders. 83–95 in A correlation of the Precambrian rocks in the British Isles. Harris, A L, Shackleton, R M, Watson, J, Downie, C, Harland, W B, and Moorbath, S (editors).Special Report of the Geological Society of London, No.6. (Edinburgh: Scottish Academic Press.)

Dunham, K C, and Dines, H G. 1945. Barium minerals in England and Wales. Wartime Pamphlet of the Geological Survey of Great Britain, No. 46.

Dwerryhouse, A R, and Miller, A A. 1930. The glaciation of Clun Forest, Radnor Forest and some adjoining districts. Quarterly Journal of the Geological Society of London, Vol. 86, 96–129.

Earp, J R. 1938. The higher Silurian rocks of the Kerry district, Montgomeryshire. Quarterly Journal of the Geological Society of London, Vol. 94, 125–160.

Earp, J R. 1940. The geology of the south-western part of Clun Forest. Quarterly Journal of the Geological Society of London, Vol. 96, 1–11.

Earp, J R. 1977. Notes on the geology, Llandrindod Wells Ordovician Inlier. Classical areas of British Geology. Institute of Geological Sciences. (London: HMSO.)

Earp, J R, and Hains, B A. 1971. British regional geology: the Welsh Borderland (3rd edition). (London: HMSO for Institute of Geological Sciences.)

Elles, G L. 1900. The zonal classification of the Wenlock Shales of the Welsh Borderland. Quarterly Journal of the Geological Society of London, Vol. 56, 370–413.

Elles, G L. 1940. The stratigraphy and faunal succession in the Ordovician rocks of the Builth–Llandrindod inlier, Radnorshire. Quarterly Journal of the Geological Society of London, Vol. 95 (for 1939), 382–445.

Elles, G L, and Slater, I L. 1906. The highest Silurian rocks of the Ludlow District. Quarterly Journal of the Geological Society of London, Vol. 62, 195–221.

Entwisle, D C. 1984. Density and porosity determinations on field samples from the Shelve inlier, Shropshire. Geophysical Laboratory Report, Engineering Geology and Reservoir Rock Properties Research Group, British Geological Survey, No. 146.

Eva, S J. 1992. Sediment deformation in the Silurian rocks of North Wales. Unpublished PhD thesis, University College of Wales, Aberystwyth.

Eva, S J, and Maltman, A J. 1994. Slump fold and palaeoslope orientations in Upper Silurian rocks, North Wales. Geological Magazine, Vol. 131, 685–691.

Evans, P R. 1957. The geology of the Longmynd–Clun area, south Shropshire. Unpublished PhD thesis, Bristol University.

Finney, S C, and Bergstrom, S M. 1986. Biostratigraphy of the Ordovician Nematograptus gracilis Zone. 47–59 in Palaeoecology of graptolites. Hughes, C P, and Rickards, R B (editors). Special Publication of the Geological Society of London, No. 20.

Forbes, E. 1848. Memorandum respecting some fossiliferous localities alluded to in Professor Ramsay and Mr Aveline’s paper, as noted on the spot. Quarterly Journal of the Geological Society of London, Vol. 4, 297–299.

Fortey, R A, Beckly, A J, and Rushton, A W A. 1990. International correlation of the base of the Llanvirn Series, Ordovician System. Newsletters on Stratigraphy, Vol. 72, 119–242.

Fortey, N J, Merriman, R J, and Huff, W D. 1994. Silurian and late Ordovician K-bentonites as a record of late Caledonian volcanism in the British Isles. (Abstract). 4 in Caledonian terrane relationships in Britain. (Keyworth, Nottingham: British Geological Survey.)

Fortey, R A, and Owens, R M. 1978. Early Ordovician (Arenig) stratigraphy and faunas of the Carmarthen district, south-west Wales. Bulletin of the British Museum, (Natural History), Geology, Vol. 30, 225–294.

Fortey, R A, and Owens, R M. 1987. The Arenig Series in South Wales: stratigraphy and palaeontology. 1. The Arenig Series in South Wales. Bulletin of the British Museum, (Natural History), Geology, Vol. 41, 69–364.

Fortey, R A, and Owens, R M. 1990. Arenig biostratigraphy and correlation in the Welsh Basin. Journal of the Geological Society of London, Vol. 147, 607–610.

Fortey, R A, and Owens, R M. 1992. The Habberley Formation: youngest Tremadoc in the Welsh Borderlands. Geological Magazine, Vol. 129, 553–566.

Goudie, A D, and Piggott, N R. 1981. Quartzite tors, stone stripes and slopes at the Stiperstones, Shropshire, England. Biuletyn Peryglacjalny (Lodz), Vol. 28, 47–56.

Grahn, Y and Bergstrom, S M. 1984. Lower Middle Ordovician Chitinozoa from the Southern Appalachians, United States. Review of Palaeobotany and Palynology, Vol. 43, 89–122.

Greig, D C, Wright, J E, Hains, B A, and Mitchell, G H. 1968. Geology of the country around Church Stretton, Craven Arms, Wenlock Edge and Brown Clee. Memoir of the Geological Survey of Great Britain, Sheet 166 (England and Wales).

Grunthal, G (editor). 1993. European macroseismic scale 1992 (up-dated MSK-scale).Cahiers du Centre European de Geodynamique et de Seismologie, Vol. 7.

Gurney, S D, and Worsley, P. 1996. Relict cryogenic mounds at Owlbury, near Bishop’s Castle, Shropshire. Mercian Geologist, Vol. 14, 14–21.

Hains, B A, and Langford, R L. 1985. Geological notes and local details for 1:10 000 sheet SO 39 SE (Norbury). (Keyworth, Nottingham: British Geological Survey).

Hains, B A, and Lynas, B D T. 1985a. Geological notes and local details for 1:10 000 sheet SJ20SE (Rorrington and Long Mountain). (Keyworth, Nottingham: British Geological Survey.)

Hains, B A, and Lynas, B D T. 1985b. Geological notes and local details for 1:10 000 sheet SO29SE (Church Stoke). (Keyworth, Nottingham: British Geological Survey).

Hall, T C F. 1922. The distribution and genesis of lead and associated ores in western Shropshire. Minerological Magazine, Vol. 27, 201–209.

Hallam, A. 1971. Re-evaluation of the palaeogeographic argument for an expanding earth. Nature, London, Vol. 232, 180–183.

Hancock, N J, Hurst, J M, and Fürisch, F T. 1974. The depths inhabited by Silurian brachiopod communities. Journal of the Geological Society of London, Vol. 130, 151–156.

Harmer, F W. 1907. On the origin of certain canõn-like valleys associated with lobe-like areas of depression. Quarterly Journal of the Geological Society of London, Vol. 63, 470–514.

Harper, J C. 1940. The Upper Valentian ostracod fauna of Shropshire. Annals and Magazine of Natural History, Series 11, Vol. 5, 385–400.

Harris, J H. 1987. The geology of the Wenlock Shales around Builth Wells. Unpublished PhD thesis, University of Cambridge.

Hede, J E. 1915. Skanes colonusskiffer. Lunds Universitets Arsskrift, N F Afd. 2, 11, (6),

Holland, C H. 1959. The Ludlovian and Downtonian rocks of the Knighton district, Radnorshire. Quarterly Journal of the Geological Society of London, Vol. 14, 449–482.

Holland, C H. 1992. The Wenlock Series. 40–42, 50–53 in Atlas of palaeogeography and lithofacies. Cope, J C W, Ingham, J K, and Rawson, P K (editors). Memoir of the Geological Society of London, No. 13.

Holland, C H, and Lawson, J D. 1963. Facies patterns in the Ludlovian of Wales and the Welsh Borderland. Geological Journal, Vol. 3, 269–288.

Holland, C H, Lawson, J D, and Walmsley, V G. 1963. The Silurian rocks of the Ludlow District. Bulletin of the British Museum (Natural History), Geology, Vol. 8, 93–171.

Holland, C H, Rickards, R B, and Warren, P T. 1969. The Wenlock graptolites of the Ludlow district, Shropshire, and their stratigraphical significance. Palaeontology, No. 12, 663–683.

Holland, C H, and Williams, E M. 1985. The Ludlow–Downton transition at Kington, Herefordshire. Geological Journal, Vol. 20, 31–41.

Holm, G. 1890. Gotland’s graptolites. Bihang till K Svenska. Vetenskapsakademiens Hardlingar, 3 and 16, Afd IV. No. 7, pp.3–34.

Howells, M F, Leveridge, B E, Addison, R, and Reedman, A J. 1983. The lithostratigraphical subdivision of the Ordovician underlying the Snowdon and Crafnant volcanic groups, north Wales. Report of the Institute of Geological Sciences, No. 83/1.

Hughes, C P. 1969. The Ordovician trilobite faunas of the Builth–Llandrindod Inlier, Central Wales. Part 1. Bulletin of the British Museum (Natural History), Geology, Vol. 18, 103.

Hughes, R A. 1989. Llandeilo and Caradoc graptolites of the Builth–Llandrindod and Shelve inliers. Monograph of the Palaeontographical Society of London. Vol. 89, No. 577, part of Vol.141 for 1987.

Hughes, T McK. 1879. On the Silurian rocks of the valley of the Clwyd. Quarterly Journal of the Geological Society of London, Vol. 35, 694–698.

Hughes, T McK. 1885. Notes on the geology of the Vale of Clwyd. Proceedings of the Chester Society of Natural Sciences. No. 3, 5–37.

Hughes, T McK. 1894. Observations on the Silurian rocks of North Wales. Proceedings of the Chester Society of Natural Sciences, Vol. 4, 141–160.

Hurst, J M. 1975. The diachronism of the Wenlock Limestone. Lethaia, Vol. 8, 301–304.

Ineson, P R, and Mitchell, J G. 1975. K/Ar isotope age determinations from some Welsh mineral localities. Transactions of the Institution of Mining and Metallurgy, Series B, Vol. 84, 7–16.

Institute of Geological Sciences. 1967. Church Stretton. 1:63 360 Geological Map Series. Sheet 166 (England and Wales)(Solid and Drift editions).

Institute of Geological Sciences. 1972. The Newtown area. 1:10 560 Special Geological Sheet. Parts of SO 08, 09 18 19. (Solid and Drift edition).

Jackson, A A. 1985. British Geological Survey 1:10 000 series SO 38 NE (Bishops Castle). Unpublished Report.

Jackson, A A. 1997. Geology of the Bishop’s Castle district: 1:10 000 sheet SO 38 NW and part of SO 38 NE. British Geological Survey Technical Report, WA/97/47.

James, J H. 1952. Notes on the relationship of the Uriconian and Longmyndian Rocks near Linley, Shropshire. Proceedings of the Geologists’ Association , Vol. 63, 198–200.

James, J H. 1956. The structure and stratigraphy of part of the Pre-Cambrian outcrop between Church Stretton and Linley, Shropshire. Quarterly Journal of the Geological Scoiety of London, Vol. 112, 315–337.

Jeppsson, L. 1990. An oceanic model for lithological and faunal changes tested on the Silurian record. Journal of the Geological Society of London, Vol. 147, 663–674.

Johnson, M E, Kaljo, D, and Rong, J Y. 1991. Silurian eustacy. 145–163 in The Murchison Symposium: proceedings of an international conference on the Silurian System. Bassett, M G, Lane, P D, and Edwards, D (editors). Special Papers in Palaeontology, No. 44,

Jones, C R. 1986. Ordovician (Llandeilo and Caradoc) beyrichiocope Ostracoda from England and Wales. Part 1. Monograph of the Palaeontographical Society of London, (No. 569, part of Vol. 138 for 1984).

Jones, C R. 1987. Ordovician (Llandeilo and Caradoc) beyrichiocope Ostracoda from England and Wales. Part 2. Monograph of the Palaeontographical Society of London, (No. 571, part of Vol. 139 for 1985).

Jones, O T. 1925. The geology of the Llandovery district: Part 1 — the southern area. Quarterly Journal of the Geological Society of London, Vol. 81, 344–388.

Jones, O T. 1947. The geology of the Silurian rocks west and south of the Carneddau Range, Radnorshire. Quarterly Journal of the Geological Society of London, Vol. 103, 1–36.

Jones, O T. 1956. The geological evolution of Wales and adjacent regions. Quarterly Journal of the Geological Society of London, Vol. 111, 323–351.

Jones, O T, and Pugh, W J. 1941. The Ordovician rocks of the Builth district. A preliminary account. Geological Magazine, Vol. 78, 185–191.

Kennedy, R J. 1988. Ordovician (Llanvirn) trilobites from SW Wales. Monograph of the Palaeontographical Society of London, (No. 576, part of Vol. 141 for 1987).

Khan, A B, and Kelling, G. 1991. Depositional environment of the Wenlock turbidite system in the northern Montgomery Trough, mid-Wales. Programme and Abstracts Welsh Basin Meeting 1991. British Geological Survey, Aberystwyth.

King, W B R. 1928. The geology of the district around Meifod (Montgomeryshire). Quarterly Journal of the Geological Society of London, Vol. 84, 671–702.

Kirk, N H. 1948. Geology of the anticlinal disturbance of Breconshire and Radnorshire, Pont Faen to Presteigne. Unpublished PhD thesis, University of Cambridge.

Kirk, N H. 1951a. The upper Llandovery and lower Wenlock rocks of the area between Dolyhir and Presteigne, Radnorshire. Proceedings of the Geological Society of London, Vol. 1471, 56–58.

Kirk, N H. 1951b. The Silurian and Downtonian rocks of the anticlinal disturbance of Breconshire and Radnorshire: Pont Faen to Presteigne. Abstracts of the Proceedings of the Geological Society of London, Vol. 1474, 72–74.

Kisslinger, C. 1980. Evaluation of S to P amplitude ratios for determining focal mechanisms from regional network observations. Bulletin of the Seismological Society of America. Vol. 70, No. 4, 999–1014.

Langford, R L, and Lynas, B D T. 1985. Geological notes and local details for 1:10 000 sheet SO 39 NE (Bridges and Stiperstones). (Keyworth, Nottingham: British Geological Survey.)

Lapworth, C. 1880. Geological distribution of the Rhabdophora. Annals and Magazine, Natural History, Series 5, Vol. V.

Lapworth, C. 1887a. The Ordovician rocks of Shropshire. Report of the 56th meeting of the British Association for the Advancement of Science, London, (for Birmingham 1886), Section C, 661–663.

Lapworth, C. 1887b. Preliminary note on the Ordovician rocks of Shropshire. Geological Magazine, Decade 3,Vol. 4, 78–80.

Lapworth, C. 1916. 36–38 in Summary of progress of the Geological Survey for 1915.

Lapworth, C, and Watts, W W. 1894. The geology of south Shropshire. Proceedings of the Geologists’ Association, Vol. 13, 297–355.

Lapworth, C, and Watts, W W. 1910. 739–769 in Geology in the field, Shropshire. Geologists’ Association Jubilee Volume.

Lawson, J D. 1955. The geology of the May Hill inlier. Quarterly Journal of the Geological Society of London, Vol. 111, 85–116.

Lawson, J D. 1956. Introduction. 563–566 in The Ludlovian rocks of the Welsh Borderland. The Advancement of Science, Vol. 12, No. 48. (London: British Association for the Advancement of Science.)

Lawson, J D. 1973. Facies and faunal changes in the Ludlovian rocks of Aymestrey, Herefordshire. Geological Journal, Vol. 8, 247–278.

Lawson, J D. 1992. The Ludlow Series. 42–47 in Atlas of palaeogeography and lithofacies. Cope, J C W, Ingham, J K, and Rawson, P K (editors). Memoir of the Geological Society of London, No. 13.

Lawson, J D, and Straw, S H. 1956. The correlation problem. 568–570 in The Ludlovian rocks of the Welsh Borderland. Lawson, J D (editor). The Advancement of Science, 12, No. 48, (London: British Association for the Advancement of Science.)

Lawson, J D, and White, D E. 1989. The Ludlow Series in the Ludlow area. 73–90 in A global standard for the Silurian System. Holland, C H, and Bassett, M G (editors). (Cardiff: National Museum of Wales.)

Leng, M J. 1990. Late Ordovician–early Silurian palaeo-environmental analysis in the Tywyn-Corris area of mid-Wales. Unpublished PhD thesis, University of Wales.

Lewis, C A. 1970. The upper Wye and Usk regions. 147–173 in The glaciations of Wales and adjoining regions. Lewis, C A (editor). (London: Longman.)

Llewellyn, P G. 1965. The graptolitic mudstones of the Howgill Fells. Geological Magazine, Vol. 102, 277–278.

Loydell, D K. 1991. The biostratigraphy and formational relationships of the upper Aeronian and lower Telychian (Llandovery, Silurian) formations of Western mid-Wales. Geological Journal, Vol. 26, 209–244.

Loydell, D K. 1993. Worldwide correlation of Telychian (Upper Llandovery) strata using graptolites. 323–340 in High resolution stratigraphy. Hailwood, E A, and Kidd, R B (editors). Special Publication of the Geological Society of London, No. 70.

Loydell, D K, and Cave, R. 1993. The Telychian (Upper Llandovery) stratigraphy of Buttington Brick Pit, Wales. Newsletters on Stratigraphy, Vol. 22, 91–103.

Loydell, D K, and Cave, R. 1994. Identification of the Wenlock–Ludlow boundary in Welsh graptolitic sequences. Berichte Geologische Bundesanstalt, Vol. 30, 142.

Loydell, D K, and Cave, R. 1996. The Llandovery–Wenlock boundary and related stratigraphy in eastern mid-Wales with special reference to the Banwy River section. Newsletters on Stratigraphy. Vol. 34, 39–64

Lynas, B D T. 1983. Two new Ordovician volcanic centres in the Shelve inlier, Powys, Wales. Geological Magazine, Vol. 120, 535–542.

Lynas, B D T. 1985a. Geological notes and local details for 1:10 000 sheet SJ30SE (Snailbeach). Report of the British Geological Survey, No. WA/DM/85/6.

Lynas, B D T. 1985b. Geological notes and local details for 1:10 000 sheet SO 29 NE (Chirbury and Priest Weston). (Keyworth, Nottingham: British Geological Survey.)

Lynas, B D T. 1986. Geological notes and local details for 1:10 000 sheet SO 28 NW (Clun Forest). (Keyworth, Nottingham: British Geological Survey.)

Lynas, B D T. 1987. Geological notes and local details for 1:10 000 sheet SO 28 NE (Mainstone). (Keyworth, Nottingham: British Geological Survey.)

Lynas, B D T. 1988. Evidence for dextral oblique-slip faulting in the Shelve Ordovician Inlier, Welsh Borderland: implications for the south British Caledonides. Geological Journal, Vol. 23, 39–57.

Lynas, B D T, and Hains, B A. 1985. Geological notes and local details for 1:10 000 sheet SJ 30 SW (Hope and Brockton). (Keyworth, Nottingham: British Geological Survey.)

Lynas, B D T, and Langford, R L. 1985. Geological notes and local details for 1:10 000 sheet SO 39 NW (Shelve and Corndon). (Keyworth, Nottingham: British Geological Survey.)

Lynas, B D T, Rundle, C C, and Sanderson, R W. 1985. A note on the age and pyroxene chemistry of the igneous rocks of the Shelve Inlier, Welsh Borderland. Geological Magazine, Vol.122, 641–647.

Macgregor, A R. 1961. Upper Llandeilo brachiopods from the Berwyn Hills, North Wales. Palaeontology, Vol. 4, 177–209.

Mackie, A H. 1993. The Ordovician and Llandovery in the Llanwrtyd Wells to Llyn Brianne area. 259–279 in Geological excursions in Powys, central Wales. Woodcock, N H, and Bassett, M G (editors). (Cardiff: University of Wales Press and National Museum of Wales. Published on behalf of the Geologists’ Association, South Wales Group.)

Manchester, RJ. 1983. The crustal structure of Wales from interpretation of gravity and magnetic surveys. Unpublished MSc thesis, University of Birmingham.

Marrow, P C, and Walker, A B. 1988. The Lleyn earthquake of 1984 July 19: Aftershock sequence and focal mechanism. Geophysical Journal, Vol. 92:3, 487–493.

Martinson, A, Bassett, M G, and Holland, C H. 1981. Ratification of standard chronostratigraphical divisions and stratotypes for the Silurian System. Episodes, No. 2, 36.

Masson Smith, D, Howell, P M, and Abernethy-Clark, A B D E. 1974. The National Gravity Reference Net 1973 (NGRN 73). Professional Papers of the Ordnance Survey, New Series No. 26.

McCabe, C and Channell, J E T. 1990. Palaeomagnetic results from volcanic rocks of the Shelve Inlier, Wales: evidence for a wide Late Ordovician Iapetus Ocean in Britain. Earth Science Planetary Letters, Vol. 96, 458–468.

McCabe, C. and Channell, J E T. 1991. Reply to comment of A Trench and T H Torsvik on ‘Palaeomagnetic results from volcanic rocks of the Shelve Inlier, Wales: evidence for a wide Late Ordovician Iapetus Ocean in Britain.’ Earth Science Planetary Letters, Vol. 104, 540–544.

Merriman, R J, Roberts, B, and Hirons, S R. 1992. Regional low grade metamorphism in the central part of the Lower Palaeozoic Welsh Basin: an account of the Llanilar and Rhayader districts, BGS 1:50K Sheets 178 & 179. British Geological Survey Technical Report, WG/92/16.

Moorbath, S. 1962. Lead isotope abundances studies. Philosophical Transactions of the Royal Society, Series A, Vol. 254, 295–360.

Morton, G H. 1869. The geology and mineral veins of the country around Shelve, Shropshire. Proceedings of the Liverpool Geological Society, Vol. 10, 1–41.

Moseley, J B. 1991. The mineralization of the Stiperstones Quartzite of the Shelve Inlier, SW Shropshire. The North West Geologist, Vol. 1, 34–38.

Moseley, J B. 1994. The origin and significance of the hematization of silicic rock of Precambrian and Ordovician age in south Shropshire. Mercian Geologist, Vol. 13, 111–115.

Murchison, R I. 1839. The Silurian System, founded on geological researches in the counties of Salop, Hereford, Radnor, Montgomery, Caermarthen, Brecon, Pembroke, Monmouth, Worcester, Gloucester and Stafford; with descriptions of the coalfields and overlying formations. (London: John Murray.)

Murchison, R I. 1859. Siluria. The history of the oldest known rocks containing organic remains, with a brief description of the distribution of gold over the earth (3rd edition).

Musson, R M W. 1993. Macroseismic magnitude and depth for British earthquakes. British Geological Survey, Global Seismology Report, No. WL/93/13.

Musson, R M W. 1994. A catalogue of British earthquakes. British Geological Survey, Global Seismology Report, No. WL/94/04.

Musson, R M W. 1996. Determination of parameters for historical British earthquakes. Annali di Geophysica. Vol. 39, 1041–1047.

Natland, M L. 1876. Classification of clastic sediments. Bulletin of the American Association of Petroleum Geologists, Vol. 60, 702.

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

Palmer, D. 1970. A stratigraphical synopsis of the Long Mountain, Montgomeryshire and Shropshire. Proceedings of the Geological Society of London, Vol. 1660, 341–6.

Palmer, D. 1972. The geology of the Long Mountain, Montgomeryshire and Shropshire. Unpublished PhD thesis, Trinity College, Dublin.

Paris, F. 1990. The Ordovician chitinozoan biozones of the Northern Gondwana domain. Review of Palaeobotany and Palynology, Vol. 66, 181–209.

Pattrick, R A D, and Boswell, R J. 1991. The genesis of the West Shropshire Orefield: evidence from fluid inclusions, sphalerite chemistry and sulphur isotopic ratios. Geological Journal, Vol. 26, 101–115.

Pauley, J C. 1990a. Sedimentology, structural evolution and tectonic setting of the late Precambrian Longmyndian Supergroup of the Welsh Borderland, UK. 341–351 in The Caledonian Orogeny. D’lemos, R S, Strachan, R A and Topley, C G (editors). Special Publication of the Geological Society of London, No. 51.

Pauley, J C. 1990b. The Longmyndian Supergroup and related Precambrian sediments of England and Wales. 5–27 in Avalonian and Cadomian geology of the North Atlantic. Strachan, R A, and Taylor, G K (editors). (Blackie and Son.)

Pauley, J C. 1991. A revision of the stratigraphy of the Longmyndian Supergroup, Welsh Borderland and of its relationship to the Uriconian volcanic complex. Geological Journal, Vol. 26, 167–183.

Pickering, K T, Bassett, M G, and Siveter, D J. 1988. Late Ordovician–early Silurian destruction of the Iapetus Ocean: Newfoundland, British Isles and Scandinavia — a discussion. Transactions of the Royal Society of Edinburgh, Vol. 79, 361–382.

Pickering, K T, Stow, D A V, Watson, M P, and Hiscott, R N. 1986. Deep-water facies, processes and models: a review and classification scheme for modern and ancient sediments. Earth Science Reviews, Vol. 23, 75–174.

Piper, J D A. 1978. Palaeomagnetic survey of the (Palaeozoic) Shelve inlier and Berwyn Hills, Welsh Borderlands. Geophysical Journal of the Royal Astronomical Society, Vol. 53, 355–371.

Piper, J D A. 1995. Palaeomagnetism of Late Ordovician igneous intrusions from the northern Welsh Borderlands: implications to motions of eastern Avalonia and regional rotations. Geological Magazine, Vol. 132, 65–80.

Pitcher, B L. 1939. The Upper Valentian gastropod fauna of Shropshire. Annals and Magazine of Natural History, Series 11, Vol. 4, 82–132.

Pocock, R W, Whitehead, T H, Wedd, C B, and Robertson, T. 1938. Shrewsbury district, including the Hanwood Coalfield. Memoir of the Geological Survey of Great Britain, Sheet 152 (England and Wales).

Potter, J F, and Price, J H. 1965. Comparative sections through rocks of Ludlovian–Downtonian age in the Llandovery and Llandeilo districts. Proceedings of the Geologists’ Association, Vol. 76, 379–402.

Powell, D W. 1956. Gravity and magnetic anomalies in North Wales. Quarterly Journal of the Geological Society of London, Vol. 111, 375–397.

Ramsay, A C. 1853. On the physical structure and succession of the Lower Palaeozoic rocks of North Wales and part of Shropshire. Quarterly Journal Geological Society of London, Vol. 9, 161–176.

Ramsay, A C. 1881. The geology of North Wales (2nd edition). Memoir of the Geological Survey of Great Britain, Vol. III.

Ramsay, A C, Bristow, H W, and Bauerman, H. 1858. A descriptive catalogue of the rock specimens in the Museum of Practical Geology. Geological Survey of Great Britain.

Ramsay, A C, and Aveline, W T. 1848. Sketch of the structure of parts of North and South Wales.Quarterly Journal of the Geological Society of London, Vol. 4, 294–297.

Rickards, R B. 1978. Silurian. 130–145 in The geology of the Lake District. Occasional Publication of the Yorkshire Geological Society. No. 3.

Ritchie, M E A, Ford, G D, and Musson, R MW. 1995. The Newtown earthquake of 17 March 1994 (3.1 ML). British Geological Survey, Global Seismology Series, Technical Report, No. WL/95/35.

Ritchie, M E A, Musson, R M W, and Woodcock, N H. 1990. The Bishop’s Castle earthquake of 2 April 1990. Terra Nova, Vol. 2, 390–400.

Ritchie, M E A, and Walker, A B. 1991. Focal mechanisms and the determination of stress directions in western Britain. British Geological Survey, Global Seismology Series, Technical Report, No. WL/91/22.

Ritchie, M E A, and Woodcock, N H. 1990. The Bishop’s Castle earthquake of 2 April 1990. British Geological Survey Gl;obal Seismology Series, Technical Report, No. WL/90/32.

Roberts, B, Merriman, R J, and Pratt, W T. 1991. The influence of strain, lithology and stratigraphical depth on white mica (illite) crystallinity in mudrocks from the vicinity of the Corris Slate Belt, Wales: implications for the timing of metamorphism in the Welsh Basin. Geological Magazine, Vol. 128, 633–645.

Rowlands, P H. 1966. Pleistocene stratigraphy and palynological investigation in the Marton, Church Stoke and Church Stretton valleys, west Shropshire. Unpublished PhD thesis, University of Birmingham.

Rundle, C C. 1984a. Rb–SR and K/Ar dating of rocks from the Shelve Inlier, Powys, Welsh Borderland. British Geological Survey Isotope Geochemical Research Group Report, No. 84/8.

Rundle, C C. 1984b. Radiometric dating of a Caradocian tuff horizon. Bulleting. Liais. Inform. IGCP project 196/3, 26–31.

Rushton, A W A. 1985. Notes on the faunas of the Shelve Inlier, their facies and their correlation. In Geological notes and local details for 1:10 000 sheet SO 39 SW (Hyssington and Lydham). Cave, R, Lynas B D T, and Langford R L (editors).British Geological Survey Report, WA/DM/85/9

Rushton, A W A. 1988. Tremadoc trilobites from the Skiddaw Group in the English Lake District. Palaeontology, Vol. 31, 677–698.

Salter, J W, and Aveline, W T. 1854. On the Caradoc Sandstone of Shropshire. Quarterly Journal of the Geological Society of London, Vol. 10, 62–75.

Sanderson, R W. 1975. Andesitic agglomerate in the Caradocian succession at Montgomery Castle, Wales. Bulletin of the Geological Survey of Great Britain, Vol. 52, 43–49.

Sanderson, R W, and Cave, R. 1980. Silurian volcanism in the central Welsh Borderland. Geological Magazine, Vol. 117, 455–462.

Sedgwick, A. 1844. An outline of the geological structure of North Wales. Proceedings of the Geological Society of London, Vol. 4, 212–224.

Shergold, J H, and Bassett, M G. 1970. Facies and faunas at the Wenlock/Ludlow boundary of Wenlock Edge, Shropshire. Lethaia, Vol. 3, 113–142.

Siveter, D J. 1980. British Silurian Beyrichiacea (Ostracoda). Monographs of the Palaeontographical Society, Vol. 1, 1–76.

Siveter, D J, Owens, R M, and Thomas, A T. 1989. Silurian field excursions: a geotraverse across Wales and the Welsh Borderland. National Museum of Wales Geological Series, No. 10.

Smith, B, and Dewey, H. 1922. Lead and zinc ores in the pre-Carboniferous rocks of West Shropshire and North Wales. Memoirs of the Geological Survey, Special Report on the Mineral Resources of Great Britain, No. 23.

Smith, N J P. 1987. The deep geology of central England: the prospectivity of the Palaeozoic rocks. 217–225 in Petroleum geology of North-West Europe. Brooks, J, and Glennie, K (editors). (London: Graham and Trotman.)

Smith, R D A, and Ainsworth, R B. 1989. Hummocky cross-stratification in the Downton of the Welsh borderland. Journal of the Geological Society of London, Vol. 146, 897–900.

Snoke, J A, Munsey, J W, Teague, A G, and Bollinger, G.A. 1984. program for focal mechanism determination by combined use of polarity and Sv/P amplitude ratio data. Earthquake Notes, Vol. 55, 15.

Soper, N J, Webb, B C, and Woodcock, N H. 1987. Late Caledonian (Acadian) transpression in north-west England: timing, geometry and geotechnic significance. Proceedings of the Yorkshire Geological Society, Vol. 46, 175–192.

Squirrell, H C, and White, D E. 1978. Stratigraphy of the Silurian and Old Red Sandstone of the Cennen Valley and adjacent areas, south-east Dyfed, Wales. Report of the Institute of Geological Sciences, No. 78/6

Stamp, L D. 1919. The highest Silurian rocks of the Clun Forest District (Shropshire). Quarterly Journal of the Geological Society of London, Vol. 74 (for 1918), 221–246.

Strachan, I. 1986. The Ordovician graptolites of the Shelve district, Shropshire. Bulletin of the British Museum (Natural History), Geology, Vol. 40, 1–58.

Strachan, I, Temple, J, and Williams, A. 1948. The age of the Neptunian Dyke at Hazler Hill. Geological Magazine, Vol. 85, 276–278.

Straw, S H. 1937. The higher Ludlovian rocks of the Builth district. Quarterly Journal of the Geological Society of London, Vol. 93, 406–456.

Straw, S H. 1953. The Silurian succession of Cwm Craig Ddu (Breconshire). Liverpool and Manchester Geological Journal, Vol. 1, 208–219.

Stubblefield, C J, and Bulman, O M B. 1927. The Shineton Shales of the Wrekin district: with notes of their development in other parts of Shropshire and Herefordshire. Quarterly Journal of the Geological Society of London, Vol. 83, 96–146.

Summerhayes, C P. 1986. Sealevel curves based on seismic stratigraphy: their chronostratigraphic significance. Palaoegeography, Palaeoclimatology, Palaeoecology, Vol. 57, 27–42.

Toghill, P, and Chell, K. 1984. Shropshire geology — stratigraphic and tectonic history. Field Studies, Vol. 6, 59–101.

Trench, A, and Torsvik, T H. 1991a. Comment on: ‘Palaeomagnetic results from volcanic rocks of the Shelve Inlier, Wales: evidence for a wide Late Ordovician Iapetus Ocean in Britain.’ by C McCabe and J E T Channel. Earth Science Planetary Letters, Vol. 104, 535–529.

Trench, A, and Torsvik, T H. 1991b. A revised Palaeozoic apparent polar wander path for Southern Britain (Eastern Avalonia). Geophysical Journal International, Vol. 104, 227–233.

Tucker, R D, and McKerrow, W S. 1995. Early Palaeozoic chronology: a review in the light of new U-Pb zircon ages from Newfoundland and Britain. Canadian Journal of Earth Sciences, Vol. 32, 368–379.

Turbitt, T, Barker, E J, Browitt, C W A, Howells, M F, Marrow, P C, Musson, R M W, Newmark, R H, Redmayne, D W, Walker, A B, Jacob, A W B, Ryan E, and Ward, V. 1985. The North Wales earthquake of 19 July 1984. Journal of the Geological Society of London, Vol. 142, 567–571.

Turner, R E. 1984. Acritarchs from the type area of the Ordovician Caradoc Series, Shropshire, England. Palaeontographica, Abteiling B, Vol. 190, 87–157.

Tyler, J E. 1987. Clastic marine facies in the Ludlow of the Central Welsh region. Unpublished PhD thesis, University of Cambridge.

Tyler, J E, and Woodcock, N H. 1987. The Bailey Hill Formation: Ludlow Series turbidites in the Welsh Borderland reinterpreted as distal storm deposits. Geological Journal, Vol. 22, 73–86.

Vail, P R, Mitchum, R M Jr, and Thompson, S, III. 1977. Seismic stratigraphy and global changes of sea level. Part 4: Global cycles of relative changes of sea level. 83–97 in Seismic stratigraphy — applications to hydrocarbon exploration. Payton, C E (editor). Memoir of the American Association of Petroleum Geologists, No 26.

Verniers, J, Nestor, V, Paris, F, Dufka, P, Sutherland, S, and Van Grootel, G. 1995. A global Chitinozoa biozonation for the Silurian. Geological Magazine, Vol. 132, 651–666.

Walmsley, V G. 1959. The geology of the Usk Inlier (Monmouthshire). Journal of the Geological Society of London, Vol. 67, 415–459.

Warren, P T, Price, D, Nutt, M J C, and Smith, E G. 1984. Geology of the country around Rhyl and Denbigh. Memoir of the British Geological Survey, Sheets 90 and 97 and parts of 94 and 106 (England and Wales).

Waters, R A, and Lawrence, D J D. 1987. Geology of the South Wales Coalfield, Part III, the country around Cardiff (3rd edition). Memoir of the British Geological Survey, Sheet 263 (England and Wales).

Watkins, R. 1979. Benthic community organisation in the Ludlow Series of the Welsh Borderland. Bulletin of the British Museum (Natural History), Geology, Vol. 31, 175–280.

Watts, W W. 1885. On the igneous and associated rocks of the Breidden Hills in east Montgomeryshire and west Shropshire. Quarterly Journal of the Geological Society of London, Vol. 41, 532–546.

Watts, W W. 1887. A Shropshire picrite. Report of the British Association, Manchester, 700.

Watts, W W. 1886. The Corndon laccolites. Report of the British Association, Birmingham, 670–671.

Watts, W W. 1925. The geology of south Shropshire. Proceedings of the Geologists’ Association of London, Vol. 36, 321–363.

Wedd, C B. 1932. Notes on the Ordovician rocks of Bausley, Montgomeryshire. Summary of Progress, Geological Survey of Great Britain, Part 2, 49–55.

Wedd, C B, Smith, B, King, W B R, and Wray, D A. 1929. The country around Oswestry. Memoir of the Geological Survey, Sheet 137 (England and Wales).

Whitaker, J H McD. 1962. The geology of the area around Leintwardine, Herefordshire. Quarterly Journal of the Geological Society of London, Vol. 118, 319–351.

Whitaker, J H McD. 1994. Silurian basin slope sedimentation and mass movement in the Wigmore Rolls area, central Welsh Borderland. Journal of the Geological Society of London, Vol. 151, 27–40.

White, D E. 1981. The base of the Ludlow Series in the graptolite facies. Geological Magazine, Vol. 118, 566.

White, D E. 1988. Biostratigraphy of the Wenlock–Ludlow Series boundary beds in Aston Dingle. 1: 50 000 Montgomery (165) Sheet. Unpublished BGS Report. WH88 289R.

White, D E, Barron, H F, Barnes, R P, and Lintern, B C. 1992. Biostratigraphy of late Llandovery (Telychian) and Wenlock turbiditic sequences in S W Southern Uplands, Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol. 82, 297–322.

White, D E, and Lawson, J D. 1989. The Přídolí Series in the Welsh Borderland. 131–141 in A global standard for the Silurian System. National Museum of Wales Geological Series, No. 9.

Whittard, W F. 1928. The stratigraphy of the Valentian rocks of Shropshire: the main outcrop. Quarterly Journal of the Geological Society of London, Vol. 83 (for 1927), 737–759.

Whittard, W F. 1931. The geology of the Ordovician and Valentian rocks of the Shelve country, Shropshire. Proceedings of the Geologists’ Association , Vol. 42, 322–344.

Whittard, W F. 1932. The stratigraphy of the Valentian rocks of Shropshire. The Longmynd–Shelve and Breidden outcrops. Quarterly Journal of the Geological Society of London, Vol. 88, 859–902.

Whittard, W F. 1938. The Upper Valentian trilobite fauna of Shropshire. Annals and Magazine of Natural History, Series 11, Vol. 1, 85–140.

Whittard, W F. 1952. A geology of south Shropshire. Proceedings of the Geologists’ Association, Vol. 63, 143–197.

Whittard, W F. 1955–1967. The Ordovician trilobites of the Shelve inlier, west Shropshire. Parts 1–9 (1955, 1956, 1958, 1960, 1961 (Parts 5 and 6), 1964, 1966, 1967). Monograph of the Palaeontographical Society, London.

Whittard, W F. 1960. Lexique Stratigraphique International. Vol. 1: Europe. Fascicule 3a: England, Wales and Scotland. Part 3A IV: Ordovician. International Geological Congress, Commission on Stratigraphy. Centre National de la recherche Scientifique.

Whittard, W F. 1979. An account of the Ordovician rocks of the Shelve Inlier in west Salop and part of north Powys (by the late W F Whittard, compiled by W T Dean). Bulletin of the British Museum (Natural History) Geology, Vol. 33,1–69.

Whittard, W F, and Barker, G H. 1950. The Upper Valentian brachiopod fauna of Shropshire. Annals and Magazine of Natural History, Series 12, Vol. 3, 553–590.

Williams, A. 1951. Llandovery brachiopods from Wales with special reference to the Llandovery district. Quarterly Journal of the Geological Society of London, Vol. 107, 85–136.

Williams, A. 1953. The geology of the Llandeilo district, Carmarthenshire. Quarterly Journal of the Geological Society of London, Vol. 108, 177–208.

Williams, A. 1963. The Caradocian brachiopod faunas of the Bala district, Merionethshire. Bulletin of the British Museum (Natural History) Geology, Vol. 8, 330–471.

Williams, A. 1973. Distribution of brachiopod assemblages in relation to Ordovician Palaeogeography. Special Papers in Palaeontology, No. 12, 241–269.

Williams, A. 1974. Ordovician Brachiopoda from the Shelve district, Shropshire. Bulletin of the British Museum of National History (Geology), Supplement, Vol. 11, 1–163.

Williams, A, Lockley, M G, and Hurst, J M. 1981. Benthic palaeocommunities represented in the Ffairfach group and coeval Ordovician incursions of Wales. Palaeontology, Vol. 24, 661–694.

Williams, A, Strachan, I, Bassett, D A, Dean, W T, Ingham, J K, Wright, A D, and Whittington, H B. 1972. A correlation of Ordovician rocks in the British Isles. Special Report of the Geological Society of London, No.3, 1–74.

Wills, L J, and Smith, B. 1922. The Lower Palaeozoic rocks of the Llangollen district with special reference to the tectonics. Quarterly Journal of the Geological Society of London, Vol. 78, 176–226.

Wilson, G V, Eastwood, T, Pocock, R W, Wray, D A, and Robertson, T, with Dines, H G. 1922. Barytes and witherite. Memoirs of the Geological Survey, Special Report on the Mineral Resources of Great Britain, No. 2 (3rd edition).

Winchester, J A, and Floyd, P A. 1977. Geochemical discrimination of different magma series and their differentiation products using immobile elements. Chemical Geology, Vol. 20, 325–431.

Wood, E M R. 1900. The Lower Ludlow formation and its graptolite fauna. Quarterly Journal of the Geological Society of London, Vol. 56, 415–492.

Wood, E M R. 1906. The Tarannon Series of Tarannon. Quarterly Journal of the Geological Society of London, Vol. 62, 644–701.

Woodcock, N H. 1973. The structure of the slump sheets in the Ludlow Series of east central Wales. Unpublished PhD thesis, University of London.

Woodcock, N H. 1976a. Structural style in slump sheets: Ludlow Series, Powys, Wales. Journal of the Geological Society of London, Vol. 132, 399–415.

Woodcock, N H. 1976b. Ludlow Series slumps and turbidites and the form of the Montgomery Trough, Powys, Wales. Proceedings of the Geologists’ Association, Vol. 87, 169–82.

Woodcock, N H. 1984a. Early palaeozoic sedimentation and tectonics in Wales. Proceedings of the Geologists’ Association, Vol. 95, 323–335.

Woodcock, N H. 1984b. The Pontesford lineament, Welsh Borderland. Journal of the Geological Society of London, Vol. 141, 1001–1014.

Woodcock, N H, and Gibbons, W. 1988. Is the Welsh Borderland fault system a terrane boundary? Journal of the Geological Society of London, Vol. 145, 915–923.

Woodcock, N H, and Pauley, J C. 1989. The Longmyndian rocks of the Old Radnor Inlier, Welsh Borderland. Geological Journal, Vol. 24, 113–120.

Woollard, G P. 1979. The new gravity system — changes in international gravity base values. Geophysics, Vol. 44, 1352–1366.

Zalasiewicz, J A. 1991. Geological notes and details of 1:10 000 sheets SO 18 NW (Cilfaesty Hill), SO 18 SE (part) (Felindre), SO 17 NW (part) (Ddol) and SO 17 NE (part) (Beguildy). British Geological Survey Technical Report, No. WA/91/29.

Ziegler, A M, Cocks, L R M, and McKerrow, W S. 1968. The Llandovery transgression of the Welsh borderland. Palaeontology, Vol. 11, 736–782.

Ziegler, A M, and McKerrow, W. 1975. Silurian marine red beds. American Journal of Science, Vol. 275, 31–56.

Appendix 1 Church Stretton No. 1 (Wentnor) borehole

Church Stretton No.1 (Wentnor). Height above OD + 167.06 m SO 38 NE [SO 3723 8978]. Complete log by Dr B A Hains, available for inspection at the British Geological Survey. Fossil numbers: B2050–B2625

Thickness m Depth m
Alluvium Sandy silt with small pebbles 1.78 1.78
Till
Clay, silty, with abundant pebbles and two layers of ill-sorted gravel up to 0.5 m Cored below 2.90 m 5.99 7.77
Wenlock
Bromleysmill Shales Formation
Siltstones, grey, argillaceous, finely micaceous, poorly bedded. Fragments of trilobites, brachiopods, bivalves, nautiloids, gastropods and monograptids 9.68 17.45
Mudstone, grey, silty, finely micaceous and unevenly bedded with fragments of trilobites, bivalves, nautiloids and monograptids 2.67 20.12
Siltstone as to 17.45 m 1.67 21.79
Mudstone, grey, silty, finely micaceous, poorly bedded. Scattered calcareous nodules. Fragments of trilobites; orthocones become less common and monograptids become more common downwards, sparse cyrtograptids 10.52 32.31
Mudstone, grey, slightly greenish in parts, silty, finely micaceous. Thin sparse limestones up to 0.15 m. Soft pale grey clay up to 0.03 m thick occurs at: 35.36, 41.73, 45.11, 46.84 and 52.65 m depth. Fossils are mainly monograptids with cyrtograptids and a few fragments of trilobites and brachiopods 21.11 53.42
Mudstone, grey, greenish in parts, silty, finely micaceous, massive to finely laminated. Burrows recorded sparsely. Scattered fragments of trilobites, brachiopods, crinoids, orthocones, monograptids and plants with pale grey clay or soft silty layers (some with calcite laminae at base), up to 0.10 m thick at 55.50, 55.52, 55.91, 56.44, 56.64, 57.43, 57.53, 57.90, 58.06, 58.37 and 58.67 m depth 6.75 60.17
Siltstone, pale grey, calcareous 0.03 60.20
Mudstone, grey, slightly silty, massive, weakly calcareous. Some calcareous nodules and widely separated limestone beds, up to 0.15 m, with much shell debris; sparse monograptids. Bioturbation strong in parts. Pale grey clay/mudstone beds at 60.45, 70.10 and 70.43 m depth 16.30 76.50
Mudstone, greyish green, massive, slightly silty, weakly calcareous. Numerous diffuse calcareous nodules, some tinted red. Abundant shelly debris, especially crinoidal near the base. Soft, pale grey clay up to 0.01 m thick with some biotite at 76.83, 77.70 and 78.59 m 2.09 78.59
Llandovery
Purple Shales Formation
Mudstone, soft, green, laminated, slickensided. Yellow-green clay with red spots at 78.64 m. Fossil debris at base 0.10 78.69
Limestone, crystalline, variegated reddish brown and green 0.03 78.72
Mudstone, reddish brown, silty, with comminuted shell debris 0.16 78.88
Clay, soft, pale greenish grey 0.04 78.92
Mudstone, reddish brown, slightly silty. Several green bands in lower part, well laminated. Some brachiopod debris 1.34 80.26
Mudstone, green, slightly silty and laminated 0.21 80.47
Mudstone, deep maroon, but green bands and spots are common; slightly silty with silt laminae. Listric surfaces common, some calcareous nodules; burrows. Polished, irregular fracture (listric) surfaces common. Thin limestones below 131.90 m. Pale grey clay up to 0.05 m thick at 101.52 , 131.37, 131.67 and 131.70 m. Brachiopods and ostracods sparse above 131.90 m, more common below together with trilobites, crinoids and bivalves. Lingula present 153.21 to 187.45 m, 244.81 and 257.40 m. Orbiculoidea sp 259.13 m, Pentamerus sp. 261.09 m, Graptolites at 105.16, 108.20, 111.61–111.81, 128.37, 130.10, 137.72, 139.70, 179.22 and 248.72 m. Much brachiopod debris in basal 0.8 m. Base uneven 184.40 264.87
Precambrian
Longmyndian Supergroup
Sandstone, purplish brown, hard and massive, medium- to fine-grained with beds of siltstone and silty mudstone, some graded beds. Small slump-fold at 265.38 m 2.21 267.08
Sandstone crushed — fault zone 1.14 268.22
Sandstone as above to terminal depth at 274.32

Appendix 2 Biostratigraphy of the Aston Mudstone Formation, Aston Dingle

Fossil inventory

To satisfy the rules and recommendations of the international codes of botanical and zoological nomenclature, authors of cited species are listed below.

Chapter 3

Brachiopods

Molluscs

Trilobites

Ostracods

Crinoid

Graptolites

Chapter 4 Caradoc

Acritarchs

Brachiopods

Mollusca

Trilobites

Ostracods

Graptolites

Chitinozoa

Chapter 5

Brachiopods

Trilobites

Ostracods

Graptolites

Chapter 6 Wenlock, Ludlow and PrŠídolí

Brachiopods

Molluscs, Bivalves

Molluscs, Gastropods

Molluscs, Cephalopods

Hyolith

‘Worm’

Trilobites

Ostracods

Echinoderm

Graptolites

Figures, plates and tables

Figures

(Figure 1) Geological sketch map of the district, including the Shelve area.

(Figure 2) Physical features of the district.

(Figure 3) Precambrian rocks of the district.

(Figure 4) Ordovician rocks of the Shelve area.

(Figure 5) Sections in Linley Big Wood (recorded by Dr A W A Rushton, 1987) and Granham’s Moor

quarry (after Lynas, 1985a) exposing the Stiperstones Quartzite and Shineton Shale Formation.

(Figure 6) Schematic representation of the relationship of the Hyssington Volcanic Member to the Stapeley Volcanic Member.

(Figure 7) Lithological log of quarry section [SO 3165 9395] in Fremes Wood exposing the second unit of the Hyssing ton Volcanic Member. See text for description of Beds 1–12.

(Figure 8) Lithological log of roadside section [SO 3193 9421] at The Llan, exposing the third unit of the Hyssington Volcanic Member. See text for description of Beds 1–11.

(Figure 9) Composite section in Tasker quarry [SO 3251 9565] and roadside to the north, Hyssington Volcanic Member. See text for description of Beds 1–26.

(Figure 10) Section through the Hyssington Volcanic Member on Pellrhadley Hill [SO 3389 5981]. See text for description of Beds 1–7.

(Figure 11) Lithological log of part of the Stapeley Volcanic Member. Upper Hurdley Quarry [SO 2961 9445]. See text for description of Beds 1–6.

(Figure 12) Palaeogeography of early Caradoc (Costonian) times. This interpretation is based on the assumption that there has been negligible strike-slip on the Pontesford–Linley and Church Stretton faults.

(Figure 13) Nemagraptus gracilis from the middle of the Spy Wood Sandstone Formation.

(Figure 14) Quartz-feldspar porphyry and breccia from the Hagley Volcanic Formation, Aldress Dingle [SO 2768 9579]. See text for description of Beds 1–4.

(Figure 15) Section through part of the Whittery Volcanic Formation at Caerbre Quarry, Marrington Dingle [SO 2741 9650]. See text for description of Beds 1–9.

(Figure 16a) Early Llandovery palaeogeography (late Rhuddanian).

(Figure 16b) Late Llandovery palaeogeography (late Telychian, griestoniensis Biozone). A Aberystwyth, B Builth Wells, Ba Bala, C Corwen, CS Church Stretton, LD Llandeilo, L Llandovery, LG Llangadfan, Ll Llangollen, N Newtown, O Oswestry, OR Old Radnor, R Rhayader, W Welshpool CSF Church Stretton Fault, SVF Severn Valley Faults, TA Tywi Anticline.

(Figure 17) Diagrammatic section through Lower Silurian formations showing the differences in thickness and the basal unconformity between Montgomery and Craven Arms.

(Figure 18) A generalised traverse through the Wenlock and Ludlow sequence, from east of Bishop’s Castle to Gregynog (modified after Cave et al., 1993).

(Figure 19) Early Wenlock sedimentary facies (approximately riccartonensis–rigidus biozones). Abbreviations: BA Brecon Anticline, CSF Church Stretton Fault, CVF Conway Valley Fault, DA Derwen Anticline, LA Ludlow Anticline, LEF Leinthall-Earls Fault, SVF Severn Valley Faults, TA Tywi Anticline. Place names A Aberystwyth, B Builth Wells, Ba Bala, C Chester, CA Craven Arms, CS Church Stretton, D Dolgellau, H Hereford, Ll Llangollen, LD Llandeilo, LS Llanidloes, LY Llandovery, M Montgomery, N Newtown, O Oswestry, OR Old Radnor, P Presteigne, R Rhayader, W Welshpool.

(Figure 20) Ludlow sedimentary facies (see facing page for key). a. nilssoni Biozone (level of Middle Elton Formation). b. tumescens Biozone (level of Bailey Hill Formation or top part of the Upper Elton Formation).c. Latest Ludlow palaeogeography (level of Cefn Einion Formation or the upper part of the Upper Whitcliffe Formation. Abbreviations: a. Tresglen Beds; b. Llety Beds; c. Basal Black Cock Beds; d. Lower Ludlow Graptolitic Shales; e. Glyndyfrdwy Stage; f. Upper Nantglyn Flags; g. Nantglyn Flags; h. Lower Forest Beds; i. Hill Gardens Formation — Llanedeyrn Formation (Eastern Avenue Member); j. Middle Elton Formation; k. Black Cock Beds: Carn Powell Facies and Trichrug Beds; l. Top Black Cock Beds and Lower Cwm Clyd Beds; m. Wilsonia Shales — finely flaggy siltstones; n. Upper Elton Formation — Lower Bringewood Formation; o. Upper Forest Beds; p. Bailey Hill Formation; q. Wilsonia wilsoni Grits; r. Nant-y-Bache Group; s. Elwy Group; t. Long Mountain Siltstone Formation: Pentre Member; u. Cennen Beds; v. ?Upper Roman Camp Beds; w. Roath Park Lake Member; a. Middle Llangibby Beds; b. Basal Holopella Grits and Shales; g. Upper Whitcliffe Formation; d. Wern Quarry Beds; ´. Dalmanella lunata Beds; u. Causemountain Formation (upper part); l. Dinas Bran Group; m. Striped Flags; B. Builth Wells, N. Newtown. References: 1. Squirrell and White, 1978; 2. Potter and Price, 1965; 3. Straw, 1953; 4. Wood, 1900; 5. Holland et al., 1963; Whitaker, 1961; 6. Holland, 1959; 7. Earp, 1938, 1940; 8. Wedd et al., 1927; Wills and Smith, 1922; 9. Warren et al., 1984; 10. Cummins, 1959a, b; 11. Walmsley, 1959; 12. Waters and Lawrence, 1987; 13. Lawson, 1973; 14. Straw, 1937; 15. Bailey, 1969; 16. Kirk, 1951a, b; 17. Palmer, 1970; 18. Lawson, 1955; Cave and White, 1971.

(Figure 21) Plot of Nb/Y against Zr/TiO2 for some Wenlock and Ludlow bentonites.

(Figure 22) Diachroneity of Lower and Middle Wenlock formations.

(Figure 23) Lithological log through the contact between the Nant-ysgollon Shales Formation and the Mottled Mudstone Member of the Nantglyn Flags Formation. Farm track near Fron Bank [SO 1795 9735]. Thicknesses in metres.

(Figure 24) Road section in the Nantglyn Flags Formation, 1 km north-west of Llanbadarn Fynydd. Thicknesses in metres.

(Figure 25) Lithological log illustrating the Gregynog Mudstone member. Stream section [SO 0910 9794] to [SO 0890 9790], 600 m north-east of Gregynog Hall.

(Figure 26) Section in Nantglyn Flags Formation. Uneven surface of thick slump sheet (debris flow deposit) [SO 1631 9833]. Vertical scale slightly exaggerated. Thicknesses in metres.

(Figure 27) Section in Lower Ludlow strata. a. Sketch map of Gyfenni Wood [SO 2171 9057] b. Section through the basal beds of the Bailey Hill Formation. Thickness in metres.

(Figure 28) Aston Dingle. a. Geological map showing biozones. 1. Biozone of G. nassa 2. Biozone of M. ludensis (with sparse G. nassa) 3. Biozone of M. ludensis (only large monograptids present) 4. Biozone of N. nilssoni b. Fossil localities a to u. See Appendix 2 for biostratigraphy c. Section across the Wenlock–Ludlow boundary in upper Aston Dingle Beds thickness in metres.

(Figure 29) Chart illustrating the chronostratigraphical mismatch at the boundary between map sheets 165 and 166.

(Figure 30) Basal Bailey Hill Formation and topmost Nantglyn Flags Formation exposed in the bank of a farm track [SO 1434 9552 to 1407 9532] near Brynrorin (Cave et al., 1993).

(Figure 31) Section base at [SO 1652 9288] in the basal part of the Bailey Hill Formation, forestry track, River Mule gorge.

(Figure 32) Basic and intermediate intrusions of the Ordovician rocks of the Shelve area.

(Figure 33) Late Devensian palaeogeography. Collage of periglacial lakes formed successively during late Devensian ice wastage across the Montgomery district and part of the adjacent Welshpool district (Sheet 151). The figure does not represent any one stage in the process, but condenses successive stages of ice decay and general westward withdrawal as a cartoon of resultant features. The ice-front levels cited are only those crucial to the maintenance of particular lakes.

(Figure 35) Silurian fold pattern in the western part of the district.

(Figure 36) Kubler indices of white mica crystallinity of 19 mudstone samples of Caradoc to Ludlow age.

(Figure 37) Bouguer gravity anomaly map of the district and surrounding area. Bouguer reduction density = 2.7 Mg/m3 Contour interval = 0.5 m Gal Annotated anomalies are referred to in text

(Figure 38) Camlad valley. a. Gravity anomaly Bouguer map including infill. Bouguer reduction density = 2.7 Mg/m3 Contour interval = 0.25 m Gal b. Interpretation of profile AB. See (a) for location. Density contrast of channel fill with surrounding bedrock set at–55Mg/m3; magnetisation of 0.085 A/m (inclination 20º, declination 180º).

(Figure 39) Regional aeromagnetic anomaly map of the district and surrounding area. Contour interval = 5nT. Annotated anomalies are referred to in the text.

(Figure 40) Interpretation of gravity and magnetic anomaly data along grid northing 2965. Polygon density (Mg/m3) / susceptibility (10-3S1 units) 1 = 2.68/0 2 = 2.70/25 3 = 2.85/10 (will remanent magnetisation 0.5 A/m background 2.70/0 4 = 2.70/0 5 = 2.74/0

(Figure 41) Fault plane solution for the Bishop’s Castle earthquakes. Equal area projection of the upper focal hemisphere. The axes of maximum and minimum compressive stress are denoted by P and T respectively. a. mainshock of 2 April 1990 at 13.46 UTC, magnitude 5.1 ML b. aftershock of 17 April 1990 at 00.52 UTC, magnitude 0.5 ML.

Plates

(Plate 1) Forden Mudstone Formation, sandstone bed, small ripple cross-bedding, both planar and trough occur in these sandstones. Base with flute casts, groove casts and prod-marks from currents left to right across the cut face. Burrow casts also common. Quarry [SO 2115 9372] 450 m east of Drainllwynellyn, Sarn (GS 432).

(Plate 2) Silurian graptolites (GS 442). Selected Wenlock (1–5) and Ludlow (6–18) Series graptolites in the BGS collections in Keyworth. Photographs are 3 6, except 11 (3 2) 1 Monograptus ludensis (Murchison). DEY 9854. Aston Mudstones Formation. Silage pit at West Dudston, Near Chirbury [SO 2432 9754]. 2 Monograptus flemingii cf. elegans Elles. CAV 2092. Nantglyn Flags Formation (Gregynog Mudstone Member). Quarry, Upper Cwm Harry, Tregynon [SO 1241 9855]. 3 Gothograptus nassa (Holm). CAV 528. Aston Mudstones Formation. Waterfall near Upper Broughton Farm, Bishop’s Castle [SO 3040 9094]. 4 Pristiograptus jaegeri Holland, Rickards and Warren. DEY 2581. Aston Mudstones Formation. Quarry, Upper Heblands Farm, Bishop’s Castle [SO 3252 9029]. 5 Monograptus cf. deubeli Jaeger. DEY 2524. Formation and locality as for Fig. iv. 6 Monograptus uncinatus orbatus Wood. CAV 1017 pars. Nantglyn Flags Formation. Bank, Cwm Badern [SO 1997 9416]. 7 Spinograptus spinosus (Wood). CAV 1017 pars. Formation and locality as for Fig. vi. 8 Saetograptus (S.) clunensis (Earp). DEY 7663. Bailey Hill Formation. Road cutting, Cwmffrydd Farm, Mainstone [SO 2514 8702]. 9, 10 Lobograptus scanicus (Tullberg). CAV 1660. Bailey Hill Formation (Dingle Mudstones Member). Trackside, near Brynrorin, Bettws Cedewain [SO 1433 9552]. 11 Bohemograptus? cf. butovicensis (Bouček). CAV 1142. Nantglyn Flags Formation. Roadside quarry, Cwmberllan [SO 2025 9252]. 12 Bohemograptus bohemicus tenuis (Bouček). DEY 7776. Bailey Hill Formation. Outcrop near Mainstone [SO 2395 8770]. 13 Monograptus aff. unguiferus Perner. DEY 9606. Bailey Hill Formation. Track near Llandyssil [SO 1931 9618]. 14 Monoclimacis micropoma (Jaekel). DEY 8123. Oakeley Mynd Formation. Track near Oakeley House, near Bishop’s Castle [SO 3434 8805]. 15 Saetograptus (S.) leintwardinensis (Lapworth) s.l. SPT 4623. Bailey Hill Formation. Track, Lower House Farm, Felindre c.552 7994]. 16 Saetograptus (S.) aff. incipiens (Wood). SPT 4956. Bailey Hill Formation. Exposure near Ddol, Felindre [SO 1185 8479]. 17 Saetograptus (Colonograptus) colonus compactus (Wood). CAV 1710. Nantglyn Flags Formation (Gyfenni Wood Shales Member). Track, Cwm Dawkin, Bettws Cedewain [SO 1437 9585]. 18 Saetograptus (Colonograptus) varians (Wood). DEY 9869. Nantglyn Flags Formation (Gyfenni Wood Shales Member). Track near Sarn [SO 2163 9040].

(Plate 3) Silurian shelly fossils (GS 443). Selected shelly fossils of the Wenlock (1–7), Ludlow (8) and Přídolí (9) Series in the BGS collections in Keyworth. Photographs are 3 6, except 8 (3 2) 1 Harpidella sp. DEY 6072. External and internal moulds, with internal mould of Raphiophorus cf. parvulus (Forbes) in bottom right corner. Aston Mudstones Formation (Edgton Limestone Member). Quarry, Lydbury North [SO 3621 8629]. 2 ‘Clorinda’ dormitzeri (Barrande). DEY 8795. Internal mould of pedicle valve. Aston Mudstones Formation, Aston Dingle c. [SO 2948 9117]. 3 Visbyella trewerna Bassett. DEY 6010. Internal moulds of conjoined valves. Bromleysmill Shale Formation. Roadside near Lydbury North [SO 3692 8754]. 4 Ananaspis aff. communis (Barrande). DEY 6322. Internal mould photographed at angle to show eye details. Bromleysmill Shale Formation. Driveway to Plowden Hall [SO 3758 8651]. 5 Strophochonetes sp. DEY 2599. Internal mould of pedicle valve with long spines. Aston Mudstones Formation. Quarry, Upper Heblands Farm, Bishop’s Castle [SO 3252 9029]. 6 Jonesea grayi (Davidson). DEY 7980. External and internal moulds. Aston Mudstones Formation. Trackside, Oakeley Farm, near Bishop’s Castle [SO 3496 8836]. 7 Harpidella sp. DEY 6158. Internal mould. Aston Mudstones Formation. Roadside, Home Farm, near Lydbury North [c.[SO 3596 8765]. 8 Microsphaeridiorhynchus nucula (J de C Sowerby), Protochonetes ludloviensis Muir Wood and Salopina lunata (J de C Sowerby), a characteristic late Ludfordian assemblage. SPT 2762. Internal moulds of pedicle and brachial valves. Cefn Einion Formation. Trackside, Black Hill, Clun [SO 3194 7860]. 9 Craniops implicatus (J de C Sowerby). DEY 7835. Concentration of internal moulds. Transitional beds between Cefn Einion and Clun Forest formations. Lane, Bryn c.[ SO 2965 8548].

(Plate 4) Bottle-nosed flute casts. Penstrowed Grits (Denbigh Grits). Penstrowed Quarry [SO 0680 9095] Newtown (GS 433).

(Plate 5) Rhythmite. Fine-grained sandstone/siltstone (pale grey): turbidite mudstone, homogeneous (mid-grey) and laminated hemipelagite (dark grey) with silt blebs (pale). Erosive bases at 17.50 m (a) and 20.313 m (b). Nantglyn Flags in Borehole B right bank of River Severn [SO 1128 9156], Newtown, depths in metres on ‘right margin’ (GS 434A, GS 434B).

(Plate 6) Gregynog Mudstone Member. Conglomerate of synbasinal mudstone clasts, mudstone matrix suspended and showing strong peridermal alternation rims. Specimen MR 31050 Photo MN 24248 from just above the Penstrowed Grit Formation [SO 0897 9792]. Stream bed (type section) 620 m north-east of Gregynog Hall (GS 435).

(Plate 7) Polished section showing typical Mottled Mudstone and slumped Mottled Mudstone. Quarry [SO 1220 9609] Bettws Cedewain (GS 436a, b).

(Plate 8) Conglomerate: large synbasinal mudstone/ siltstone clasts, largely matrix supported. Matrix consists of calcareous mudstone, highly charged with exotic phosphatic pelloids and shell debris. Stream bed [SO 1020 9288] 290m 354° from Llwyn-derw, Nantglyn Flags (Specimen MR 31051) (GS 437).

(Plate 9) Convolute bedding in the lower element of a fine sandstone–mudstones couplet, Bailey Hill Formation. An ‘erosive’ reactivation surface meets the base of the couplet at the position of the arrows. It has no expression on the base of the bed which is flute cast, currents flowed from left to right almost parallel with the long cut face but towards the viewer, at right angles to the shorter face. Quarry [SO 2288 9005] in upper Hopton Dingle (GS 439).

(Plate 10) Bailey Hill Formation. Broad flute casts of low relief. Quarry [SO 2288 9005] Upper Hopton Dingle (GS 440).

(Plate 11) Mule Gorge above Rock Mill, looking south-east [SO 1637 9391] (GS 441).

(Front cover) The Ordovician rocks of Shelve; the dolerite sill of Corndon Hill forms the highest ground seen on the skyline. View looking east from Montgomery Castle [SO 2215 9680] (GS 431). Photographer Audrey A Jackson.

Tables

(Table 1) Geological succession in the Montgomery district.

(Table 2) Stratigraphy of the Longmyndian Supergroup on either side of the main Longmynd Syncline.

(Table 3) Classification of the Ordovician rocks and comparison with nomenclature used by other authors.

(Table 4) Correlation of Caradoc formations.

(Table 5) Llandeilo and early Caradoc trilobites and graptolites of Shelve, Llandeilo, Builth and Berwyn.

(Table 6) Correlation of Wenlock sequences in North Wales and the Welsh Borderland.

(Table 7) analysis of Wenlock and Ludlow k-bentonites (made by Caleb Brett Laboratories Ltd).

(Table 8) Range chart of Wenlock graptolites.

(Table 9) Range chart of Ludlow graptolites.

(Table 10) Ludlow nomenclature and comparison with neaby areas.

(Table 11) Oakeley Mynd Formation: fauna from a temporary trench, north of Oakeley Mynd. Localities 1–10 are numbered in sequence south-south-east from a datum point at the intersection of the trench with a field boundary [SO 3494 8816]. Locality 10 is adjacent to the southern corner of Oakeley Wood [SO 3483 8805].

(Table 12) Classification of the Clun Forest Formation (Přídolí).

(Table 13) Modal analyses of Silurian microgranites. percentages derived from 500 counts (after Sanderson and Cave, 1980).

(Table 14) Kubler indices of white mica crystallinity of mudstone samples of Caradoc to Ludlow age, determined by X-ray diffraction analyses.

(Table 15) Synopsis of European Macroseismic Intensity Scale (EMS-92). A complete description of the EMS-92 scale is given in Grunthal (1993).

(Table 16) Hypocentral parameters of Bishop’s Castle mainshock and aftershocks. EPI Date Time Grid reference Depth Error Depth EPI event (hr/min/secs) (km) ± km error mag. no. ± km (ML)

Tables

(Table 2) Stratigraphy of the Longmyndian Supergroup on either side of the main Longmynd Syncline

WESTERN LIMB EASTERN LIMB

Longmyndian Supergroup

Wentnor Group

 Bridges Formation Bridges Formation
Bayston — Oakswood Formation Bayston-Oakswood Formation

Stretton Group

 (Faulted contact) Portway Formation

Linley Formation (equivalence uncertain)

 Lightspout Formation
Synalds Formation
Burway Formation
Stretton Shale Formation

(Table 7) analysis of Wenlock and Ludlow k-bentonites (made by Caleb Brett Laboratories Ltd)

Var.\ ID: RC-1541 RC-1542 RC-1543 RC-1545 RC-1546 RC-1547 RC-1548 RC-1553 RC-1554
SiO2 34.98 59.32 59.53 52.25 58.21 45.86 55.07 62.84 69.55
A12O3 15.69 20.89 23.62 25.42 24.08 23.08 24.44 23.04 14.45
TiO2 0.47 1.13 0.80 1.06 0.41 0.48 0.74 0.30 0.54
Fe2O3 1.44 3.81 2.58 4.18 3.06 1.94 2.35 3.01 4.53
MgO 1.82 2.31 1.88 1.61 1.95 1.78 2.70 0.68 2.52
CaO 20.62 0.40 0.13 0.90 0.14 8.47 0.54 0.05 0.41
Na2O 0.50 1.89 0.65 1.19 1.09 0.59 0.54 0.54 0.58
K2O 4.64 5.40 6.12 5.92 5.92 6.02 7.22 5.39 3.36
MnO 0.20 0.03 0.01 0.01 0.03 0.12 0.02 0.06 0.02
P2O5 0.12 0.13 0.11 0.08 0.04 0.04 0.08 0.03 0.17
LOI 19.71 4.32 4.48 7.19 4.89 11.44 5.10 4.16 3.56
Total 100.18 99.63 99.92 99.82 99.81 99.82 99.81 100.09 99.64
Ba 613 553 385 822 588 697 272 1074 261
Ce 111 119 180 35 56 88 128 103 115
Co 6 11 5 7 5 5 7 < 2 7
Hf 10 10 17 16 19 12 23 6 11
La 48 49 67 14 8 25 35 46 32
Nb 18 21 27 19 44 12 29 32 19
Ni 14 31 17 9 30 19 21 10 11
Rb 141 193 206 194 199 210 273 226 120
Sm 13 24 16 < 1 16 17 18 7 7
Sr 445 86 38 151 49 388 51 108 54
Ta 4 3 4 3 6 <2 <2 4 4
Th 11 10 37 16 45 28 17 41 22
Y 53 68 67 58 68 87 99 40 42
Yb 5 5 3 6 7 7 6 6 3
Zr 389 479 557 600 356 363 877 196 378

Samples

Sample no. Map ref. Series and biozone Sample no. Map ref. Series and biozone
RC 1541 [SO1797 9276] Ludlow, nilssoni RC 1547 1417 9540 Ludlow, scanicus-incipiens
RC 1542 [SO1785 9000] Ludlow, probably nilssoni RC 1548 2222 9021 Ludlow, probably scanicus
RC 1543 [SO1737 9583] Ludlow, scanicus RC 1553 0164 7740 Wenlock, riccartonensis
RC 1545 [SO1415 9533] Ludlow, scanicus RC 1554 2129 9007 Ludlow, nilssoni-scanicus
RC 1546 [SO1894 9132] Wenlock-Ludlow ludensis-nilssoni

(Table 8) Range chart of Wenlock graptolites

STAGES

SHEINWOODIAN

HOMERIAN
BIOZONES M. riccartonensis/ C. rigidus C. linnarssoni C. ellesae C. lundgreni G. nassa M. ludensis
Mcl. flumendosae x
M. antennularius x
P. dubiusis.l. x cf. x
M. flemingiis.l. x c c
M. flexilis cf: flexilis x
P. cf. meneghini x
C. ellesae? x
Mcl. flumendosae cf. kingi x
Plectograptus x
P. pseudodubius cf. x
C. hamatus x
C. lundgreni x
C. sp. x
·         nassa x x
M. flemingii elegans x
M. flemingii flemingii c
P. dubius dubius x
P. jaegeri cf. x c
P. cf. lodenicensis x
P. spp. x c c
P. cf. dubius ludlowensis x x
·         (Bal.) lawsoni (cf.)
Pl. macilentus (cf.)
M. auctus (cf.)
M. deubeli x
M. gerhardi x
M. ludensis c

Key to genera: C. Cyrtograptus, G. Gothograptus, H. (Bal.) Holoretiolites (Balticograptus) , Mcl. Monoclimacis, M Monograptus, P. Prisiograptus, Pl. Plectograptus, c = well represented, x = recorded. Qualification of entire identification within brackets; qualification of species identification without.

(Table 9) Range chart of Ludlow graptolites

STAGES

GORSTIAN

LUDFORDIAN
BIOZONES nilssoni1 nilssoni u L. scanicus P. tumescens (S. incipiens) S. leintwardinensis B. bohemicus
B. bohemicus bohemicus x
M. ludensis (x)
M. uncinatus orbatus c (x)
M. aff. unguiferus x
N nilssoni x (x)
Pl. macilentus x (x)
P. dubius dubius x x cf.
P. dubius s.l. x
P. vicinus x x c x
P. spp. x x x x
S. (C.) colonus colonus c c x
S. (C.) colonus compactus c c
S. (C.) varians c c x
S. spp. x x x x
Sp. spinosus x (x)
B. bohemicus s.l. x
L. crinitus x x
L. progenitor c
L.     scanicus x c
Mcl. micropoma x x
M.   cf. auctus x
S. (C.) roemeri x cf.
S. (S.) chimaera salweyi x x
L. cf. simplex x
Mc/. cf. haupti x x
P. dubius cf: ludlowensis x
P. tumescens x c
S. (S.) chimaera chimaera c
S. (S.) chimaera semispinosus c c
S. (S.) chimaera s.l. c c
S. (S.) clunensis ? ?
S. (S.) incipiens cf. c
P. cf. tumescens minor x
S. (S.) aff. incipiens x x
S. (S.) leintwardinensis leintwardinensis c
S. (S.) leintwardinensis c
Primus
S. (S.) leintwardinensis s.l. c
B. bohemicus tenuis x

Key to genera: B. Bohemograptus. C. Colonograptus, L. Lobograptus, Mcl. Monoclimacis, M. Monograptus, N. Neodiversograptus, P. Prisiograptus, Pl. Plectograptus, S. Saetograptus, Sp. = Spinograptus, c = well represented, x = recorded (x) in lower beds only, ? = Biozone uncertain.

(Table 11) Oakeley Mynd Formation: fauna from a temporary trench, north of Oakeley Mynd

Localities 1–10 are numbered in sequence south-south-east from a datum point at the intersection of the trench with a field boundary [SO 3494 8816]. Locality 10 is adjacent to the southern corner of Oakeley Wood [SO 3483 8805].

Locality no. 1 2 3 4 5 6 7 8 9 10
Brachiopoda
chonetid x
J. grayi x
Bivalvia
cardiolid x
lunulacardiid x x
Cephalopoda
Radnoroceras cf. dimidiatum x x
‘ Orthoceras' mocktreense x x x
‘ O. 'aff. nicholianum x
‘ O. 'cf. originale x
' O. ' recticinctum x cf.
‘ O. 'sp. x x
Ostracoda
Beyrichia s.l. x
Entomozoe sp. x
Graptolithina
B. bohemicus x
Lobograptus sp. x
N. nilssoni x
Pristiograptus sp. x x
S. (C.) colonus cf. cf. x cf.
S. (C.) varians x x x cf. x
S. (S.) incipiens cf.
S. sp. x x x x
indet. monograptids x x x x x x x x x x

Localities

Locality Distance from datum (m) BGS Fossil registration number
1 36 m HN 637–642
2 52 m HN 643–657
3 65 m HN 658–667
4 66m HN 668
5 67 m HN 669–670
6 68m HN 671–672
7 73 m HN 673–679
8 95m DEY 8017–8026
9 98 m HN 680–693
10 145m DEY 7987–8016

Localities 1–10 are numbered in sequence south-south-east from a datum point at the intersection of a trench with a field boudary [SO 3494 8816]. Locality 10 is adjacent to the southern corner of Oakeley Wood [SO 3483 8805].

(Table 13) Modal analyses of Silurian microgranites. percentages derived from 500 counts (after Sanderson and Cave, 1980)

Owlbury Quarry

(E 40657)

 Upper Heblands farm quarry (E 40646) Upper Heblands northern quarry (E 40651)
Quartz 19 25 15
Plagioclase 52 39 48
K-feldspar 14 24 25
Mica 14 12 10
Accessories 1 < 1 2

(Table 14) Kubler indices of white mica crystallinity of mudstone samples of Caradoc to Ludlow age, determined by X-ray diffraction analyses

Sample Grid Detailed locality Stratigraphical details White mica
no. reference crystallinity
BRM SO 0°2q
1264 [SO 131 905] 2 km ESE of Newtown Mudstone, very feeble spaced cleavage; Bailey Hill Fm (basal facies) ; Ludlow ( nilssoni — scanicus Biozone) 0.45
1265 [SO 207 911] Sarn, 6 km SSW of Montgomery Silty mudstone (hemipelagite), Wenlock (probably ludensis Biozone) 0.37
1266 [SO 207 911] Sarn, 6 km SSW of Montgomery Mudstone (turbiditic mud), Wenlock (probably ludensis Biozone) 0.36
1267 [SO 233 932] 3.5 km SE of Montgomery Blocky mudstone (uncleaved), Ludlow (nilssoni Biozone) 0.43
1268 [SO 222 964] Montgomery Castle Olive-grey mudstone (uncleaved), Caradoc (Soudleyan) 0.79
1269 [SO 239 986] Field outcrop 2.5 km NE of Montgomery Brown mudstone; Lower Wenlock 0.49
1270 [SO 231 985] Field outcrop 2 km NNE of Montgomery Olive-grey mudstone; very feeble, spaced cleavage; Caradoc (Soudleyan) 0.91
1271 [SO 202 995] River-bed outcrop, River Severn 3.5 km NW of Montgomery Grey mudstone (uncleaved) ; Caradoc, Soudleyan 0.83
1272 [SO 194 992] Garthmyl 4 km NW of Montgomery Silty mudstone (uncleaved) ; Caradoc, Soudleyan 0.83
1273 [SO 183 988] 2 km S of Berriew Blocky mudstone (uncleaved), Llandovery? 0.41
1274 [SO 173 988] 2.5 km SW of Berriew Calcareous mudstone; Wenlock (?ludensis Biozone) 0.38
1275 [SO 154 987] 3.5 km WSW of Berriew Shale (hemipelagite) , very feeble cleavage; Ludlow (nilssoni Biozone) 0.39
1276 [SO 136 987] 5.5 km WSW of Berriew Mudstone (turbiditic), very feeble cleavage; Wenlock (ludensis Biozone) 0.42
1277 [SO 125 981] 7 km WSW of Berriew Mudstone (no cleavage) ; Wenlock (lundgreni Biozone) 0.42
1278 [SO 095 988] Tregynon, 7 km NNW of Newtown Mudstone (hemipelagite) ; very feeble cleavage; Wenlock (?ludensis Biozone) 0.39
1279 [SO 105 916] River Severn, Newtown Mudstone; very feeble cleavage; Ludlow (nilssoni Biozone) 0.42
1281 [SO 068 909] Quarry, 4 km W of Newtown Mudstone, very feeble spaced cleavage; Lower Wenlock 0.61
1282 [SO 152 892] 1 km SE of Kerry near Bryn-llywarch Mudstone, very feeble spaced cleavage; Ludlow Ludlow (?nilssoni Biozone) 0.38
1283 [SO 193 901] Lane exposure, near Llancowrid Silty mudstone (uncleaved) ; Upper Wenlock (?ludensis Biozone) 0.33

(Table 15) Synopsis of European Macroseismic Intensity Scale (EMS-92). A complete description of the EMS-92 scale is given in Grunthal (1993)

(Table 16) Hypocentral parameters of Bishop’s Castle mainshock and aftershocks

EPI Date Time Grid reference Depth Error Depth EPI event (hr/min/secs) (km) ± km error mag. no. ± km (ML)

EPI event no. Date Time (hr/min/secs) Grid reference Grid reference Depth(km) Error ± km Depth(km) Error ± km
1 02 04 90 13 46 34.2 3298 2826 14.1 3.8 1.1 5.1
2 02 04 90 22 04 14.3 3300 2825 17.5 4.4 2.3 1.0
3 03 04 90 05 18 42.6 3298 2834 15.6 3.5 1.5 1.5
4 05 04 90 02 13
5 17 04 90 00 52 34.0 3309 2843 15.3 2.1 0.6 0.5
6 29 04 90 05 52 37.1 3302 2842 15.7 0.4 0.6 0.0
7 05 05 90 18 16 24.9 3301 2836 16.0 0.9 0.5 0.4
8 27 07 90 02 12 34.3 3301 2832 16.1 0.3 0.3 0.2