Geology of the Perth and Dundee district. Memoir for 1:50 000 geological sheets 48W, 48E, 49

By M. Armstrong, I. B. Paterson and M. A. E. Browne

Bibliographical reference: Armstrong, M., Paterson, I. B., And Browne, M. A. E. 1985. Geology of the Perth and Dundee district. Memoir British Geological Survey, Sheets 48W, 48E, 49, 108 pp.

British Geological Survey Scotland

Geology of the Perth and Dundee district Memoir for 1:50 000 geological sheets 48W, 48E, 49

M. Armstrong, I. B. Paterson and M. A. E. Browne

Contributors: P. J. Brand, J. I. Chisholm, A. L. Harris and D. Stephenson

1835 Geological Survey of Great Britain 150 Years of Service to the Nation. 1985 British Geological Survey. Natural Environment Research Council

London: Her Majesty's Stationery Office 1985 © Crown copyright 1985 First published 1985. ISBN 0 11 884368 0. Printed in the UK for HMSO Dd 737400 C20 8/85

• Authors:
• M. Armstrong, BSc, PhD I. B. Paterson, BSc M. A. E. Browne, BSc British Geological Survey, Murchison House, West Mains Road, Edinburgh EH9 3LA
• Contributors P. J. Brand, BSc, J. I. Chisholm, MA and D. Stephenson, BSc, PhD British Geological Survey, Edinburgh
• L. Harris, BSc, PhD University of Liverpool

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

Books

• British Regional Geology
• The Midland Valley of Scotland, 3rd Edition, 1985
• One-inch Sheet Memoirs
• Geology of the Stirling District, 1970
• The Geology of East Fife, 1977

Geological maps

• 1:625 000
• Geological (North) Quaternary (North) Aeromagnetic (Sheet 1)
• One-inch to one mile (1:63 360) and 1:50 000 (Scotland)
• Sheet 39W (Stirling): Solid 1982; Drift 1982
• Sheet 39E (Alloa): Solid 1982; Drift 1982
• Sheet 40 (Kinross): Solid 1971; Drift 1973
• Sheet 41 (North Berwick): Solid 1970; Drift 1971
• Sheet 48W (Perth): Solid 1983; Drift 1985
• Sheet 48E (Cupar): Solid 1982; Drift 1982
• Sheet 49 (Arbroath): Solid 1980; Drift 1981

Preface

The district described in this memoir is covered by Sheets 48 and 49 of the 'one-inch' Geological Map of Scotland. These sheets were originally surveyed by J. Geikie, H. H. Howell, B. N. Peach, H. M. Skae and J. Young. Sheet 48 was published in 1883 and Sheet 49 in 1884.

A resurvey of Sheet 49 commenced in 1961 and was completed in 1964. The surveyors were Dr M. Armstrong, Messrs J. I. Chisholm, I. H. Forsyth and I. B. Paterson under Dr J. R. Earp and Mr T. R. M. Lawrie as District Geologists. The resurvey of Sheet 48 followed immediately and was completed by 1974. The surveyors were Dr M. Armstrong, Messrs M. A. E. Browne, J. I. Chisholm, Dr A. L. Harris and Mr I. B. Paterson under Mr T. R. M. Lawrie and Mr E. G. Poole as District Geologists.

A series of 20 Geological Survey boreholes was drilled in 1971–72 in the Tay–Earn area to examine the succession of the Lower Carboniferous rocks in the district and, as a secondary aim, to investigate the stratigraphy of the Quaternary deposits. The results of this work are already published. The cores from boreholes sunk to investigate foundation conditions for bridges, roads and buildings in the Perth and Dundee areas and for water supply in the district were examined by Dr M. Armstrong, Messrs M. A. E. Browne, J. I. Chisholm and I. B. Paterson. Most of the fossils from field localities were collected by Mr P. J. Brand and from boreholes by Mr D. K. Graham. During the course of the survey, staff of the Geophysical Department of IGS gave valuable assistance in the investigations of intrusive bodies in the Dundee area. Miss K. E. Babington, Messrs J. M. Dean, D. N. Halley and K. I. G. Lawrie assisted in the field in magnetometer and levelling surveys, etc.

Sheet 49 (Arbroath) was published in Solid and Drift editions in 1980 and 1981 respectively. Sheet 48 was divided into Sheet 48W (Perth) and 48E (Cupar). Sheet 48E Drift and Solid editions were published in 1982 and 1983 respectively. Sheet 48W Solid was published in 1983 and the Drift edition is in press.

Neither of the original one-inch sheets was accompanied by a sheet explanation but some southern parts of the district were described by A. Geikie in his memoirs on Central and Western Fife and Kinross, published in 1900 and on Eastern Fife published in 1902. The Carboniferous rocks on Sheet 49 and the Quaternary deposits in the St Andrews area have been described recently in the memoir on the Geology of east Fife by Messrs I. H. Forsyth and J. I. Chisholm, published in 1977. Several papers were published on various topics concerning the district and adjacent areas both during and after the resurvey and these are referred to in the text. The most notable of these are the study of the Lower Old Red Sandstone of Strathmore by Dr M. Armstrong and Mr I. B. Paterson, published in IGS Report No. 70/12, and Report 81/7 on the Quaternary estuarine deposits of the Tay–Earn area by Mr I. B. Paterson, Dr M. Armstrong and Mr M. A. E. Browne.

The present memoir was written mainly by Dr M. Armstrong, Messrs I. B. Paterson and M. A. E. Browne. Dr A. L. Harris (University of Liverpool) has contributed the chapter on the Dalradian rocks, Dr D. Stephenson wrote the chapter on late-Carboniferous dykes and Mr J. I. Chisholm contributed details for the descriptions of the Devonian and Quaternary rocks. Mr P. J. Brand compiled the appendix on the Devonian fossils, the arthropods being identified by Dr C. D. Waterston of the Royal Scottish Museum and the fish by Dr R. S. Miles of the British Museum (Natural History). Dr R. Neves (University of Sheffield) identified the Carboniferous miospores and the Carboniferous macrofossils were determined by Dr R. B. Wilson. The Quaternary macrofaunas were identified by Mr D. K. Graham. The microfaunal determinations were done mainly by Miss D. M. Gregory and Dr. J. E. Robinson (University College, London). The photographs were mainly taken by Mr A. Christie, Mr W. D. Fisher, Dr A. L. Harris and Mr F. I. MacTaggart. The memoir was edited by Dr R. B. Wilson.

G. M. Brown Director, British Geological Survey Keyworth, Nottingham NG12 5GG. 5 June 1985

Six-inch maps

The six-inch to one mile field maps covering, wholly or in part, the 1:50 000 sheets 48W, 48E and 49 are listed below, with the names of the surveyors (M. Armstrong, M. A. E. Browne, J. I. Chisholm, I. H. Forsyth, A. L. Harris and I. B. Paterson) and the date of survey.

The maps are not published but most of them are available in the British Geological Survey office, Murchison House, Edinburgh where photocopies can be purchased.

Notes

• In this memoir the word 'district' means the area of land included on the Scottish 1:50 000 geological sheets 48W, 48E and 49.
• Numbers in square brackets are National Grid references within 100 km square NO.
• Numbers preceded by the letters ED and S refer to the Sliced Rock Collection of the British Geological Survey.
• Numbers preceded by the letters D, MNS, PMS and TS refer to the British Geological Survey Photograph Collection. A full list of geological photographs taken in the district is available on application to British Geological Survey, Murchison House, Edinburgh, EH9 3LA.

Geology of the Perth and Dundee district—summary

The district described in this memoir includes the ground to the north and south of the Firth of Tay, the valleys occupied by the lower reaches of the rivers Tay and Earn, and parts of the Sidlaw and Ochil hills. This is the first comprehensive account of the district, and covers the area included in sheets 48W, 48E and 49 of the 1:50 000 geological map of Scotland.

The introductory chapter outlines the physical features of the district. The Dalradian rocks in the north-west corner of the district are described, with particular reference to their structure, and there then follow detailed descriptions of the Lower and Upper Devonian rocks which underlie most of the district. The thick developments of Lower Devonian volcanic rocks are dealt with separately. The small areas of poorly exposed Lower Carboniferous sediments are then described, mainly from borehole evidence.

Successive chapters then deal with the late-Carboniferous dykes and the structure of the district. The Quaternary deposits are treated in detail as the sediments of this age blanket much of the district, and the history of their deposition in late-Glacial and post-Glacial times is highly complicated and has been the subject of much research. The final chapter is on economic geology, and details are given of the various natural resources of the district.

Geological sequence

The geological formations occurring within the district are summarised below. For a description of the strata within the divisions marked * in that part of Sheet 49, south-east of the Dura Den Fault, the reader is referred to Forsyth and Chisholm (1977).

SOLID FORMATIONS

Chapter 1 Introduction

Area and physical features

The district described in the following account contains those parts of the Tayside and Fife regions which lie within the confines of three sheets on the 1:50 000 scale of the Geological Survey map of Scotland. The sheets in question are 48W (Perth), 48E (Cupar) and 49 (Arbroath). The district is bounded to the north by a line extending westwards from the North Sea coast immediately north of Arbroath to the Highland Border at Dunkeld. The western boundary runs southwards to the Forteviot area of lower Strathearn and the southern boundary passes eastwards to the neighbourhood of Cupar and Stratheden. The area south–east of the River Eden, where the Carboniferous strata are continuous with those in the area of Sheet 41 (one-inch-to-one-mile scale), has already been described in the Geological Survey memoir relating to east Fife (Forsyth and Chisholm, 1977). The main features of the geology and topography of the district are shown in (Figure 1).

The small tract of Highland ground occupied by Dalradian rocks in the north–west corner of the district is deeply trenched by the River Tay. Below Dunkeld the river enters and traverses an extensive area of lower ground on Lower Devonian sedimentary rocks at the western end of Strathmore, flowing at first eastwards towards the confluence with the River Isla near Meikleour, and thereafter southwards towards Perth, receiving from the west the Shochie Burn and the River Almond. The NE-trending depression of Strathmore is succeeded on its south-east side by the Sidlaw Hills which extend to the north-east from Perth. This high ground is composed of both volcanic and sedimentary rocks of Lower Devonian age and culminates north of Dundee in the principal summits of King's Seat (377 m), Auchterhouse Hill (426 m) and Craigowl Hill (455 m). To the north-east of Dundee the southern and eastern slopes of the Sidlaw Hills give way eastwards to gently falling ground underlain mainly by Lower Devonian sedimentary rocks but broken by isolated hills and ridges related to extrusive and intrusive igneous rocks.

South-west of the narrow valley traversed by the River Tay at Perth the Sidlaw Hills are continued by volcanic rocks forming the ridge of Moncreiffe Hill which to the south overlooks the broader valley of lower Strathearn. The confluence of the eastward-flowing River Earn with the River Tay, north of Abernethy, occurs at the head of the Firth of Tay. The low ground of Strathearn is continued north of the Firth of Tay by the extensive lowland known as the Carse of Gowrie. Both areas are underlain by downfaulted Upper Devonian and Carboniferous rocks which are largely concealed by thick Quaternary deposits. North of the Carse of Gowrie rises the steep southern fault-face of the Sidlaw Hills, known as the Braes of the Carse. South of Strathearn and the Firth of Tay the ground rises abruptly across a north-facing fault scarp which delimits the Lower Devonian volcanic rocks of the Ochil Hills. The most prominent summit in this area is Norman's Law (285 m). To the south and east the Ochil Hills are succeeded by the low ground of the Howe of Fife, Stratheden, and the coastal area of Tentsmuir.

Almost the whole of the district, with the principal exception of Stratheden, drains into the Firth of Tay. The River Tay itself has the largest discharge in Britain. It is tidal as far upstream as the mouth of the confluent River Almond. The firth has long provided access by sea to the Perth area as evidenced by the recently excavated Roman port at Carpow. By land, however, the River Tay and its wide firth impeded communications and the sites of bridges became places of importance. Until relatively recently the lowest bridge across the Tay was at Perth where successive structures were built after recurrent destructive floods. The building of the famous Victorian railway bridge across the Firth near Dundee changed this situation, and more recently the choice of routes across the Tay has been further widened by the construction of new road bridges at Dundee and at Friarton, east of Perth.

The district is mainly given over to agriculture, with livestock rearing predominating on the higher ground and mainly arable farming elsewhere. Soft fruit, chiefly in the Carse of Gowrie, and seed potatoes are specialities of the district. Commercial and sporting salmon fishing on the rivers Tay and Earn are of considerable importance.

The principal centres of population are Dundee, Perth, Cupar, Arbroath, Carnoustie, Coupar Angus, Newburgh and Dunkeld. Dundee, the third city of Scotland in terms of population, and the administrative centre of the Tayside Region, is a seaport and manufacturing centre. Perth, an ancient capital of Scotland, is notable for its livestock sales and distilling industry. Arbroath is a fishing port and manufacturing centre while Carnoustie is perhaps most notable for its association with golf. Cupar, the former county town of Fife, with a similar function in relation to Fife Region, is the centre for a large agricultural area.

Outline of geological history

The oldest rocks in the district belong to the Dalradian Supergroup and crop out to the north-west of the Highland Boundary Fault-zone. They were deposited as sandstones and mudstones which were later metamorphosed to cleaved grits and slates respectively. They are thought to have been deposited as proximal turbidites. They are probably of Cambrian age and were folded and metamorphosed during the Caledonian Orogeny. Before the start of Lower Devonian times, the Dalradian rocks were elevated into a mountainous area higher and more extensive than the present Highlands. Contemporaneous erosion of the mountains was extremely active, especially as no significant cover of vegetation had yet evolved, and large rivers carried great volumes of detritus into a subsiding area of low ground in the area of the present Midland Valley. Coarse detritus was deposited in large conglomerate fans, mainly at the foot of the mountains but also at times extending far into the basin. Sandstones were laid down on extensive alluvial plains by braided rivers flowing generally to the south-west. This dominantly arenaceous and commonly red, non-marine succession, long known as the Lower Old Red Sandstone, is about 9000 m thick to the north-west of the present district. There, near Stonehaven, the mainly Devonian sequence passes down into basal beds of Silurian age.

Silurian rocks are not known to occur in the district where the oldest exposed post-orogenic strata rest unconformably on the Dalradian rocks in the Highland Boundary Fault zone, and represent a Lower Devonian horizon high in the sequence in the Stonehaven area. Considerable overlap against the contemporaneous Highlands is therefore apparent.

Volcanism of andesitic character was periodically active and thick developments of lava are intercalated within the sedimentary fill of the basin. It is probable that a period of reduced sedimentation is marked by a thin but widespread zone containing concretionary limestone (calcrete) in the upper part of the sequence.

The Lower Devonian rocks of the Midland Valley have been regarded as the fill of the post-orogenic molasse-basin which developed on the site of a fore-arc basin that lay between the continental massif of the Scottish Highlands and the subduction complex generated in the area now occupied by the Southern Uplands during closure of the Iapetus Ocean (Leeder, 1982). The Devonian volcanism may have involved melting of mantle material above a slab of oceanic lithosphere within a north-westerly-dipping subduction zone possibly with an admixture of subducted oceanic sediments (Thirlwall, 1981; 1983).

Sedimentation was terminated by the onset of earth-movements after the end of the Lower Devonian. Folds of NE-trend, the Strathmore Syncline and the Sidlaw Anticline, developed, and along the parallel Highland Boundary Fault-zone overthrusting towards the south-east took place on steep dislocations. The newly uplifted Lower Devonian rocks became an area of upland undergoing vigorous erosion and consequently no Middle Devonian strata exist in the district. Eventually, during the Upper Devonian, sediment began to accumulate, and Upper Devonian strata rest with marked unconformity on the denuded and irregular land surface cut into the folded Lower Devonian rocks.

The Upper Devonian strata are mainly arenaceous and represent fluvial sediments laid down by rivers draining generally eastwards towards the contemporaneous sea somewhere to the east of the present outcrops. A development of concretionary limestones (calcrete) in fluvial sandstones at a higher stratigraphical level (the Kinnesswood Formation) is considered to mark a general reduction in the rate of sedimentation, probably under the influence of climatic change at or near the Devonian–Carboniferous boundary. In early Dinantian times the succeeding Ballagan Formation, consisting of mudstone with thin beds of dolomitic limestone, was laid down conformably upon the Kinnesswood Formation. It probably represents a widespread transgression, with deposition in shallow water on a coastal shelf area characterised by a very restricted marine fauna of ostracods and quasi-marine bivalves (Modiolus). Gypsum and halite are associated with the sediments. Fully marine conditions were distant at this period. Later in the Dinantian, younger sediments of the Calciferous Sandstone Measures, as developed in east Fife, were laid down in a cyclic deltaic sequence.

Coal-forming material accumulated periodically, and restricted marine incursions took place. During the deposition of the Lower Limestone Group marine influences became much stronger and some marine limestone with richer faunas accumulated. In early Namurian times, thick coals were deposited and marine incursions were rare, the faunas being confined mainly to Lingula.

The Upper Devonian and Carboniferous rocks of the district were affected by earth-movements, initiated perhaps as early as mid-Dinantian times. The most striking result was the subsidence of the Tay Graben between the large NE-trending North Tay and South Tay faults. This structure is somewhat oblique to the axis of the earlier Middle Devonian Sidlaw Anticline, the crest of which is now largely concealed by unconformable, downfaulted Upper Devonian and Carboniferous rocks underlying the Carse of Gowrie. Other major faults, on an easterly trend, occur both within and outside the graben, as for example at Errol and Newburgh and in the Flisk–Wormit area. Late-Carboniferous quartzdolerite dykes of similar trend are probably later than the main fault movements.

The geological history of the district in Mesozoic and Tertiary times is unknown but the principal present-day physical features may have existed by the end of the Tertiary. In common with the rest of Scotland, the district was glaciated on a number of occasions during the Quaternary and the landscape was modified in the process. Glacial deposits appear to relate exclusively to the last (Devensian) glaciation. Till was laid down extensively. During the retreat of the ice, meltwaters deposited spreads of sand and gravel, mainly near the ice-margins. Marine deposits laid down during the glacial retreat now occur well above present sea level as a consequence of the recovery of the land from glacio-isostatic depression imposed by the weight of the ice-sheet. A series of raised shorelines mark stages in this recovery. The sea fell below its present level at about the period of the readvance of glaciers in the west of Scotland between 11 000 and 10 000 years BP. Subsequently the sea rose, but a later fall is recorded by a peat layer formed approximately 8000 years ago. A later marine transgression culminated about 6000 years ago, the deposits of this episode forming the widespread carselands of lower Strathearn and the Carse of Gowrie. Subsequently the sea gradually withdrew to its present position.

Chapter 2 Dalradian

Introduction

Regionally metamorphosed psammitic and pelitic sediments belonging to the Dalradian Supergroup (Harris and Pitcher in Harris and others, 1975) occupy an area of about 16 km^2, lying to the north-west of the Highland Boundary Fault-zone, in the neighbourhood of Dunkeld. The metasediments, which are within the greenschist facies of metamorphism, are in places seen to be overlain by volcanic and sedimentary rocks of Lower Devonian age.

Previous work in the area was carried out by Anderson (1947), Stringer (1957) and Shackleton (1958). The results of the present survey, which was mainly undertaken in the period 1967 to 1969, have been reported or referred to in several publications (Harris, 1972; Harris and Fettes, 1972; Harris and others, 1976; Harris and others in Bowes and Leake, 1978).

Stratigraphy and sedimentology

All the metamorphic rocks of the area are placed within the Southern Highland (Upper Dalradian) Group (Harris and Pitcher in Harris and others, 1975). Although direct fossil or radiometric evidence of their age is lacking in the Dunkeld area, they are almost certainly of Cambrian (probably Lower Cambrian) age. Three locally significant lithostratigraphical formations have been distinguished, the Birnam Grits, the Birnam Slates and the Dunkeld Grits, the Birnam Slates being subdivided on the basis of their colour (Figure 2). It has been suggested (Shackleton, 1958) that the Dunkeld and Birnam grits are of the same general age and lie on opposite limbs of a major fold, the core of which is occupied by the Birnam Slates. It is apparent, however, from abundant structural and sedimentary evidence (Harris, 1972; Harris and Fettes, 1972) that the Dalradian formations in the area form part of a generally right-way-up sequence which youngs towards the north-west, the Birnam Grits being the oldest. Local reversal of the stratigraphy occurs on the inverted limb of intermediate and minor-scale folds.

Birnam Grits

Fine- to coarse-grained sandstones, commonly occurring as graded beds believed to have been laid down as proximal turbidite deposits, make up the bulk of the Birnam Grits. Many of the sandstone beds, which are from 0.3 to 1.5 m thick, are overlain by a layer of siltstone a few centimetres thick. This layer is commonly absent, however, and the arenaceous basal parts of up to 12 depositional units may coalesce to form composite sandstone bodies. Clasts derived by the penecontemporaneous erosion of the siltstone layers are common in the basal parts of sandstone units, as well as clasts, in some cases more than a centimetre long, of quartz and feldspar. The matrix, consisting of white micas and chlorite with fine-grained quartz and feldspar, is characterised by a tectonically imposed planar structure which is commonly deflected around the larger clastic grains. A feature of many of the sandstone beds is the presence of closely spaced, ramifying silty laminae set at a high angle to the bedding. In some cases, the base of the sandstone units is marked by shallow load-cast sructures, especially where the underlying siltstone is thicker than average, as on Birnam Hill [NO 038 405].

Within the outcrop of the Birnam Grits on the Obney Hills and on Newtyle Hill, there are several developments, up to 50 m thick, of purple silty slate, most of which were considered by Shackleton (1958) to lie within the cores of folds. The present study confirms that the slates worked at Newtyle Quarry [NO 046 413] lie within the core of an antiform, and probably belong to the Birnam Slates, and that the purple slates which crop out on the southern slopes of the Obney Hills and in the Garry Burn (Figure 2) occupy the core of a synform. However, all other purple slate units within the outcrop of the Birnam Grits are in genuine stratigraphical sequence. The slates are commonly closely laminated and contain lenticular beds of ripple-laminated, fine- to medium-grained sandstone, nowhere more than a few centimetres thick. These structures are well displayed at localities [NO 021 374] in the Obney Hills and in Newtyle Quarry.

The base of the Birnam Grits is not seen, being concealed beneath Lower Devonian rocks. Stratigraphical passage into the succeeding Birnam Slates is seen only in Newtyle Quarry, the junction being more usually a fault.

Birnam Slates

A lower (purple) member and an upper (green) member of the Birnam Slates have been recognised, the contact between them being fairly sharp. The Lower Birnam Slates, which owe their colour to their content of finely disseminated hematite, consist mainly of homogeneous well-cleaved siltstones, with rare laminae and thin lenticular beds of fine-grained, in some cases ripple-laminated, sandstone. The slates were much sought after and have been worked in numerous quarries along their outcrop. Their post-deformation thickness is about 250 m.

The Upper Birnam Slates are somewhat coarser grained and less well-cleaved than the Lower Birnam Slates and contain abundant chlorite. There are numerous thin beds of sandstone which are commonly graded but in some cases are ripple-laminated. The transition to the overlying Dunkeld Grits is gradational, with an increase in the proportion of sandstone. The junction, arbitrarily taken at the base of the lowest sandstone bed encountered in a north-westwards traverse, may be diachronous. The post-deformation thickness of the Upper Birnam Slates exceeds 700 m.

Although the rocks assigned to the Dunkeld Grits within the area show increased complexity and deformation towards the north-west, accompanied by a rise of metamorphic grade, their original sedimentary characters are still clearly visible. As in the Birnam Grits, fine- to coarse-grained sandstones predominate, usually in the form of graded units 0.5 to 1.0 m thick which pass up into siltstone. The proportion of sandstone to siltstone in the Dunkeld Grits varies considerably. At some localities, green siltstone predominates; at others, the sandstone components of several depositional units coalesce to form thick composite sandstone bodies. As is well seen in exposures [NO 052 422] near the top of Newtyle Hill, these are commonly of lenticular form and apparently were deposited in asymmetric channels, several metres across.

Structure

It has been argued (Harris, 1972) that all the Dalradian metasediments in the Dunkeld area lie within the NW limb of a synform — the downbent hinge of a regional foldstructure — the Tay Nappe. The SE limb and hinge of the synform is presumed to lie to the south of the Highland Boundary Fault-zone, where it is concealed beneath Lower Devonian rocks. The increase of tectonic complexity and metamorphic grade north-westwards in the Dalradian metasediments is a reflection of increasing depth within the nappe.

Cleavage

In both the Birnam and Dunkeld grits the coarse-grained sandstones commonly carry two cleavages, the relative ages of which are uncertain. One of the cleavages (Slp of Harris and others, 1976) ((Plate 2) and (Figure 3)) consists of spaced partings marked by thin ( <1 mm) micaceous laminae which transect the sandstone beds at intervals of approximately 1 cm and which in detail ramify and coalesce. The orientation of this cleavage varies about the axial plane of intermediate- and large-scale folds of bedding. On the shallowly inclined, short, inverted limbs of such folds the cleavage is much steeper than the bedding, commonly being nearly vertical; on the long, steeply inclined, right-way-up limbs the cleavage is much shallower than the bedding, with inclinations of only a few degrees. The cleavage fans about the hinge of the folds, converging downwards across antiform hinges. Microscopic examination (eg. (S53660)) shows that in the micaceous laminae which define the cleavage, elastic grains are almost absent and that both white mica and chlorite are present.

The micaceous Slp partings commonly truncate elastic grains of quartz or quartz aggregates, thereby inducing a new lithological lamination in the sandstone beds, which is independent of the original bedding planes and which probably originated by pressure solution during the early stages of deformation, the solubility of silica in hydrous fluids being greatly enhanced at the elevated temperatures and confining pressures characteristic of the lower greenschist facies of metamorphism. The micaceous laminae which formed on the site of these zones are considered largely to represent the recrystallised insoluble residue formed from the mica-rich matrix of the sandstones. The volume of this residue and the extent of the apparent displacements which have occurred across the laminae suggest that as much as 20 per cent by volume of the sandstones has been removed in this way.

The other cleavage (S1) in the sandstones occurs in the lithons, the margins of which are defined by the S1p laminae. This is a thoroughly penetrative cleavage (Plate 2) and (Figure 3) in that all the elements which make up the sandstones are affected. The elastic quartz grains have a moderately well defined preferred dimensional orientation and the crystal cleavage of mica flakes, which largely constitute the matrix, is similarly orientated. The resulting fabric is approximately parallel to the axial plane of minor- and intermediate-scale folds, and is analogous with and parallel to the slaty cleavage which is a feature of the fine-grained tops of the graded sandstone units and of the occasional silty slate bands which interrupt the sandstone succession. Detailed measurements of the Sl, Slp and bedding relationships in several large- and intermediate-scale folds, on Birnam Hill [NO 039 402] and on Obney Hill [NO 023 374] indicates the configuration summarised diagrammatically in (Figure 3).

Slaty cleavage is the common tectonic planar structure in the fine-grained sediments of the Lower and Upper Birnam slates and in the slaty bands within the sandstone formations. Frequently, the slaty cleavage planes, which are approximately parallel to the axial plane of tectonic folds, are marked by a lineation of which two types are distinguished. One is formed by the trace of bedding lamination on the cleavage planes, a trace which is parallel to the hinge of local minor folds. The other lineation consists of a very fine ribbing, parallel to the true dip of the cleavage planes, which is emphasised by small elongated bodies of altered pyrite. This lineation approximates to the direction of finite extension suffered by the slates during deformation. The slaty cleavage planes dip west-north-west at variable but moderate-to-steep angles, and hence the extension lineation plunges variably in this direction.

Refraction of the cleavage planes is a common feature of graded beds (Plate 2).

Folds

Fold style in the area to a great extent reflects the competency of the rocks undergoing deformation. With very few exceptions, the only surfaces folded are those of bedding, the tectonic cleavages imposed during the folding being subsequently unmodified, and it is likely therefore that only one episode of penetrative deformation affected these rocks. In the silty slates, folds on a small scale (a centimetre to a few metres) (Plate 2) are tight, with interlimb angles as low as 20°. The rocks in the fold limbs are commonly attenuated and there is thickening into the hinges. In the sandstones, the folds tend to be more open, interlimb angles of 40°–50° being normal, and on a scale of several metres, in places several tens of metres. Although less noticeable than in the slates, there is commonly some attenuation of fold limbs in the sandstones. The axial planes of folds usually dip at moderate angles to the west-north-west.

The plunge of minor- and intermediate-scale folds varies systematically across the area, a culmination of plunge occurring approximately in the valley of the Tay where fold hinges are sub-horizontal. Plunge is generally northeastwards on the north-eastern side of the valley, and southwestwards on its south-western side. In some cases variation of plunge on a smaller scale must occur and at several localities eg. [NO 056 418] the plunge of minor folds or of the cleavage-bedding lineation has a reclined attitude, fold plunge approaching parallelism with the true dip direction of the axial plane. Such variation in the plunge of the folds is extremely unlikely to be the result of renewed folding because the axial plane of such curving folds varies very little despite the change of plunge.

Invariably where a pair of folds (synform-antiform) is recorded, the sense of overturning or vergence deduced from its geometry indicates that a major synform, on which the fold pair is parasitic, lies to the south-east. Because the axial trace of this major synform is not encountered within the present Dalradian outcrop, it is inferred to lie below the Upper Palaeozoic cover to the south-east of the Highland Boundary Fault-zone.

The folds of the area and their associated cleavages face downwards (Shackleton, 1958; Harris and Fettes, 1972). Thus, whereas the formations throughout the area are steeply inclined to the north-west and are right-way-up, the fold structures throughout are downward-facing. As a result, their axial planes, although dipping north-west, do so at a shallower angle than the regional dip of the formation.

Two major folds of note were traced in the field: (a) on the south-west slopes of Obney Hill; (b) from the south-east slopes of Birnam Hill into Newtyle Quarry [NO 046 413].

(a) In the core of a synform, purple silty slates contain thin discontinuous siliceous laminae and carry lenticular ripple cross-laminated fine siliceous sandstones. The fold axis trends approximately ENE and plunges at a shallow angle (5°–10°) variably towards the west and west-south-west. By means of abundant but small exposures, laminated and cross-laminated siliceous beds can be traced from steep to shallow angles around the downward-closing hinge of the fold where the bedding can be seen to be inverted, and at a high angle to cleavage which dips at 60°–70° north-north-west. The sandstones in the envelope of the synform young away from the core, thus confirming the downward-facing nature of the structure. Between the crop of this synform and the main crop of the Birnam Slates to the north-west, there intervenes an antiform and synform, the last being broken by a reversed fault in the immediate vicinity of the slate-grit junction.

(b) From the vicinity of [NO 035 395] the trace of a shallowly plunging antiform can be drawn into Newtyle Quarry [NO 046 413], where it is completely exposed. Here the core of the antiform, which plunges northwards at about 30°, contains purple slate, probably Lower Birnam Slates. Because the plunge of the fold axis is reversed at the Tay valley culmination, the slate is missing from the core of the antiform at exposures [NO 040 404] on the lower slopes of Birnam Hill.

Faults

Minor faults showing features characteristic of brittle deformation, including crushed and brecciated rocks, clay gouge, slickensides and quartz fibres are commonly seen in the old quarries. The major faults represented on the map are nowhere exposed, although their course is marked in many places by topographic features. The presence of many of these faults is inferred, not only from this topographic evidence, but also from the distribution of the lithostratigraphic units and the pattern of 'younging' recorded. This is particularly true of the faults which pass through the Obney Hills, Birnam Hill and Newtyle Hill (Figure 2). These are probably reversed faults dipping to the north-west at moderate angles (Harris, 1972).

Structural history

From regional considerations it is known that the Dalradian rocks were deformed during the early Caledonian (Grampian) orogeny (c.520 to 490 Ma). This produced folds and penetrative and spaced cleavages which, because of subsequent downbending along the Highland Border, now face steeply downwards to the north-west. Some of the faulting, especially the moderately inclined reversed faults along the Birnam Grit–Birnam Slate junction, may have been generally contemporaneous with the first episode of folding. Much of the deformation in the area must have occurred before the deposition of the Devonian rocks because the latter rest unconformably on the downward-facing structures in the Dalradian.

Metamorphism

Metamorphic grade is low throughout the area, although there is perceptible increase in the grain size of pelitic rocks (slates to phyllites) in a north-westerly direction. Large chlorite flakes occur widely in the Upper Birnam Slates and in the grit formations, but clearly predate some of the tectonism, being wrapped by cleavage traces and deformed. Whether these chlorite flakes formed as porphyroblasts or are large clastic grains of sedimentary origin is uncertain. In the north-west part of the area small garnets, of uncertain composition and now partially converted to chlorite, occur in thin pelitic laminae and further testify to rising metamorphic grade in that direction.

Chapter 3 Lower Devonian

Introduction

In the district the rocks of Lower Devonian age (Figure 4) are entirely of terrestrial origin, consisting mainly of fluvial sandstones and conglomerates with subordinate, although substantial, developments of lavas and associated volcaniclastic sediments. Long referred to as Lower Old Red Sandstone, they form part of a large outcrop which extends along the northern side of the Midland Valley of Scotland from the Firth of Clyde near Helensburgh to the Kincardineshire coast. As a result of earth movements of Middle Devonian age, the rocks are disposed in major NE-trending folds, the Sidlaw Anticline and, farther to the north, the Strathmore Syncline, which are aligned parallel to the Highland Boundary Fault-zone. Fractures within this zone commonly form the northern margin of the Lower Devonian outcrop but in the present district the oldest Devonian rocks extend to the north of the fault-zone and near Dunkeld rest unconformably on Dalradian metasediments. To the south, the outcrop of the Lower Devonian is terminated by the unconformable base of the Upper Devonian strata.

The previous literature on the Lower Devonian rocks in the general area was reviewed by Armstrong and Paterson (1970) who reconciled earlier classifications of the strata in Angus and Kincardineshire by Hickling (1908) and Campbell (1912; 1913) respectively. Six major lithostratigraphical divisions of the Lower Old Red Sandstone, recognised by Campbell, were modified and applied on a regional basis. These are, in ascending sequence, the Stonehaven, Dunnottar, Crawton, Arbuthnott, Garvock and Strathmore groups. Following the decision by the Committee on Stratigraphy of the International Union of Geological Sciences to place the Silurian–Devonian boundary at the base of the Monograptus uniformis Zone, at least the lower part of the Stonehaven Group must be regarded as Silurian on the basis of the faunas found at Cowie Harbour outside the present district. No zonally diagnostic fossils have been obtained from the strata lying between this horizon and beds low in the Arbuthnott Group which have yielded assemblages of fish, arthropods and spores considered to be of Gedinnian age (Westoll, 1951; Westoll in House and others, 1977; Richardson, 1984). It is suggested that the Devonian base be taken at the incoming of coarse conglomerate into the sequence, that is, at the base of the Dunnottar Group.

Numerous radiometric age determinations have been carried out on rocks of volcanic origin in the Crawton and Arbuthnott groups. The most reliable (Thirlwall, 1981; 1983) suggest that the base of the Devonian dates from at least 408 Ma and may be as old as about 412 Ma, a figure in good agreement with the time-scale proposed by Mackerrow and others (1980).

Classification

The oldest Devonian rocks to crop out in the district are assigned to the Crawton Group. They occur in a narrow strip in the Highland Border area, north-east of Dunkeld, where they consist mainly of partially welded vitric tuff, here termed the Lintrathen Ignimbrite. This is the most westerly occurrence of an important marker horizon which has been traced in discontinuous outcrops, mainly on the north side of the Highland Boundary Fault-zone, for a distance of 80 km as far as Kincardineshire (Paterson and Harris, 1969).

The rocks of the Arbuthnott Group crop out mainly in the axial zone of the Sidlaw Anticline where they have been subdivided into a number of markedly diachronous sedimentary and volcanic formations (Figure 5). The oldest rocks seen in the area north of Dundee consist of andesitic lavas of the Montrose Volcanic Formation. These are overlain by the Dundee Formation which is composed mainly of grey and drab-coloured, cross-bedded, fluviatile sandstone but contains important intercalations of fine-grained, flat-laminated sandstones ('flagstones'), siltstones and mudstones, laid down, at least in part, in lacustrine conditions. In the course of quarrying operations for the valuable flagstones in these fine-grained members, 19th century geologists, including Agassiz and Powrie, discovered the famous fish and arthropod faunas.

On both limbs of the Sidlaw Anticline, the Dundee Formation is succeeded in general by the lavas and associated volcaniclastic sediments of the Ochil Volcanic Formation, which form the high ground of the Ochil and Sidlaw hills. The onset of volcanic activity was markedly diachronous, being earliest in the south-west part of the area where, in the Ochil Hills of north Fife, the Arbuthnott Group is composed almost entirely of rocks of volcanic origin. The lavas are predominantly pyroxene-andesite with subordinate olivine-basalt and trachyandesite. Volcaniclastic sediments, mainly conglomerates, are widespread. These beds are generally composed of detritus of basic composition, resembling that of the exposed lavas. In the western Sidlaw Hills, however, there are thick lenticular bodies of conglomerate in which the pebbles and boulders are mainly of acid or intermediate lava types, almost unknown at outcrop. In the eastern Sidlaw Hills and in the Ochil Hills in north Fife, greenish grey mudstones and Baggy sandstones, typical of the Dundee Formation, are intercalated in the volcanic sequence at several horizons.

Rocks of the Arbuthnott Group occur also in the Highland Border area, on the northern limb of the Strathmore Syncline. They are affected by many large faults and the detailed correlation is uncertain. Lavas are developed at several horizons in a sequence which is composed mainly of conglomerates. In most of these the clasts consist predominantly of basic volcanic rocks but some beds contain mainly fragments of metasedimentary rocks. For convenience, the top of the Arbuthnott Group is taken at the top of the conglomerate sequence.

In the axial zone of the Sidlaw Anticline, the strata of the Arbuthnott Group are cut by numerous intrusions which are chemically related to the lavas and are considered to be of the same general age. The lavas and associated volcaniclastic sediments of the Arbuthnott Group and the minor intrusions are described in the next chapter.

Strata assigned to the Garvock Group crop out on both limbs of the Sidlaw Anticline and in a narrow strip in the Highland Border area where they are much affected by faulting. They consist for the most part of cross-bedded sandstones, usually red-brown or purple-grey in colour, but there are also conglomerates in the Carnoustie to Arbroath area and along the Highland border. The fossiliferous argillaceous developments characteristic of the Arbuthnott Group are lacking and the sandstones of the Garvock Group differ from those of the older division in that they commonly contain many clasts of limestone and calcareous mudstone. These were evidently produced by the penecontemporaneous erosion of beds of mudstone in which calcareous concretions had been formed. The base of the Garvock Group is variously taken at the top of the highest argillaceous member of the Dundee Formation or above the highest lava flow or lava-conglomerate of the Ochil Volcanic Formation, the boundary being to some extent diachronous and transitional.

In the coastal area between Monifieth and Arbroath, the Garvock Group comprises in upwards sequence the Red Head Formation, the Auchmithie Conglomerate and the Arbroath Sandstone. The topmost beds of the group are not represented on land on the south-east limb of the Sidlaw Anticline, but north-west of the anticlinal axis the group is present in its entirety. In the Perth area the Garvock Group consists for the most part of sandstone and the bulk of these arenaceous strata is assigned to the Scone Formation. A thin but distinctive formation at the top of the group, the Campsie Formation, includes basal conglomerates and sandstone succeeded by sandstone and mudstone. The basal coarse beds are persistent and contain lenticular bodies of concretionary limestone considered to have originated as a result of soil-forming processes during a period of reduced sedimentation. Widely recognised in the Strathmore region (Armstrong and Paterson, 1970), this development has been variously termed the Stanley Limestone or the Pittendreich Limestone. On the northern limb of the Strathmore Syncline, a large part of the Garvock Group is eliminated at outcrop by the Spittalfield Fault, the most southerly dislocation in the Highland Boundary Fault-zone. The basal beds which remain consist of red-brown, pebbly, cross-bedded sandstones.

The youngest Lower Devonian strata, referred to the Strathmore Group, are preserved only in the axial zone of the Strathmore Syncline. At the base is a thick development of red-brown, silty mudstone, termed the Cromlix Formation, which extends in almost continuous outcrop for a distance of more than 130 km in the Strathmore region (Armstrong and Paterson, 1970). The mudstone is overlain by and partly interdigitates with a thick series of purple-grey, cross-bedded sandstones with beds of red-brown mudstone assigned to the Teith Formation. Elsewhere in the Strathmore region, the strata of the Strathmore Group are known to become coarser grained in a north-west direction, the Cromlix Formation becoming arenaceous or even conglomeratic. In the present district, however, the oldest beds of the Strathmore Group are not seen on the northern side of the Strathmore Syncline, being eliminated at outcrop by the effects of the Spittalfield Fault. On the basis of plants and spores, a late-Siegenian to early or middle Emsian age has been suggested for the Strathmore Group (Westoll in House and others, 1977).

Conditions of deposition

The Lower Devonian rocks accumulated in a major basin aligned NE–SW and centred on the Midland Valley of Scotland. During deposition of the very thick Lower Devonian sequence, movements on fracture-zones at the basin margins probably helped to maintain highland areas on both flanks. Throughout the Midland Valley, but especially along the Highland border and on both limbs of the Strathmore Syncline in Kincardineshire, the sedimentary sequences in marginal areas are predominantly conglomeratic. It has been inferred that coarse detritus from the Highlands was laid down along a mountain front in a series of large overlapping alluvial fans which passed out south-eastwards into finer sediments on lower ground (Waterston in Craig, 1965; Francis and others, 1970), a view supported by studies of palaeocurrent directions (Bluck in Bowes and Leake, 1978). However, in the Strathmore region as a whole, the contemporaneous sandstones of the Arbuthnott and Garvock groups are characterised by a dominant palaeocurrent trend parallel to the basin-axis and directed to the south-west (Armstrong and others in Friend and Williams, 1978) away from the major conglomerate fans in Kincardineshire. Similar palaeocurrent directions have been obtained for Lower Devonian strata in other parts of the Midland Valley and adjacent areas (Bluck in Bowes and Leake, 1978; Morton, 1979). It would seem that much of the material introduced into the Midland Valley basin, at least during the periods of the Arbuthnott and Garvock groups, was redistributed by a large south-westerly-flowing, mainly braided river system, and the drainage pattern strongly suggests that in Lower Devonian times a basin margin or a positive area lay to the east of the present coastline of Angus and Kincardineshire. The evidence provided by the distribution of heavy minerals in the Lower Devonian sediments in the Strathmore region (Braithwaite and Jawad Ali, 1978) is consistent with this view.

During deposition of the Lower Devonian, and especially of the Arbuthnott Group, thick piles of lavas were eruptedmainly andesites of calc-alkaline affinities, many having relatively high concentrations of potash (Gandy, 1975). On the evidence of spatial variation in trace element concentrations and ratios, in certain isotope ratios and in some major elements, the volcanic activity has been related to subduction associated with closure of the Iapetus Ocean.

The lavas were erupted on to the alluvial plains of the Lower Devonian river system, building upstanding volcanic terrains capable, locally, of generating small scale alluvial fans which filled valleys in the underlying lava surface (Francis and others, 1970; Bluck in Bowes and Leake, 1978). Nevertheless, as indicated by palaeocurrent directions obtained from cross-bedded sandstones in the Dundee Formation (Table 1), the south-westerly flow was not in general deflected by the volcanic pile but apparently found its way through it, depositing quartzose sand and silt within cavities in the lavas. From time to time, however, the drainage may have been impeded as a result of the volcanic activity and this may partly explain the common occurrence of the argillaceous lacustrine intercalations in the dominantly arenaceous development of the Arbuthnott Group in the area about Dundee. Basin-subsidence in the areas of maximum volcanic activity appears to have been somewhat greater than in areas where the sequence lacks a significant volcanic component. The lavas are considered to form an integral component of the fill of a single broad basin which occupied the full width of the Midland Valley of Scotland during the Lower Devonian. Bluck (in Bowes and Leake, 1978) envisaged a more significant role for the volcanic pile, considering that it separated two elongate sedimentary basins. Even if his view is correct, however, this situation could have existed only during part of the limited period when the Arbuthnott Group was being deposited.

The cross-bedded sandstones forming the bulk of the Arbuthnott and Garvock groups probably represent the bed-load of a powerful, braided, river system. Mudstone clasts, derived by the destruction of argillaceous overbank sediments, occur commonly in the basal parts of channel-sandstone units. Characteristically, in the Garvock Group sandstones, these are accompanied by abundant pebbles and even boulders of fine-grained limestone which had formed as concretions within the mudstone of the overbank deposits. Some of the concretions retain traces of the mudstone which originally enclosed them. It is possible that during the deposition of the Garvock Group the rainfall became more markedly seasonal and fluctuations of the water-table height caused the formation of CaCO₃ concretions as the floodplain sediment periodically dried out. The concentration of beds and nodules of pedogenic carbonate within the Campsie Formation at the top of the Garvock Group suggests a further development of this tendency and marks an episode associated with or following pronounced degeneration of the river system, possibly at a time of climatic change. Modern calcretes are restricted to climatic zones with a mean annual temperature in the range 16–20°C and a seasonally distributed rainfall of 100 to 500 mm.

A significant change of environmental conditions is indicated by the character of the basal deposits of the Strathmore Group as represented by the ill-sorted, red-brown mudstones of the Cromlix Formation. These sediments, which elsewhere in the Strathmore region pass to the north-west into sandstones and conglomerates, may have been laid down in the most distal parts of alluvial fans emerging from highland valleys. They do not appear to have been subjected to appreciable fluvial reworking, implying that either the south-westerly axial drainage was less active or was displaced towards the south-east. Higher strata in the Strathmore Group consist of grey and purple-grey sandstones and red-brown silty mudstones, disposed in upwards-fining fluvial cycles and it would appear that the axial flow within the basin was restored during this period. The common occurrence of mudstone layers, implying preservation of the overbank component of the cycles, may mean that the rivers were meandering rather than braided, as was characteristic of earlier periods.

Crawton Group

Rocks assigned to the Crawton Group occur only in a narrow outcrop north-east of Dunkeld where they consist almost entirely of the Lintrathen Ignimbrite (p.36). At West Cult [NO 0631 4210] this deposit is seen resting upon 0.75 m of coarse conglomerate with large boulders of cleaved grit which in turn rests unconformably on Dalradian metasediments.

Arbuthnott Group

Sidlaw Anticline

Montrose Volcanic Formation

Lavas assigned to the Montrose Volcanic Formation occur in small inliers along the axis of the Sidlaw Anticline at Tealing House [NO 413 380] and north of Newton [NO 455 419]. A lenticular body of lava, intercalated in the strata of the Dundee Formation in the neighbourhood of Middle Brighty [NO 445 388], is also referred to the formation. The lavas of all three outcrops are basic pyroxene-andesites, occurring in several textural varieties.

Dundee Formation

The Dundee Formation is composed mainly of medium- to coarse-grained cross-bedded sandstone with intercalations, commonly more than 30 m thick, of flaggy sandstone interbedded with siltstone and mudstone (Figure 6). As defined by Armstrong and Paterson (1970), the formation incorporates the Sidlaw, Carmyllie and Cairnconnan divisions of Hickling (1908; in Robson, 1948), and part of his Red Head Series. The base of the formation is taken at the top of the lavas of Tealing House and Newton; its top is placed above the highest of the characteristic argillaceous developments. The Dundee Formation crops out only in the axial zone of the Sidlaw Anticline and is best developed in the area around Dundee where it was extensively quarried for paving stone. Towards the south-west on both limbs of the anticline, the sedimentary sequence gives way by interdigitation to the lavas of the Ochil Volcanic Formation. In the Sidlaw Hills and the Ochil Hills of north Fife, beds of sandstone, siltstone and mudstone can be traced for several kilometres into the main lava-pile before being replaced by sediments of volcanic origin (Figure 5). On the south-east limb of the Sidlaw Anticline, north-east of Dundee, the highest argillaceous member lies some distance above the uppermost lavas of the Ochil Volcanic Formation. The converse is true on the north-west limb of the anticline. Isolated lava-flows within the sedimentary sequence at a number of places are, with the exception of those of Middle Brighty, assigned to the Ochil Volcanic Formation.

The cross-bedded sandstones, which form the bulk of the Dundee Formation, are lithic or sub-lithic arenites containing angular or subangular quartz grains and subordinate feldspar and mica in variable amounts of clay or silt grade matrix. According to Jawad Ali and Braithwaite (1977), pyroxenes and amphiboles are uncommon. The majority of the rock clasts are of intrabasinal origin, consisting either of basic lava or of grey or red-brown mudstone or siltstone derived by erosion of the overbank deposits. Well-rounded pebbles of more distant origin occur sporadically in the sandstones. They include quartzite and acid igneous rocks, some of which may have been derived from lavas and others from intrusions in the Highland source rocks.

The sandstones are usually light grey or light olive-grey in colour but in the area around Guynd [NO 564 418] reddish hues predominate. This coloration is probably of secondary origin and may relate to the prolonged period of erosion, during the Middle Devonian, when the land surface upon which the Upper Devonian rests was formed. It is possible that the Arbuthnott Group strata at Guynd were formerly situated only a short distance below this surface. The general absence of ‘red bed’ conditions during deposition of the Dundee Formation may mean that the water table within the floodplain sediments was normally at a high level. There is evidence that there was a considerable development of vegetation at this time and decaying organic material probably helped to maintain reducing conditions. The sandstones which overlie lavas in the Newton area [NO 454 419] and in the Monikie Burn, near Panmure Gardens [NO 543 377], and in the Elliot Water below Arbirlot [NO 602 406] are also red-brown in colour but this may reflect a high content of weathered volcanic detritus.

The flagstone-mudstone members of the Dundee Formation are exposed in numerous quarries, as well as in a number of stream sections. They are considered to have been laid down during periods when fluvial deposition was interrupted by the formation of extensive shallow lakes.

Sediment introduced into these lakes was laid down in the upwards-coarsening sequences which are characteristic of prograding deltas. In a few examples, the best being in the now-infilled Duntrune Quarry [NO 4375 3529], the delta-deposit rests upon a thin layer made up of alternating laminae of light grey calcareous mudstone and darker grey carbonaceous siltstone (cf. Armstrong and others in Friend and Williams, 1978). This 'rhythmite' clearly accumulated below the wave base in the deepest parts of the lake, farthest from any sediment source. The couplets, which have an average thickness of little more than one millimetre, possibly represent annual increments to the lake sediment.

The basal part of the lacustro-deltaic deposit more usually consists of silty mudstone containing only a few thin developments of closely laminated mudstone. Laterally persistent tabular beds of fine-grained sandstones, ranging in thickness from a few centimetres to about one metre, are commonly intercalated in the silty mudstones, particularly in their upper part. These beds, which usually have a sharp base, are in some cases apparently structureless, in others they show flat- or ripple-lamination. The silty mudstones are considered to have accumulated on the delta-front or in embayments between major distributaries, the thin sandstones representing crevasse-splay deposits. The delta-front deposit is overlain, in some cases gradationally, in others with a sharp or erosional contact, by multistorey sandstone bodies up to 6 m thick. These are thought to have been laid down within the channels of the principal distributaries under moderately high energy conditions. They show a variety of sedimentary structures, including planar bedding, small scale cross-bedding and, especially in their upper part, ripple-lamination. Convolute bedding and 'ball and pillow' structures, on all scales up to one metre in diameter, are especially common. They were probably caused by the fluidisation of the underlying argillaceous sediment (Potter and Pettijohn, 1977) and can occur in any sandstone bed which rests on silty mudstone.

The upward-coarsening delta deposits are commonly associated with fluvial cycles in which, unusually for the Dundee Formation, overbank sediment is preserved. It is possible that these were laid down while meandering rivers replaced the more usual braided system because the existence of a lake farther downstream reduced the gradients on the thalweg.

With few exceptions, the fish and eurypterid fossils, for which the Lower Devonian of the Dundee area is noted, were recovered from the lacustro-deltaic deposits, no doubt partly as a consequence of the selective nature of the quarrying activities. In the laminated mudstones, which were laid down in the deepest water farthest from any sediment source, the fauna consists mainly of acanthodian fish, most likely to have been free swimmers. The amount of carbon present in these sediments suggests that the water conditions at depth may have been somewhat anaerobic at times, thereby excluding bottom-living species, such as the cephalaspids and eurypterids which make up the rest of the fauna. These latter forms are most commonly recovered from the silty mudstones, siltstones and sandstones of the delta-front and the distributary channels. The acanthodians inhabited the higher energy environments also, but their more delicate construction means that their preservation potential is small and in general only the massive dorsal spines have been found. Plant debris occurs in abundance in the delta sediment, being most abundant in the siltstones and finer sandstones. The enigmatic plant Parka decipiens has been found in all lithologies from laminated mudstone to moderately coarse sandstone.

Details

Although the cross-bedded sandstones represent much the greater part of the Dundee Formation, they are rather poorly represented in natural exposures and were quarried less extensively than the flagstones. However, they are typically developed in the area around Balluderon Hill where the attitude of the cross-bedding was determined at a number of localities (Table 1). Cross-bedding measurements were taken also in a quarry [NO 379 423] north-east of Dryburn. Typical cross-bedded sandstone was formerly seen in the extensive Leoch quarries [NO 360 360], in Pyotdykes Quarry [NO 346 343] at Muirhead, and also in other quarries about 2 km farther west. Similar sandstone, associated with mudstones and flagstones, is exposed in a stream section south-east of Fowlis and also in the Blacklaw Burn due south of Berryhill. Farther upstream, in a disused quarry [NO 3055 3316], the sandstone is dull brown in colour and is composed largely of volcanic detritus. Sandstones rich in lava debris occur also in an exposure [NO 3318 3882], north of Bonnyton (S54428) and in a quarry [NO 3380 3979] near Scotston (S54444).

A roadside exposure [NO 315 296] and ploughed-up debris in fields to the west provide evidence of sediments rich in lava detritus in the area south-east of Longforgan. Farther east, Pilmore Quarry [NO 334 294] formerly displayed grey, red-weathering sandstone with lenses of lava detritus.

Massive cross-bedded sandstones were formerly worked in extensive quarries north of the Perth–Dundee railway at Kingoodie. Farther east, between Kingoodie and Invergowrie Station, the railway line traverses another large sandstone quarry. North-east of Invergowrie a smaller quarry shows medium- to coarse-grained, cross-bedded sandstone with a few pebbles. Similar rocks, which yielded Parka, were exposed temporarily during construction work [NO 360 308] north of the main buildings of the Ninewells Hospital. Several sandstone quarries, now filled up, formerly existed in the Menzieshill–Lochee area on the north-west side of Dundee. Sandstones of the Dundee Formation, cut by intrusions of porphyrite and acid porphyrite, are exposed along the railway cutting and in disused quarries between Ninewells and a point [NO 382 297] southeast of Harris Academy. East of the centre of Dundee, sandstones are exposed on the north side of the Dundee–Broughty Ferry railway line 700 to 900 m W of the bridge [NO 433 310] at Stannergate.

Exposure of the Dundee Formation in the area north and northeast of Dundee is generally poor but cross-bedded sandstone was formerly quarried at Lumley Den [NO 402 417], at Coral Den [NO 407 382] and on Dodd Hill [NO 431 397]. Measurements of the attitude of the cross-bedding were made at the last of these localities and also at sections in the Fithie Burn [NO 438 539], near Duntrune House (Table 1). Good sections are also available in the Pitairlie Burn, at Denfind Plantation, and in the Crombie Burn, north-west of Gallow Hill [NO 589 399]. In the Monikie Burn, north-west of Panmure Gardens, brown sandstones with abundant volcanic detritus overlie lavas of the Ochil Volcanic Formation. They are associated with siltstones and silty mudstones, which may represent overbank sediment, and contain beds of conglomerate, up to 0.6 m thick, composed of well rounded lava pebbles. The section is repeated, as a result of faulting, south-east of Panmure Gardens. Similar strata occur downstream from the top of the lavas in the Elliot Water at Arbirlot. Cross-bedded sandstones, which are pale red or purple-grey in colour but are otherwise typical, are exposed in sections in the Elliot Water at Guynd. In the Rottenraw Burn, about 700 m west-southwest of Kellyfield [NO 587 406], a bed 1.3 m thick of hematite-stained, calcareous siltstone (S50097) rests on purple-grey, cross-bedded volcanic-detrital sandstone (S51220), (S51221). It is overlain by some 5 m of conglomerate (S50098), containing well rounded pebbles mainly of altered rhyolite or rhyodacite lava, of a type not known at outcrop in the neighbourhood. The pebbles, which are heavily coated with hematite, are set in a sandstone matrix composed largely of fragments of basic lava of more local origin, with subordinate quartz grains. Eastwards, purple-grey 'ashy' sandstones are exposed in the stream as far as Kellyfield, disposed in an asymmetrical anticline. One of the rare small scale folds recognised in the entire area, this structure may have been caused by drag on an unmapped fault lying to the north of the burn.

The fish-bearing mudstone and flagstone developments which characterise the Dundee Formation are distributed over an area in excess of 1000 km² in the axial region of the Sidlaw Anticline. Where best developed, in the area around Dundee, they occur at some eight or ten different stratigraphical levels within the dominantly arenaceous sequence. Individual beds may well have a considerable persistence but, because of inadequate exposure and structural complication, this cannot be demonstrated. In some cases, however, beds of flaggy sandstone and siltstone can be traced for several kilometres, particularly within the zone where the sediments of the Dundee Formation give way south-westwards by interdigitation to the lavas of the Ochil Volcanic Formation. The most south-westerly of the lacustro-deltaic deposits in the area are at Fingask Castle [NO 228 274] in the Sidlaw Hills and at Clatchard Craig Quarry [NO 243 179] in the Ochil Hills of north Fife.

The argillaceous members of the Dundee Formation were formerly quarried at numerous locations in the area around Dundee. The characteristic features of the flagstone-mudstone developments, such as planar bedding, parting lineation and various water-expulsion structures, are still visible at some localities, and only a selection of the measured sequences is detailed here or illustrated graphically on (Figure 6).

One of the most informative sections, consisting of the lower part of a single upward-coarsening, lacustro-deltaic cycle, was formerly available in the now infilled Duntrune Quarry [NO 4375 3529], north of Dundee.

A similar, probably strictly equivalent sequence (Figure 6) was recently proved in a series of shallow boreholes drilled to investigate a rock fall which took place during excavation of a deep cutting in the course of the re-alignment of the Dundee to Forfar (A94) trunk road at Powrie Brae [NO 419 349]. It is of interest that beds of 'rhythmite', only a few centimetres thick, composed of alterating laminae of pale, greenish grey mudstone and dark grey, carbonaceous, silty mudstone, could be correlated over a distance of several hundred metres. The stratal dip at the site of the road cutting was no more than 12° and the rock slippage apparently was brought about by a combination of small faults, extremely flat bedding planes, commonly filmed with mica, and high hydrostatic pressures following a period of heavy rain.

Other good sections of the lacustro-deltaic deposits of the Dundee Formation are available in the neighbourhood in Balluderon Hill, particularly in Balluderon Quarry [NO 366 392] and Mount Quarry [NO 363 394] (Figure 6), where parts of several upwards-coarsening cycles are developed. At a disused quarry [NO 351 384] at Auchterhouse, some distance to the south-west, only the upper portion of a single deltaic cycle is now exposed, but discarded slabs in the spoil heaps show many excellent examples of 'ball and pillow' structure (Plate 4). Typical lacustro-deltaic sediments are also well exposed in the dens of Rossie [NO 292 311], Balruddery [NO 308 331] to [NO 317 323], and Fowlis [NO 321 344] to [NO 328 326]. ^-Balruddery Den is notable as the locality from which were obtained the fossil remains identified by Agassiz as parts of the giant arthropod Pterygotus anglicus.

In the area close to Dundee, there is a general tendency for the argillaceous developments to be thicker. Many of the disused flagstone quarries have been or are being infilled, but good sections are still available in quarries at Dodd Hill [NO 452 395], Craig Hill [NO 432 358], where the strata are intruded by a cross-cutting body of Acid porphyrite, Tealing Hill [NO 416 398] and near Wellbank [NO 474 377]. There are also natural sections in Murroes Burn [NO 4595 3549] to [NO 4615 3479], south of Hole of Murroes, in the same stream [NO 463 335] to [NO 412 342] north-east of Pitkerro House and on the north side of the Dighty Water [NO 466 328] east of Linlathen. Strata of similar character have been located by drilling between central Dundee and Broughty Ferry. Farther north there are exposures in Pitairlie Burn [NO 4862 3791] to [NO 4933 3720], in the Crombie Burn [NO 5291 4039] to [NO 5344 4006] and in Monikie Burn [NO 5410 3794] to[NO 5530 3740]. At the last of these localities, the mudstone–flagstone bed, the outcrop of which is repeated twice by faulting, lies at a horizon above the lavas of the Ochil Volcanic Formation. Its top is taken as the top of the Dundee Formation in the area. Similar beds, believed to represent the same horizon repeated by faulting, are exposed in the Panlathymill Burn [NO 560 381] and [NO 564 376] north-northwest of Muirdrum and in the Elliot Water [NO 606 404] at Arbirlot.

The most persistent of the lacustro-deltaic deposits extend for considerable distances south-westwards as intercalations within the rocks of the Ochil Volcanic Formation on both flanks of the Sidlaw Anticline. Thus, on the north limb of the fold, a development of flaggy sandstone and shale, of typical Dundee Formation character, has been traced into the neighbourhood of Fingask Castle, being exposed [NO 229 278] in the Craig Burn. Farther north, greenish grey sandstones and siltstones, overlain by a feldsparphyric pyroxene-andesite, are exposed in a disused quarry [NO 2536 3454], near Southballo Hill.

On the south-east limb of the Sidlaw Anticline in north Fife, argillaceous rocks typical of the Dundee Formation are exposed in association with volcanic rocks in a quarry [NO 3867 2574] and on the shore around the Fishing Lodge [NO 3851 2582], near Peacehill Point.

The general sequence consists of at least 15 m of grey and green siltstones, with bands of volcanic detritus, resting on grey, cross-stratified, feldspathic sandstone, probably more than 60 m thick. A 3-m thick band of grey shale among the sandstones just west of the Fishing Lodge contains a fauna (Geikie, 1902, pp. 37, 355 and 358) typical of the Dundee Formation. Green siltstones, possibly of the same general age, are exposed in the cliffs [NO 376 255] and on the shore north of Kilburns, where more than 10 m of siltstone and mudstone, with bands of volcanic detritus, are seen below feldsparphyric lava.

In a nearby quarry [NO 3443 2258], about 5 m of shattered and jointed aphyric lava, with a thin autobrecciated layer at its base, rests on about 5 m of the green siltstone–mudstone lithology. In a borehole [NO 3331 2185] about 1 km SW of the Corbiehill quarries, 50 m of sediments recorded by the driller between layers of 'whinstone' may represent the full thickness of the sedimentary sequence at that point. About 3 km farther to the south-west, in a quarry [NO 3107 2016] not far from Norman's Law, the section is:

The most westerly known occurrence of argillaceous strata of Dundee Formation type lies just to the south of the present area in Pitmedden Forest, where the sequence in a natural section [NO 2100 1394] is:

The proximity of lava outcrops to the north suggests that the section may represent almost the whole of the Corbiehill beds, which must be much thinner than at the borehole.

North and north-east of Corbiehill the outcrop is broken up by faults but rocks, considered to be of the same general age, are exposed at several places along the foot of the hill between Birkhill Home Farm [NO 338 229] and Coarsebrae. A good section is seen in a quarry [NO 3528 2372], not far west of Coarsebrae:

All these strata probably lie above the feldspathic sandstones at Corbiehill. Not far to the west of the section just described, about 6 m of pale grey feldspathic sandstone, with siltstone bands, is exposed in a small quarry [NO 3506 2372], and about 1 km farther to the south-west, in a quarry [NO 3423 2315] by Tenacre Well, about 5 m of the green siltstone–mudstone lithology are seen.

In Clatchard Craig Quarry [NO 2425 1788], on the north side of Ormiston Hill, argillaceous sediments are exposed beneath a flow of hypersthene-andesite considered to correlate with the lava of similar composition at Norman's Law. The full sequence is:

The following sequence is exposed in a long-disused quarry [NO 2495 1855] at Parkhill, which is notable as the type locality of the fossil plant Parka decipiens.

The most westerly known occurrence of argillaceous strata of Dundee Formation type lies just to the south of the present area in Pitmedden Forest, where the sequence in a natural section [NO 2100 1394] is:

Ochil Volcanic Formation

The rocks of the Ochil Volcanic Formation occur on both limbs of the Sidlaw Anticline, where they form the high ground of the Sidlaw and Ochil hills, and also in the Highland Border area (Figure 4). They consist mainly of lavas but include thick and persistent intercalations of volcanic-detrital sediments. Rocks of pyroclastic origin are uncommon.

The volcanic pile reaches its maximum development to the north-west of Cupar, where its estimated thickness is at least 2400 m. In the western Sidlaw Hills, where the lower part is not exposed, the thickness of the Ochil Volcanic Formation does not exceed 1500m. North-eastwards, on both limbs of the Sidlaw Anticline, the lavas give way by interdigitation to the sediments of the Dundee Formation. The youngest flows tend to persist farthest, suggesting a north-eastwards migration through time of the main focus of volcanic activity.

Highland Border area

The Arbuthnott Group sediments in this area consist predominantly of conglomerate, mainly composed of volcanic detritus. In the basal beds, however, as seen in a track-side exposure [NO 067 418], the clasts are well rounded boulders, up to 20 cm in diameter, of quartzite, schistose grit and Lintrathen Ignimbrite. The conglomerates interleaved with the lavas on the north side of the Stralochy Fault can be seen in a small stream [NO 086 426] north-west of Craigend and on crags [NO 097 429] and [NO 103 433] on Craig of Clunie. There the well rounded boulders, up to 60 cm in diameter, include some quartzite and schist but are mostly of basic lava.

South of the Stralochy Fault, in conglomerates exposed in the Garry Burn [NO 026 365] to [NO 034 362] and in the unlined railway tunnel [NO 056 391] at Bynes of Murthly, the clasts are mainly of Highland origin, subangular and generally less than 15 cm long. Conglomerate higher in the sequence in the cutting east of the railway tunnel and in a stream [NO 066 393] north of Kingswood, contains angular, mainly basic, lava fragments. At both localities, the conglomerate is interbedded with coarse, lava-detrital sandstone. There are good exposures of coarse, angular lava-conglomerate on the crags at Caputh Hill [NO 080 406], in a stream [NO 097 419] at Culthill and on the Hill of Gourdie [NO 112 425].

Garvock Group

The base of the Garvock Group is defined at the top either of the youngest lava-flow assigned to the Ochil Volcanic Formation or of the youngest lacustro-deltaic member of the Dundee Formation, whichever is at the highest stratigraphical level, and is diachronous to some extent. The top of the group is taken at the base of the red-brown mudstones of the Cromlix Formation. The constituent formations of the Garvock Group are composed mainly of cross-bedded sandstones, containing clasts of metamorphic and igneous rocks and also concretionary limestone detritus of intra-basinal origin, the last component being especially characteristic of these strata. In the Arbroath area conglomerates are intercalated in the sequence, and in Strathmore the topmost formation contains a persistent conglomeratic zone with lenticular beds of nodular limestone (calcrete). In north Fife, the lower part of the Garvock Group contains several flows of basaltic lava.

South-east limb of Sidlaw Anticline

Red Head Formation

A sequence mainly of red-brown to purple-grey, fine- to coarse-grained sandstone, intervening between the top of the Kelly Den Member of the Dundee Formation and the base of the Auchmithie Conglomerate, is assigned to the Red Head Formation. This division, about 455 m thick, was originally described by Hickling (1908) from the coast section south of Red Head, in an area outwith the limits of the district. Inland, the Red Head Formation is in general poorly exposed, the best section [NO 613 399] being along that part of the valley of the Elliot Water below Arbirlot known as Kelly Den. Here the Kelly Den Member of the Dundee Formation is succeeded downstream by younger strata, mainly cross-bedded fine- to coarse-grained sandstone, which dip southeast at about 20 degrees. Red mudstones and siltstones, some beds containing small calcareous concretions, form an important subordinate component of the lower part of the formation. In higher beds exposed farther to the south-east, the sandstones contain more pebbles, and mudstone and limestone clasts of intraformational origin are more common.

Strata of the Red Head Formation are also exposed in the Panlathymill Burn downstream from the neighbourhood of Panlathy Mill [NO 565 375] and in the Monikie Burn downstream from a point north of Heughhead [NO 552 372]. In these sections the sandstones are generally purple-brown to red-brown, with grey sandstones more in evidence in the older part of the sequence in the Monikie Burn. Limestone clasts and pebbly bands are largely confined to the higher strata.

Auchmithie Conglomerate

A conglomeratic sequence exposed in the cliffs at Auchmithie, north of Arbroath, and named Auchmithie Conglomerate by Hickling (1908), has been traced to the south-west as far as the ground north of Carnoustie. At Auchmithie, where the formation includes three successive developments of conglomerate separated by sandstone, the overall thickness is approximately 245 m. The conglomerate beds contain rounded cobbles of quartzite, reddish jasper and other metasediments.

Within the ground under description there are few exposures of the Auchmithie Conglomerate and it is probable that the formation thins to the south-west, passing laterally into sandstones. On the south-west side of Arbroath, conglomerate with interbedded sandstone dips to the south-east at about 20°, forming low reefs on the foreshore in the neighbourhood of the Football Ground [NO 637 402] and the Swimming Pool [NO 634 402]. These rocks are terminated about 200 m W of the Swimming Pool at a N-trending fault, west of which is sandstone thought to form part of the younger Arbroath Sandstone.

Outcrops of conglomerate, probably representative of the Auchmithie Conglomerate, occur in the valley of the Panlathymill Burn [NO 5150 3650] about 700 m NNW of Craigmill. The conglomerate is at least 2 m thick with included clasts up to 0.15 m across. In exposures along the Monikie Burn, within 500 m of its confluence with Panlathymill Burn, there are conglomerate layers up to 1 m thick in grey and red-brown cross-bedded sandstones.

Arbroath Sandstone

Cross-bedded, fine- to medium-grained sandstones which overlie the Auchmithie Conglomerate constitute the highest known strata of Lower Devonian age on the south-east limb of the Sidlaw Anticline and were named Arbroath Sandstone by Hickling (1908). The formation is known to be at least 365 m thick, and on the coast near Arbroath, where it is overlain with marked unconformity by the Upper Devonian, the sandstones are bright purple-red in colour. Near Carnoustie, however, no more than 3 km to the south-west, sandstones of the division are generally green or purplish grey. The red coloration at Arbroath may be due to deep weathering during the period prior to the deposition of the Upper Devonian and may relate to the proximity of the unconformity.

The Arbroath Sandstone dips to the south-east at about 20° on extensive wave-cut platforms north-east of Carnoustie and at Arbroath. Trough-cross-bedding is well displayed in these sections and inspection of the exposures gives the immediate impression that the elongate troughs lie parallel to the strike of the rocks, and that the dominant palaeocurrent trend is to the south-west. The impression is confirmed by measurement (Table 1).

In the Arbroath and Carnoustie areas the sandstones contain at many levels abundant pebbles and even boulders (up to 0.3 m across) of nodular limestone. These clasts are considered to have originated as calcareous concretions within argillaceous overbank deposits which subsequently were almost completely destroyed as the result of river channel migration. The limestone clasts commonly have the appearance of slightly abraded concretions and may not have been transported far before being incorporated in sandy channel deposits. The limestone of certain clasts appear to enclose mudstone which presumably represents part of the original host rock. Intact mudstone bands are rare in the Arbroath Sandstone but one example containing concretions may be observed at the foot of the cliffs [NO 662 412] 200 m E of Whiting Ness. It has undergone partial penecontemporaneous erosion and has yielded a 'trail' of limestone clasts to an adjacent channel sandstone.

The largest of the limestone clasts tend to occur at the base of the sandstone co-sets and may represent lag-deposits; the smaller clasts tend to lie in the cross-sets. Many individual co-sets are relatively hard and calcareous in their upper parts which project on the abraded wave-cut platforms. Their top surfaces in places display polygonal jointing which does not penetrate below the hard zone. These features, the formation of concretions in mudstones and the hardened upper zones in sandstones, are attributed to pedogenic processes in response to fluctuations in the height of the water-table.

A borehole [NO 5372 3192] at Buddon Ness showed that at least part of the promontory is underlain by sandstones referable probably to the Arbroath Sandstone. Away from the coast the Arbroath Sandstone is exposed in a stream section [NO 574 363] to [NO 580 353] north-east of Carnoustie. MA

North Fife

On the SE limb of the Sidlaw Anticline in north Fife, rocks assigned to the Garvock Group are poorly exposed. Basaltic lava was formerly worked at Morton Quarry [NO 4677 2568] where three flows with a total thickness of 37.8 m were penetrated by the Morton Quarry Borehole drilled by IGS. The lavas were underlain by dull brown and greenish grey,

The Arbroath Sandstone dips to the south-east at about 20° on extensive wave-cut platforms north-east of Carnoustie and at Arbroath. Trough-cross-bedding is well displayed in these sections and inspection of the exposures gives the immediate impression that the elongate troughs lie parallel to the strike of the rocks, and that the dominant pebbly sandstone with beds of brown mudstone and siltstone. Sandstones and pebbly sandstones were encountered also in further boreholes at Morton Loch [NO 4606 2623], Morton Farm [NO 4647 2600] and Shanwell [NO 4801 2768].

North-west limb of Sidlaw Anticline

Scone Formation

Strata of the Scone Formation constitute most of the Garvock Group within that part of the district which lies northwest of the axis of the Sidlaw Anticline. The rocks of the formation are mainly sandstone and attain a thickness of about 2000 m. Usually purple-brown or purple-grey, they are mainly cross-bedded and like the sandstones of the Garvock Group elsewhere are characterised in general by their content of limestone and mudstone clasts, although these constituents are not universally present.

Details

In the area west of Perth the Scone Formation is poorly exposed but the sandstones were worked in quarries near Dupplin Castle [NO 065 196], Crossgates [NO 045 208] and Lamberkine [NO 070 214]. The lowest beds of the formation are exposed in streams on the hill slopes south of Forteviot, such as the Water of May [NO 054 163] to[NO 074 162], the Bog^-ton Burn [NO 050 155] to [NO 055 152], the Garvock Burn [NO 038 150] to [NO 040 139] and the Dunning Burn [NO 020 144] to [NO 025 135]. These deposits are atypical, being mainly muddy siltstones and silty mudstones with thin beds of fine-grained sandstone and conglomerate with some 'acid' pebbles. Some beds have hardened tops. Trace-fossils are present including sinuous feeding trails and simple pipes, and there are large discoidal structures, which may also be trace-fossils or small sand volcanoes. Root traces are also present and at exposures in the Water of May [NO 067 161] and in the Garvock Burn [NO 042 146], well branched plant stems up to 35 cm long have been found. In this area a pyroxene-andesite lava flow, 30 m thick, occurs within the main body of sandstones about 300 m above the top of the Ochil Volcanic Formation.

Sandstones of the Scone Formation were formerly worked in quarries in the neighourhood of Huntingtower Castle [NO 084 251], but most of these are now infilled.

North of Perth, strata of the Scone Formation dip in general to the north-west at about 20°. They are exposed along the Annaty Burn both upstream and downstream from the road bridges at New Scone. The sandstones are grey, reddish brown and purple, fine- to coarse-grained and generally display cross-bedding. Pebbles of metasediments, lava and intraformational mudstone and limestone are of common occurrence. In the lower part of the sequence south-east of New Scone a tributary ravine [NO 143 256] shows thin-bedded, brown and grey calcareous sandstone and siltstone underlain by cross-bedded sandstone along a NW-trending dislocation.

North-west dipping cross-bedded sandstones containing mudstone and limestone clasts are exposed in the valley of the Cramock Burn [NO 123 275] west of Balboughty and also north-east of that place in the Whiggle Burn [NO 130 281]. Sporadic exposures of fine- to medium-grained grey and purple-brown sandstone with associated impersistent mudstone bands, some over 0.5 m thick, occur along the course of a small stream west and north-east of Scones Lethendy [NO 130 285]. To the north-west a deep ravine descending to the south-west [NO 124 290] to [NO 125 285] shows mainly fine- to medium-grained cross-bedded sandstone. Here also there are intercalated mudstone bands. Sandstone appears in the walls of a disused quarry [NO 120 292] opened in a quartz-dolerite dyke south of Ardgilzean.

Sandstones of the Scone Formation are exposed sporadically, in places as substantial cliff-sections, along the St Martin's Burn between Colenden [NO 110 299] and the St Martins [NO 155 305] area. Limestone and mudstone clasts are locally common and there is a small proportion of interbedded mudstone. Sandstone near the top of the Scone Formation is exposed in the Cambusmichael Burn [NO 120 321] east of Cambusmichael. Sandstones are exposed in association with quartz-dolerite dykes at Craigmakerran [NO 142 322] and in a disused quarry [NO 140 338] north-west of Wolfhill farmhouse.

Sandstones were worked in quarries [NO 284 404] near Kinpurnie Castle, probably the Newtyle and Auchtertyre localities of the old accounts, and on the northern slopes of Kinpurney Hill.

Campsie Formation

The Campsie Formation which constitutes the upper division of the Garvock Group in the Perth district includes two conglomerate beds with intercalated sandstones overlain by sandstones with interbedded mudstones near the top. Concretionary limestone (cornstone), which occurs at several levels mainly in the lower part of the sequence, is believed to have been formed by pedogenic processes during episodes of restricted sedimentation. The basal calcareous zone within the Campsie Formation is remarkably persistent near the top of the Garvock Group throughout the Strathmore outcrop (Armstrong and Paterson, 1970) and the beds have been worked as a source of lime at several places, mainly to the north-east of the Perth area.

Details

An almost complete section of the Campsie Formation is exposed in the bed and banks of the River Tay at the sharp bend [NO 124 327] south of Campsie. This constitutes the type section, and is as follows:

The conglomerates and intervening limestone-bearing beds in the Campsie river section constitute a small inner in which they are delineated for clarity as a single conglomerate bed on 1:50 000 Geological Sheet 48W (Solid). In other exposures of strata of the Campsie Formation only one conglomerate has been proved, and although more than one coarse bed may well exist, generally the basal part of the division is represented on the map as undivided conglomerate.

South-west of the Campsie river section, conglomerate crops out in the Cambusmichael Burn [NO 119 321] and in the immediate neighbourhood of Berryhills [NO 115 317]. At Berryhills the conglomerate contains quartzite boulders up to 0.3 m in diameter. During the original geological survey J. Geikie (unpublished information) recorded also greenstone, felstone, porphyrite, syenite and jasper from this locality and deduced a southwards direction of transport from the imbrication of the clasts. He noted also that cornstone was formerly worked at Berryhills. In river exposures [NO 106 312] on the east bank of the Tay, south-east of Gowrie, strata of the Campsie Formation are traversed by a quartz-dolerite dyke of ENE-trend. Here there are up to 3 m of conglomerate underlain by 1.8 m of coarse sandstone with limestone concretions.

Near Almondbank, strata of the division are exposed on the east bank 1065 262] of the River Almond, downstream of the road bridge at Almondbank. Here a bed of concretionary limestone is overlain by several metres of cross-bedded, pebbly sandstone which is succeeded in turn by at least 6 m of conglomerate containing relatively small well rounded clasts.

In the ground north-east of the Campsie riverside sections there are a number of small exposures, mainly of conglomerate. For 100 m E of the site of a disused limekiln [NO 1298 3305] north of Byres there are exposures of conglomerate with some concretionary limestone. Conglomerate occurs in the faces of an elongate quarry [NO 138 338] in a quartz-dolerite dyke north-west of Wollhill while subjacent sandstones typical of the Scone Formation crop out along the quarry to the east. Another quarry [NO 141 341] west of Knowehead shows 3.6 m of conglomerate containing clasts of Highland derivation up to 0.3 m in diameter. Concretionary limestone, up to 0.3 m thick, occurs near the base. A kiln formerly existed here.

Highland Border area

Garvock Group strata occur in a narrow strip along the northern side of the Spittalfield Fault. They consist of purple-grey or red-brown, medium- to coarse-grained, pebbly sandstone with clasts and rare impersistent beds of silty mudstone. There are few exposures, the best being in the Garry Burn [NO 037 361] south-south-east of Nether Obney, in a stream [NO 103 417] at Millhole and in a disused quarry [NO 116 424], north-east of Middle Gourdie.

Strathmore Group

On the south-east limb of the Strathmore Syncline, the Strathmore Group comprises a lower division, which consists mainly of mudstones, the Cromlix Formation, and an upper predominantly arenaceous division termed the Teith Formation. The two formations, which were first named in the Stirling district (Francis and others, 1970), interdigitate with each other on a regional scale with the result that, within the limits of the present area, the Cromlix Formation appears as two developments separated by sandstones of the Teith Formation. The lower and more widespread of the occurrences of the mudstones is about 355 m thick, the upper one is approximately 115 m thick.

On the north-west limb of the Strathmore Syncline, movement on the Spittalfield Fault has entirely eliminated the Strathmore Group at outcrop.

Cromlix Formation

The rocks assigned to this formation consist largely of reddish brown, green-streaked and spotted, ill-sorted, silty and in some cases sandy mudstones. They generally appear to be massive but at the excellent exposures along the River Almond upstream of Almondbank, a crude stratification is emphasised by differential weathering, presumably reflecting subtle changes of grain or of degree of cementation. At intervals through the sequence, lenticular bodies of fine-grained, commonly cross-bedded sandstones occur (Plate 8), which are considered to represent the fill of former river channels. In all cases the transition from the Cromlix Formation to the Teith Formation is gradational, by increase in the frequency of channel sandstones.

In addition to the exposures along the River Almond between the road bridge at Almondbank and the weir [NO 0476 2722], downstream of Craigenhall, there are good sections of the lower part of the Cromlix Formation along the River Tay between Craighall [NO 108 321] and Mains of Stobhall [NO 131 353]. Details of the upwards transition to the Teith Formation are visible in a tributary stream at Burnside, a short distance upstream of its confluence [NO 138 356] with the Tay.

The mudstones of the upper development of the Cromlix Formation are exposed in the bed of the River Almond, upstream of the weir [NO 0393 2802]. Sections [NO 0585 3017] to [NO 0546 3040] in the Shochie Burn, upstream of Pittendyne, show beds of mudstone up to 25 m thick intercalated in sandstones of the Teith Formation.

Teith Formation

The Teith Formation consists largely of fine- to medium-grained cross-bedded, purple-grey sandstones, commonly showing parting lineation, the average of 25 measurements taken at a number of exposures in the River Almond and the Shochie and Ordie burns being 072°/252°. The cross-bedding indicates that the palaeocurrents flowed generally towards the west. The beds of red-brown siltstone and silty mudstone are considered to represent overbank sediment laid down on the floodplain. In many cases these have been destroyed by erosion, as a result of channel migration, and the debris incorporated as angular clasts in the basal part of channel units, within thick, multistorey, sandstone bodies. Plant fossils were obtained from an exposure [NO 0424 3066] in the Shochie Burn near Millhole, from a disused quarry [NO 0941 3894] near Gellyburn, possibly the Murthly locality of Dawson (1890) and from a section [NO 1433 3595] on the east bank of the River Tay near Balholm.

There are good exposures of the Teith Formation in the River Almond between Craigenhall [NO 048 272] and Dalcrue Bridge [NO 045 280], where the mudstones and sandstones are baked in the vicinity of W–E quartz-dolerite dyke, and upstream of Lynedoch Cottage. Rocks typical of the Teith Formation are also exposed in the Shochie Burn upstream of Moneydie [NO 066 297] and in the Ordie Burn [NO 044 340], near Hill of Logiebride. The sandstones of the formation were formerly worked in Den Quarry [NO 082 338], north of Westwood, at Hilton Quarry [NO 075 344], about a kilometre south-south-east of Bankfoot, and at Gellyburn [NO 044 389], where they are cut by a W–E quartz-dolerite dyke. There are good exposures of the lowest beds of the Teith Formation in the River Tay for about 2.5 km downstream from the railway bridge [NO 149 372] near Cargill. The strata dip steeply in places and are cut by faults. In the river [NO 135 358], east of Innernytie, the sandstones are intruded by a quartz-dolerite dyke.

Chapter 4 Lower Devonian volcanic and related rocks

Introduction

There are two main developments of Lower Devonian lavas in the district. The Montrose Volcanic Formation, in the lower part of the Arbuthnott Group, occurs as small inliers along the axis of the Sidlaw Anticline but also includes a large lenticular body within the sediments of the Dundee Formation in the Middle Brighty area. The Ochil Volcanic Formation, in the upper part of the Arbuthnott Group, occurs on both limbs of the Sidlaw Anticline and also in the Highland Border area. Its relationship with the Dundee Formation is strongly diachronous. The lavas are best displayed on the high ground of the Sidlaw and Ochil hills, where they contain thick beds of volcaniclastic sediments, believed to be mainly of local origin. Pyroclastic rocks are uncommon in the Lower Devonian of the area but, close to Stenton, the thin local representative of the Crawton Group consists almost entirely of partially welded vitric tuff, part of the widely developed Lintrathen Ignimbrite. The general distribution of volcanic and associated rocks is shown on (Figure 7).

Numerous minor intrusions in the axial area of the Sidlaw Anticline are considered to be broadly contemporaneous with the lavas.

The lavas and associated intrusions in the district are part of a major volcanic suite which developed widely in Scotland during the late Silurian to early Devonian period (Stillman and Francis in Harris and others, 1979; Elliott in Sutherland, 1982; Fitton and others in Thorpe, 1982). Numerous chemical analyses (Gandy, 1972; 1975; Thirlwall, 1981; 1983) indicate that the volcanic province as a whole is of calcalkaline type. High potash concentrations in relation to silica, in the Sidlaw Hills (Gandy, 1975) and other areas, suggests affinities with the 'shoshonitic' suites characteristic of active continental margins rather than with island-arc volcanic activity. It has been proposed (Thirlwall, 1981; 1983) that magma genesis was associated with a north-westerly dipping subduction zone which was operative during closure of the Iapetus Ocean.

Lavas

The lavas of the district consist principally of varieties of andesite and basalt, with some flows of trachyandesite. They are classified primarily on their mineralogy but their chemical composition has also been taken into account (Table 2). Thus, it is rarely possible to identify the basalts in the field as they differ from the more basic andesites only in having a lower silica content (less than 53 per cent) and commonly a higher proportion of olivine pseudomorphs. The more acid lavas are usually distinctive, however, being lighter in colour and often fluxioned. Basic pyroxene-andesites, with or without conspicuous feldspar phenocrysts, greatly predominate over all other lava types.

The lava-flows are commonly amygdaloidal in their upper part. The amygdales are mainly filled with calcite, quartz and chloritic minerals but agate occurs commonly and various zeolites have been recorded. Autobrecciation is common, in some cases affecting the whole thickness of the lava-flow but more usually occurring in zones above and below a compact centre, as is well seen in cuttings along the southern approach road to the Tay Road Bridge [NO 426 287] to [NO 425 274]. As appears to be characteristic of Lower Devonian volcanic centres in Scotland, the lava-flows are commonly invaded by sediment, generally fine-grained sandstone or siltstone, which fills fissures and cavities in the rock, including the interstices between lava blocks in autobrecciated zones.

The lavas are moderately well exposed, especially on the high ground. Although individual flows can rarely be traced, groups of flows of similar composition and texture can be correlated over considerable distances. This is particularly true in the eastern Sidlaw Hills where the structure is relatively simple and the sequence contains sedimentary beds at several horizons (Figure 5). Correlation of the much faulted rocks in the Ochill Hills of north Fife is less secure and relies to a considerable extent on the presumed lateral persistence of the distinctive hypersthene-andesite which forms crags at Norman's Law. Although this flow has a maximum known thickness of about 130 m and a flow of similar composition, worked in quarries near Ethiebeaton, is possibly even thicker, the average flow-thickness in the Ochil Volcanic Formation probably does not exceed 15 to 20 m.

Montrose Volcanic Formation

The lavas of this formation crop out as small inliers along the axis of the Sidlaw Anticline at Moatmill [NO 413 380] and north of Newton [NO 455 419] and as a lenticular body of lava intercalated in the strata of the Dundee Formation in the neighbourhood of Middle Brighty [NO 445 388]. The lavas at all three outcrops are basic pyroxene-andesites occurring in several textural varieties. They are poorly exposed and their outcrops have been delimited partly on the basis of geomagnetic evidence. In the Moatmill House and Newton areas, where the lavas are almost horizontal, only their uppermost part is seen and their full thickness cannot be determined. At Middle Brighty, the lavas dip at a low angle to the south-east and have a maximum thickness of about 150 m. They appear to die out towards the south-west and the north-east but the possibility that their outcrop is terminated by undetected faults in unexposed ground cannot be excluded.

Details

The lavas in the Moatmill outcrop were shown on the first edition of Geological Sheet 49 as intrusive. Their relation with the overlying sedimentary rocks of the Dundee Formation is seen only in a stream section [NO 4128 3832], where the contact appears to be faulted. The occurrence within the body of two distinct lithological varieties, both common in the local extrusive sequences, supports their reclassification as lavas. Exposure is now limited to shallow stream sections at Moatmill [NO 411 367] and near Tealing House but the lavas were formerly seen in a disused quarry [NO 4207 3703], now in-filled, 400 m W38°N of West Shielhill. At Moatmill and Shielhill Quarry, the lava is fine-grained and lacks obvious phenocrysts. The labradorite feldspar laths, which form the bulk of the rock, are commonly fluxioned. Clinopyroxene occurs as ophitic grains (S49213), which in an analysed sample from Shielhill Quarry is pseudomorphed in chloritic mineral, and orthopyroxene is present in small amounts. Olivine is common as small subhedral grains, replaced by brown or yellow iddingsite with hematite rims.

The Newton outcrop lies mainly outwith the district. Parts of two flows, which are approximately horizontal but are displaced by a number of small faults, are seen in exposures in the Corbie Burn and its tributaries.

Ochil Volcanic Formation

The rocks of the Ochil Volcanic Formation occur on both limbs of the Sidlaw Anticline, where they form the high ground of the Sidlaw and Ochil hills, and also in the Highland Border area (Figure 4). They consist mainly of lavas but include thick and persistent intercalations of volcanic-detrital sediments. The volcanic pile reaches it maximum development to the north-west of Cupar, where its estimated thickness is at least 2400 m. In the Sidlaw Hills, where the lower part is not exposed, the thickness of the Ochil Volcanic Formation does not exceed 1500 m. Northeastwards, on both limbs of the Sidlaw Anticline, the lavas give way by interdigitation to the sediments of the Dundee Formation. The youngest flows tend to persist farthest, suggesting a migration to the north-east through time of the main focus of volcanic activity.

The location of this source is unknown, however, and it is not clear whether the lavas are derived from a single large centre, possibly situated to the south of the present outcrops, or were extruded, perhaps from fissures, at a number of localities.

The lavas are considered to have been extruded on to the floodplains of a major river system which flowed towards the south-west. Although this may have caused temporary lakes to form, for the most part the drainage made its way through the lava-pile. Quartz-rich sediment, derived from a source outwith the volcanic province, was deposited in fissures and cavities in the lavas. As this is a feature of the lavas as a whole it is inferred that an approximate balance between basin subsidence and growth of the volcanic pile was maintained.

Details

Basalt

Basalts are most common in the eastern Sidlaw Hills where they occur at several levels and form perhaps 5 per cent of the local lava-sequence. The lowest flow recognised, a feldsparphyric rock with a few olivine pseudomorphs only, is exposed [NO 2410 2814] near Craighead Quarry.

On the hillslopes above Kinnaird, there are three separate developments of basalt. The lowest of these, possibly a single flow, has a coarse-grained doleritic texture and is exceptionally rich in olivine pseudomorphs. ((S52595), (S59270), (S59272), (S59274)). Augite is generally ophitic with respect to the feldspar laths of the ground-mass but occurs as small prisms in a sample (S59262) collected at an outcrop [NO 2010 2624], 360 m N43°W of Evelick Castle. The basal part of this flow, resting on tuffaceous sandstone, is exposed in a small quarry [NO 2326 2875], 360 m N19°E of Woodwell. The basalts of the middle of the three developments are finer-grained and may be less basic, as they contain only sparse olivine pseudomorphs.

They have been traced to the north-east from near Kirkton into poorly exposed ground where their outcrop is affected by faulting and by the emplacement of large sill-like intrusions of basic porphyrite. In the uppermost of the three developments, which has been traced to the north-east from a point west of Beal Hill to near Woodburnhead [NO 231 298], the basalts are mainly fine-grained, generally fluxioned and richer in olivine pseudomorphs.

Basalt also occurs at several levels in the attenuated lava-sequence of the Newtyle [NO 297 414] area. The lowest of these, which crops out [NO 335 415] in the low ground north of Henderston, contains microphenocrysts of feldspar together with rather sparse olivine pseudomorphs. A single flow of basalt was formerly worked in the large quarry [NO 305 400], 700 m W of Millhole. This rock contains microphenocrysts of plagioclase feldspar and from 3 to 4 per cent of olivine pseudomorphs. The rock is generally compact and fresh but is autobrecciated and invaded by sediment in the south-west part of the quarry. Basalts occur also on the north slopes of the Sidlaws from Keillor Hill to Kinpurney Hill. They are well exposed in a stream section [NO 310 417] to [NO 313 415] south-east of Denend, where there are at least six flows. All are fine-grained and contain 2 to 4 per cent of olivine. IBP

A flow of feldsparphyric basalt is exposed [NO 448 310] in the Firth of Tay shore-section east of Stannergate. At a lower stratigraphical level a basalt flow, associated with tuffs below a mudstone-flagstone zone in the Dundee Formation, has been located by foundation boreholes, in the area [NO 436 315] of Craigie House. Other basalt lavas, also proved by drilling, occur between Craigie House and Dundee Law. MA

In north Fife, a series of basalt flows about 100 m thick has been traced from Fliskmillan Hill [NO 305 205] to Corbiehill Plantation [NO 343 226] where their outcrop terminates at the Fincraigs Fault. The lowest flows, which are markedly olivine-phyric, are well seen on Lindamus Hill [NO 303 211] and in the steading at Pittachope [NO 312 211]. The upper flows of the group, exposed to the north of Norman's Law summit, are feldsparphyric. Basalts exposed in Hell Den [NO 3397 2363] and Low Wood on the north side of the Fincraigs Fault may lie at the same horizon as the Fliskmillan basalts. Basalts near the top of the volcanic pile at Airdit Hill [NO 407 200] can be traced to the north-east as far as Balmullo. Basalts have been recognised at other levels in the largely undifferentiated lava-sequence in north-east Fife, for example on the shore 820 m S57°W of Wormit Railway Station, but it was not possible to delimit their outcrop.

Basic pyroxene-andesites

Lavas classed as basic pyroxene-andesite greatly predominate over all other rock types in the Ochil Volcanic Formation. Most are porphyritic to some extent but in general only those which contain an abundance of large (greater than 3 mm) feldspar phenocrysts have been differentiated on the maps. The lavas in the areas of undifferentiated andesite and basalt in north Fife east of Luthrie are probably mainly basic pyroxene-andesites, but they are known to include hypersthene-andesites as well as basalts. IBP

Perth area

The oldest lavas in the area between the Kinnoull and Moncreiffe faults are exposed in the general neighbourhood of Rhynd [NO 156 203] east of Moncreiffe Hill. Predominantly aphyric basic pyroxene-andesites, they include some feldsparphyric lavas exposed in a pipeline trench [NO 165 203] east of Rhynd. Younger feldsparphyric basic pyroxene-andesites form crags on the south side of Moncreiffe Hill and their outcrop extends east of the summit to be repeated by faulting in the ground south of Grange of Elcho [NO 143 214]. The feldsparphyric lavas are overlain by aphyric basic andesite lavas which form the summit and north-west slopes of Moncreiffe Hill. The contact of aphyric lavas upon feldsparphyric lavas is visible in the crags on the south side of the hill and on the east side [NO 200 122] of the M90 motorway cutting south of Perth. The aphyric lavas which are well exposed in the cutting rest on a markedly uneven surface of feldsparphyric lava.

The upper part of the lava-sequence in this area consists for the most part of aphyric basic andesites. There are a number of sedimentary intercalations, chiefly thick conglomerate beds containing clasts of acid igneous rocks, but also thinner zones containing sandstone and siltstone. The aphyric lava worked in Friarton Quarry [NO 115 208] overlies such a zone. The only important occurrence of feldsparphyric lavas is around Craigclowan [NO 121 208] on the west side of the motorway.

The lavas in the general area of Moncreiffe and Friarton hills were described by Davidson (1932b) as olivine-basalt, but although here, as elsewhere in the district, the lavas commonly contain olivine, their content of this mineral is insufficient to justify the name basalt as it is used in this account.

In the ground between the Kinnoull Fault and the belt of W–E faults which crosses the Sidlaw Hills between Kilspindie and New Scone [NO 137 263] the oldest rocks are aphyric lavas which crop out near Pepperknowes [NO 192 223]. They are succeeded by feldsparphyric lavas on Pans Hill [NO 185 216] and on Glencarse Hill [NO 185 228]. The crags on the south side of the latter hill are of amygdaloidal lavas containing agates. Aphyric basic pyroxene-andesite overlies the feldsparphyric lavas in Pepperknowes Quarry [NO 184 223], forms the hill south-east of Goukton [NO 175 224] and has been traced as far north as quarries [NO 184 246] north-east of Craignorth. Feldsparphyric lavas at a higher stratigraphical level occur north-west of Craignorth, both above and below a prominent belt of acid conglomerate. These lavas are confined mainly to the ground north of Kinfauns but feldsparphyric flows interbedded with the lower part of the conglomerate are exposed in the neighbourhood of Kinfauns Home Farm [NO 148 224].

Aphyric lavas with subordinate feldsparphyric flows crop out on the steep slopes below the crags on the south side of Kinnoull Hill. They extend to the north-east and become thicker into the neighbourhood of Arnbathie. The succeeding basic pyroxene-andesites of Kinnoull Hill are conspicuously feldsparphyric and form high rock faces on the south side of the hill. Several flows of feldsparphyric lava are probably present. The lowest one, separated from the rest by a bed of acid conglomerate, is exposed in a cutting [NO 126 223] at the side of the Perth–Dundee road on the south-west side of Kinnoull Hill. Two flows, separated by a boulder bed, were formerly exposed in Muirhall Quarry [NO 138 240]. Aphyric lavas overlie the feldsparphyric lavas of Kinnoull Hill and form the uppermost rocks of the Ochil Volcanic Formation as far north as the ground east of New Scone.

Sidlaw Hills

Lavas near the base of the exposed sequence, on the hillslopes west of Rait and Kilspindie, are mainly non-feldsparphyric. They are commonly slaggy and autobrecciated. The lower members of this series of flows can be traced on the slopes above the Carse of Gowrie to the neighbourhood of Castlehill, where they are well exposed in the Baledgarno Burn, downstream of Milton [NO 264 309]. The higher flows, which in the area north-east of Rait are separated from the lower flows by a sedimentary intercalation, appear to die out in the area north of Outfield [NO 247 302]. Somewhat higher in the sequence, at Beal Hill [NO 204 273], there are two developments of feldsparphyric pyroxene-andesite, separated from each other by a bed of volcanic conglomerate, exposed at Swirlhead [NO 205 278], and also, in places, by a basaltic flow. The feldsparphyric flows can be traced to the south-west into the neighbourhood of Goddens [NO 186 255]. Northeastwards, the lower series of flows apparently die out near Woodwell [NO 231 284]. The upper series thickens to the north-east to a maximum at Pitmiddle Hill [NO 237 303], where topographic features indicate that there are at least six flows, before apparently dying out near Dundriven [NO 251 343].

The feldsparphyric flows are overlain by a persistent sedimentary horizon which is succeeded in turn by a further series of lavas of similar composition and texture. These thin to the south-west from a maximum in the area [NO 305 905] west of Millhole and have not been traced beyond Lochindores [NO 267 357]. In the area between Millhole and Long Loch [NO 288 388] they are overlain by a thin development, possibly a single flow, of aphyric pyroxene-andesite, which thickens to the south-west at the expense of the underlying porphyritic flows. Fresh, compact lava of this group of flows was formerly worked at Tullybaccart Quarry [NO 264 363]. In this area the lavas are commonly fresh and compact, as is also the case in the area to the north and west of Arnbathie.

The aphyric lavas are succeeded by a further series of feldsparphyric basic pyroxene-andesites which extends from the Perth area to just north of South Ballo [NO 260 357]. Lavas of this group form the crags of Kinnoull Hill [NO 135 230]. Features suggest that there are at least seven or eight flows in this area of maximum development between Frankleyden [NO 205 295] and at Pitmiddle Wood [NO 228 309].

Pyroxene-andesites predominate in the series of lavas which lie at the top of the volcanic sequence in the area between Kirkton of Col-lace [NO 197 320] and Kinpurney Hill [NO 320 419]. Fresh, sparsely porphyritic lava, apparently belonging to a single flow at least 40 m thick, is currently being worked at Collace Quarry [NO 208 316], on the west side of Dunsinane Hill. Lavas of similar type form crags on Gask Hill [NO 236 348] and on Westerkeith Hill [NO 280 377], and were formerly worked in a quarry [NO 247 354] on the west side of Northballo Hill. Throughout this area, the sparsely porphyritic lavas rest upon a group of abundantly feldsparphyric flows, mostly basic pyroxene-andesites but including basalts on the escarpment [NO 217 316] at Black Hill, where parts of several flows are exposed. At Lundie Craigs, on Westerkeith Hill, a single, thick lava-flow of this series is penetrated by a remarkable system of interconnecting tunnels filled with sediment. Feldsparphyric pyroxene-andesites crop out also in the area between Keillor Hill [NO 283 387] and Kinpurney Hill [NO 323 417] and are exposed in a stream section [NO 3120 4150], south-east of Denfind.

Ochil Hills

On the south-east limb of the Sidlaw Anticline in north Fife between Dunning and Luthrie [NO 020 146], the lavas consists mainly of basic non-feldsparphyric pyroxene-andesites but include some flows of basalt which, however, could not be mapped. In the much faulted ground west of Glen Farg [NO 163 145], there are good exposures on the upper slopes of Glenearn Hill [NO 108 155], on West Dron Hill [NO 115 155], on Dron Hill and on Balmanno Hill where the lava is extensively autobrecciated. East of Glen Farg, basic pyroxene-andesites are seen at Drumvain [NO 180 145], on Tarduff [NO 200 150] and Pitcairlie [NO 210 160] hills, and on Ormiston Hill where there is good trap featuring. At Loanhead Quarry [NO 1900 1572], a single lava-flow, at least 30 m thick, was worked. This has a very irregular top and an autobrecciated basal zone containing much infiltrated green siltstone.

The feldsparphyric pyroxene-andesites occur at two main strati-graphical levels in the area west of Luthrie, the lower lying close to the base of the exposed sequence. The flows of this series are seen in the burns on the lower slopes of Dron Hill and Balmanno Hill. Farther east, they are exposed on hills at Kilknockiebank [NO 150 146] and Barclayfield [NO 170 145] and in a large quarry [NO 233 180] south of Newburgh. On nearby crags, a macroporphyritic flow was seen which contained at least five flow-units up to 2 m thick, each with a scoriaceous top and pipe amygdales at their base. These were up to 20 cm long and had their upper parts bent over towards the east. The section also shows two thin dyke-like bodies of calcitised andesite, one of which is truncated at the base of a flow-unit and may fill a lava-tunnel.

The upper series of porphyritic flows which are associated with volcaniclastic deposits, occupies the high ground between Lumbennie Hill [NO 215 155] and Dunboghill [NO 290 165]. Some of the flows are rich in olivine pseudomorphs and may be basaltic.

In the poorly exposed lava outcrop in north Fife east of Luthrie, it has been possible to trace only the most distinctive rock types, the remainder being classed as andesite and basalt, undivided. The rocks in this category consist of aphyric and microporphyritic types, mainly basic pyroxene-andesites but including hypersthene-andesites as well as basalts. Markedly feldsparphyric pyroxene-andesites have usually been distinguished, and in some cases provide valuable marker horizons, as, for example, the series of coarsely porphyritic flows exposed at Darklaw Hill. Towards the southwest these flows can be traced for a few hundred metres only, but to the north-east they can be recognised at Round Hill [NO 382 224], and, displaced westwards by a fault, at Shambleton Rocks [NO 377 230], whence their outcrop extends through Priory Farm [NO 384 238]. Feldsparphyric pyroxene-andesites exposed on the shore at Nether Kirkton [NO 3595 2530] and just north of Wormit House [NO 3920 2598] may lie at a similar horizon. Mainly feldsparphyric lavas, with a single 10-m thick macroporphyritic flow, in a fault-bounded area around Birkhill House [NO 336 234], may also correlate with the lavas at Darklaw Hill.

Macroporphyritic pyroxene-andesites, about 80 m thick, crop out on the northern and eastern slopes of Mount Hill. They are well exposed in a large quarry [NO 3327 1604], west of Mount Farm, and terminate against the Fernie Fault farther south. Westwards from Mount Hill they can be traced over Lindifferon Hill to Cunnoquhie Mill [NO 311 163]. The lavas at Mount Hill and Darklaw Hill are of rather similar appearance and may lie at approximately the same level in the sequence.

Basic pyroxene-andesites, including porphyritic varieties, are seen interbedded with volcaniclastic sediments on the shore near Tayport [NO 4518 2936] to [NO 4384 2929], and on the north-facing slopes of Hare Law [NO 443 285].

Aphyric and sparsely porphyritic varieties of pyroxene-andesite occur widely within the area. Typical members of the class were formerly worked in Bishop Sharp's Quarry [NO 4385 2811] and at Brackmont Quarry [NO 431 224] where the sequence exposed in 1962 was:

Several flows of non-feldsparphyric pyroxene-andesite, most of which have autobrecciated zones above and below a compact centre, are exposed on the shore [NO 4067 2699] to [NO 4250 2881] at Newport-on-Tay.

Broughty Ferry–Arbirlot area

Basic pyroxene-andesites, generally although not entirely nonfeldsparphyric, constitute the greater part of the volcanic sequence in the Broughty Ferry area. The rocks are commonly amygdaloidal and autobrecciated, these characteristics being displayed by the lowest lava [NO 437 309] in the Firth of Tay shore-section east of Stannergate, by lavas exposed along the course of the Dighty Water downstream from Balmossie Mill [NO 476 326], and by rocks in cuttings on the disused Broughty Ferry–Monikie railway line between Ethiebeaton and Grange of Monifieth [NO 488 328]. Thick acid andesite lavas, which occur midway in the sequence north-east of Broughty Ferry, die out northwards towards the Pitairlie Burn [NO 512 364], beyond which the lava-sequence consists entirely of basic pyroxene-andesites. The lava formerly worked at a large quarry [NO 5127 3762] is a fresh compact non-feldsparphyric variety, as are most of the rocks exposed in a good section in the Monikie Burn [NO 5240 3852] to [NO 5383 3815]. At an exposure [NO 5434 3769] in the burn, due south of Panmure Gardens, the highest flow of the volcanic sequence is feldsparphyric. Its scoriaceous top is seen to be overlain by lava-rich pebbly sandstone. North-eastwards, the aphyric pyroxene-andesites interdigitate with and are replaced by feldsparphyric varieties, which make up the entire lava-sequence exposed in the Elliot Water [NO 600 405] at Arbirlot. IBP, MA

Highland Border area

Basic pyroxene-andesites occur also in the Highland Border area, mainly in the broad outcrop around Snaigow Farm [NO 028 430]. The lavas here are poorly exposed and have been delimited with the help of geomagnetic traverses. They are generally aphyric and commonly have a platy fracture. The rock exposed in a disused quarry [NO 092 425] near Craigend is also a basic pyroxene-andesite.

Andesite

In the ground east-north-east of Dundee there are acid andesite lavas in which the proportions of clinopyroxene and orthopyroxene are similar. In the Firth of Tay shore-section [NO 443 310] to [NO 447 311] a flow of aphyric acid andesite displays spectacular autobrecciation. Inland, similar, but non-autobrecciated, lava occurs in outcrops [NO 472 314] north-west of Reres Hill, Broughty Ferry. Northwards from the Dighty Water [NO 478 325] to [NO 482 325] aphyric nonautobrecciated andesite forms a thick median zone within the Ochil Volcanic Formation. The rock is well exposed at Roman Hill [NO 477 334], south-east [NO 485 338] of Ethiebeaton, north-west [NO 492 343] of Ardownie and immediately north [NO 503 354] of Downie Mill. It is remarkably free from amygdales or any obvious parting and conceivably represents a single flow up to 140 m thick. MA

Hypersthene-andesite

Sidlaw Hills

On the north-west limb of the Sidlaw Anticline, lavas assigned to this category are comparatively rare, and are mainly feldsparphyric with large pseudomorphs after rhombic pyroxene. Lava of this type is intercalated in the sediments of the Dundee Formation in the neighbourhood of Old Balkello [NO 367 383]. Exposure is poor and the lava outcrop has been delimited mainly on geomagnetic evidence. Lithologically similar lavas occur at several levels within the area west of Kilspindie. The oldest of these developments is exposed in a small quarry [NO 2115 2525], west of Pitroddie, where parts of two flows are seen, the lower one having an autobrecciated top. Microporphyritic hypersthene-andesite, probably a single flow, appears to extend from the vicinity of Evelick Castle [NO 2035 2600] for a distarice of about 1 km to the north-east. Lava of similar type and stratigraphical position is exposed on the hillside [NO 196 254] about 800 m east of Goddens.

Aphyric, well fluxioned hypersthene-andesite, apparently representing a single thick but laterally restricted flow, crops out on the north-facing hillslopes around Hoole [NO 1208 310].

Ochil Hills

In north Fife, a valuable marker horizon is provided by a widely developed flow of aphyric hypersthene-andesite best exposed at Norman's Law, where it forms a strong dip and scarp feature and is about 130 m thick. A hypersthene-andesite lava-flow, up to 90 m thick and believed to be at the same stratigraphical level, is exposed on the high ground above Glen Farg [NO 165 145]. In a quarry [NO 1835 1525] south of Castle Law, the lava is cut by curved joint planes. The flow is currently being worked at Clatchard Craig Quarry [NO 245 177], where a basal autobrecciated zone, up to 4 m thick, was exposed with green sediment infilling the interstices between blocks. The sediment appears to be somewhat baked and its bedding is disturbed, suggesting that it was introduced while the lava was still hot and slightly mobile. There are good exposures also on the high ground at Parkhill [NO 250 185] and at Higham [NO 176 186], in Glenduckie Quarry [NO 282 190] and on Denmuir Hill [NO 300 195].

To the north-east of Norman's Law, the continuity of the outcrop of the hypersthene-andesite is interrupted by a number of NW-trending faults, but it forms a strong craggy feature wherever it comes to the surface, as at Loch Hill [NO 314 207], Craigsimmie [NO 318 212] and Craigancroon [NO 334 219]. Beyond the last locality it thins rapidly and is only rarely exposed, but can be seen at a quarry [NO 3445 2254] in Muir Den Covert. About 1 m of autobrecciated lava at the base of the flow is seen resting on sediment with bands of lava detritus in a small quarry [NO 3106 2016] near Norman's Law, and at the top of the flow a thicker zone of autobreccia has been mapped between Black Craig [NO 312 199] and Brunton [NO 323 208]. The thick central part of the flow is a compact pale grey rock with a marked feldspar-foliation and well developed columnar jointing.

An aphyric hypersthene-andesite exposed on the hillslopes between Balhelvie [NO 309 217] and East Flisk [NO 325 224] closely resembles the Norman's Law flow.

Hypersthene-andesites, some containing abundant phenocrysts of feldspar, occur at various stratigraphical levels within the largely undifferentiated lava-sequence above the Norman's Law horizon, being most common among the flows near the top of the volcanic pile. Typical of the feldsparphyric varieties are the rocks formerly worked in a quarry [NO 265 174] near Lindores, where fresh hypersthene phenocrysts are common in a microporphyritic lava (S54857). Similar lava, at a lower stratigraphical level, is exposed in a roadside quarry [NO 3414 2139], north-east of Drumnod, in a quarry [NO 3681 2286] by the steading at Fincraigs and in the small quarry [NO 3914 2264] at Bell Craig, north-east of Kilmany. Feldsparphyric hypersthene-andesites occur also in Black Wood [NO 4015 1950], in Brackmont Quarry [NO 4309 2236] (the lower of the two flows), and on the shore [NO 4492 2935] between Tayport and Tayside.

The aphyric hypersthene-andesites, usually fine-grained and well fluxioned, occur mainly at levels above the macroporphyritic pyroxene-andesites of Darklaw Hill. Rocks typical of the class are exposed [NO 388 173] near Foodie House. Similar rocks were worked in a quarry [NO 3850 1604], between Foxton and Kingask, and in a quarry [NO 4031 1788] west of West Craigfoodie.

Other examples occur in the strongly trap-featured ground between Newport-on-Tay and Tayport, notably in quarries on the north side of Knock Hill [NO 4229 2567], near Scotscraig [NO 4447 2835] and on Laverock Law [NO 4330 2804]. At the last of these, the lava is strikingly fresh and contains abundant grains of virtually unaltered rhombic pyroxene, probably hypersthene. Some, and possibly all, of the well-fluxioned flows exposed in cuttings on the southern approach to the Tay Road Bridge are of hypersthene-andesite. At least five flows are seen in the most northerly cutting [NO 4257 2797] to [NO 4269 2843], each with an autobrecciated zone above and below a compact centre, which is generally affected by steep undulose flow-jointing.

Highland Border area

Several of the lava-flows in the Highland Border area are of hypersthene-andesite. Most are felsic, very fine-grained, well fluxioned and aphyric, as in the outcrops which extend north-north-east from Craig Tronach [NO 053 402] and Craig of Stenton [NO 066 406]. Feldsparphyric varieties are also present in the area west of Whins of Fordie [NO 0768 4112], in the vicinity of Linn of Stenton [NO 0725 4039] and in the railway cutting [NO 0485 3965] south-south-east of Ringwood.

Hornblende-andesite

Lavas in which hornblende predominates over pyroxene are rare but occur at a few localities in the Ochil Hills in north Fife. Just south of the present area, at Rossie Quarry [NO 249 120], near Auchtermuchty and around Leckiebank [NO 230 121], two flows of hornblende-andesite, close to the top of the lava-pile, are notable for the xenoliths which they contain. These are mainly of two types, the first consisting of dioritic rock with amphibole and plagioclase, the second being a layered gabbroic' rock with olivine, orthopyroxene and clinopyroxene, and plagioclase (Thirlwall, personnel communication). There are also at least two rounded pebbles of quartzite. Hornblende-andesites at a similar horizon are developed around Binzian Hill [NO 066 142]. They are colour banded, in shades of grey and purple, and contain phenocrysts of fresh hornblende, clinopyroxene and orthopyroxene, magnetite and apatite in a fine-grained, mainly altered, glassy matrix.

Farther east, at a quarry [NO 3788 1627] near Kingask, lava with pseudomorphs after hornblende, as well as scattered phenocrysts of plagioclase feldspar and altered orthoclase, was formerly worked.

Trachyandesites

North of the River Tay, the only lavas referred to this group form Kirkton Hill, north of Abernyte, where parts of two flows or flow-units are present. The rock, well exposed at Kirkton Craig [NO 2613 3139] and in a quarry [NO 2572 3173] to the north-west, is fine-grained, well fluxioned and contains phenocrysts up to 2 mm long of plagioclase and alkali feldspar. The trachyandesites have a greatest thickness of about 110 m but appear to thin rapidly to the south-west and the north-east. IBP

In north Fife a trachyandesite, the compact central portion of which forms a conspicuous crag [NO 373 247] at Naughton, is pale and feldsparphyric. The rock can be traced by its characteristic appearance for a short distance to the east and west but it does not persist as far as Wormit. The trachyandesite, exposed at a quarry [NO 4009 2441] near Newton Farm, is also porphyritic, with phenocrysts of acid plagioclase, magnetite, biotite and pseudomorphs after orthopyroxene. It probably correlates with the very similar Naughton trachyandesite.

Two flows of pale grey, flinty, aphyric trachyandesite, which are separated by autobreccia and have a total thickness of about 35 m, are exposed on the shore [NO 4470 2935] at Tayside.

Acid lavas

A pink, fluxioned rhyolitic lava dipping steeply southwards at Peacehill Point [NO 3842 2586] was believed by Geikie (1902, pp. 36–37) to be at the base of the sequence exposed in the Wormit area, but is now thought to have been faulted down into its present position from a higher stratigraphical level. It may lie in a small volcanic vent. JIC

Sedimentary veins and infillings

Sedimentary fillings of small fissures and the interstices between blocks in autobreciated flows occur at most lava outcrops. Generally the sediment consists of pale greyish green, calcareous, fine-grained sandstone and siltstone, commonly thinly bedded. In some cases (S52034), the deposit consists mainly of material derived by erosion from the lavas, in others ((S55186), (S55187)) a more distant source is indicated by the presence of angular quartz grains, together with brown and white micas and clasts of metamorphic rocks. At a few localities, the sedimentary bodies extend laterally for many metres and appear to be the infillings of lava-tunnels. The most spectacular example is at Lundie Craigs [NO 2787 3770] where interconnected sheet-like bodies up to 100 m long of well bedded, mainly fine-grained sandstone occur at several levels within a single thick flow of feldsparphyric pyroxene-andesite. Some of the bedding planes carry impressions left by patches of froth (Plate 13), suggesting that the sedimentary input was intermittent, with periods when drying out took place. Strong turbulent current flow is indicated by the occurrence of flute marks. Other examples of lava-tunnel infillings are exposed in Milton Den [NO 2654 3094] and in Guildy Den [NO 5309 3837].

Pyroclastic rocks

Rocks of undoubted pyroclastic origin are uncommon in the Lower Devonian of the district. They include the Lintrathen Ignimbrite, a partially welded vitric tuff which forms the bulk of the Crawton Group in the Highland Border area, and a number of smaller bodies in the Dundee Formation around Dundee. In north Fife, several areas of fragmental rocks composed mainly of clasts of various types of lava, for example at Kilmaron Hill and Myrecairnie Hill, are shown, with reservations, as vents on the published map (Sheet 48E). A pyroclastic origin for areas of fragmental rocks at Over Durdie near Glen Carse (Geikie, 1897, p. 307) is not now accepted, the rocks being regarded as volcanic conglomerate.

Highland Border area

The Lintrathen Ignimbrite, which extends to the north-east in discontinuous outcrops in the Highland Border area for a distance of about 80 km, was originally described as an intrusion (Teall, 1888; Geikie, 1897) under the name Lintrathen Porphyry and was so shown on the first edition of Geological Sheet 48. Considered subsequently to be a dacitic lava (Campbell, 1913; Allen, 1928), it was shown to consist of partially welded vitric and lithic tuffs by Paterson and Harris (1969), who believed it to be of the same general age as the distinctive macroporphyritic lavas which lie at the top of the Crawton Group in Kincardineshire. A Rb/Sr age determination of 411.4 ± 5.8 Ma for the ignimbrite (Thirlwall, 1983) is compatible with the suggestion by Paterson and Harris (1969) that the Lochnagar granite complex, recently shown to be an annular structure (Oldershaw, 1974), was a possible source of the ignimbrite.

The Lintrathen Ignimbrite, which has a maximum thickness in the Dunkeld area of about 40 m, contains corroded and broken crystals of feldspar, quartz and biotite, along with devitrified glass shards and clasts of pumice and Dalradian metasedimentary rocks, set in an irresolvable cryptocrystalline siliceous or quartzofeldspathic matrix. Eutaxitic texture is developed in places, its distribution through the body indicating that parts of at least two ignimbrite flows are represented. The base of the deposit is exposed only at West Cult [NO 0631 4210], where the tuff, heavily charged with flakes of slate of local origin, rests upon 0.75 m of coarse conglomerate with large boulders of cleaved grit.

Dundee area

Dark grey tuff, overlain by 'shales and flags' was recorded by J. Geikie during the original geological survey from a subsequently obscured shore-section at a point [NO 432 311] west of Stannergate, Dundee. More recently, foundation boreholes in the neighbourhood of Craigie House, to the north-east, penetrated basalt lava and showed this rock to be both underlain and overlain by grey vitric tuff. The upper tuff layer, possibly as much as 80 m thick, is overlain by a mudstone-flagstone zone of the type characteristic of the Dundee Formation. It seems probable that the tuff recorded by Geikie was the lower bed of this sequence. The lower tuff was exposed again in a recent excavation [NO 431 311] on the north side of Stannergate Road.

A sequence of lava-rich sediments, formerly exposed in a quarry [NO 411 311] on the north side of the Dighty Water west-north-west of Mill of Mains, showed an upward sequence from coarse sandstone into grey and purple-brown, poorly-bedded rocks of mudstone grade, 3 m thick, below a flow of feldsparphyric basic pyroxene-andesite. The mudstone contained rounded lava clasts with a spongy internal texture and is possibly a tuff.

East of Mill of Mains, on the line of a disused mill-lade on the south side of the Dighty Water [NO 417 332], there is an isolated outcrop of a purple-brown detrital rock, full of volcanic material and tentatively identified as a tuff. It contains a mass of grey siltstone 0.60 m long.

In the Pitkerro Burn [NO 451 337] to [NO 452 342] there are outcrops of a rock of igneous origin which appears to be in a general way concordant with the adjacent top of the zone of silty mudstone and flaggy sandstones exposed in the Murroes Burn to the east. Under the microscope the rock has a vesicular texture, somewhat reminiscent of the Stannergate tuff. However, no layering is observable in outcrops and there is a suggestion of baking of the adjacent sediments both below and above the mass. The body is also somewhat discordant in detail at its base and has included masses of sedimentary rock. It is possible that it is actually an intrusive tuff or vesiculated igneous intrusion. MA

Possible vent-structures

Large poorly exposed areas of bleached, brecciated and silicified country rock in the areas of Kilmaron Hill and Myrecairnie Hill may lie in or near volcanic necks. The breccias are of characteristic pale mauve and cream colours, and contain blocks of various lava types of all sizes up to at least 2 m across. They are best exposed in quarries at Toss Hill 1353 170], in a quarry [NO 3655 1670] north-west of Hilton and in a quarry [NO 3673 1849] south-east of Hillcairnie. Carbonate veining accompanies the more usual silicification. The possibility that the breccias are actually altered volcaniclastic sediments in the metamorphic aureoles of concealed intrusions cannot be excluded.

An area in which the lava outcrops are very irregular, with partial bleaching and some brecciation, is to be found in the fields [NO 377 206] west of Wester Forret, and there is a poorly exposed area of bleached and altered material in the fields [NO 347 160] north of Wester Balgarvie. Both areas may lie in volcanic necks but have not been so distinguished on the published map (48E). JTC

Possible tuffisite intrusion

In a cutting [NO 427 311] on the north side of the Dundee–Broughty Ferry railway west of Stannergate, the general south-east dips of sandstones of the Dundee Formation in this area are disturbed. Over a distance of 7 m in the wall of the cutting there is a zone of brecciation bounded by two steep surfaces trending about north-north-east, which limit relatively undisturbed sandstones to the east and west. The breccia has a grey to purplish sandstone matrix which incorporates rock fragments, mainly identifiable as mudstone set at all angles. There are also discrete masses of sandstone and a large steeply-dipping mass of mudstone. The breccia has diffuse vertical elements and there are veins of calcite, baryte and hematite. It is suggested that this occurrence is related to volcanic activity, the breccia being in the nature of a non-igneous tuffisite. MA

Volcaniclastic sediments

Deposits composed almost exclusively of lava debris can be traced as intercalations within the volcanic pile on both limbs of the Sidlaw Anticline for distances of several kilometres, in some cases apparently passing laterally into sediments typical of the Dundee Formation. In the Highland Border area, the Arbuthnott Group consists largely of 'volcanic' conglomerates.

North Fife

The oldest volcaniclastic sediments in north Fife occur within the sequence which lies beneath the hypersthene-andesite of Norman's Law. At Castle Law [NO 1700 1525] to [NO 1875 1568], approximately 40 m of thickly bedded fine- to coarse-grained volcanic conglomerate rests on at least 18 m of brown, thinly bedded, fine- to coarse-grained sandstone with siltstone layers. The conglomerate is composed mainly of pebbles of basic lava types, but there are some beds in which pale, acid lava fragments predominate. Eastwards, the proportion of detritus of volcanic origin decreases and the sequence appears to give way laterally to sediments typical of the Dundee Formation.

Volcanic sediments at a somewhat higher stratigraphical level are exposed on Carpow Hill [NO 208 166] and at Craigsparrow [NO 218 165]. The sequence at Carpow Hill is:

Volcanic conglomerates, believed to be high in the sequence, crop out on Drumfin Hill [NO 083 165]. These may be of the same general age as the thick volcaniclastic sediments which, divided into a lower and an upper series by a development of feldsparphyric pyroxene-andesites, crop out in the area between Pitmedden Forest [NO 220 150] and Lindifferon Hill. The deposit which underlies the lavas has a maximum thickness of about 9 m in the area south of Dunbog. It is generally coarse-grained, with angular to well-rounded blocks mainly of basic lava but with some acid pebbles. The main development of volcaniclastic sediments, up to 300 m thick, consists largely of coarse conglomerate with rare beds of sandstone and siltstone. Acid pebbles are uncommon. Bedding is usually obscure or absent although visible in the core obtained from a borehole drilled for water at Lochiehead [NO 2518 1319], just to the south of the present district. Some of the clasts in this deposit are so angular that it is unlikely that they were transported any great distance and it is possible that the sediments were laid down at least in part as debris flows. Fragmental rocks at Kilmaron Hill and Myrecairnie Hill which are classed as tuff and agglomerate on the map and considered to lie within vents, may actually be lateral equivalents of the Lindifferon Hill volcaniclastic deposits, heavily altered within the metamorphic aureole of a large, concealed igneous intrusion.

In the Gauldry–Wormit area, the continuity of outcrop is interrupted by large W–E faults. It is possible, however, that volcanic conglomerates in the Gauldry area, which contain fragments of pink acid material on Shambleton Hill [NO 381 230], are similar in age to the conglomerates of Lindifferon Hill. Volcaniclastic rocks in the area between Balmerino and Wormit Bay, which at exposures in the fields [NO 377 251] south of Kilburns contain angular felsic clasts, may also be of the same general age. The eastward continuation of the Kilburns acid breccia can be recognised in outcrops on the hilltop [NO 3845 2555] south of Peacehill Point, but farther east it cannot be detected and no definite connection can be established between it and the central acid portion of the sequence of volcaniclastic sediments of the Wormit area. These deposits, which are well exposed on the shore and hillsides at Wormit, have been described in some detail by Durham (1886) and Geikie (1902, pp. 35–47). Both authors misunderstood the pattern of faulting, however, and neither account of the sequence is entirely correct. The lowest 40 m or so (Geikie's bed 8), exposed on the shore beneath and to the west of the abutment [NO 3956 2635] of the Tay railway bridge, is a basic lava-conglomerate with rounded pebbles and boulders up to 30 cm long set in a matrix of fine lava debris. It is overlain, at a small reef called Long Craig [NO 3961 2639], by a 70-m thick sequence of volcanic sediments, most of which contain fragments of pink acid material. The outcrop of these beds extends almost as far as the Woodhaven Fault [NO 4010 2667] but is complicated in one area [NO 3985 2650] by small faults. Finely banded volcanic-detrital sediments of acid composition exposed at the top of the cliff by the railway bridge abutment lie near the base of the acid division. They are succeeded, at the boathouse [NO 3963 2637], by two thin lava-flows, above which lies a thick series of mixed acid and basic conglomerates and breccias. Near the top of these there is a conspicuous band of green-weathering volcanic-detrital sandstone (Geikie's bed 14), 1.2 m thick, well seen about 200 m E of Long Craig, and in the cliff above there is a very coarse unbedded breccia (Geikie's bed 16) which contains boulders of pink flow-banded lava up to 2 m in diameter. This bed is also exposed beyond the zone of minor faulting, in the cliffs [NO 4000 2654] south-west of the Woodhaven Fault, and is underlain, as before, by a bed of green sediment. A block of rhyodacite glass, obtained by Durham (1886, p. 423) and subsequently analysed by Shand (1929), probably came from this locality. The breccia is overlain, at the cliff top, by a lava-flow.

The volcaniclastic sequence of the Wormit area may correlate in general with a zone of lavas and lava-rich sediments exposed discontinuously on the shore near Tayport, between Roofert Rock [NO 4185 2936] and a point [NO 4384 2929], 150 m farther east. TIC

Dundee area

In the Firth of Tay shore-section [NO 443 310] east of Stannergate there is a bed of volcanic conglomerate about 6 m thick. It underlies a conspicuously autobrecciated acid andesite. The lava clasts are rounded and there are traces of stratification.

Perth area

In the Perth area, between the Moncreiffe Fault and the zone of faulting extending eastwards from New Scone, there is a conspicuous development of volcanic conglomerate within the Ochil Volcanic Formation. The conglomerate, which occurs at more than one stratigraphical level, is composed almost entirely of rounded clasts of acid igneous rocks of types which do not occur in the local extrusive sequence. In a description of the conglomerates on Friar-ton and St Magdalene's hills south of Perth, Davidson (1932b) found that 'In no case does the percentage of acid and sub-acid rocks reach less than 95 per cent, and in all the boulder analyses made about 98 per cent of the specimens belong to such types'. Davidson further stated that the dominant rock-type is a biotiteandesite with porphyritic feldspars, and records the occurrence of lelsites' resembling those of Peacehill Point and Lucklaw Hill in Fife, and also specimens petrographically indistinguishable from the acid porphyrite of Ninewells, Dundee. The apparently abrupt transition from 'acid' to 'basic' conglomerates to the north of the Perth area is probably more apparent than real and there may be a zone of transition in the relatively poorly exposed conglomerates near Arnbathie.

The wide distribution and uniform character of the acid volcanic conglomerates in the Perth area suggests that these sediments were derived from a relatively large contemporaneous source area covered almost exclusively by acid rocks. An acid volcanic cone or raised extrusive dome is envisaged. From such a source, acid detritus might be expected to spread widely and interdigitate with near-contemporaneous basic andesites, which is the observed relation. The actual acid extrusives, however, might reasonably be expected to be much more restricted in distribution and thus not reach the area of the exposed volcanic sequence in the form of lava. The possible source rocks to which Davidson (1932b) drew attention are in themselves too small in area to be a credible source for the acid conglomerates, but they may conceivably have related to former acid extrusives at the top of the volcanic sequence in north Fife. If such rocks ever existed, however, they were destroyed by erosion at the level of the present outcrops before the deposition of the Garvock Group. Such an eastern source seems in any case to be too distant and high in the Ochil Volcanic Formation to account for the observed stratigraphical and areal distribution of the acid volcanic conglomerates, which is better explained by a more western source, possibly a site between Perth and the Highland Border, where its eroded remnant would now be deeply buried below strata of the Garvock and Strathmore groups.

The acid volcanic conglomerates in the ground south of Perth are well exposed on St Magdalene's Hill, where their interdigitation with basic lavas can be traced. Conglomerate at a lower level was located in a borehole [NO 1147 2136] north-west of Friarton Quarries [NO 110 212]. Grey siltstone exposed in the floor of the quarries passes down into coarse volcanic detritus. A lower bed, containing clasts of acid igneous rocks up to 0.25 m in diameter, occurs below feldsparphyric lavas on the south side [NO 116 205] of Borden Hill. A thick zone of acid volcanic conglomerate south-east of Scoonieburn, proved by drilling, was formerly exposed during the construction of the M90 motorway and was cut in the Moncreiffe railway tunnel. The upper part of this bed consists of sandstone which is still exposed in the motorway cutting [NO 122 205].

North of the River Tay acid volcanic conglomerate is present in a thick development below the lavas of Kinnoull Hill. It is well exposed near the foot of the slopes due south of the summit of the hill. At a higher level conglomerate beds are intercalated near the bottom of the feldsparphyric lavas of Kinnoull Hill, and one within these lavas is exposed in a road cutting [NO 125 223]. Still higher in the sequence a conglomerate up to 0.8 m thick was formerly exposed between successive flows of feldsparphyric lava in the now infilled Muirhall Quarry [NO 240 139]. East of Kinnoull Hill the main body of the acid volcanic conglomerate extends into the neighbourhood of Kinfauns where there are intercalations of lava. Northwards, conglomerates rich in pink acid lava-detritus have been traced into the ground north-west of Commonbank [NO 178 246]. Farther east there is an extensive development of acid volcanic conglomerate between Balthayock Wood [NO 190 235] and Over Durdie. This development may represent a faulted repetition of the conglomerate zone near Commonbank. Whether or not this is so, the conglomerate bed within the area north and north-east of Bathayock Wood is certainly extensively repeated by faulting.

Throughout the area north of the River Tay the acid volcanic conglomerates are commonly coarse. Rounded boulders up to 1 m in diameter have been recorded in exposures [NO 146 224] west of Kin-fauns Home Farm, in a roadside outcrop [NO 230 167] west of Balthayock House, from a deposit below feldsparphyric lavas a little to the south-west of Hollowdub [NO 168 235] and in forestry road cuttings [NO 191 245] south-east of Oliverburn. In some places the deposit is less coarse-grained and the included fragments are not obviously rounded. It is possible, though not proven, that pyroclastic beds are included.

Sidlaw Hills

In the Sidlaw Hills, there are four main developments of volcaniclastic sediments within the lava-sequence. The lowest can be traced from the neighbourhood of Goddens [NO 186 254] to the north-east to Whitehills. At a small quarry north-west of Evelick Castle [NO 2010 2620] the following sequence is exposed:

Coarse volcanic conglomerate is exposed also by Shanry [NO 2033 2686] and at Ladywell [NO 2125 2763] 'ashy' sandstone about 3 m thick, with thin beds of reddish brown mudstone, is overlain by lava. At exposures in the neighbourhood of Sunnyhall [NO 2390 2919], the mudstone beds are thicker and contain plant fragments, and the sandstones are less tuffaceous. At Whitehills Quarry [NO 233 308], the sandstones resemble those in the Dundee Formation, being quartzose and rich in mica.

At a somewhat higher stratigraphical level, coarse-grained, mainly basic, volcanic conglomerates are well exposed on the southern slopes of Pole Hill [NO 195 261], where they enclose a thin flow of basalt, and of Beal Hill [NO 205 272]. At exposures near Swirlhead [NO 206 278], the conglomerate contains rounded boulders of basic lava up to 0.6 m in diameter. The deposits thin to the north-east and contain a higher proportion of sandstone, as in exposures near Woodwell [NO 231 234], and near Whitehill Quarry.

The third development of volcaniclastic sediments, composed of pale purple, coarse-grained conglomerate, is exposed [NO 1280 2565], a little to the south-east of Arnbathie. For some distance to the northeast, its outcrop is complicated by faulting, but there are good exposures of coarse-grained conglomerate, with lenses of tuffaceous sandstone, to the north-west and north of Westlaws [NO 2280 2919], and also [NO 2335 2969] to the east of Woodburnhead. The sediments are not exposed in the area to the north, and when next seen at a small exposure [NO 2403 3352], south of Glenbran, the deposit consists of pebbly, tuffaceous sandstone, which gives way to typical Dundee Formation flaggy sandstones and siltstones in quarries [NO 2505 3463] and [NO 2537 3453], north of Dundriven.

The uppermost of the sedimentary intercalations within the lava pile can be traced to the north-east from Bandirran House [NO 201 306] to the northern slopes of Henderston Hill. It consists mainly of cross-bedded, pebbly sandstones which are only slightly tuffaceous and little different from the ordinary sandstones of the Dundee Formation, for example at a quarry [NO 2355 3355], west-south-west of Glenbran, and another [NO 2839 3803], near Wester Keith.

Petrography of the lavas

Basalts

The basalts form a continuous series with the pyroxene andesites. They have been distinguished on the basis of their lower silica content. Where this is not known, the rocks with the greatest abundance of olivine pseudomorphs have been allocated to the basalt category. However, the correlation between silica percentage and olivine content is not close and some rocks with abundant olivine prove on analysis to be pyroxene-andesites. Conversely, rocks mapped as pyroxene-andesites probably include some basalts.

Olivine usually occurs as euhedral phenocrysts generally from 0.1 to 1.0 mm long but in some cases exceeding 2 mm ((S49246), (S52595), (S59272) and(S60581)). It is almost never fresh being replaced by green, foxy-red or straw-coloured chloritic or serpentinous minerals, rimmed with iron oxide. In more altered rocks, the olivine may be pseudomorphed in quartz or carbonate with iron oxide, or in iron oxide alone. In the absence of chemical analysis, lavas with more than 5 per cent of olivine are classed as basalt.

Plagioclase, with an average composition of An_55-60, makes up the bulk of the rock. It commonly occurs as phenocrysts from 2 to 5 mm long ((S51630), (S54455)), which may be zoned from bytownite to andesine or oligoclase, in some instances in oscillatory fashion (S52064). Only those rocks which are conspicuously porphyritic have been distinguished on the maps. Laths and microlites of plagioclase form the bulk of the groundmass of the basalts ((S59259), (S59262), (S60601)), being commonly arranged parallel to the flow ((S52614), (S59266)).

Augite is occasionally present as colourless or pale brown phenocrysts (S56394) but more usually occurs as small prismatic crystals ((S59262), (S59266)) or as minute granules ((S49246), (S52564), (S59268)) in the groundmass. In a few cases, as noted by Harry (1956), augite has an ophitic ((S52595), (S59270), (S59272), (S59274)) or subophitic ((S53271), (S54446), (S54455)) relationship with the feldspar laths of the groundmass.

Many of the basalts are essentially holocrystalline but some contain small amounts of partially devitrified glassy residuum interstitial to the plagioclase laths ((S52565), (S54455), (S59262), (S60701)). Rarely, the glassy material exceeds 10 per cent ((S52604), (S59263)).

Basic pyroxene-andesites

The rocks of this class contain from 53 to 56 per cent of silica. Most contain euhedral phenocrysts of olivine, generally but not always forming less than 5 per cent of the rock. The olivine is invariably represented by pseudomorphs, usually in serpentinous mineral rimmed by iron oxide. Plagioclase feldspar, in the andesine-laboradorite range, commonly occurs as phenocrysts ((S49928), (S60585)) which are often corroded and zoned ((S51222), (S56391)). In a few rocks, almost colourless augite also occurs as phenocrysts ((S40571), (S48738), (S48750), (S60579), (S60585)), but is more usually present as small prisms or granules ((S49204), (S52337), (S52622)). Rarely, augite is in ophitic relation to the plagioclase laths of the ground-mass ((S49939), (S52361), (S56389), (S60587)).

Small prismatic crystals of orthopyroxene, probably hypersthene, sometimes fresh ((S49262), (S51644)) but more usually pseudomorphed in green chloritic mineral (S49259), are generally present in small amounts in the basic pyroxene-andesites.

The groundmass consists of laths and microlites of anclesine-labradorite feldspar, which in many cases are flow-aligned to some extent ((S52597), (S56378), (S60582), (S60584)). Many rocks also contain turbid devitrified glassy material, either interstitial to the groundmass feldspars ((S51227), (S59269)) or enclosing them ((S51211), (S59271)).

Hypersthene-andesites

This category includes andesites of more acid composition (silica content from 56 to 60 per cent), in which the predominant mafic mineral is orthopyroxene, probably mainly hypersthene. In a few cases this is fresh ((S52938), (S54857), (S59253), (S59254)) but is more usually pseudomorphed in chloritic or serpentinous mineral. The rocks fall into two groups. The first consists of fine-grained, essentially aphyric, well fluxioned lavas, light grey in colour. They may contain rare pseudomorphs after olivine, and a few phenocrysts of andesine-labradorite feldspar. Clinopyroxene, probably augite in most cases, occurs as small prisms and granules but is markedly subordinate to orthopyroxene ((S49260), (S52336), (S52920), (S52960), (S56254)). The groundmass consists of laths of andesine-labradorite feldspar. Small amounts of interstitial quartz may be present. Lavas classed as andesites at Ethiebeaton and Stannergate [NO 445 309] in the Dundee area are also hypersthene-andesites of this type.

Members of the second group of hypersthene-andesites are conspicuously porphyritic and are commonly pale purplish grey in colour. They contain numerous phenocrysts of andesine feldspar, in some cases zoned and rimmed with oligoclase. There are usually a few pseudomorphs after olivine ((S56263), (S59260)) and all contain well-shaped phenocrysts of orthopyroxene, occasionally fresh ((S51623), (S54308)) but more generally replaced in serpentinous or chloritic material ((S49926), (S51596), (S51625), (S56248), (S56252)). The groundmass consists of andesine feldspar laths with variable amounts of devitrified glass ((S52612), (S56263)).

Hornblende-andesites

In a few rocks, probably towards the acid end of the andesite range (61 to 65 per cent of silica), hornblende ((S49921), (S54315)) may predominate over pyroxene. Magnetite and apatite also occur as phenocrysts.

Trachyandesites

Following Francis and others (1970), the term trachyandesite is applied to leucocratic, well fluxioned, pale-coloured rocks containing alkali feldspar as well as acid plagioclase. Within this category is the essentially aphyric rock (S52007) which forms crags to the north of Kirkton. The trachyandesite at Newton [NO 410 244] is porphyritic, with phenocrysts of acid plagioclase, magnetite, biotite and pseudomorphs after orthopyroxene, set in a fine-grained matrix with fluxioned feldspar laths.

The silica content (54.01 per cent) of the rock formerly worked in Craighead Quarry [NO 2410 2809] suggests that it is basic andesite rather than a trachyandesite, as stated by Harry (1956).

Acid lavas

A single flow of pink acid lava is exposed at Peacehill Point. It consists (S49941) of a fine mosaic of alkali feldspar and quartz with small crystals of decomposed feldspar (Flett in Geikie, 1902, p. 387).

Minor intrusions

Numerous minor intrusions cutting the sediments and lavas of the Arbuthnott Group in the axial area of the Sidlaw Anticline around Dundee and in north Fife are chemically similar to the lavas and are considered to be of the same general age. The intrusions range in size and form from thin, near vertical dykes, through boss-like masses more than a kilometre across to the large, irregular near-concordant sheets of the area west of Dundee. Many of the bodies are poorly exposed and their field relations are unknown.

In most of the district, the minor intrusions are subdivided on the basis of their composition and mineralogy into olivine-dolerites, basic porphyrites, porphyrites and acid porphyrites, corresponding to the main divisions of the lava suite. In north Fife, further categories have been distinguished on the basis of texture. A general distribution of the minor intrusions is shown on (Figure 8).

Details

Olivine-dolerite and basalt

As in the case of the basalt lavas, intrusions of this composition are uncommon. They occur mainly in the area north-west of Dundee, the largest being an irregular sill-like body which forms high ground near Clushmill [NO 301 372]. The rock here is generally fresh and blue-grey in colour. A small body [NO 270 325], intruded into the flagstones and shales of the Dundee Formation east of Newtongray, also appears to be a sill although no actual contact is exposed. The W–E-trending body a little to the south may be a dyke. The form of the intrusions near East Adamston [NO 334 357], at Camperdown House [NO 355 330] and near the Infirmary [NO 399 305], Dundee is obscure but they may be bosses. The East Adamston mass, in particular, has no surface expression and is known only from excavations [NO 3308 3586]; [NO 3363 3607] for pylon foundations.

In north Fife, dykes near Corbiehill [NO 343 224] and a kilometre north of Kilmany [NO 387 228] are of feldsparphyric basalt with abundant pseudomorphs after olivine.

Basic porphyrite

The basic porphyrites constitute by far the most voluminous group of intrusions particularly in the area around Dundee. Chemical analyses show them to be broadly equivalent to the basic andesite lavas. The intrusions are commonly somewhat more basic at their margins, tending towards olivine-dolerites in some cases. A feature of many of the bodies is the occurrence of veins and diffuse patches of more acid material.

The most prominent of the basic porphyrite intrusions are the large, irregular sheet-like bodies, in places more than 150 m thick, which extend from Rossie Priory [NO 285 308] to Adamston Wood [NO 320 365], a distance of about 8 km. The north-east part of the complex is poorly exposed and its outcrop is based partly on geomagnetic evidence. There are good examples of acid segregations in a disused quarry [NO 282 318] at Hilltown of Knapp. The basic porphyrite intrusion at Castle Huntly may also be sill-like, the steep contact with the country rock exposed in the west wall of a quarry [NO 3040 2966], north of the castle, probably being of local significance only.

A number of small bodies in the area around Kellas also appear to be sills. The upper surface of one of these masses, exposed [NO 4605 3551] in Murroes Burn near Mill of Murroes, is concordant with the bedding of the overlying Dundee Formation sediments. Sills at Powrie Brae [NO 434 345] and Duntrune Hill [NO 447 352] may well be parts of the same body, displaced by an unmapped W–E- trending fault. Duntrune Hill forms the crag of a 'crag-and-tail' structure and it is of interest that recent boreholes drilled upon it penetrated many metres of completely decomposed sandy material before encountering solid rock. In Glack Quarry [NO 449 353] northeast of Duntrune Hill, there are good examples of acid segregation patches.

The body of fine-grained basic porphyrite formerly quarried at Kingennie House [NO 476 356] is probably also mainly concordant but may include a vertical boss at its south-west end. The very similar rock quarried at Cunmont [NO 487 370] forms a body of elliptical cross-section, at least 50 m thick, inclined to the south-east at an angle of about 20 degrees. At the time of the survey, sandstones of the Dundee Formation were visible above, and in places at the base of, the working face.

A number of basic porphyrite intrusions in the area around Auchterhouse and Balluderon hills are probably also sills. The only unfaulted contact seen, however, is at Linn of Balluderon [NO 3706 3910] where the concordant top of a 30-m thick sill is exposed. A small group of irregular bodies of basic porphyrite [NO 340 404] north-east of Scotston may be connected at depth. Intrusions within Dundee near Balgay Park [NO 385 308], in the area of the University north of the Tay railway bridge and in the neighbourhood of Caird Park [NO 405 323] and [NO 410 330] are probably sills but their contacts with the country rock are nowhere seen.

Many other bodies of basic porphyrite, such as the group around Kirkton of Strathmartine [NO 377 353] are poorly exposed but their restricted outcrop suggests that they may include bosses. A small nearly circular body [NO 274 358] near Lochindores may connect at depth with the Rossie Priory sill-complex.

The most prominent of the rather rare basic porphyrite dykes north of the River Tay cut the sandstones of the Dundee Formation on the north side of Auchterhouse Hill. A W–E dyke [NO 295 336], west of Binn, contains analcime and may actually be of late-Carboniferous age.There are few bodies of basic porphyrite south of the Firth of Tay. One N–S-trending dyke, 15 m wide, cuts the lavas on the shore [NO 4493 2935] near Tayport. A small body of basic porphyrite at Horselaw [NO 356 148] may be a boss.

Porphyrites and related rocks

The intrusive equivalents of the less basic andesite lavas have been classified generally as 'porphyrites'. South of the Firth of Tay, subdivision of this category has been carried out on the basis of texture. Coarse-grained 'doleritic' varieties have been classed as microdiorite; fine-grained rocks have been named after the lava-type which they most closely resemble. The porphyrites occur as irregular sill-like masses, dykes and small bosses.

The largest porphyrite bodies are intruded into the Dundee Formation within the city of Dundee. They comprise a petrographically uniform series of augite-porphyrites, commonly fresh and blue-grey in colour, at Dundee Law [NO 390 313], Baxter Park [NO 420 317], Longhaugh [NO 422 328] and Whitfield [NO 435 335]. Their relations with the country rock are not well known but they are probably in the form of irregular sills. The heavily altered, rather more felsic porphyrite of Gallow Hill [NO 393 414] is probably also a sill.

The Dundee Law intrusion contains coarse leucocratic patches which bear a general resemblance to certain of the acid porphyrite intrusions. These coarse patches occur north-west of the summit of the Law and also on the south-west side of the intrusion, where several discrete areas of coarse light-coloured rock within the dark grey porphyrite of the Law are exposed near the west end [NO 390 307] of Dudhope Terrace. A little farther to the north-west a mass of the coarser material [NO 388 308] is emplaced within the siltstones which in turn are enclosed by the normal porphyrite of Dundee Law. MA

A few porphyrite dykes in the district mostly have a northerly trend. They include a thin, irregular dyke cutting a body of acid porphyrite on the west wall of a disused quarry [NO 441 351], west of Duntrune House and one of the two dykes exposed in the railway cutting [NO 370 296] to [NO 382 296], east of Ninewells. Chemical analysis suggests that the easterly one is actually a basic porphyrite. South of the River Tay, dykes at Fincraigs [NO 369 229], Lewes Wood [NO 348 222] and west-north-west of Wester Forret [NO 386 207] are of andesitic texture. A 2-m wide pyroxene-andesite dyke at Drybrae Scalp [NO 4252 2876] is very altered. The dyke in the road cutting [NO 4251 2748], near Newport, dips westwards at 55°. An arcuate body of microdiorite intruded into the northern part of the microgranodiorite boss at Forret Hill is thought to be dyke-like in form. A probable eastward extension of the Forret Hill microdiorite is poorly exposed among the lavas in a burn [NO 4035 2067] just north of Logic village. Smaller bodies of microdiorite on the southern side of Forret Hill, on Kedlock Hill [NO 375 189] and on Kilmaron Hill [NO 359 165] may be bosses. IBP, MA, JIC

Acid intrusions

A varied group of intrusive rocks of generally acid composition have been classified in terms of their texture and mineralogy. They are usually pale coloured, fine-grained and commonly conspicuously feldsparphyric. Phenocrysts of quartz, biotite or hornblende may be visible in hand specimen. The available chemical analyses indicate that the acid intrusions range in composition from trachyandesite to rhyolite. IBP

North of the Firth of Tay, the acid intrusions have been grouped together as 'acid porphyrites'. They appear generally to be irregular sheets, as in the case of the body emplaced in strata of the Dundee Formation on the western side of Dundee. In the railway cutting [NO 366 297] to [NO 375 296] east of Ninewells the intrusion exhibits crosscutting contacts in places, for example at a point on the north wall of the cutting [NO 370 297] east of the railway bridge at Ninewells. MA

The well exposed basal contact of the small body of acid porphyrite [NO 433 358], west of Craighill is mainly concordant but in places it can be seen cutting the bedding in the country rock. Fragments of this distinctive flesh-coloured, feldsparphyric rock with ragged cavities partially infilled with chloritic mineral can be identified in a glacial 'trail' extending several kilometres to the east. The large masses of acid porphyrite at Dronley Wood [NO 340 385] may also be sills but their relations with the country rock are nowhere exposed.

South of the Firth of Tay, the largest intrusion is the rhyolitic felsite of Lucklaw Hill. Thought to be a lava flow by Geikie (1902), it is now considered to be an intrusion, probably a laccolith. It occupies over a square kilometre of hilly ground between Lucklaw Hill and the Lucklaw Hill Fault and is best exposed in a large quarry [NO 420 213] at the south side of the mass. Just west of the quarry, there are northerly trending veins containing quartz and baryte.

North of the main mass, at Crumblie Hill [NO 410 226] lies a body of brecciated felsite similar to the Lucklaw Hill rock. Its intricate contact with the country rock is well exposed 500 to 900 m W of South Straiton [NO 420 229]. The breccia contains fragments of basic lava in places and this feature suggests that the mass is an intrusion-breccia which originally lay near the roof of the Lucklaw Hill felsite and was let down to its present level by the Lucklaw Hill Fault. It is cut off on the north side by the Straiton Fault.

The microgranodiorite, 1.5 by 1.0 km across, at Forret Hill is believed to be a boss. A near-vertical contact with the country rock is exposed at a point [NO 3878 2009] on the western side of the intrusion. Chemical analysis shows it to be generally equivalent to the trachyandesite lavas.

Smaller bodies, probably of similar composition and classed as felsic alkaline rocks', form a well marked group of small dykes and sills in the area between Dunbog [NO 285 178] and Balmerino. The dykes trend in general between north-west and north-north-west. A larger body of rock, near Kedlock [NO 385 194], of similar composition but better crystallised, is probably a boss. This is cut by a 25-m wide, NE-trending acid porphyrite dyke, one of several in the area. A second such dyke, also about 25 m wide, is exposed in a quarry [NO 4074 2116] near Ardlogie House.

Petrography of intrusive rocks

Olivine-dolerite and basalt

In the large olivine-dolerite intrusions of the Dundee area, the rock is usually fresh and well crystallised. Well shaped olivine crystals, 1.5 mm or more in length, occur in some abundance ((S52154), (S54143), (S60476)) almost invariably pseudomorphed in brown and green serpentinous mineral. In one sample (S54419) from the large Clushmill sill, fresh olivine occurs in the cores of some of the phenocrysts. Feldspar phenocrysts, consisting of labradorite plagioclase, are sporadically represented in some rocks. Clinopyroxene, probably augite, is alway present and is almost always fresh. It may occur as well shaped crystals (S52061) or as large grains, up to 2 mm across, in ophitic relationship with the labradorite laths of the groundmass (S54424). Iron oxide is common as rods, octahedra and skeletal crystals.

The basalts ((S51595), (S51636)) of the dykes on the south side of the Firth of Tay, are feldsparphyric and contain abundant olivine pseudomorphs.

Basic porphyrites and related rocks

Intrusions in this group occur in several textural varieties. In the sheet-like bodies in the area around Dundee, the rock tends to be medium-grained and of 'doleritic' aspect, blue-grey when fresh but commonly altered to some extent. The category includes rocks termed quartz-hypersthene dolerites. They usually contain phenocrysts of labradorite-andesine plagioclase, commonly zoned with less basic rims (S52005). Olivine, pseudomorphed in brown or green serpentinous mineral, is present in some cases ((S52053), (S52066)), particularly in samples from marginal parts of intrusions. Clinopyroxene is probably always present, in some cases as fresh, well-shaped crystals ((S50111), (S52044), (S52053), (S52056)), or as small grains (S52202), but more often partially or wholly replaced by chloritic mineral (S52043). Rhombic pyroxene, pseudomorphed in chloritic mineral, occurs in a number of samples ((S50118), (S50119), (S52005), (S52051)). In the rock (S52180) from Balgay Hill [NO 3774 3087], fresh orthopyroxene is intergrown with clinopyroxene. Phenocrysts of iron oxide as octohedra, flakes and skeletal crystals are common (S48882). The groundmass consists of stubby laths of plagioclase, labradorite-andesine when fresh but commonly sericitised or albitised to some extent. Quartz is generally present in small amounts as interstitial patches ((S52005), (S52053)). Many rocks contain some glassy residuum, partially devitrified with microlites of feldspar and iron oxide ((S52053), (S52056)).

The rock of the acid segregations, which are a characteristic feature of many basic porphyrite intrusions, is coarse-grained and consists predominantly of feldspar laths, up to 7 mm long (S50114). These are usually turbid, albitised, dusted with hematite and partially replaced by chlorite or carbonate (S52041). Small amounts of olivine pseudomorpbs may be present ((S48848), (S52032)) and there may be some fresh clinopyroxene (S52058) but this is generally replaced by chloritic mineral. Elongate phenocrysts of orthopyroxene (more than 1.0 mm long), pseudomorphed in chlorite and carbonate, occur in a few rocks ((S48882), (S50114)), in some cases associated with fresh clinopyroxene (S50120). Interstitial patches of quartz are generally present; apatite grains and iron oxide are common accessories.

Typical members of the coarser-grained group of basic porphyrites are the masses at Rossie Priory and Castle Huntly, at Tinkletop, Thrawparts, Linn of Balluderon, Powrie Brae and Duntrune Hill and the small sills in the Fithie Burn [NO 445 345] and in Murroes Burn. The group of intrusions of assorted sizes in the area around Birkhill, Caird Park and Kirkton of Strathmartine are all coarse-grained. Of these, the bodies at Gallow Hill [NO 377 360], Hillhouses [NO 388 358] and South Achray [NO 365 350] contain abundant clinopyroxene, in one case (S53312) as large grains ophitic to the groundmass feldspar, and sufficient olivine to suggest that they are transitional to olivine-dolerites. This is true also of the small body on the north side of the Dighty Water [NO 423 333], near Fintry.

The second main group of basic porphyrites, represented by the bodies at South Kingennie, Cunmont and East Skichen, are significantly finer-grained. They contain scattered phenocrysts of basic plagioclase, commonly corroded, partially sericitised and chloritised, and conspicuously zoned ((S50104), (S50105)). There are generally a few pseudomorphs after olivine and abundant fresh clinopyroxene as small grains and prisms, usually distributed through the rock but in some cases forming aggregates up to 2 mm long (S50102). Small amounts of rhombic pyroxene, pseudomorphed in chloritic mineral (S50103), are generally present. The groundmass consists of labradorite-andesine plagioclase laths and there is usually some partly devitrified glass with microlites of feldspar and rhombic pyroxene (S50101). In several rock slices from the Cunmont mass, this material occurs in thin veins (S50102).

The small boss at Horselaw [NO 355 147], in north Fife, is of the above type but is holocrystalline.

Porphyrites and related rocks

The largest intrusions of andesitic composition are the irregular sills at Dundee Law, Craigie (Baxter Park) and Whitfield. These are composed of fine-grained holocrystalline rock, termed augite-porphyrite. Olivine is absent except in the somewhat more basic mass at Craigie (S48847) which contains a scattering of well-shaped crystals pseudomorphed in serpentinous mineral and calcite. Clinopyroxene is present as abundant granules, small prisms and occasional larger crystals up to 0.5 m long (S45974). Orthopyroxene occurs in appreciable quantities as prismatic crystals, in some cases fresh or only partly replaced by chlorite (S48863). The feldspar laths of the groundmass are mainly andesine-labradorite. There are small amounts of interstitial quartz. Iron oxide, as small octahedra and rods, and apatite needles are present as accessory minerals.

Acid segregations are less common a feature than in the basic porphyrites but have been noted in the case of the Dundee Law body, where the rock is medium-grained, holocrystalline and consists mainly of feldspar in two generations with minor amounts of ferromagnesian minerals, including scraps of fresh clinopyroxene and pseudomorphs after orthopyroxene, and interstitial quartz ((S53327), (S53328)).

The porphyrite of the Gallow Hill sill [NO 413 417] is somewhat altered. It consists of rare phenocrysts in a fluxioned matrix of feldspar laths with a little quartz. All ferromagnesian minerals are replaced by chlorite and carbonate (S52008).

In north Fife, a number of bodies of medium-grained 'doleritic' rock of andesitic composition have been classed as microdiorite, the largest being the dyke-like body on the north side of Forret Hill. In addition to plagioclase in the andesine-labradorite range, commonly with alkali feldspar rims, these contain clinopyroxene and orthopyroxene, both pseudomorphed in serpentinous mineral and iron oxide ((S49233), (S49943)). The small microdiorite [NO 395 195] near Craigsanquhar (S49943) is very similar to the Forret Hill rock. The fresh rock on Kilmaron Hill contains pseudo-morphs after olivine, fresh or partially replaced clinopyroxene and orthopyroxene, commonly intergrown with one another, biotite flakes and interstitial quartz (S51620). The generally similar Kedlock Hill microdiorite is finer-grained (S51606).

The feldsparphyric porphyrite of the irregular mass [NO 321 184], north-west of Parbroath, contains abundant large crystals of orthopyroxene, often completely unaltered, and subordinate amounts of clinopyroxene, set in a turbid, brown glassy matrix (S54307).

Acid intrusions

Most of the rocks in this group are fine or very fine-grained, buff or orange in colour. A few are aphyric and are composed of flow-aligned acid plagioclase laths with some interstitial quartz and patches of chloritic material replacing ferromagnesian minerals (S51632). More commonly the rocks are feldsparphyric, containing phenocrysts of acid plagioclase. These may be sporadic, as in the sill-like bodies of Duntrune House (S50109) and Craighill (S50089) and in most of the suite of small dykes and sills classed as 'felsic alkaline rocks' in the Dunbog to Balmerino area ((S52591), (S52946), (S52947), (S52956)), or abundant, as in the small sill [NO 337 226] near Corbiehill ((S52933), (S52934)(S52935)) and the dyke-like body [NO 213 151] north-west of Raemore farm steading (S54829). Amounts of recognisable ferromagnesian minerals are usually small. The Duntrune House sill contains a few large, partly pseudomorphed clinopyroxene grains (S50121). The well fluxioned Dronley rock [NO 340 365] contains a little orthopyroxene, pseudomorphed in chloritic mineral, and rare flakes of biotite (S54414). The large sill-like body at Ninewells contains a little greenish brown biotite, partially replaced by iron oxide ((S52161), (S52169)). Biotite, rimmed with iron oxide, occurs in some abundance in a sample (S52950) from Emily Hill [NO 3242 1979].. The large rhyolitic felsite at Lucklaw Hill contains rare biotite flakes as well as scattered oligoclase-albite plagioclase phenocrysts in a fine-grained quartz and feldspar groundmass (S48730). Hornblende is associated with biotite in acid porphyrite dykes at Ayton House ((S56065), (S56066), (S56067)), Kedlock Hill (S51608), Ardlogie House (S48734) and north of Foodie Hill (S49930).

A few of the acid intrusions are coarser-grained. The pyroxene-microgranodiorite of Forret Hill is holocrystalline and consists of phenocrysts of zoned labradorite and andesine, pseudomorphs after pyroxene and magnetite, set in an abundant matrix of quartz and alkali feldspar (S50662). There is at least 12 per cent of modal quartz and about 14 per cent of quartz in the norm. A segregation vein (S49946) consists of intergrown sanidine and quartz with tridymitic texture. The small body of felsic alkaline mck south-west of Forret Hill is of similar appearance but is Tess acid. It consists of plagioclase, partly as phenocrysts, orthopyroxene and magnetite set in an abundant matrix of alkali feldspar with some quartz (S51604), (S51651). IBP, MAEB, JIC

Chapter 5 Upper Devonian

Introduction

Early interest in the Upper Devonian rocks (formerly Upper Old Red Sandstone) of the Firth of Tay and Stratheden areas centred on the fossil fish remains first discovered, as reported by Fleming (1831), during 1827 at Drumdryan Quarry [NO 3845 1320] near Cupar, just outwith the present area, and at Clashbenny Quarry [NO 2130 2120], north of the River Tay. Anderson (1837; 1845) subsequently found the remains of similar fish at Dairsie [NO 418 167] in Stratheden, at Birkhill [NO 327 230] and at Parkhill [NO 259 194] by the Firth of Tay. Anderson (1841) added Abernethy [NO 189 160] and Glen Farg [NO 164 153] in lower Strathearn to the list of fossiliferous localities, and most importantly he announced the discovery, by workmen in 1836, of the richly fossiliferous fish bed at Dura Den [NO 416 146], which he later described in detail (Anderson, 1859). More recent descriptions of the vertebrate faunas and discussions of their stratigraphical significance include those by Horne and others (1915), Jarvik (1950), Westoll (1951) and Waterston (in House and others, 1977).

Early accounts of the general geology and stratigraphy of the Upper Devonian of Forfarshire (Angus) and Fife include those by Powrie (1861; 1862), Geikie (1878; 1900; 1902) and MacNair (1908). Some features of the rocks have been described by Craig and Balsillie (in Paton and Millar, 1912), Davidson (in Melville, 1939), Innis (in Mackie, 1939), Walker (1961; 1963) and Ramsay (in Jones, 1968). There is a description of the more famous localities in the area in the excursion guide by MacGregor (1968). The stratigraphy and sedimentology of the Upper Old Red Sandstone in Stratheden and the area adjacent to the Firth of Tay have recently been discussed by Chisholm and Dean (1974) and by Browne (1980a) respectively.

Classification

Chisholm and Dean (1974) described the Upper Old Red Sandstone in the Stratheden area. The sequence was divided into formations with type areas and lithological characters defined. The succession is as follows:

Kinnesswood Formation Mainly sandstone with mudstone bands; rubbly beds of nodular, concretionary dolomite ('cornstone').

Knox Pulpit Formation ( = Kemback Formation) Mainly white or cream-coloured, weakly cemented, fine- to coarse-grained sandstone, commonly closely laminated, with sharp differences of grain-size between adjacent laminae (Chisholm and Dean, 1974, pl. 10); commonly cross-bedded and flat bedded.

Dura Den Formation Mainly fine-and very fine-grained sandstone, ripple-laminated, with thin beds of argillaceous siltstone. In upper part, the sandstones show lamination of Knox Pulpit Formation type and Skolithos-like burrows; in lower part, the sandstones are of Glenvale Formation type. Mudcracks are common.

Glenvale Formation Mainly brown, red, yellow or purple troughcross-bedded sandstones; mudstone clasts are common and also fish scales; argillaceous siltstone bands in places.

Burnside Formation Mainly cross-bedded sandstones with pebbles of 'Highland' origin and, near the base, of Lower Devonian lavas.

The fish faunas obtained from the Dura Den and Glenvale formations prove them to be of Upper Devonian (Famennian) age (Westoll in House and others, 1977). Neither the Knox Pulpit Formation nor the cornstone-bearing Kinnesswood Formation have yielded identifiable fossils, but the latter formation is conformably succeeded in a large part of the Midland Valley of Scotland by grey mudstones and thin beds of dolomitic limestone ('cementstone), considered from their miospore content to be of probable late Tournaisian (Tn3c) age. The Devonian–Carboniferous boundary must therefore lie within the strata of the Knox Pulpit and Kinnesswood formations. Waterston (in Craig, 1965) suggested that some of the strata assigned to the Upper Old Red Sandstone might be of Carboniferous age, and it is possible that the early Tournaisian is represented by some or all of the Kinnesswood Formation, which accordingly is included in the description of the Carboniferous. The Knox Pulpit Formation, on the other hand, shows stronger links with the Upper Devonian, in that sandstones with the characteristic lamination, 'herringbone' cross-stratification and Skolithoslike burrows are intercalated in the upper part of the Dura Den Formation.

Cornstone-bearing sandstones in the area adjacent to the Firth of Tay were assigned to the Kinnesswood Formation by Browne (1980a), who considered the underlying Upper Devonian strata to be indivisible and referred them to the Clashbenny Formation. It would appear that the Dura Den and Knox Pulpit formations are not represented in the Firth of Tay area.

Conditions of deposition

Following earth movements during the Middle Devonian period, the Lower Devonian rocks were uplifted and subjected to severe erosion, particularly on the crest of the Sidlaw Anticline where strata up to 4000 m thick were removed. When sedimentation was resumed in Upper Devonian times, the sandstones of the Burnside, Glenvale and Clashbenny formations were laid down, mainly in the channels of a large river system, which flowed generally eastwards, as shown by cross-bedding measurements (Chisholm and Dean, 1974). Subsequently, the distinctive sandstones of the Knox Pulpit Formation were deposited by currents which flowed mainly towards the west. On the evidence of alternating W–E cross-bedding measurements, possibly of tidal flow origin, Chisholm and Dean interpreted the environment as shallow marine, although fluvial and, in particular, aeolian processes could not be ruled out. The Dura Den Formation was considered by them to have been laid down in the course of the transition from fluviatile to marginal marine or aeolian conditions.

Stratheden

Burnside Formation

The formation consists mainly of cross-stratified sandstone with siliceous pebbles of 'Highland' origin and pebbles of Lower Devonian lavas, as well as mudstone clasts of intraformational origin. The strata vary in colour but in the Cupar area are generally yellow, white or more rarely purple. Where the lower part of the formation is seen in the Fernie Burn, near the Bow of Fife [NO 321 129], the lowest 30 m or so are reddish brown sandstones with beds of conglomerate composed mainly of lava pebbles. Above, conglomeratic beds within the sandstones contain mainly siliceous pebbles. Intact beds of silty mudstone, which are considered to represent overbank sediment, are uncommon. Foreset dip measurements from the Burnside Formation as a whole (Chisholm and Dean, 1974, p. 18) indicate eastward transport. No fossil fish remains have so far been found in the formation. The thickness given for the Burnside Formation by Chisholm and Dean (1974, p. 18) of up to 60 m may be in error if, as Foster and others (1976) suggested, there is a fault within the outcrop of the formation in the Cupar area.

Glenvale Formation

The formation is composed mainly of trough-cross-stratified sandstones, which in the Cupar area are usually white or yellow in colour and contain no siliceous or lava pebbles. Pellets of red or green, silty mudstone are very common, however, and fish fragments have been found at a number of places. Bands of silty mudstone, which are considered to represent overbank deposits, are preserved at the top of some upwards-fining sequences. Foreset dip measurements, as in the Burnside Formation, are towards the east (Chisholm and Dean, 1974, p. 19) and the general sedimentary character indicates deposition in a braided or meandering river system. The thickness of 280 m quoted for the formation by Chisholm and Dean, as in the case of the Burnside Formation, may be an underestimate. The top and base of the Glenvale Formation are nowhere seen in the area. Typical sections of this formation may be seen in the Lydox Mill quarries [NO 418 167] and in a stream section at Balass [NO 392 136].

Dura Den Formation

About 30 m of sediments exposed in Dura Den have long been noted for their abundant fish remains. For this reason and because they are of distinctive facies, Chisholm and Dean (1974, p. 19) placed them in an independent formation, although the unit has not been recognised anywhere else. Much of the formation consists of beds of red, cream and green micaceous siltstones which alternate irregularly with fine-grained cream sandstones. These sediments are mainly flat-bedded but ripple-lamination occurs in the sandstones. Polygonal desiccation cracks are common in the lower part of the unit, in association with thin, red and green clay-partings. In the upper part of the sequence there are coarser bands of flat-bedded sandstone, up to 20 cm thick, which show marked variations in grain size between adjacent laminae. Skalithos-like burrows are also present near the top of the formation. In addition there are thicker cross-stratified sandstone bands at intervals through the sequence. They are of two kinds. The first, which is more common in the lower part of the formation, resembles the clay-pellet bearing, trough-cross-stratified fossiliferous sandstones of the Glenvale Formation. The second kind, which predominates in the upper part of the formation, resembles the sandstones of the Knox Pulpit Formation; it shows marked variation in grain size between laminae and contains few mudstone clasts. Foreset dip measurements in the former indicate eastward transport of the sediment and in the latter westward transport.

Only the upper part of the formation is exposed in Dura Den where in descending sequence the succession is as follows:

The fossiliferous horizons at Dura Den were described by Chisholm and Dean (1974, pp. 19–20) as follows:

'The famous Holoptychius bed (Anderson, 1859, p. 52), which contained exceptionally well preserved fish remains crowded together, lay not far above river level by a house named 'The Laurels' (Horne and others, 1915), at a horizon near the base of the exposed part of the formation. A bed containing abundant Bothriolepis hydrophila (Miles, 1968, pp. 11–12), and here named the Bothriolepis bed, lay about 15 m above the Holoptychius bed (Anderson, 1859, p. 52). More recent finds, about 10 m above the Holoptychius bed, have been reported by Attridge (1956) and by Wattison (1958). Slabs of the Holoptychius bed in the Royal Scottish Museum (No. 1957.1.709) and the Manchester Museum (No. L 10867) consist of poorly bedded sandstone with irregular undulating bedding planes partly coated with greenish, argillaceous material; the lithology closely resembles that of sandstones which form part of the fine-grained facies as exposed, for example, in the river bed [NO 4166 1468] upstream from Yoolfield Mill. The description of the bed made at the time of the most recent excavations (Home and others, 1915, p. 121) mentions the presence of 'sun-cracks' 5 to 10 cm below, and 2 cm above, the fossiliferous layer. There can be little doubt, therefore, that the fossils were preserved in the fine-grained facies rather than in either of the cross-stratified sandstone facies. MAEB, JIC

Knox Pulpit Formation

White and cream coloured sandstones in the Kemback area show many of the characteristic features of the Knox Pulpit Formation of the Lomond Hills. They were referred to the Kemback Formation by Chisholm and Dean (1974, p. 20) but it is now considered that the resemblances between the deposits in the two areas outweigh the differences and it is proposed that the term Kemback Formation be dropped. The formation consists of weakly cemented, yellow, white and buff sandstones, usually cross-stratified but ripple- and flat-bedded in places. Low-angle foresets and reactivation surfaces are present. Marked differences in grain size between adjacent laminae, a feature already described from sandstone beds in the Dura Den Formation, is particularly characteristic of the Knox Pulpit Formation but is less prevalent in the Kemback area than in the type area in the Lomond Hills. Most of the sandstones are very fine- to medium-grained but there are beds of coarse grit. The largest clasts are usually pink felsitic fragments up to 5 mm across but some small ochreous clasts also occur. Mudstone clasts, however, are virtually absent. 'Millet seed' quartz grains are common. At some localities in the Lomond Hills area, foreset dip directions have a bimodal distribution, and may display a herringbone pattern of opposed foresets, but in the Kemback area the direction of transport of the sandstones is consistently westwards. No Skalithos-like biirrows or fish scraps have been reported from the Knox Pulpit Formation in the Kemback area.

Approximately 90 m of sandstone can be ascribed to the Knox Pulpit Formation in the Kemback area. The base of the formation, which is gradational, can be seen in the cliff [NO 4169 1476] opposite Yoolfield Mill, 20 m above the road. Recent unpublished data suggest that at least 60 m of the upper part of the unit is eliminated by the major Dura Den Fault on which Carboniferous strata are thrown down to the south.

Firth of Tay area

Clashbenny Formation

Rocks of Upper Devonian age, contained in a graben-like structure in the area adjacent to the Firth of Tay, cannot be subdivided and have been referred to the Clashbenny Formation (Browne, 1980a, p. 4). They consist mainly of reddish brown, cross-bedded sandstones which, towards the base of the formation, contain pebbles of quartz, quartzite and Lower Devonian lavas. There are a few thin beds of reddish brown, silty mudstone but the overbank sediments have more usually been destroyed and are represented by angular clasts within the sandstones. The base of the Clashbenny Formation was formerly exposed at Quarry Hall [NO 112 184], near Bridge of Earn, where MacNair (1908) recorded conglomeratic sandstone, with abundant quartz and lava pebbles, resting on Lower Devonian lava. At Flisk Wood [NO 328 231] quartz pebble-bearing sandstone of the Clashbenny Formation is exposed in close proximity to lavas of the Ochil Volcanic Formation but the actual contact is not visible. The thickness of the inclined Clashbenny Formation in the Firth of Tay is not known but must be considerable, perhaps from 600 to 900 m.

Fossil fish remains have been found at a number of localities, the most notable being Clashbenny Quarry [NO 2130 2118], where the fauna includes Phyllolepis, Bothriolepis and Holoptychius, indicative of an Upper Devonian (Famennian) age. Fish remains have also been obtained near Balmerino [NO 3555 2463], at Parkhill [NO 2593 1943], in an excavation [NO 131 194] to [NO 133 192] on the line of the M90 motorway, near Moncrieffe House, and at a depth of about 64 m in a borehole [NO 1675 1749] near Culfargie.

In the area south of the Firth of Tay, old quarries at Pitkeathly [NO 1145 1605], [NO 1160 1605] are typical of the bulk of the formation but at Glenearn Quarry [NO 1080 1610] and in the bed of the River Farg [NO 1641 1531] sandstones with quartz pebbles are seen. In the Ballo Burn at Abernethy the sandstones of the Clashbenny Formation may be seen dipping steeply northwards and faulted by the South Tay Fault against Lower Devonian lavas [NO 1890 1582]. Farther north in the Tarduff Burn a small fault appears to bring in the Kinnesswood Formation above [NO 1886 1602].

Bright reddish brown sandstones typical of the Clashbenny Formation were exposed in an excavation [NO 1325 1934] on the line of the M90 motorway, south of Perth. In the Carse of Gowrie area, the sandstones were worked in quarries at Wester Ballindean [NO 258 294] and at Inchture [NO 2790 2865], now infilled, as well as in the presently flooded Clashbenny Quarry. The first of these was reopened to provide stone for the reconstruction of the boundary wall of Rossie Priory grounds, which was displaced during the re-alignment of the Perth–Dundee (A85) trunk road. Bright red sandstones with beds containing quartz pebbles are exposed a little to the south of the North Tay Fault in the Baledgarno Burn [NO 275 204], south-west of Castlehill, and in the Rossie Burn [NO 292 208] east of Rossie Church. Farther north-east, red sandstones with thin beds of conglomerate with quartz and quartzite pebbles are exposed in Balruddery Den [NO 3165 3220], Fowlis Den [NO 328 375] and Gray Den [NO 3320 2762].

In addition to the surface exposures the sandstones of the Clashbenny Formation were penetrated in seven boreholes drilled by IGS in the Carse of Gowrie area (Browne, 1980a).

Arbroath area

East of Arbroath, in the general neighbourhood of Whiting Ness, a sequence of mainly red-brown and yellow conglomerates with subordinate sandstone beds and basal and marginal breccias rests with striking unconformity (Plate 15) on the Lower Devonian Arbroath Sandstone (Hickling, 1908; 1912). No fossils have been recorded from the rocks above the unconformity but because the sandstones bear a general resemblance to those in the fossiliferous Upper Devonian sequences in Stratheden and the Carse of Gowrie an Upper Devonian age seems probable. The absence of cornstone, suggesting a stratigraphical position below that of the Kinnesswood Formation of Fife, is consistent with this conclusion.

The angular discordance between the Upper and Lower Devonian at Whiting Ness is marked. The Lower Devonian dips to the SE at about 25° whereas the Upper Devonian is inclined approximately ESE at 10°. Upper Devonian sediments were deposited against steep slopes forming part of the sub-Upper Devonian land-surface, as can be seen in the cliffs [NO 6510 4100] on the east side of the Horse Shoe east of Whiting Ness. It is probable, however, that these observed irregularities of the unconformity in the immediate neighbourhood of Whiting Ness constitute only a small part of the palaeo-relief in this area. The western limit of the Upper Devonian outcrop on the foreshore [NO 6510 4100] 900 m W of Whiting Ness is also an unconformable junction, and an ascending sequence broken by only small faults can be followed on the intervening shore. On the assumption that the Upper Devonian stratification was originally horizontal and on the basis of the prevailing dip (10° to the ESE), the breadth of outcrop perpendicular to the strike in the ground between the two emergences of the unconformity, about 450 m, implies about 100 m of relief on the ancient land-surface.

Most of the Upper Devonian sediment appears to have accumulated in channels and bars in the active part of an alluvial plain covered with sand and gravel. White, flat-bedded, fine- to medium-grained sandstone of floodplain type was laid down preferentially in areas close to the steeper slopes on the surface of unconformity near Whiting Ness (Ramos and Friend, 1982, pp. 313–314). A general direction of transport towards the south-east is apparent (Ramos and Friend, 1982, p. 307). Conspicuous breccias, composed of angular fragments of Arbroath Sandstone, and clearly derived from the ancient bedrock slopes, accumulated at the bases of these features as lenticular deposits. In places the breccias rest on the unconformity but elsewhere they occur at a somewhat higher level and are intercalated within, or occupy channels cut in, Upper Devonian sandstone. In the area of the Steeple Rock [NO 6585 4095] south-south-west of Whiting Ness, blocks of Arbroath Sandstone up to 2.5 m in length occur immediately above the unconformity. Ramos and Friend (1982, p. 307) have deduced a south-westerly direction of transport for the breccias. This relates to the slope of the ancient land-surface against which impinged the south-easterly flowing rivers of the contemporaneous alluvial plain.

Chapter 6 Lower Carboniferous

Introduction

The interesting occurrence of Carboniferous strata in the Firth of Tay area was commented on by Geikie (1900, pp. 26, 39). He was referring to the presence of cementstones near Dron, first recorded by Anderson (1845), and Geikie correlated these beds with the cementstones of Ballagan Glen near Glasgow. Underlying these cementstones in the Firth of Tay area, with apparent conformity, are red, white and purple sandstones with beds of concretionary limestone ('cornstone') which were recorded by Anderson (1837) and Grierson (1845). Browne (1980a, pp. 4–11), using successions proved by boreholes, correlated the underlying sandstones with the Kinnesswood Formation of Fife, which he considered to be, in part at least, of Carboniferous age; the cementstones were placed in a new division, the Ballagan Formation.

Kinnesswood Formation

Strata typical of the Kinnesswood Formation were formerly worked in quarries, now overgrown, near Jock's Lodge [NO 2557 1916] and Clunie Bleachfield [NO 221 176] (Anderson, 1837; 1841; 1845) and at Murie Quarry [NO 2333 2260], near Errol (Grierson, 1845), where grey, purple and reddish brown, cross-bedded sandstones, some 6 m thick, with irregular veins and nodules of concretionary limestone are still visible. A similar sequence is exposed in a stream section [NO 1886 1602] near Abernethy, and at least two beds of concretionary limestone were quarried [NO 129 161] in the neighbourhood of West Dron. In addition, rocks of the formation were penetrated in the East Dron, Beautyfield and Mains of Errol boreholes (Figure 9) drilled in 1971–72 by IGS to investigate the relationship of the Devonian and Carboniferous rocks in the area (Browne, 1980a).

As shown by the boreholes, the Kinnesswood Formation consists of white, grey, purple, reddish brown and greenish grey, cross-stratified, channel-sandstones with subordinate beds of reddish brown and greenish grey silty mudstone, disposed in upwards-fining, fluvial cycles. Nodules and veins of concretionary limestone, in some cases coalescing to form beds up to a greatest recorded thickness of 2.58 m, have developed in the sandstones and mudstones as a result of pedogenic processes during periods of reduced sedimentation (Allen, 1960; 1974). As a result of penecontemporaneous erosion, angular clasts of cornstones and of over-bank mudstones are commonly incorporated in the basal part of many of the channel-sandstones, which are generally less than 0.6 m thick. The sandstones are composed mainly of subangular to well rounded grains of quartz and quartzite with subordinate feldspar, which decreases in amount in the upper part of the formation.

Staining techniques and refractive index determinations show that the cornstone nodules penetrated in the East Dron Borehole consist predominantly of calcite and dolomite (Browne, 1980a). Smaller amounts of iron-rich and iron-poor calcite are also present, but these tend to be more common in the coarser-grained carbonate matrix which encloses the nodules. The carbonate in the nodules generally occurs as a microcrystalline mosaic, commonly with a brecciated or 'clotted' appearance due to variations of grain size. Veins of carbonate, in some cases dolomite, commonly cut the nodules and interstitial patches of unaltered sediment. In the upper parts of cornstone beds, where replacement by carbonate is almost complete, the original sediment may be represented only by thin irregular veins and patches of illite and chlorite, with a little calcite, dolomite and quartz.

The Mains of Errol Borehole (Browne, 1980a) is considered to have penetrated the full thickness of the Kinnesswood Formation, (Figure 9) between depths of 45.5 and 78.0 m. The junctions with the Clashbenny Formation below and the Ballagan Formation above appear to be transitional, with no representative of the Dura Den or Knox Pulpit formations of Fife. No fossils of any kind have been recovered from the Kinnesswood Formation at any locality in the district.

Ballagan Formation

Anderson (1845) first reported the existence of grey mudstones and cementstones near Dron and Carboniferous plant fragments, shells and fish remains were recorded from the sediments by Geikie (1900). Thomson (1903) compared the strata with those in Ballagan Glen and MacNair (1908) illustrated the section formerly exposed in a railway cutting near East Dron. These beds were placed in the Ballagan Formation by Browne (1980a, p. 7).

Strata of the Ballagan Formation are now exposed only in the Dron Burn and its tributary [NO 137 153] to [NO 141 158], near Dron, and in Grant Wood [NO 1357 1598]. During 1971–72, a borehole was sunk by IGS close to East Dron in order to obtain a complete sequence through the Lower Carboniferous rocks and to examine their relationship with the Upper Devonian. Further boreholes were sited at Newburgh and near New Farm, in a successful attempt to confirm the presence in these areas of Carboniferous strata suspected from the records of trial boreholes for water. Of other boreholes drilled in order to delimit the outcrops of the Ballagan Formation, only those at Mains of Errol and West Dron penetrated Carboniferous strata.

On the evidence of the succession found in the boreholes, the Ballagan Formation has been subdivided into three units (Figure 10), namely, in ascending sequence, the Mains of Errol Member, the Dron Member and the Newburgh Member (Browne, 1980a).

Mains of Errol Member

Strata between depths of 20.6 and 45.5 m in the Mains of Errol Borehole and between 193.0 and 224.8 m in the East Dron Borehole, consisting of reddish brown, purplish grey and greyish purple silty mudstones and siltstones, have been assigned to the Mains of Errol Member. The mudstones are usually unbedded but the siltstones commonly show ripple-lamination. There are a few thin nodular beds of cementstone, in some cases showing traces of lamination which have generally been disrupted by desiccation cracks. The mudstone beds may also contain numerous, sub-spherical nodules of dolomite, less than 2 mm in diameter. At several horizons, compact dolomitic mudstones with a brecciated appearance which is emphasised by a colour-mottling in shades of reddish brown and purplish grey, may represent incipient soil profiles (cf. Francis and others, 1970, p. 120, pl. iv, fig. 7). Gypsum nodules and veins occur at numerous levels in the East Dron Borehole but are less common at Mains of Errol.

Ostracods, Spirorbis, and fish remains have been found at several horizons in the Mains of Errol Member, but modiolids were recovered only in the Mains of Errol Borehole at a depth of 32.70 m. Spores indicative of the CM Zone (Neves and others, 1973) were recovered from a single sample at a depth of 29.3 m in the Mains of Errol Borehole.

Dron Member

Strata between depths of 42.0 and 193.0 m in the East Dron Borehole are referred to the Dron Member, as are all the rocks penetrated by the New Farm A Borehole and by the West Dron Borehole below 9.50 m (Figure 10). The Dron Member is therefore about 151 m thick. The sediments consist mainly of grey mudstones and silty mudstones with numerous nodules and beds of dolomitic limestone (cementstone) and a few beds of fine-grained, usually ripple-laminated, grey sandstone. Below 137 m in the East Dron Borehole, the cementstone beds are thicker than average (up to 1.64 m) and rather widely spaced. More generally, the cementstones are from 0.15 to 0.30 m thick and are usually unbedded although there are traces of lamination in a few cases. They consist usually of microcrystalline ferroan dolomite, sometimes with a turbid appearance due to slight variation of grain size. Some beds are composed mainly of calcite.

The silty mudstones are usually bedded, the lamination in numerous instances being disrupted by mudcracks, which are infilled by coarser-grained sediment, commonly with ostracods. The mud-clasts produced by the sun-cracking are usually flat but some are concave. Pseudomorphs after salt occur at a number of levels. Nodules and veins of pink and white gypsum are present at numerous levels throughout the Dron Member in the East Dron Borehole. The nodules tend to be pink and to post-date the veins in most cases. At similar stratigraphical levels in the West Dron Borehole, the Dron Member contains little or no gypsum. It is possible that the gypsum at West Dron was mobilised during movements of the very large South Tay Fault, a short distance to the south.

A rich but restricted fauna, consisting of ostracods, estheriids, Spirorbis, Modiolus latus (Portlock), Sanguinolites sp. and fish scraps, is present at numerous levels in the Dron Member but is of little diagnostic value. Miospore assemblages, however, characteristic of the CM Zone (Neves and others, 1973) and probably indicative of an Upper Tournaisian (Tn3c) age, were recovered from a number of samples from the East Dron, West Dron and New Farm A boreholes.

Newburgh Member

Strata cut by the Newburgh B Borehole are assigned to the Newburgh Member. Its thickness is unknown but exceeds 50 m. The beds consist for the most part of grey mudstones and siltstones with beds up to 2.5 m thick of fine-grained, in some cases, rooty sandstones. Nodules and beds of cementstone are rare. Strata above a depth of about 42.0 m in the East Dron Borehole and above 9.50 m in the West Dron Borehole are also referred to the Newburgh Member (Figure 10). Gypsum nodules occur in the member only in the East Dron Borehole.

The fauna recovered from the Newburgh Member is similar to that in the Dron Member. Samples from the West Dron and Newburgh boreholes yielded CM Zone miospore assemblages but, in addition, a single sample at a depth of 30.31 m at Newburgh contained miospores typical of the overlying Pu Zone.

Conditions of deposition

The sediments of the Ballagan Formation are considered to have been laid down in extremely shallow water, coastal sabkha-like conditions, established when subsidence permitted the sea to encroach westwards on to the broad alluvial plains of the large river system which had deposited the underlying Kinnesswood Formation. The area was subject to periodic drying out as is shown by the prevalence of mud-cracks and the abundance within the deposit of the evaporite minerals gypsum, dolomite and salt. During deposition of the Mains of Errol Member, emergence for more prolonged periods is indicated by the occurrence of incipient soil profiles. The salinity of the shallow marine waters probably varied considerably and only the most tolerant of organisms could survive. The presence of sandstones with rooty horizons in the Newburgh Member suggests transition to the deltaic conditions which are characteristic of the Carboniferous as a whole in the Midland Valley of Scotland.

Chapter 7 Late-Carboniferous dykes

Introduction

The extensive swarm of W–E and WSW–ENE quartzdolerite and tholeiite dykes which crosses the Midland Valley and the southern Highlands of Scotland is well represented in the western part of the district. The dykes are described in definitive works by Walker (1934; 1935) who includes a comprehensive list of previous work in his1935 paper. Subsequently, Francis and others (1970) and Forsyth and Chisholm (1977) have described equivalent rocks from adjoining areas and the geochemistry of the whole suite has been investigated by Walker (1965) and Macdonald and others (1981). The area falls within the 'North Dyke Swarm' of Walker (1935) which has a general ENE trend, but in detail many of the dykes trend W–E, particularly in the southern half of the area (Figure 11). Individual dykes may be traced westwards for up to 130 km into the SW Highlands, and the swarm is further extended west-northwest to the southern Outer Hebrides. Eastwards the swarm may be traced beneath Permian and younger rocks of the North Sea by magnetic and seismic methods, which have revealed many broad WSW-ENE dykes, extending at least as far as the Central Graben (Russell and Smythe, 1983). The swarm thus defines an arcuate fracture system, which may be continued through the Oslo Graben of southern Norway, and which has been attributed to extensional stress associated with lithospheric separation in a proto-North Atlantic rift zone north-west of the British Isles (Russell and Smythe, 1983).

Field relationships

The dykes are almost always near vertical and commonly exceed 10 m in width, with a maximum recorded width of 35 m in St Magdalene's Quarry, Perth [NO 115 210]. Thin chilled margins are usually present and contact metamorphism of country rocks is commonly observed. In sandstones the metamorphism is generally shown by a hardening and change of colour up to 30 cm from the dyke. In mudstones of the Cromlix Formation the characteristic purple-brown colour is changed to grey for up to 1 m from the contact, as is conspicuously displayed around the dykes of Campsie Linn [NO 125 340]. Polygonal cross-joints are commonly developed perpendicular to the dyke margins and are well seen in Crossgates Quarry [NO 046 207]. Late-stage acid segregations, of coarser grain size than the host dolerite, occur in some dykes and thin basaltic veins are present in places (eg. immediately east of Campsie Linn).

Although some dykes follow fault lines for short distances (eg. south-west of Gauldry, Pitroddie Den and in the Highland Boundary Fault-zone), the majority are discordant to the principal faults. Field mapping, supplemented in some areas by flux-gate magnetometer traverses, reveals that the dykes are liable to abrupt lateral shifts of course, giving rise to dyke-echelons, often consisting of segments less than 500 m in length (Figure 11). Larger segments appear to be continuous for up to 7 km. Minor lateral shifts, of magnitude less than the dyke width, are commonly displayed in quarry sections and in upstanding natural outcrops, eg. Corsiehill [NO 134 234], Craigmakerran [NO 142 322] and Campsie Linn (Plate 19). In some cases the shifts affect only one wall of the dyke, or the shifts on opposite walls are offset, leading to considerable variations in dyke width (eg. quarries at [NO 073 244] where the dyke-width changes from 13 to 26 m).

The extensive complex of quartz-dolerite sills in east-central Scotland (Francis, 1982), which is genetically related to the dykes, extends north-eastwards throughout Fife towards St Andrews (Balsillie, 1922; Walker and Irving, 1928). The north-eastern termination occurs south-east of Kemback on the southern margin of Sheet 49 as a complicated set of intrusions of variable shape and size. The intrusions, which are the only late-Carboniferous sills within the district, are described in detail in the east Fife memoir (Forsyth and Chisholm, 1977).

Age

The quartz-dolerite and tholeiite dykes are intruded into all the principal formations of the district so that their age may only be inferred from field relationships of the swarm in other areas (Walker, 1935; Forsyth and Chisholm, 1977; Cameron and Stephenson, 1985). Much of this information comes from east Fife where quartz-dolerite dykes cut Westphalian B strata and where a variety of inconclusive evidence for an upper age limit is obtained from relationships of quartz-dolerites with late-Carboniferous or early-Permian volcanic vents and plugs. Better field evidence from the western Highlands and northern England supports a late-Carboniferous or early-Permian age, and K-Ar whole rock dates of 295 to 290 Ma indicate a late-Westphalian to early Stephanian age which is now generally accepted (Fitch and others, 1970; de Souza, 1979).

Details of distribution

The distribution of dykes within the area is described from north to south and from west to east in the following section.

A few dykes of quartz-dolerite occur within the Highland Boundary Fault-zone, where their trend is deflected from ENE to NE on entering the fault-zone (eg. on the north side of the Hillocks of Gourdie). Immediately to the south of this zone, a group of mainly short (0.5 to 1.5 km) quartz-dolerite and tholeiite dykes have a W–E trend around Waterloo and Murthly, but two longer dykes of this group swing into a more normal ENE trend between Cargill and Coupar Angus. The area between here and Perth is cut by many dykes of both quartz-dolerite and tholeiite, with trends which vary from ENE in the north to almost W–E in the south around Dalcrue and Almondbank. This group continues in a general ENE direction, with diminishing intensity, as far as the eastern edge of Sheet 48E (eg. Lumley Den, [NO 402 417)). Many of the dykes can be traced for 2 to 7 km, although there are also several short lengths. A particularly good example of en-echelon dyke segments, 3 to 5 km in length, extends for a total of 21 km through Bankfoot and Coupar Angus at the northern edge of this group. Many of the localities described by Walker (1934) as typical of his tholeiite types are within this group: dykes at Campsie Linn (NE dyke) [NO 125 340], Pitroddie Quarry [NO 202 253], Newmiln [NO 128 304] and Corsiehill Quarry [NO 134 234] are described as 'Corsiehill Type'; and the dyke at Craigmakerran [NO 142 322] and a 4 m-wide ENE dyke at Campsie Linn are of 'Craigmakerran Type'. The wide W–E dyke at Campsie Linn (Plate 19), which can be traced eastwards through Wolfhill Farm, is a quartz-dolerite. To the east-north-east of Perth, short dyke segments, mostly of tholeiite, occur in the Pitroddie Fault and close to the North Tay Fault at Kinnaird. Other en-echelon tholeiites occur north-west of Dundee between Balruddery and Birkhill.

South of the Tay Estuary and south of Perth, most of the dykes have a W–E trend. They are fewer in number than in the area to the north, but most of them are wide and may be traced for long distances, either as near-continuous dykes or as interrupted lengths, often with en-echelon offsets. Four large dykes and several smaller ones crop out intermittently from the western edge of Sheet 48W to Friarton Hill, Perth. These dykes form the eastern end of a group which may be traced almost continuously westwards to Loch Fyne. Good exposures are found in quarries on Friarton Hill, especially in St Magdalene's Quarry [NO 115 210] which shows continuous exposure across a 35 m-wide quartz-dolerite with a tholeiite margin and a 50 cm-wide chill against basalt lava (ED1690), (ED1691), (ED1692), (ED1693), (ED1694), (ED1695), (ED 1696). Farther east are four segmented dykes, three tholeiite and one quartz-dolerite, which have been traced, partly by magnetometer, almost to the east coast. The northern dyke of this group is the type example for Walker's (1935) 'Newton Hill Type' tholeiite. The Glenduckie Hill–Luthrie–Forret Hill tholeiite exhibits changes in texture along its length and is more like a quartz-dolerite at its western end. The most southerly dyke-echelon consists of several segments of quartz-dolerite and tholeiitic andesite, up to 14 km long and 15 to 30 m wide, extending from the southern end of Glen Farg (Sheet 40) to Kemback and Brownhills, near St Andrews. The eastern segments are described in detail in the east Fife memoir (Forsyth and Chisholm, 1977).

Petrography

The suite has been divided on textural grounds into quartzdolerites and tholeiites and Walker (1934; 1935) further subdivided the tholeiites into several named types based mainly on the proportion and nature of the glassy mesostasis. Current investigations show that a spectrum of textures exists from quartz-dolerite through the various types of tholeiite and it is often difficult to fit individual rocks precisely into Walker's classification. There is no geographical distribution pattern, which is contrary to the belief of Walker (1934; 1935) that tholeiites predominate north of Perth. Both quartz-dolerites and tholeiites can occur as long persistent dykes though there is a slight tendency for tholeiites to occur as thinner dykes, and also as marginal facies of thicker quartz-dolerite intrusions. This was recognised by Walker (1935, pp. 144–145) and by Francis and others (1970, pp. 236–237) who also reported changes along the length of a dyke through various tholeiite types to quartz-dolerite. Lateral variations are also observed in some of the dykes of the present district.

The quartz-dolerites and tholeiites consist essentially of plagioclase, augite, pseudomorphs after olivine or hypersthene, occasional pigeonite, Fe-Ti oxides and a quartzo-feldspathic or glassy residuum with accessory apatite, pyrite and secondary amphibole and biotite (Plate 20). Chlorophaeite may be present locally. The tholeiitic nature of the magma is well illustrated by the presence of two pyroxenes, the unstable nature of the olivine, where present, and the siliceous nature of the mesostasis. The distinction between quartz-dolerite and tholeiite may be made on the following criteria, not all of which may be evident in any one sample:

1. The quartz-dolerite mesostasis consists of small, intersertal, crystalline areas containing fine, micropegmatitic intergrowths of quartz and alkali feldspar ((Plate 20).1). The tholeiite mesostasis consists of glass, often microlitic and usually devitrified, which is generally slightly more abundant than in the quartz-dolerites ((Plate 20).2 and 20.3).
2. Augite occurs as large, ophitic to subophitic crystals in the quartz-dolerites but is generally smaller and granular to subophitic in the tholeiites.
3. Fe-Ti oxides generally occur as equidimensional crystals in the quartz-dolerites suggesting a single phase. In the tholeiites both squarish magnetite and long skeletal ilmenite crystals are visible.
4. Olivine pseudomorphs are rare or absent from quartz-dolerites but are more common in the tholeiites and are particularly abundant in dyke margins. Either pseudomorphs after hypersthene or fresh pigeonite are normally present in quartz-dolerites but are rarely found in tholeiites.

Many of the tholeiite dykes are texturally similar to the quartz-dolerites and grade into the latter. These tholeiites are relatively coarse grained with subophitic augite and only small areas of glassy mesostasis (eg. (S52591), (S56974), (S56987), (S58331): (ED1677), (ED1679). Such dykes probably equate with the 'Corsiehill Type' of Walker (1934; 1935, pl. 5, fig. 3). A more distinctive group of tholeiites is characterised by a greater proportion of interstitial microlitic glass which is crowded with a very fine, reticulate mesh of ilmenite needles. Similar glass also occurs in perfect spherical ocelli enclosed by small, curved, tangential laths of plagioclase ((Plate 20).3). Good examples of this type are found in the Newton Hill and Luthrie dykes of Fife ((S49280), (S51602), (S51624), (S51626)), representing Walker's 'Newton Hill Type' (1935, pl. 5, fig. 5), which he equated with the 'Craigmakerran Type' (Walker, 1934) ((S56983), (S56986), (S56988)). Almost all tholeiite dykes of the district fall within the range of these types apart from a few in which the plagioclase and augite show slight porphyritic tendencies ((S51646), (S60577)). The quartz-dolerite dykes are remarkably uniform in texture apart from slight variations in grain size. Good examples are to be found in slices (S8934), (S57364), (S60567), (S60568), (S60760), (S68336) and (ED1691).

Chilled rocks at dyke margins and in cross-cutting veins develop a microporphyritic, basaltic texture as the proportion of mesostasis increases. Euhedral microphenocrysts of plagioclase, augite and pseudomorphs after olivine or rarely hypersthene are set in either a glassy mesostasis or a very fine, holocrystalline aggregate of plagioclase, augite and magnetite ((S68330), (ED1696)).

The close spatial association between quartz-dolerites and the various tholeiite types, occasionally within the same dyke, indicates a genetic connection. This is supported by the petrographical evidence in that all types have a very similar mineralogy in which similarities in mineral relationships outnumber the subtle differences. Macdonald and others (1981) do not differentiate between quartz-dolerites and tholeiites on geochemical grounds and regard almost all as having crystallised from a quartz-tholeiite magma. Textural differences were most likely induced by conditions of crystallisation, such that the glassy, quenched textures of the tholeiites contrast strongly with the crystalline, intersertal intergrowths of the quartz-dolerites, which suggest slower cooling in the larger intrusions, possibly under the influence of trapped volatiles.

Macdonald and others (1981) quote many analyses from the present district (op. cit., analyses 19–35, 70–73), including most of the type localities of Walker (1934; 1935) and some of the other better-known dykes (Broadgreen, Campsie Linn–Wolthill, Craigmakerran, Dalcrue, Cross-gates, Friarton, Corsiehill, Newton Hill, Luthrie and Kemback). Individual dykes reveal only slight chemical variation along their length, eg. the Loch Fyne–Perth (Friarton) dyke, although fractionation is recorded between margin and core of some thicker dykes (eg. the Craigmakerran dyke). The trace element chemical variation within individual dykes is much less than that observed between dykes so that geochemical 'fingerprinting' is possible in some cases (eg. the Campsie Linn–Wolfhill dyke).

Macdonald and others (1981) show that the magmas were of 'High Fe-Ti' tholeiite type such as tend to occur in areas of active lithospheric spreading and that they are typical of continental tholeiite suites in terms of rare-earth and other element contents. There is no evidence of crustal contamination and all the geochemical features can be related to a mantle source. Between-dyke variations in certain trace element ratios (eg. Zr/Nb) suggest that the mantle source was inhomogeneous and that the dykes represent a 'plexus of small, partly independent magma reservoirs' rather than a single regional source. Low-pressure fractionation of olivine, plagioclase, pyroxene and oxides has resulted in a few dykes of tholeiitic andesite composition (eg. the Kemback dyke). More evolved compositions occur only as aplite veins and patches in the thicker sills of the Midland Valley and hence are rarely seen in the present district. However, analyses of glassy groundmass separated from a tholeiite dyke (Walker, 1935) demonstrate the high concentrations of SiO_2, K₂O and volatiles in the residuum. The presence of high Fe and Ti in residual liquids until relatively late in the crystallisation process is indicated by the abundance of ilmenite in intersertal glass and also by late-stage chlorophaeite in some rocks.

Details

Plagioclase (37 to 55 per cent) occurs as randomly-orientated, euhedral laths ranging from 0.5 to 2.5 mm in length. The laths are often tightly packed, but where space permits they commonly form poorly developed stellate clusters. Crystals usually show normal zoning with compositions within the labradorite range although Walker (1935) records more basic compositions of up to An₇₀ in some tholeiites and outer zones are commonly of oligoclase (An₂₀). Albitisation and alteration to sericite is observed in some rocks, but sieve-texture replacement is rare.

The earlier-formed mafic minerals, olivine and/or hypersthene (3 to 8 per cent) are always heavily altered to carbonate, serpentine, bowlingite and chloritic aggregates and are represented by pseudomorphs, commonly rimmed by magnetite. Subhedral crystals are common, enabling tentative distinction to be made between possible hypersthene (rectangular) and olivine (lozenge-shaped or rounded, oval with cross-fractures). Fresh hypersthene was detected in only two samples of quartz-dolerite (S8934) and (ED1691). Pigeonite is sparsely distributed but is relatively abundant in a few very fresh samples (eg. (S68336)). It occurs as long, narrow, colourless crystals, commonly with a median twin plane, which occur separately and as cores to augite crystals or possibly as rims on hypersthene. Augite (25 to 35 per cent) varies from granular clusters to ophitic plates, only locally having euhedral tendencies (eg. (S61919)). It is usually post-plagioclase in crystallisation, though the two phases exhibit near-coprecipitation in some tholeiites. Most samples contain fresh, pale brown, non-pleochroic augite, which commonly exhibits twinning (100 and sector twins) and normal zoning. It is commonly the last phase to alter, usually to carbonate. Iron-titanium oxides (6 to 11 per cent) vary greatly in habit, but in most of the tholeiites two distinct phases may be identified, both of early crystallisation. Magnetite occurs as small, equidimensional grains with subcubic tendencies, whereas ilmenite forms larger, skeletal crystals or a reticulate mesh which may be partly altered to leucoxene.

The amount of mesostasis varies from 6 per cent in some 'Corsiehill Type' tholeiites to 20 per cent in the 'Newton Hill' tholeiite with intermediate values of 8 to 12 per cent in quartzdolerites. There is a progression from slightly turbid brownish glass, through deeper brown, devitrified glass and microcrystalline quartzo-feldspathic aggregates to the intergrowths of quartz and alkali-feldspar in the quartz-dolerites. In addition to micropegmatite, quartz and alkali-feldspar also occur as pinkish spherulites of delicate, fibrous cryptopegmatite. Relatively abundant additional areas of primary quartz can be quite coarse-grained. The glass can be quite homogeneous with only small specks of iron oxide, but more commonly it is microlitic with small crystallites of plagioclase, augite and ilmenite in varying proportions. In many rocks the glass is altered to a pale green, chloritic aggregate, and intersertal chlorophaeite may be present.

Apatite is the most common accessory mineral, occurring as small prisms included by other phases. It is conspicuous in most quartz-dolerites, but is rarely detected in tholeiites. A small amount of pyrite is usually present. Hornblende commonly occurs as thin reaction rims on augite and sporadic fringes of biotite occur around magnetite. Very thin needles of rutile are common as inclusions within quartz, both in the mesostasis and in vesicles.

Irregular vesicles are filled by intergrowths of chlorite, shadowy quartz and calcite. Some vugs contain prismatic quartz (S56984). In addition to, or in place of these vesicles, many rocks contain very distinctive, perfectly spherical 'ocelli', which have sharp margins, commonly rimmed by small, tangential or curved plagioclase laths ((Plate 20).3) (S51602). These are particularly common in tholeiites having a relatively high proportion of glass but are also occasionally found in other tholeiites and quartz-dolerites. The ocelli are almost entirely filled with material identical to the intersertal groundmass of the rock but many contain a core of quartz and/or calcite. It has been suggested that they originated as 'steam cavities' which were subsequently filled by the still-liquid residuum of the magma and by later secondary minerals (Flett in Peach and others, 1910). DS

Chapter 8 Structure

Introduction

The rocks within the district have undergone three principal episodes of earth movement, namely: Lower Palaeozoic deformation and metamorphism of the Dalradian rocks; folding and faulting of the Lower Devonian rocks during the Middle Devonian, accompanied by major overthrusting along the Highland Boundary Fault-zone; folding and faulting of the Upper Devonian and Carboniferous strata. Each episode was followed by a period of uplift during which the raised terrain was eroded. Both the Lower and Upper Devonian strata rest unconformably on the eroded surfaces of older rocks.

The principal structural features of the district are shown on (Figure 12). The multiphase deformation and metamorphism affecting the Dalradian rocks have already been described in Chapter 2.

The principal folds affecting the Lower Devonian strata, the Strathmore Syncline and the Sidlaw Anticline, have a NE trend parallel to the Highland Boundary Fault-zone. This zone constitutes the present approximate north-west boundary to the Lower Devonian outcrops. It also probably formed the contemporaneous limit of the area of thickest Lower Devonian sedimentation south of the Highlands. The structures imply shortening of the crust in a NW–SE direction and this is manifest in overthrusting on steep dislocations in the fault-zone (Ramsay, 1964). The function of the Highland Boundary Fault-zone as a basin boundary suggests the probability of contemporaneous faulting on this line during the Lower Devonian, but of this there is no direct evidence. Furthermore the Lower Devonian sequence appears particularly poor in penecontemporaneous deformation structures which might be ascribed to seismic activity. The steep vertical limb of the Strathmore Syncline close to the Highland Boundary Fault-zone is known to involve the youngest strata of the Strathmore Group in the Edzell area (Armstrong and Paterson, 1970). This demonstrates that whether or not deformation of the basin began during the Lower Devonian, the principal folding took place after the deposition of the Strathmore Group which is, in part, of Emsian age (Westoll in House and others, 1977).

There is some evidence (p. 63) for intra-Upper Devonian faulting on the coast north of Arbroath, but apart from this, there is no reason to think that sedimentation of Famennian to Dinantian age was broken by any major earth-movement, and the large faults which affect the Upper Devonian are probably of post-Dinantian age. Post-Dinantian faults dominate the structure of the Carse of Gowrie and Firth of Tay areas, where Upper Devonian and Carboniferous strata have subsided into the Tay Graben, and they also occur in north Fife. The period of activity on these faults was probably similar to that on the Ochil Fault, a structure which was active over a considerable part of the Carboniferous period (Francis and others, 1970, p. 247). The Tay Graben is a somewhat complex structure compounded of faults of W–E and NE–SW trend. The generally smaller faults of NW–SE trend which are common in the volcanic rocks of the Ochil and Sidlaw hills adjacent to the graben may be earlier and related possibly to the Middle Devonian folding, but it is not improbable that they were reactivated and modified while the boundary faults of the graben were active.

The youngest structural event identifiable in the rocks of the district is the general late-Carboniferous N–S extension which opened the tension fractures occupied by the W–E quartz-dolerite dykes.

Highland Boundary Fault-Zone

The Highland Boundary Fault-zone has a long and complex history, and the existing fault-belt, across which the Lower Devonian and Dalradian rocks are juxtaposed near Dunkeld, was probably developed on a pre-existing structural line. It may be inferred moreover, from the great thickness of the Lower Devonian sequence in the Strathmore Syncline, that penecontemporaneous faulting must have taken place along the Highland Border during deposition.

It is presumed that the dominant dislocations of NE- SW trend are thrust-faults and there is some local evidence at Birnam (Harris, 1972) which suggests that the fault planes dip to the north-west although overall downthrow is to the south-east. The nature and position of the presumed earlier penecontemporaneous faulting is unknown and these structures may have been overlapped and concealed in the course of sedimentation.

A principal fracture in the fault-zone near Stenton [NO 066 405], the Thornton Fault, has long been identified as the 'Highland Boundary Fault' (Allan, 1928; Anderson, 1947). This dislocation brings together different horizons within the Arbuthnott Group. The Stralochy Fault, somewhat smaller, lies a little to the south-east. The largest fracture, the Spittalfield Fault, is even farther to the southeast. Its existence close to the axis of the Strathmore Syncline, between Waterloo and Spittalfield, may be deduced from the fact that strata high in the Strathmore Group near Murthly, south of the axis, appear to be in juxtaposition with strata of the Garvock Group near Caputh.

Strathmore Syncline and Sidlaw Anticline

The Strathmore Syncline is a markedly asymmetrical structure with a steep NW limb dipping away from the Highland Border in the Berry Hill-Caputh area. North-west of Caputh, volcanic conglomerate of the Arbuthnott Group on the steep limb dips to the south-east at up to 75°. In the Pass of Birnam, dips are up to 58° while north of Spittalfield dips of up to 60° have been recorded. The steepest dips in the Garry Burn north-east of Berry Hill are, however, no more than 35°. The axial zone of the syncline, apart from an area of variable dips near Tullybelton, is not well exposed.

The SE limb of the Strathmore Syncline, also the NW limb of the Sidlaw Anticline, displays dips gradually increasing away from the synclinal axis. North-west of a line through Almondbank, Luncarty and Stanley, north-westerly dips in strata of the Strathmore Group are generally in the range 10° -15° whereas farther to the south-east, dips in strata of the Garvock Group mostly lie between 20° and 25°. The NW limb of the Sidlaw Anticline is well defined as far as the North Tay Fault and its extension north of Dundee. The axial zone of the anticline is only well known north of the Castle Huntly Fault, south of which the Lower Devonian is concealed by younger strata in the Tay Graben, but the anticline is again recognisable in the volcanic rocks beyond the South Tay Fault south of Dron. Between the North Tay Fault and its extension north-east of Rossie Priory and a SSW-trending fault on the west side of Dundee Law there is a broad area in which strata of the Dundee Formation dip generally at less than 10° towards the south-west and south-south-west. It has been inferred (Armstrong and Paterson, 1970, p. 19) that these dips may constitute a south-westerly axial plunge of the Sidlaw Anticline, but a more probable explanation is that the area of south-westerly dips, bounded as it is by faults, is related to post-Dinantian subsidence into the Tay Graben.

The entire outcrop of the volcanic rocks of the Arbuthnott Group south of the Firth of Tay lies on the SE limb of the Sidlaw Anticline. Although dips in the lavas are difficult to determine, it is evident from measurements on the intercalated sediments south of Balmerino and from the overlying Garvock Group sandstones near Dairsie that the general SE dip in this area is about 20°. North of the Firth of Tay south-easterly dips are characteristic of the Dundee–Arbroath area. Between Tealing and the outcrop of the lavas of the Arbuthnott Group between Broughty Ferry and Arbirlot, dips are generally less than 10° whereas towards the Angus coast they are generally between 15° and 25°. At Whiting Ness near Arbroath the unconformity of gently dipping Upper Devonian strata upon Arbroath Sandstone of Lower Devonian age, dipping at 25°, provides proof of the pre-Upper Devonian age of the Sidlaw Anticline.

Faults of NW to NNW Trend affecting mainly Lower Devonian strata

Faults of approximately NW-SE trend appear to be particularly common in parts of the Lower Devonian outcrop, especially in the volcanic rocks of the Ochil Hills between Newburgh and Cupar and in the Sidlaw Hills north-east of Perth. These are generally small dislocations. Most are inferred from the juxtaposition of differing rock types, and many coincide with topographic features. Away from the outcrops of the volcanic rocks and the larger intrusions, faults appear to be conspicuously less common. This may be ascribed to the pronounced tendency of igneous rocks, more rigid than sedimentary rock, to deform by fracturing, although the greater extent of their exposure may be relevant in this context. It seems probable that many of the faults originated during the mid-Devonian folding of the Sidlaw Anticline and Strathmore Syncline, but it is equally plausible to regard many as having been reactivated by, if not initiated by, post-Dinantian deformation. In most cases, however, only a post-Lower Devonian age can be proved. The best example of reactivation of an earlier fault occurs just outside the district at the Dark Cave, on the north side of Carlingheugh Bay, north of Arbroath. There, Upper Devonian strata are faulted against Lower Devonian. Fissures in the Lower Devonian in the fault-zone are infilled with Upper Devonian material which encloses cleavage fragments of baryte veinstuff. This probably implies an Upper Devonian renewal of movement on a pre-Upper Devonian Fault.

Post-Dinantian faults associated with the Tay Graben

The existence of early Dinantian strata within the Tay Graben shows that the associated faults are at least partly of post-Dinantian age. There is no clear evidence in the district, apart from the isolated occurrence at Arbroath (p.63), that faulting occurred during the Upper Devonian, although this cannot be discounted.

The structure within the graben is not well known and the simple pattern of W–E faults is hypothetical. The dip of the Carboniferous strata is known from surface exposures only at Dron, Clashbenny and Mains of Errol and the structure has been drawn to conform with these. It is known from IGS boreholes at Inchcoonans and at New Farm (Bore B) and from others farther north (Browne, 1980a, fig. 1) that the area of Carboniferous rocks at Errol does not persist north of the line of the W–E fault tentatively drawn here. It is suspected that the Carboniferous strata at Errol lie within a N–NE-trending syncline and the same may be true of the Newburgh development. The line of the Castle Huntly Fault is drawn south of the outcrop of basic porphyrite at Castle Huntly and north of the IGS Burnside Borehole which proved Upper Devonian at rockhead.

The magnitude of the subsidence within the graben is probably considerable. In the ground north of the Castle Huntly Fault the Arbuthnott Group dips south-west for 10 km from Tealing, where the rocks are very low in the group, to the Kingoodie area, where the stratigraphical level is much higher. South of the Castle Huntly Fault, even if the southwest dip flattens out below the Upper Devonian, the southward downthrow of the Castle Huntly Fault itself and of the northern boundary fault of the Carboniferous at Errol, combine to make it probable that the strata concealed below the Upper Devonian immediately north of the South Tay Fault are relatively high in the Arbuthnott Group. As the strata south of the South Tay Fault and its associated dislocation in the Fliskmillan area are low in the Arbuthnott Group, and particularly when allowance is made for the fact that the lowest exposed Arbuthnott Group strata at Fliskmillan are geodetically higher than the concealed Arbuthnott Group strata immediately north of the fault zone, a large northerly downthrow is indicated, probably in excess of half the full thickness of the Arbuthnott Group. A figure of 2 km N of Fliskmillan is suggested for this throw in the Lower Devonian, and if, as previously suggested, the faults are essentially younger than the Carboniferous rocks in the graben, then a similar throw in the Upper Devonian is indicated. This implies that the Upper Devonian and Carboniferous strata are together between 1000 and 1500 m thick. The figure of 1000 m suggested for the thickness of the Upper Devonian in the Kinross-Howe of Fife area (Foster and others, 1976) is in accord with the thicknesses suggested above.

North-eastwards from the area of presumed greatest subsidence in the graben north of Fliskmillan the throw on the South Tay Fault system is diminished by the southward downthrowing Errol Fault and by a N–S fault of westward downthrow south of Dundee. The throw on the South Tay Fault is further diminished to the north-east by the cumulative effects of the coalescent north-downthrowing W–E trending faults in the lavas of north Fife outwith the graben. On the other side of the graben the throw of the North Tay Fault is diminished northwards by the effects of the Errol and Castle Huntly faults. The progressive diminution to the north-east of the amount of subsidence in the graben is shown in the ground between Dundee Law and the North Tay Fault by the extensive area of SW dips which signifies an important post-Carboniferous modification of the axial zone of the pre-Upper Devonian Sidlaw Anticline.

Compressional structures of post-Carboniferous age

The post-Carboniferous structures already described are believed to be mainly normal faults and there is little evidence to challenge this view but equally there is not a great deal of information relating to the hades of the various dislocations. It would appear, however, that from exposures at Dron and at the foot of Balruddery Den that the post-Carboniferous faults are generally steep but normal. Northeast-trending zones of complex structure, such as the Ceres-Maiden Rock Fault-zone in north-east Fife (Forsyth and Chisholm, 1977, p. 227), have been associated with a period of Upper Carboniferous W–E compression. The Ballumbie Fault on the north side of the Firth of Tay is of NE trend. It appears to be associated on its south-east side in the Pitkerro area with a north-plunging fold in siltstones and mudstones of the Dundee Formation and may be of similar type and age to the structures in Fife. MA

Chapter 9 Quaternary

Introduction

The glacial features and deposits within the district are attributed mainly to the last major Devensian ice-sheet which at its maximum development, about 18 000 years ago, reached as far south as the Wash and covered a large part of the North Sea Basin. No positive evidence of previous glaciations or inter-glacial deposits has been recognised in the district. It is probable, however, that ice-moulded features of the landscape owe their form to the accumulated effects of more than one glaciation.

The evidence of glacial striae, erratics and drumlins shows that in general the late-Devensian ice, advancing from a centre in the west Highlands, fanned out across east-central Scotland and moved approximately eastwards over the district, eventually laying down extensive deposits of till. During the subsequent recession, the emergence of high ground, as the upper surface of the ice wasted down, eventually confined the bulk of the ice to the major valleys, where active glaciers were sustained for some time by a remnant ice-cap in the Highlands.

Elsewhere, such as in coastal areas of Fife, parts of the ice-sheet, cut off from their Highland source by the emergence of intervening hills, melted out while in a stagnant condition. Fluvioglacial deposits were laid down by glacial meltwaters in a variety of associations with the melting ice, and much sediment of glacial derivation was carried into the late-Glacial sea. In addition, various sediments were deposited during the subsequent fluctuations of sea level in late- and post-Glacial times. All of these processes resulted in a highly complex distribution of deposits in the district. Much research has been done on the interpretation of the origins of these deposits and the topographical features associated with them. The changes in relative sea level and the classification of the late- and post-Glacial deposits are shown on (Figure 13). The distribution of the main features and deposits, excluding till, is depicted on (Figure 14).

Outline of Quaternary history

Early retreat stages

At an early stage in the deglaciation when the ice-front was still some distance east of the coast and the coastal hills were beginning to emerge from the ice-cover, ice-controlled drainage moving eastward cut numerous meltwater channels on the hillslopes north of the Firth of Tay (Figure 14). The sand and gravel deposits, which occur at high levels on the southern slopes of the Sidlaw Hills between Kirkton of Auchterhouse and the Rottenraw Burn near Arbirlot, may also be referred to early stages in the deglaciation. At Arbroath, somewhat younger fluvioglacial sand and gravel, which includes an esker, probably relates to meltwaters passing around the eastern end of the Sidlaw Hills and thereafter moving southwards through coastal ice. West of Monifieth, the valley of the Dighty Water provided an eastward route for meltwaters from the area of the Sidlaw Hills. The deposits here include an esker which can be traced from Kirkton of Strathmartine to the neighbourhood of Ethiebeaton.

Englacial meltwaters from the southern margin of ice occupying the Firth of Tay made their way through the gap in the Ochil Hills south of Wormit. Great quantities of glacial sediment were laid down between Wormit and Leuchars within ice, already stagnant, which at this stage still occupied the lower lying coastal areas of north-east Fife.

During deglaciation, relative sea level was high because the land was still deeply depressed as a result of ice-loading. Recession of the ice in the Firth of Tay and the coastal areas was accompanied by deposition of marine clays and sands–the Errol Beds (Figure 13). A fauna of foraminifera, bivalves and ostracoda of arctic affinities indicates that the climate was still markedly severe and deglaciation was presumably caused by diminution of precipitation over the Scottish ice-cap. In areas such as those around Arbroath, Leuchars, Monifieth and Benvie, where masses of dead ice survived in contact with the encroaching late-Glacial sea, complex associations of the marine deposits and fluvioglacial sediments were formed. The marine deposits now occur at high levels as a result of glacio-isostatic rebound during and after deglaciation. A sequence of raised shoreline features (Figure 15) marks stages in the general fall of relative sea level. In the absence of radiometric dates, it provides a means of relating the glacial retreat phenomena in different areas.

Deglaciation of the coastal Lowlands and the Firth of Tay

With the recession of the ice-front into the Firth of Tay the isolation of remnant areas of dead ice north of Monifieth and in the Leuchars- Wormit area south of the Tay probably became complete. Fluvioglacial deposition in the latter area ceased when the Firth of Tay ice-front receded west of Wormit, but persistent dead ice near Leuchars subsequently had a long association with successively younger late Devensian marine deposits and features (Rice, 1962; Chisholm, 1966). Farther south, the recession of the ice in Stratheden is marked by sporadic deposits of sand and gravel laid down on either side of the valley as far west as Cupar. In Stratheden, which expands south-westwards into the wide low-lying area of the Howe of Fife, the course of deglaciation was controlled by the progressive isolation of the ice in the Howe from the ice-sheet to the west, an isolation which began as the ice thinned over the Ochil Hills. It is probable that the ice in the Howe of Fife became stagnant while an active glacier still occupied the Firth of Tay and in these conditions deglaciation of the Howe may have been relatively rapid as compared with that in the Firth, thus allowing relatively early access by the late-Glacial sea, as postulated by Browne and others (1981).

After the recession of the ice-front in the Firth of Tay from a position east of Wormit, glacial meltwater deposits were laid down in the Invergowrie area and it is probable that a considerable area of dead ice was cut off near Benvie. Fluvioglacial sand and gravel at Abernyte, Inchture, Kinnaird, Rait and Pitroddie, north of the Firth of Tay, and at Fliskmillan south of the Firth, was laid down in positions marginal to the former glacier. While the glacier extended east of Newburgh, meltwaters were constrained to carry sand and gravel into the Lindores area and on through the Ochil Hills eastwards and southwards. During subsequent recession of the Firth of Tay glacier, fluvioglacial sand and gravel was laid down between Newburgh and Abernethy.

Deglaciation of the Tay valley and western Strathmore

A glacier occupying the lower Tay valley below Perth was confluent with one in Strathearn but eventually it lost contact with its southern neighbour when the Firth of Tay ice-front receded into the neighbourhood of the present Tay–Earn confluence. This lower Tay glacier would have been sustained by ice moving eastwards from the valley south of Methven and from Glen Almond. Its south-easterly movement down the Tay valley below Perth probably co-existed at an early stage with the ENE ice-flow which moulded the drumlins north of Perth.

During the glacier's retreat up the Tay valley towards Perth, fluvioglacial sand and gravel was laid down in deltas on the north side of the valley at Kinfauns and Barnhill, where meltwater discharged along the margin of the ice into the late-Glacial sea. The deposits at both localities are characterised by large inclined foresets, capped by flat-lying sediments. The foresets at Barnhill extend up to 37 m OD which indicates a minimum level of the late-Glacial sea at this locality.

At a time when the flow of ice from the west into the neighbourhood of Perth was weakening, as the lower Tay glacier receded towards Perth, it is probable that the contemporaneous input of ice from the west into that part of the upper Tay valley east of Dunkeld was also diminishing. The ice occupying Strathmore was becoming progressively stagnant, as was the ice occupying the Tay valley between the confluences of the rivers Almond and Isla.

Meltwater channels in the Tay valley north-west of Balbeggie, in the area south-west of Bankfoot and on the southern slopes of the depression south of Methven, were cut by ice-directed drainage, which ultimately escaped through the Perth area into the Firth of Tay. The channels on the east bank of the River Tay near Cargill, however, appear to relate to drainage moving northwards into the contemporaneous ice occupying western Strathmore, implying that a meltwater divide existed within the ice in the lower Tay valley between Stanley and Cargill. It is surmised that at the time of the cutting of the channels, the surface of the ice sloped away from the Stanley–Cargill area to the north-east into Strathmore and to the south-east down the valley towards the Firth of Tay.

At a somewhat later stage, meltwaters, flowing eastwards through ice in the upper Tay valley, and confined to the southern margin possibly because of the continued presence of active ice in the middle of the valley, laid down a belt of moundy sand and gravel, including a well-defined esker, along the south side of the valley between the Pass of Birnam and the Tay–Isla confluence. Continuation of the esker east of the River Tay in the area of sand and gravel south of Links indicates that, at the time of formation of this feature, the drainage from the upper Tay valley was unable to pass southwards into the lower Tay valley. This is consistent with the inferred meltwater divide in the ice south of Cargill.

Meltwaters from the upper Tay valley initially continued farther eastwards through the area of stagnant ice occupying Strathmore and spilled over a watershed, now at about 60 m OD near Forfar, north of the district. Kettled sand and gravel at the Pass of Birnam south of the River Tay may have been laid down by meltwaters passing to the north-east into the Tay valley, but the moundy deposits which extend still further southwards through Bankfoot to Luncarty and those in the valley west of Luncarty and in the area east of Almondbank are related to drainage flowing eastwards and southwards into the ice occupying the lower Tay valley. As this ice became increasingly pervious, the drainage from Strathmore and the upper Tay valley began to find its way through the Cargill–Stanley area towards Perth. The ice occupying the lower Tay valley disintegrated, the post-glaciation course of the river south of Cargill was established, and the spillway near Forfar was abandoned. A consequence was that the water-level fell rapidly in the area of low ground east of Dunkeld, still largely occupied at this stage by dead ice. In the Meikleour area, sediment entering western Strathmore from the north was deposited in a large fan, now pitted by numerous kettleholes. The upper surface of this deposit descends near Kinclaven to about 46 m OD, a level related to the new outlet down the lower Tay valley, and significantly lower than the levels reached by the older moundy deposits associated with the esker at Links. Fluvioglacial sand and gravel terraces in the Tay valley below Birnam, and isolated flat-topped kames near Spittalfield, are probably of the same general age as the Meikleour fan.

Late-Glacial marine incursion in the Tay and Earn valleys

Following the general clearance of ice from the lower Tay valley the lower ground north of Perth was flooded by the late-Glacial sea. Clays and silty clays attributable to the Errol Beds were laid down as far north as the neighbourhood of Stormontfield, and arctic marine faunas in these deposits are known to occur as far north as the area east of Almondbank. The Errol Beds are overlain near Scone and Stormontfield by sand and gravel terraces which were laid down during the later advance of coarse sediment down the valleys of the River Tay and its west bank tributaries. The coarse material was repeatedly reworked, with the formation of lower terraces, in reponse to the general fall of the late-Glacial sea level.

The deglaciation in Strathearn was accompanied by the deposition of the higher fluvioglacial sand and gravel terraces in the Abernethy and Forgandenny areas. A large part of the sand and gravel in the area west of Forgandenny, however, overlies and post-dates Errol Beds which were laid down during the late-Glacial marine inundation which accompanied deglaciation. The sand and gravel occurs in fans originating from valleys in the Ochil Hills, and includes terraces of various ages formed at successively lower levels in response to the fall in level of the late-Glacial sea (Browne, 1980b).

The recession of the Earn ice-front allowed the sea to penetrate beyond the western margin of the district and eventually to enter from the west the valley south of Methven, depositing Errol Beds there. Access of the late-Glacial sea to the Almondbank area was probably still impeded, however, by a complex area of dead ice with associated englacial gravel which had already been deposited during the deglaciation at a point where the meltwater drainage from the Almond valley entered low ground. The subsequent advance of gravel down the Tay valley in response to the fall of the late-Glacial sea from the marine limit was paralleled by the advance from the Almond valley of a large gravel fan which spread out both to the south-west and south-east, burying not only most of the early ice-gravel complex at Almondbank but also Errol Beds which had been deposited upon its flank to the east.

The gravel terrace underlain by Errol Beds east of Almondbank has been correlated with the prominent marine feature known as the Main Perth Shoreline (Sissons, 1966; Cullingford in Gray and Lowe, 1977). This raised shoreline was so named (Sissons, 1966) because of its supposed association with a postulated glacial readvance known as the Perth Readvance (Simpson, 1933). The concept of a re-advance was subsequently discredited (Paterson, 1974) and the Main Perth Shoreline is now thought to have formed during the climatic amelioration recorded at about 13 000 to 13 500 years BP by Ruddiman and McIntyre (1973).

Windermere Interstadial and Loch Lomond Stadial

The amelioration initiated an interstadial period when climatic conditions were considerably less frigid than those existing during the deposition of the Errol Beds and was probably responsible for the final melting of buried ice-masses, thus producing kettleholes in the surfaces of gravel fans at Almondbank and elsewhere. The period of improved climate is represented in the late-Glacial marine sequence in the Tay-Earn area by the Powgavie Clay which was discovered by drilling through younger deposits in the Carse of Gowrie (Paterson and others, 1981). A marine fauna in the Powgavie Clay resembles that found in the part of the Clyde Beds laid down in the west of Scotland before the later arctic period of the Loch Lomond Stadial, and both deposits correspond in a general way to the Windermere Interstadial described from a lacustrine sequence in the Lake District (Coope and Pennington in Coope, 1977).

In the lower reaches of both the Tay and Earn valleys, sediment was carried progressively seawards, as pre-existing terraces were reworked in response to the fall of sea level below the Main Perth Shoreline. Deltaic sand and silt deposits known as the Culfargie Beds were laid down in the late-Glacial sea. They probably pass seawards into the Powgavie Clay.

Successively lower estuarine terraces, the Lower Perth Shorelines of Cullingford (1972; in Gray and Lowe, 1977), formed during minor marine transgressions which interrupted the general fall of sea level. After the abandonment of the lowest of these shorelines, the sea is thought to have fallen to about the level of an erosion surface below the Port Allen Gravel, which is considered to represent the approximate equivalent in the Tay-Earn area of the Main Lateglacial Shoreline of the Forth valley (Sissons, 1974a). Channels cut in the older late-Glacial deposits to still lower levels are considered to have formed during a period when the sea may have descended below the Main Lateglacial Shoreline, probably during the recurrence of arctic conditions between 11 000 and 10 000 years BP known as the Loch Lomond Stadial.

Post-Glacial deposits and sea level changes

During the subsequent rise of the sea to levels marked by the top of the estuarine Carey Beds (Figure 13), gravel deposits filled the channels and formed extensive spreads in the lower Tay valley and in lower Strathearn (the Friarton Gravel and the Earn Gravel). Later peat deposits, collectively termed the Sub-Carse Peat, formed on the exposed surfaces of the Carey Beds as the relative sea level fell again. This fall was interrupted by minor marine transgressions which correspond to the Main and Low buried shorelines of the Forth valley (Sissons, 1966). It was succeeded by a pronounced transgression to the level associated with the upper surface of the post-Glacial marine deposits (Carse Clay) and termed the Main Postglacial Shoreline (Sissons, 1967). The final fall to the present sea level was again probably punctuated by minor transgressions during which the post-Carse estuarine deposits were laid down, forming marine terraces which were described under the name Lower Carse Shorelines by Cullingford (1971).

Glacial erosion features

Successive Quaternary glaciations have modified the topography of the district and glacial erosion has appreciably lowered rockhead in places. Thick drift deposits conceal buried rockhead valleys in the Tay valley east of Dunkeld and around Perth, in lower Strathearn, along the Firth of Tay and in Stratheden. Specific evidence of glacial overdeepening, such as has been obtained in the Forth and Devon valleys (Francis and others, 1970, p. 263), is known only from the Dunkeld area. Here rockhead has been proved by drilling to descend at least as low as Ordnance Datum on the new line of the A9 trunk road, whereas lower down the Tay valley between Kinclaven and Stanley rockhead is no lower than 15 m above OD.

Erosion by ice, moving generally to the south-east, transverse to the strike of the Lower Devonian lavas, probably deepened SE-trending depressions across the Sidlaw Hills, notably the gap holding Lochindores [NO 270 355] south-east of Coupar Angus and the depression known as the Glack of Newtyle [NO 310 400] (Linton, 1951). The elongate rock basins occupied by Long Loch and Lundie Loch north of Lundie [NO 290 365] were however eroded along the NE-trending, outcrops of sedimentary zones in the lavas. Here it is probable that the disposition of the less resistant beds rather than the direction of ice-movement has governed the orientation of the basins.

Upstanding rock features have almost certainly suffered some glacial modification throughout the district but the orientated landforms known as 'crag-and-tail' are mainly developed in association with outcrops of intrusive igneous rocks in the Dundee area. Dundee Law, with a westward-facing crag of porphyrite, is a prominent example.

Well preserved glacial striae have, in a few localities, been observed on unweathered rock surfaces exposed by the removal of till. However, the weathered surfaces more commonly encountered do not usually retain recognisable striae. Few records, therefore, were obtained during the resurvey and on (Figure 14) they are supplemented by observations made, mainly by J. Geikie, during the primary geological survey. Observed mainly on the high ground formed by Lower Devonian lavas and Dalradian rocks, the glacial striae trend generally about south-east over most of the district. They possibly relate to an ice-movement earlier than that related to drumlins on adjacent lower ground. North of Dundee, however, the few records available show a change to an easterly orientation which accords with the trend of the crag-and-tail features in that area. South of the Firth of Tay large scale flutings on Lower Devonian lavas west of the Eden Estuary (Forsyth and Chisholm, 1977, fig. 27) indicate an ice-movement somewhat south of east.

Till, drumlins and erratics

Till, the most widely distributed of the glacial deposits, consists of ice-transported material which was either laid down at the base of the moving ice-sheet or was lowered on to bedrock as the ice melted. Generally a sandy stony clay, and commonly containing clasts of considerable size, till occurs at surface over much of the district and is also known to be present beneath younger superficial deposits. On the Ochil and Sidlaw hills and also along the Highland Border it has been mapped in places close to the highest summits, although it is generally thin, patchy or absent on the higher ground.

Whereas ground covered by till is commonly featureless, particularly at high levels, drumlins, ice-moulded till ridges aligned in the direction of ice-movement, occur in certain areas below the 150 m contour. Drumlins are numerous in the neighbourhood of the lower Tay and Almond valleys north-west of the Sidlaw Hills, and several prominent examples are developed on the area of till which emerges through a cover of late-Glacial marine deposits in the Carse of Gowrie west of Errol. Others occur in lower Strathearn, in the Dundee–Longforgan area, in lower Stratheden and south-east of St Andrews (Forsyth and Chisholm, 1977, pp. 233–234).

Ice-movements to the east-north-east, conforming more or less to the trend of the principal valleys, may be inferred from the dominant orientation of the long axes of drumlins in the areas both north-west and south-east of the Sidlaw Hills. These movements are markedly oblique to the SE-trend of glacial striae on the hills, from which it might be deduced that 'two distinct episodes of ice-movement have occurred. A different explanation has been put forward for similar situations in the areas west of Stirling and around Auchterarder (Francis and others, 1970, p. 256); namely that movements within the ice-sheet took place simultaneously in divergent directions at high and low levels.

The probability is, however, that any drumlins formed in such circumstances, while the ice still covered the high ground, would in any case be subject to further moulding at a later stage by ice confined to the low ground after the emergence of the hills. It is very likely therefore that these drumlins, at least in their present form, are younger than the high-level glacial striae.

The dominant ENE-trend of drumlins north-west of the Sidlaw Hills is locally modified south-west of Bankfoot by a SE-orientation which may relate to contemporaneous ice invading the area from the north-west.

In lower Strathearn, near Bridge of Earn, the ESE drumlin-trends indicate ice-movement again conforming to the valley trend which here is roughly parallel to the trend of striae on the ridge of Moncrieffe and Friarton hills to the north. Farther east, near St Andrews, drumlins provide evidence of ice-movement to the east-south-east conforming to the general trend of the coastline which is subparallel to major glacial fluting of rock surfaces at higher levels (Forsyth and Chisholm, 1977, fig. 27). However, as in other places where valley trends differ from the orientation of adjacent high level ice-movement indicators, drumlins a short distance away in lower Stratheden follow the NE-trend of that valley.

In the area around Dundee the trend of topographical depressions, crag-and-tail features and also glacial striae conforms to the easterly orientation of the few drumlins which exist here. The high ground of the eastern Sidlaw Hills, immediately to the north, carries striae of easterly as well as SE-trend. It appears therefore that even on high ground within this area there are indications of a general departure of ice-movement from the SE-trends exhibited by striae farther west.

It is probable that over much of the district the till is between 2 and 5 m thick. The deposit thins in general towards high ground but areas shown on the 1:50 000 drift sheets as lacking drift are known in places to contain small unmapped patches of till, even in ground close to rocky summits. On low ground the till is known to be very thick in places. In the Carse of Gowrie, 30 m of till, presumably forming part of a buried drumlin, was penetrated in the IGS Muirhouses Borehole [NO 2699 2493] below post-Glacial marine deposits. In lower Strathearn, 24 m of till was proved below late-Glacial marine deposits in the IGS Beautyfield Borehole [NO 1547 1567].

Although far-travelled clasts of resistant rock types are characteristic components of the till throughout the district, the bulk of the deposit is apparently of more local derivation, its composition at any point usually reflecting the nature of bedrock either beneath the deposit or at no great distance from it. In the area north-west of the Highland Boundary Fault-zone near Dunkeld, the till incorporates much Dalradian material and has a yellow-brown colour. This deposit has to a limited extent been carried to the south-east on to the outcrop of the Lower Devonian rocks close to the fault-zone. Away from-the Highland Border the till is mainly a drab red-brown, stony, sandy clay derived mainly from the Lower Devonian sandstones which constitute the dominant bedrock lithology over much of the district. A notable variant, the till associated with the outcrop of the Lower Devonian Cromlix Formation north-east of Stanley, is full of the debris of the characteristic purple-brown or chocolate-coloured mudstone of that stratigraphical division. This easily recognisable material, is however not conspicuous in the till south-east of the outcrop of the formation and it is presumed that the mudstone does not readily survive glacial transport, although it probably imparts coloration to the deposit.

The Upper Devonian sandstones in lower Strathearn and the Carse of Gowrie give rise to a relatively bright red, sandy till which in places has been transported eastwards beyond the limits of the outcrop. Thus a till-sequence encountered in the IGS West Dron Borehole [NO 1348 1614] comprises red till underlain by a basal layer of grey till derived from the local Carboniferous mudstones. West of Invergowrie, red till of Upper Devonian derivation has been transported eastwards on to the outcrop of Lower Devonian sandstones which are distinctly drab-coloured in comparison. In the area west of St Andrews, reddish brown till, characteristic of the neighbouring Devonian areas to the north and west, has been carried on to the Carboniferous outcrop by several kilometres (Forsyth and Chisholm, 1977, p. 233 and fig. 27).

Within the areas mapped as till-covered, there are at a number of localities deposits which differ from the normal, sandy, stony clay. Along pipeline trenches cut during recent years within the district, open textured deposits with a high content of rounded clasts, but nevertheless lacking obvious bedding, were encountered in a few places. The relation of this material to more normal till is obscure and it is possible that pre-existing or subglacial waterlain deposits have been incorporated in the till. The lack of any recognisable associated landform is consistent with this view and is the reason that separate mapping of the deposits is impracticable. A gravel deposit below a considerable thickness of till was penetrated south-west of Errol by the IGS Mains of Errol Borehole [NO 2380 2190] in an area where the till has been moulded into drumlins. This deposit may be older than the late-Devensian glaciation which formed the till, but there is no independent evidence of its age.

Transported blocks which occur within till and also as isolated erratics throughout the district provide impressive evidence of glaciation. They are most readily examined on intertidal rock platforms such as that north-east of Carnoustie, where numerous large boulders presumably represent a lag-deposit formed by the marine erosion of till. The general occurrence of schistose grit and other Dalradian metamorphic rocks as erratics throughout the district is testimony to the Highland source of the glaciation. Quartz-dolerite, mainly derived from the numerous late-Carboniferous dykes which traverse the Lower Devonian rocks west of the Sidlaw Hills, is also common as an erratic, and conglomerate transported from the Highland Border is also found. The Lower Devonian volcanic and sedimentary rocks appear in general to have yielded few very larger erratics, although as smaller clasts they are common in the drift deposits. An erratic resembling the pink acid porphyrite west of Craighill [NO 436 357] north of Dundee, has been found on Pitlivie Moor [NO 556 396] north of Carnoustie, from which may be inferred an ENE transport of over 12 km, in general conformity with the ice-movements deduced from other evidence in this area.

Glacio-isostasy and raised shorelines

The recovery of the Scottish landmass from the glacioisostatic depression imposed by the weight of the late-Devensian ice-sheet was probably initiated as that load began to diminish during deglaciation. The land was therefore already undergoing active uplift as it emerged from its ice-cover, but because the recovery was at this stage very incomplete, the land being still deeply depressed, newly deglaciated ground was flooded for a considerable distance above present sea level by the late-Glacial sea. The extent of the initial inundation was enhanced by the eustatic rise of sea level caused by the return of glacial meltwaters to the oceans during this period. The eustatic rise was, however, outpaced throughout the greater part of the late-Devensian (late-Glacial) and Flandrian (post-Glacial) periods by glacioisostatic uplift, resulting in an overall fall in relative sea level. The late-Glacial sea reached its highest local level, the marine limit, at different times in different areas as ground was progressively uncovered by the receding ice. Only rarely, if ever, can the marine limit be precisely identified in the district; usually no more may be inferred than that the limit lies no lower than the highest local indication of a marine presence, whether this be a marine deposit or a shoreline feature.

Below the marine limit, shoreline features were successively formed and abandoned in response to changes in relative sea level. The raised shorelines of the Forth and Tay–Earn areas and of the coastal areas of Fife and Angus have been the subject of detailed study in recent years and their relations and distribution have been summarised in a series of shoreline diagrams (Sissons and Smith, 1965; Cullingford and Smith, 1966; Sissons, Smith and Cullingford, 1966; Cullingford, 1971; Cullingford in Gray and Lowe, 1977). A summary diagram (Figure 15) based mainly on the work of Cullingford is generally applicable to the district. In this diagram, and in the published shoreline diagrams relating to specific areas, the salient feature is the general eastward tilt of the shorelines which is a consequence of the differential nature of glacio-isostatic recovery; the amount of uplift increasing towards the former centre of the ice cap in the western Highlands. The tilt which affects the Main Post-glacial Shoreline, a feature post-dating the late-Devensian glaciation by several thousand years, demonstrates that glacio-isostatic recovery continued long after the disappearance of the ice.

The interpretation of the height-data used in the construction of the shoreline diagrams does to some extent depend on the assumption that an older shoreline would be expected to have a higher gradient than a younger shoreline because of its longer exposure to differential isostatic uplift. The assumption is supported by good evidence relating to the two best defined of the shorelines. The Main Postglacial Shoreline is a particularly clear feature in the Tay–Earn and Forth areas where it is represented by raised tidal flats, the measured altitudes of which enable the eastward tilt to be determined with something approaching precision. The Main Perth Shoreline, although less well marked than the Main Postglacial Shoreline, is by fortunate circumstances particularly well isolated in altitude from other late-Glacial raised marine features, with which it might otherwise have been confused, and can be traced with relative ease on the diagrams. It has a distinctly higher gradient than the Main Postglacial Shoreline, as is consistent with its greater age. These two shorelines therefore appear to provide a secure basis for the shoreline diagrams although some reservations might be expressed concerning the gradients of certain of the older shorelines deduced from the height data in east Fife.

Certain coastal features, such as the post-Glacial sand ridges on the seaward fringes of Tentsmuir, are probably related to shortlived depositional or erosional events. The principal raised shorelines in the district are, however, fewer in number than would be expected if such were their origin and it is therefore believed that each shoreline may constitute the aggregate result of marine activity during a more prolonged episode when relative sea level was stable. Such episodes would imply contemporaneous balance between eustatic sea level rise and glacio-isostatic uplift.

As glacio-isostatic uplift in the period after deglaciation generally exceeded the rate of eustatic sea level rise, it follows that episodes during which the two processes were in balance imply accelerations of eustatic sea level rise. In some cases this may have produced marine transgressions particularly in areas peripheral to the centre of glacio-isostatic uplift.

The age of the Main Postglacial Shoreline is known from radiometric dating of relevant material from the district (Cullingford in Gray and Lowe, 1977). Of the other shorelines, the Main Buried Shoreline has been dated in the Forth valley (Sissons and Brooks, 1971) while the Main Perth Shoreline (Paterson, 1974) has been associated with the climatic amelioration recorded in ocean cores at about 13 500 years BP (Ruddiman and McIntyre, 1973). No available radiometric dates relate to the older shorelines in the district, and at present resort must be made to the evidence of gradients in the assessment of the ages of the shorelines.

Variation in gradient within a sequence of raised shorelines is the consequence of the operation of differential isostatic uplift over the varying periods which have elapsed since each shoreline was formed. The gradients are independent of subsequent eustatic movements and therefore provide some measure of the ages of the features. The available information on the ages and gradients of the three shorelines mentioned above was used by Andrews and Dugdale (1970) to calculate from relevant gradients the ages of certain of the group of raised shorelines known as the East Fife Shorelines (Cullingford and Smith, 1966). Ages ranging from 18 250 to 15 100 years BP were obtained for shorelines EF1 to EF6. However, the accuracy of these, and of revised and somewhat younger dates obtained by the use of more recent data, has been questioned (Sissons, 1974b) because deglaciation of part of eastern Fife is implied at a time when the late-Devensian ice-sheet was at its maximum extent in England. The apparently unacceptably, high dates obtained by Andrews and Dugdale (1970) may cast some doubt on the accuracy of the recorded gradients of the oldest shorelines in east Fife. This supposition may be justified by the fact that in a contemporaneous series of shorelines north of the Firth of Tay (Cullingford and Smith, 1980) the highest recorded gradient is markedly lower than the gradients of East Fife shorelines EF1 and EF2. Nevertheless, even north of the Tay the highest gradient implies an age in excess of 16 000 years BP, and pending satisfactory dating of the shorelines by other methods, a deglaciation date of this order must be provisionally accepted.

Glacial meltwater channels

Although glacial meltwaters utilised to a great extent the lines of existing drainage, and may in places have initiated the courses of present streams, they also cut many channels which are clearly anomalous in relation to present topography. These are now either dry or carry very small, misfit streams. Some of them follow oblique downhill courses, and not uncommonly occur in channel-systems with individual members cut at successively lower levels. It is inferred from such occurrences that downhill limit of penetration of meltwater below the ice, marked by an individual channel, was progressively lowered with the elapse of time. Although in a few cases channels of this type may have been cut precisely at the contemporaneous ice-margin, it is probable that most were cut below the marginal zone of the ice, thus falling into the category of sub-marginal channels. Other channels of generally steeper gradient descend hillsides more or less directly and correspond to sub-glacial chutes. They occur both as individual channels and as the steep downhill extensions of sub-marginal channels and presumably formed where abruptly increased penetration of meltwater below the ice became possible.

Details

The meltwater channels (Figure 14) in the area east of the Sidlaw Hills have a general easterly trend and gradient, which is consistent with their cutting by meltwater draining eastwards from an ice-sheet which thinned away from the Highlands. At an early stage of deglaciation, meltwaters, which originated near Skichen [NO 510 415] at a low point on the Sidlaw watershed, were impeded in their descent by ice which still occupied low ground on the seaward side of the hills. The meltwaters were constrained to flow eastwards (Rice, 1962, p. 8) thus cutting the prominent channel [NO 550 397] occupied by the Pitlivie Burn south of Cairncortie. Subsequently, as the ice receded from the hills, conspicuous channel-systems were cut in the Craigend [NO 584 395] area and in the ground around Balmachie [NO 546 369]. Many of the component channels are no more than a metre or two deep and are incised only into till, but in places bedrock was cut, most notably between Craigend and Fauldiehill [NO 570 395] where one exceptionally large channel is about 15 m deep.

Farther west, in the ground north-east of Dundee, a number of eastward-draining channels were successively formed as the ice receded from the hills. The southernmost two of these, descending the seaward slopes east of Claypots Castle [NO 453 319], are unusual in that they are thought to represent spillways cut at a late stage in deglaciation by drainage from the valley of the Dighty Water which was blocked by sand and gravel and dead ice in its lower reaches. The channel west of Ballumbie [NO 443 337] may owe its form in part to earlier glacial scouring around the porphyrite hill immediately to the north. This channel, and another to the north, presumably carried meltwaters later and at a lower level than the meltwater stream which deposited the esker in this neighbourhood.

In the ground east of Kirkton of Auchterhouse a number of channels of relatively low gradient descend eastwards but give way southwards, towards Kirkton of Strathmartine, to channels descending southwards and directly down the slopes. A channel of unusual origin occurs immediately west of Jeanfield [NO 366 370]. Here an E-trending till-ridge was formed during glaciation, producing a closed basin on its north side. During deglaciation meltwaters spilled over the southern rim of the basin, initiating a channel, the level of which later stabilised as bedrock was encountered and which formed the outlet of a loch. Artificial deepening of the outlet subsequently drained the loch.

The emergence of the summits of the Sidlaw Hills during deglaciation modified the general eastward flow of meltwaters and this is most apparent in the area immediately west of the hills. On the east side of the Tay valley, south-east of Cargill, a considerable number of channels fall to the north and north-west towards the area of the Tay–Isla confluence. The meltwaters which cut this channel system presumably drained thereafter to the north-east through Strathmore. South of a W–E line passing between Cargill and Stanley, however, meltwater channels lying east of the River Tay descend with few exceptions to the south-west, consistent with ice-controlled drainage down the lower Tay valley.

Most channels west of the River Tay were cut by meltwaters which ultimately drained to the south-east into the lower Tay valley at Perth, but there is a considerable divergence in their orientation. A number drain to the south-west and south-south-west in the ground to the east of the Ordie–Garry valley which extends from the River Tay near Luncarty to Bankfoot. A single channel [NO 080 340] in this neighbourhood drains to the south-east to the west of Five Mile Wood and is, unlike the others, of the submarginal type. The numerous channels which exist west of the Ordie–Garry valley drain generally eastwards: their courses may be governed by the orientation of drumlins in this ground. Numerous channels, mainly sub-glacial chutes, descend the southern slopes of the depression south of Methven.

Glacial meltwater deposits

Throughout the period of recession of the late-Devensian ice-sheet in the district, elastic sediments of glacial origin were transported and deposited by meltwater. Where these fluvioglacial sediments accumulated within decaying glacier ice, or close to its receding margin, any terrace-like forms which may have been originally produced are liable to have been either destroyed or modified when the associated ice melted, resulting in a moundy topography of kames with kettleholes. In a few places, as at Arbroath, in the ground north of Dundee, in the Wormit- Leuchars area, east of Lindores and near the confluence of the rivers Tay and Isla, the fluvioglacial deposits are partly in the form of eskers, prominent narrow ridges of gravel which probably mark the former courses of subglacial or englacial meltwater streams flowing eastwards away from the area of thickest ice. In situations at or beyond positions of the principal ice-margin, where in general little or no dead ice was present at the time of deposition, the original flat or terraced landforms have been largely preserved.

In various parts of the district, and at various stages of the deglaciation, fluvioglacial sediment was deposited in close association with the late-Glacial sea which at this period stood appreciably higher than at present. In certain places the sediment was discharged directly at the ice-front into the sea, and seaward-dipping foresets were formed in deltaic deposits of sand and gravel, as at Kinfauns and Barnhill, east of Perth. Elsewhere, fluvioglacial deposits exhibiting the characteristic kame and kettle landforms associated with the melting out of dead ice, occur in coastal areas at altitudes which are below the marine limit in immediately adjacent ground. This situation exists in the Arbroath, Barry–Monifieth, Wormit–Leuchars and Benvie areas where the deposits are thought to have accumulated within glacier ice up to levels related to the englacial water table which was in turn controlled by the level of the late-Glacial sea into which meltwater eventually drained through the ice. Subsequently, in these areas, extensive patches of dead ice enclosing and underlying large volumes of fluvioglacial sediment were isolated during the recession of the principal ice-front into the Firth of Tay and were probably eroded at their outer edges by the late-Glacial sea while coastal features, now raised, were cut in adjacent ground. Such tracts of sediment-charged ice survived in the contemporaneous cold climate long enough and in sufficient extent to exclude the late-Glacial sea from the areas where the existing low-level fluvioglacial landforms have clearly not been subjected to marine action and it is concluded that sea level had fallen below the altitude of such landforms before the associated glacier ice had melted.

Along the lower Tay and Almond valleys and in Strathearn, sand and gravel terraces of late-Glacial age were laid down after the greater part of the district was deglaciated. In places such deposits are known to be younger than late-Glacial marine deposits which they overlie. Nevertheless certain of the older terraces notably in the Almond-bank and Forgandenny areas clearly predate the final melting of long-lasting isolated masses of dead ice, as is indicated by the existence of kettleholes in the terrace surfaces. Although in a broad sense proglacial, and classed on the 1:50 000 maps as meltwater deposits, these terraces might be regarded as river alluvium of late-Glacial age. The late-Glacial rivers which deposited them were almost certainly fed by meltwater from the Highlands, but a considerable proportion of the sediment forming the terraces was probably supplied by other streams, being eroded from the newly-deglaciated terrain which at this early stage lacked a cover of vegetation.

Details

Wormit–Leuchars (St Fort) area (Figure 16)

The large triangular area of sand and gravel between Wormit, Tayport and Leuchars attracted attention at an early date (Chambers, 1848, pp. 53–54). Laid down by meltwaters from the Tay glacier passing to the south-east through the Wormit Gap, the deposits have given rise to a topography characterised over most of the area by mounds and kettleholes, linear ridges and flat-topped plateaus. In central and eastern parts of the area, however, subsequent marine reworking and deposition has locally produced a more subdued topography. The close association of glacial and marine landforms has given rise to differing opinions on the history and origin of the deposits (Durham, 1877; Geikie, 1902, pp. 297–301; Rice, 1961; Chisholm, 1966; Cullingford, 1971).

Deposits exhibiting glacial landforms

Within the area of fluvioglacial deposits a central zone, about 0.3 km broad, consisting of linear mounds and hollows, extends from Wormit to St Michael's Inn (Figure 16). It contains the coarsest gravels, with blocks up to 1 m across, as may be seen in several gravel pits. Individual ridges or eskers run roughly parallel to the length of the central zone as a whole. The moundy sand and gravel flanking the central linear zone shows no well marked linear elements and the material is less coarse-grained. Particularly noticeable in these areas are the steep-sided plateaus with flat tops at levels mostly between 34 and 40 m OD (Cullingford, 1971).

Associated marine deposits and features

A sandy deposit of marine origin, described as valley sand by Rice (1961) and Chisholm (1966) occupies depressions in the central, southern and eastern parts of the area. The surface of the deposit is gently undulating, almost flat in places, and its margins lie at levels between about 27 and 20 m OD. The deposit is more fine-grained than the fluvioglacial material, consisting mainly of sand with subordinate silt and gravel. To the south it merges into the late-Glacial marine sands and clays of Stratheden.

Raised beach deposits, consisting of sand and fine gravel, and forming a sloping terrace backed in places by a cliff, are present along the eastern side of the area from Tayport to Leuchars, and occur also at Wormit, Newport and Balmullo. Where the raised beach deposits abut against a back-feature cut into fluvioglacial material, their upper limit lies between 20 and 25 m OD but in other situations away from the gravel areas, where the raised beach deposits lie on till or rock, the upper limit reaches 36 m in places.

Sequence of events

Both valley sand and raised beach deposits are of marine origin. It is relevant to discuss them here because the glacial meltwater deposits at St Fort were laid down in environments partly controlled by sea level, and the later undoubted marine deposits have an important bearing on the history of sea level changes. The development of the St Fort deposits, as deduced from the present evidence (Rice, 1961; Chisholm, 1966; Cullingford, 1971), may conveniently be divided into four overlapping stages, as follows:

First stage During deglaciation an ice-mass, occupying the low ground between Wormit and Leuchars and part of Tentsmuir, was traversed by meltwater flowing to the east and south-east in tunnels or channels (Rice, 1961, p. 122). At this stage, during which the late-Glacial sea probably fell from shoreline EF5 ((Figure 15)) to below the level of shoreline EF6, the Firth of Tay Glacier, which provided the meltwaters, extended east of Wormit. The initial deposits were the coarse gravels of the central linear zone, and although these gravels were disturbed by collapse-structures, induced by subsequent melting of the associated ice, they do not seem to have been glacially overridden or pushed. The ice-mass must therefore have become immobile and was probably effectively detached from the glacier by this time. In areas away from the central belt, less coarse deposits were laid down in association with the melting ice, and the flat-topped plateaus were formed. This occurred under the control of an englacial water-level related to the level of the contemporaneous sea, which is presumed to have been in contact with the eastern and southern sides of the mass of dead ice (Rice, 1961, pp. 121–122). The actual level of the late-Glacial sea at the time of formation of the plateaus has been the subject of differing opinions. Rice (1961) and Chisholm (1966) ascribed the construction of the plateaus to a period of high sea level during which raised beaches developed at about 36 m above OD both north and south of the present area of fluvioglacial deposits, at

Scotscraig and Balmullo respectively. Cullingford (1971), however, found that the upper surfaces of the plateaus are inclined, the surface of the Gallowhill plateau descending southwards to about 27 m above OD. The implication is that near Leuchars the late-Glacial sea had fallen below this level, which corresponds to the position of shoreline EF6 (Figure 15), during the period of formation of the plateaus.

Cullingford (1971) concluded that the plateaus were contemporaneous with shoreline EF6 but this cannot be true because when the sea stood at this position, the Firth of Tay ice-front lay to the west of Abernethy (Armstrong and others in Gemmell, 1975). It is concluded, therefore, that the sea fell to a level below 27 m OD at Leuchars after the formation of shoreline EF5 and while the ice-front lay east of Wormit. The sea, however, subsequently rose to the level of shoreline EF6 but only after the ice-front had withdrawn west of Wormit, cutting off the supply of sand and gravel to the Leuchars area.

Second stage After recession of the glacier, reworking of the glacial deposits at Wormit into a raised beach at about 30 m OD began (Chisholm, 1966, p. 166). This feature corresponds roughly to the sea level marked by shoreline EF6. The mass of dead ice around Leuchars continued to melt slowly in the frigid climate, and kettleholes began to form, but marine erosion was still confined to the seaward margins of the complex of dead ice and sediment beyond the present eastern limits of the fluvioglacial deposits.

Third stage The level of the sea near Leuchars fell gradually to about 20 m OD. The ice-mass continued to melt, and the sea inundated some of the hollows that were produced, penetrating further into the centre of the gravel area and depositing valley sand. The margins of the earliest valley sand in the Motray Valley, in the area of Strathburn [NO 4300 2352], St Michael's Sandpit and Brackmont [NO 4285 2207], lie at levels between 24 and 27 m OD, but later deposits around Pickletillem were deposited only up to a level of about 21 m OD. Meanwhile, marine erosion at the seaward side of the gravel continued, producing an area of raised beach deposits with back-features varying in height between about 20 and 26 m OD, depending on when cutting took place during this period of falling sea level. At Leuchars Lodge a beach was cut into an area containing buried ice which, after the sea had withdrawn, melted to produce kettleholes (Chisholm, 1966).

Fourth stage The sea fell from 20 m to below 10 m OD and the last remnants of buried ice melted out; the glacial morphology developed to its final form, while the sea withdrew eastwards and southwards from the raised beaches and inlets of valley sand.

The best exposures are in working or recently abandoned sand and gravel pits. Three of these, at St Fort, East Links Wood and St Michaels, lie in the central zone of linear gravel mounds; one, at Wormit, lies in moundy sand to one side of the central zone; and one, at Straiton, has been excavated into the side of a plateau.

Stratheden area (Figure 16)

In Stratheden the glacial meltwater deposits occur principally to the west of Kemback and form narrow strips at levels up to 64 m above OD on both sides of the valley. There is a small esker at Cupar. The sand and gravel in these deposits was for the most part carried to the north-east by meltwaters from ice occupying the Howe of Fife, an extensive basin of low ground which drains to the north-east through Stratheden. At the time of deposition of the fluvioglacial sand and gravel at, and east of, Cupar, the ice extended seawards of that place, the meltwaters ultimately draining into the expanded Eden estuary of late-Glacial times. Part of the deposits on the south side of the valley, immediately west of Kemback, was clearly derived by way of meltwater channels from the Pitscottie–Ceres area to the south of the district. A short distance beyond the southern limits of the Cupar (48E) Sheet the upper surface of a gravel terrace on the south side of Stratheden descends to the north-east from an elevation of 43 m above OD south of Cupar.

Arbroath–Carnoustie area

Fluvioglacial sand and gravel exhibiting moundy and kettled topography occurs in the ground adjoining the lower part of the valley of the Brothock Water at Arbroath. Glacial landforms, unmodified by marine action, occur here at levels distinctly lower than the marine limit in neighbouring coastal areas.

It is presumed that with the emergence of high ground farther west, the fluvioglacial deposits were laid down by meltwaters draining eastwards and southwards through a belt of coastal ice extending some distance seawards of the present shore. The most notable feature at Arbroath is a series of discontinuous gravel ridges which extends along the floor of the Brothock valley (Rice, 1959, fig. 5). South of St Vigeans, and north of the limit of 1:50 000 Sheet 49, this esker changes direction and, leaving the Brothock valley, climbs through the valley of the Hercules Den Burn into the western outskirts of Arbroath. Here it borders the western limit of the area now mapped as late-Glacial marine deposits and forms the prominent Keptie Hills [NO 634 413]. There is a terminal southern outlier at Dishland Hill [NO 637 405]. South of the High School [NO 636 409] small exposures show gravel with some included blocks over 0.3 m in diameter. West of the Keptie Hills the surface relief is more subdued and this is also true of the area of sand and gravel east of the Brothock Water, between Arbroath and Whiting Ness [NO 660 412].

The sand and gravel deposits near Mains of Kelly [NO 597 392] were probably laid down within the coastal ice by meltwaters from north of the Craigend ridge passing through the system of submarginal channels [NO 584 395] in the Craigend area. Kame and kettle topography is well developed in the area of Three Mile Wood [NO 602 392]. The deposits along the Rottenraw Burn west of Arbirlot probably relate to a later phase of ice-recession when the gravels adjoining the lower Elliot Water below Arbirlot may also have been laid down.

Spreads of sand and gravel exhibiting subdued relief occur on the ground between Muirdrum and Carnoustie and also north of Pitlivie [NO 552 384]. South of the Carlogie Hotel a terrace [NO 561 360] near the upper limit of the deposit and with a mean elevation of 32.9 m above OD has been interpreted by Cullingford and Smith (1980, table 1) as a raised shoreline fragment, and it is possible that marine reworking of fluvioglacial material has taken place at a higher level than the mapped upper limit of late-Glacial marine deposits in this area.

Dundee–Monifieth area

Extensive areas of fluvioglacial sand and gravel occur in and north of the valley of the Dighty Water west of Monifieth. The deposits include an esker which can be traced, as a chain of gravel ridges either in isolation or in association with broader spreads of sand and gravel from west of Kirkton of Strathmartine to Balmuir [NO 402 343]. Thereafter the esker leaves the valley of the Dighty Water, continuing eastwards to the north of Whitfield [NO 437 334] to join the broad area of sand and gravel east of the Fithie Burn [NO 452 342], and terminates in the ground west of Ethiebeaton. During and after the episode of sub-glacial drainage represented by the esker, meltwaters passed east of Ethiebeaton to lay down sand and gravel within the coastal ice north-east of Monifieth. A fan of sand and gravel now represented by terraces developed on both sides of the embayment north-east of Balmossie probably relates to a final stage in the use of this meltwater route. The fan gives way eastwards to a broad belt of sand and gravel extending as far east as Barry [NO 536 344]. This ground is undulating with a few closed hollows and represents dead ice topography unmodified by marine action except in a few places along its southern edge. As at Arbroath and St Fort the evidence here points to a co-existence of glacier ice over the considerable range of levels of the late-Glacial sea represented by raised shorelines and littoral deposits north-east of Carnoustie. The sand and gravel deposits are currently worked west of Cotside in a pit [NO 528 342] where a working face, up to 12 m high, displays fine-grained sand with silty clay bands overlain by coarse gravel.

Melting of ice in the valley of the Dighty Water eventually allowed meltwaters to pass by this more southerly route. The kettled spreads of sand and gravel between Linlathen [NO 462 329] and Monifieth, and the terraces which flank the Dighty Water between Mill of Mains and Baluniefield [NO 449 324], were laid down probably in association with immobile ice in the bottom of the valley. Dead ice in the Monifieth area probably checked marine access thus protecting the glacial landforms around Balmossie Mill [NO 476 325] from destruction. Consistent with this view, channels draining southwards east of Claypots Castle [NO 452 319] are thought to represent an escape route for the contemporaneous drainage of the Dighty valley, impounded by the ice at Monifieth.

Invergowrie area

Of the deposits in the ground west of Dundee, a poorly exposed spread on the north side of Balgay Hill [NO 377 307] is probably the oldest. The fluvioglacial sand and gravel on lower ground nearer Invergowrie was probably laid down after the earlier meltwater route north of Balgay Hill was abandoned, as the ice receded from the high ground. The spread between Ninewells [NO 365 300] and the neighbourhood of Mains of Gray [NO 336 321] has given rise to moundy topography in places. A temporary section at a point

[NO 3632 3051] north-east of Invergowrie showed 0.6 m of coarse gravel on 0.9 m of red sand resting on coarse bouldery drift. A short distance uphill from the limit of raised marine deposits east of Invergowrie a sewer trench [NO 3539 3049] revealed 2.1 m of coarse gravel with included blocks up to 1 m in diameter near the top.

Between Invergowrie and Huntly Farm [NO 324 311], sand and gravel gives rise to a belt of undulating ground south of the Foulis Burn. Farther west, south of Carmichael's Farm, the upper surface [NO 305 310] of a plateau of sand and gravel is broken by large kettleholes and gives way towards Huntly Farm to moundy, dead ice topography. An embayment holding the poorly-drained depression known as Marl Mire [NO 302 308] on the western side of the plateau probably marks the general eastern limit of Firth of Tay ice when sand and gravel was being laid down in association with dead ice cut off in the Benvie depression. The plateau surface, at 37 m above OD is almost certainly related to a former position of the late-Glacial sea at a lower level in the Invergowrie area, but the view of Cullingford (in Gray and Lowe, 1977) that the plateau forms part of a fragmented fluvioglacial terrace descending eastwards and seawards to as low as 21 m above OD cannot be substantiated in the absence of clear measurable features in the Benvie area. The existence within the sand and gravel north-west of Invergowrie of a small closed depression [NO 343 308] with a rim at rather over 15 m above OD suggests that dead ice may have survived here until the late-Glacial sea had fallen below the Main Perth Shoreline which stands at 19 to 20 m above OD east of Invergowrie (Cullingford in Gray and Lowe, 1977, p. 30).

Glencarse–Perth area (Figure 17)

A small area of fluvioglacial sand and gravel near the eastern end of Glen Carse has been worked intermittently in a pit [NO 191 225] west-south-west of Glencarse House. North-west of Glencarse village there are long disused workings in another small spread [NO 193 217].

A considerable proportion of the deposit of fluvioglacial sand with subordinate gravel, which originally formed an extensive undulating plateau south-east of Kinfauns Church, was removed for use in foundations during the reconstruction of the Perth–Dundee road. Made ground which now replaces the original sand and gravel consists largely of Carse Clay excavated from the line of the new carriageways. At one stage in the removal of the sand and gravel a thin intercalated bed of till-like material was observed within the fluvioglacial deposit. Towards the eastern side of the former sand and gravel deposit, eastwards-dipping foresets capped by beds of low inclination were observed.

On the east side of the River Tay south of Barnhill, near Perth, a terrace of coarse fluvioglacial sand and gravel has been extensively quarried. Notable features of this deposit are, as at Kinfauns, eastward-dipping foresets capped by flat-lying sand and gravel. The scale, however, is larger and the foresets are as much as 9 m high. The deposit is interpreted as a kame-terrace developed on the margin of the Tay glacier where coarse meltwater deposits were discharged directly into the late-Glacial sea. The height of the foresets is possibly a measure of the minimum contemporaneous depth of water and the upper limit of the foresets at 37 m OD, is probably a short distance below the marine limit.

Lower Tay valley (Figure 18)

Deposits represented on the geological map as fluvioglacial sand and gravel in the ground north and north-west of Perth include two distinct categories which relate to two phases of deposition separated locally by a marine incursion of late-Glacial age. Initially, during the melting of the great bulk of the glacier ice in these areas, moundy accumulations were laid down subglacially and englacially near Almondbank and Balmblair [NO 072 282], in the ground west of the confluence of the Ordie and Shochie burns, and in the Bankfoot–Luncarty area. The sediment was probably in large part transported by the ice-controlled drainage which cut the series of south- and east-draining meltwater channels west of the lower Tay valley. Subsequently when the ice had largely disappeared, the moundy deposits were in part submerged by the late-Glacial sea which flooded the lower parts of the Almond and Tay valleys, extending as far north as the Stormontfield- Luncarty area (Figure 14). Fine-grained marine sediment of glacial derivation accumulated to considerable thickness, filling the original marine inlets, and was eventually covered by spreads of younger sand and gravel carried downstream by proglacial rivers fed by glaciers then still existing in the upper Almond and Tay valleys. The entry of much of the coarse sediment into the Tay valley below Stanley was probably delayed until final disintegration of glacier ice which had blocked the valley between Stanley and Cargill. The drainage of the Upper Tay valley and of Strathmore, formerly constrained to flow eastwards through a spillway near Forfar (Sissons, 1963; Paterson, 1974), was thereafter able to pass southwards towards Perth. As the confluent Almond and Tay deltas advanced into the sea, the original upper arms of the late-Glacial Tay estuary were progressively replaced by alluvial plains. Sand and gravel overlying marine clay in the Huntingtower and Scone areas probably passes southwards into the higher parts of the sequence of late-Glacial marine clays and fine-grained sand preserved within the city of Perth, mainly west of the River Tay. As later erosion has largely destroyed the late-Glacial delta at Perth the development of the delta in relation to the late-Glacial sea is not known in detail. However, farther upstream, successively lower sand and gravel terraces in the lower Tay valley and in the Almond, Ordie and Shochie valleys were formed in response to the progressive fall in late-Glacial sea level, during which the coarser sediments of the pro-grading delta doubtless continued to advance downstream.

The fluvioglacial terrace deposits laid down in the second phase of deposition are in general lacking in dead-ice features but around Almondbank, in an area which is of outstanding interest (Paterson, 1974), a terrace surface is extensively modified by kettleholes and other dead-ice embayments. In a section formerly exposed on the north bank of the River Almond near Berthapark the sand and gravel of this terrace overlies laminated marine clays which in turn rest on till. On the basis of this tripartite sequence, and of comparable sequences of sand and gravel upon marine clay in Strathearn, Simpson (1933) postulated a glacial readvance, the 'Perth Readvance', which was thought to post-date the marine clays of the lower Almond valley and Strathearn. The concept was adapted and extended by Sissons (1963) and the existence at Almondbank of kettleholes in deposits which undoubtedly post-date the marine clays appeared to constitute good supporting evidence. However, during the resurvey of ground within the limits of the supposed readvance near Stirling, Francis and others (1970) failed to detect a separate till which might be ascribed to any such episode. They pointed out also that the superposition of sand and gravel on marine clays and silts may be adequately explained by the progressive seaward extension of a late-Glacial marine delta. The general concept of the Perth Readvance was examined and rejected by Paterson (1974) who ascribed the formation of the kettleholes in the terrace east of Almondbank to the melting-out of residual dead-ice masses enclosed within the marine deposits, rather than within the terrace deposits. Browne (1980b) later related the kettleholes to the melting of ice enclosed in the still older moundy deposits of sand and gravel, a view followed in the present account. A similar explanation may be applicable to apparent dead-ice hollows in terraces in the Marlehall area north-west of Luncarty.

Whatever the age of envelopment of dead-ice masses near Almondbank may be, it is clear that final melting of the ice took place after the deposition of the kettled terrace, a feature which is probably contemporaneous with the Main Perth Shoreline (Sissons and Smith, 1965). A younger terrace near Berthapark, some 3 m lower, is not pitted by kettleholes. It is therefore presumed that the dead ice may have not long survived the fall of the sea below the level of the Main Perth Shoreline, a conclusion which is in general accord with evidence from the Wormit–Leuchars and Benvie areas. Fluvioglacial deposits in lower Strathearn and in the upper Tay valley and western Strathmore have been described by Browne (1980b) and Paterson (1974).

Late-Glacial marine deposits

During and after deglaciation large volumes of sediment were carried into the late-Glacial sea, either by glacial meltwaters or by normal streams actively eroding a post-glaciation land surface which was only slowly acquiring a cover of vegetation. The marine deposits laid down at this time include clays and silts of intertidal to subtidal origin, coarser littoral deposits associated with raised beaches and deltaic deposits. There is commonly a passage landwards into fluvioglacial or fluvial deposits.

In the more seaward areas two principal divisions have been recognised within the marine clays, the older Errol Beds (Peacock in Gemmell, 1975) and the younger Powgavie Clay (Paterson and others, 1981). The Errol Beds are characterised by the arctic fauna long known from Errol and elsewhere on the east coast of Scotland (Jamieson, 1865; Brown, 1867; Brady and others, 1874). The Powgavie Clay contains a significantly less cold late-Glacial fauna comparable with that contained in the older part of the Clyde Beds of the west of Scotland which predates the Loch-Lomond Stadial (Peacock and others in Gray and Lowe, 1977; Graham and Wilkinson, 1978; Peacock and others, 1978). The marine sequence in the Tay–Earn area becomes progressively coarser when traced westwards. In this direction the Errol Beds and particularly the Powgavie Clay pass laterally into the deltaic sands and silts of the Culfargie Beds (Paterson and others, 1981).

Errol Beds

The Errol Beds are an assemblage of marine clays, silts and sands which are widely distributed in coastal areas of east-central Scotland and which contain a distinctive arctic fauna. They are thought to have been laid down between the late Devensian maximum at 18 000 years BP and about 13 500 years BP. The latter date marks the approximate age of the change from arctic conditions to boreal conditions in near-shore marine sediments in Scotland, a transition which is believed to coincide with that recorded about 13 000 to 13 500 BP by the planktonic foraminifera in deep cores west of the British Isles and by the Coleoptera on land (Ruddiman and McIntyre, 1973; Coope, 1977).

Especially characteristic of the Errol Beds are the laminated, generally red-brown marine clays, the 'plastic clays' of Chisholm (in Forsyth and Chisholm, 1977, p. 237), which were laid down in intertidal to subtidal conditions. These are locally intercalated with coarser sediment where conditions gave rise to an input of sand into the late-Glacial sea. As a consequence of the generally falling relative sea level, younger littoral sands of this period were not uncommonly laid down with discordance on older subtidal clays, sometimes occupying channels cut in them (cf. Chisholm in Forsyth and Chisholm, 1977, p. 246). Within the fill of the Tay–Earn estuary, however, there is a general and gradual upward-coarsening within the Errol Beds which reflects a major input of glacially derived sediment into the late-Glacial estuary.

The Errol Beds are best known in the Carse of Gowrie where the older records relating to the former claypits near Errol have been supplemented in recent years by information obtained in a programme of drilling by the British Geological Survey. The Errol Beds were formerly exposed at the Inchcoonans claypit [NO 241 233], where the sequence and fauna of the 'Arctic Clay of Errol' were described by Davidson (1932a). Although the section has been obscured and eventually lost by infilling in recent years, Davidson's general lithological sequence has been confirmed by deep augerholes put down during the resurvey of the district and the succession has been subdivided (Paterson and others, 1981, p. 10). A revised statement of the Inchcoonans sequence is as follows:

Scattered, glacially-striated stones, which occur here and also near Gallowflat [NO 213 207] in the only currently open claypit in the district, are considered to have been dropped into the clay from floating ice-masses in the late-Glacial sea. Intercalations of gravel within the clay in an area of aberrant development at the western end of the pit at Inchcoonans (Paterson and others, 1981, fig. 5), where Davidson (1932a) recorded large-scale disturbance of the bedding, may have resulted from the release by an iceberg of a large quantity of debris of glacial origin.

The arctic fauna recorded from Inchcoonans claypit by Davidson (1932a, pp. 55–68) appears to have been largely confined to the two lower divisions (Graham in Paterson and others, 1981, p. 27; Graham and Gregory, 1981, pp. 215–222) and there is no evidence that any part of that fauna was derived from the uppermost division C2. The arctic molluscan fauna comprises the gastropods Buccinum cyaneum and Lunatia pallida?, and the bivalves Astarte borealis, Hiatella arctica, Macomia calcarea, Musculus latvigatus, M. niger, Palliolum groenlandicum, Portlandia arctica, Thracia distorta and T. cf. septentrionalis. Recent investigations at Inchcoonans detected a microfauna in lithological divisions A, B and Cl but not in C2 (Paterson and others, 1981, fig. 9). The arctic ostracods Krithe glacialis, Rabilimus mirabilis and Cytheropteron montrosiense are particularly characteristic of this sequence.

In the absence of fauna, possibly due to leaching of carbonate from that part of the sequence close to the surface, the affinities of the uppermost division C2 at Inchcoonans are uncertain. It is, however, known (Paterson and others, 1981, p. 11) that from elsewhere in the Errol area, probably from now disused claypits near Fala, Brady and others (1874, p. 617) listed a microfauna which, besides the species recorded from Inchcoonans, also included a considerable number of ostracods indicative of more temperate conditions. Most of these boreal ostracods are known to occur in the Clyde Beds of the west of Scotland and are compatible with the micro-fauna of the Powgavie Clay (p.86). It has therefore been suggested (Paterson and others, 1981, p. 11) that in the old Fala claypits the Errol Beds were overlain by a local correlative of the Powgavie Clay, and that the fauna recorded from Errol by Brady and others (1874) was a mixed one including both arctic and boreal assemblages. On the basis that the presumed equivalent of the Powgavie Clay at Fala may be represented at Inchcoonans by the uppermost division C2, it is proposed to exclude this division from the Errol Beds. The existence of an erosion surface below C2 is consistent with this view.

The distribution of the elements of the Errol Beds micro-fauna within the three lithological divisions A, B and C 1, which were closely sampled in augerholes, permits the recognition of three faunal divisions within the Errol Beds (Paterson and others, 1981, p. 11). These faunal divisions approximate to the lithological divisions. The lower faunal unit is characterised in relation to the foraminifera by a dominance of Elphidium clavatum over E. bartletti. In the middle faunal unit E. clavatum yields dominance to E. bartletti. There is a marked fall off in the ostracods Rabilimis mirabilis and Cytheropteron arcuatum at the top of this unit. The ostracod Krithe glacialis, characteristic of the lower and middle units, is absent from the upper unit, while the foraminifer E. clavatum regains dominance over E. bartletti.

Reddish brown clay, closely resembling the basal division (A) of the sequence at Inchcoonans, and corresponding to the 'plastic clay' of the St Andrews area, is widely distributed in the district. The deposit rests upon bedrock, till or fluvioglacial sand and gravel and probably formed as a veneer, now missing locally due to erosion or slumping, but which potentially covered the late-Glacial sea-floor up to the marine limit. The clay occurs at altitudes ranging from 51 m below OD in the IGS Culfargie Borehole up to about 40 m in above OD near Abernethy (Armstrong and others in Gemmell, 1975). This evidence suggests that the late-Glacial sea was more than 90 m deep in parts of the Tay-Earn estuary at the time of deglaciation.

It is probable that deposition of the reddish brown clay took place by precipitation from a thin freshwater surface layer maintained by glacial meltwaters from the retreating glacier-front as McManus. (1972) suggested. Subsequent shallowing of the estuary was caused mainly by the general fall of relative sea level but also in partbytheaccumulation of sediment. The change of colour at the top of the basal, lithological division at Inchcoonans may be related to the recession of the Strathearn ice-front towards the Highlands, which caused increased input into the estuary of material of Highland origin, replacing the hitherto dominant red sediment of Devonian derivation. The further upward change to the silty clay of division C1 at Inchcoonans probably reflects the contemporaneous advance of deltas into the upper parts of the late-Glacial estuary.

Powgavie Clay

A thick deposit of marine clay, containing subangular ice-rafted clasts, but otherwise lithologically and faunally distinct from the Errol Beds, was encountered in the IGS Powgavie boreholes, and was accordingly named Powgavie Clay (Paterson and others, 1981, p. 13). Lithologically the Powgavie Clay is characterised throughout by the occurrence of graded units, each comprising thin basal layers of pale coloured silt or sand passing up into dark brownish grey clay. In general the deposit is somewhat foetid, with sporadic laminae or thicker beds stained black by iron sulphide, and this suggests that at the time of deposition there was a higher organic content in the late-Glacial estuary than when the Errol Beds were laid down. The fauna of the Powgavie Clay is appreciably more abundant and diverse than that of the Errol Beds and as a whole is indicative of significantly less frigid climatic conditions. The fauna resembles an even more diverse fauna (cf. Brady and others, 1874; Peacock and others in Gray and Lowe, 1977; Peacock and others, 1978; Graham and Wilkinson, 1978) recorded in the West of Scotland from the Clyde Beds, a deposit which probably in-cludes time-equivalents of the Powgavie Clay.

The molluscs of the Powgavie Clay include the gastropods Lora sp., Omalogyra atomus, Onoba semicostata, Retusa obtusa, Skeneopsis planorbis and Trophonopsis clathratus and the bivalves Astarte (Tridonta) montagui, Macoma calcarea, Mytilus edulis, Nucula tenuis, Nuculana pernula, Spisula elliptica, Tellina sp., Thyasira sp. and Yoldiella lenticula. Although richer in species and numbers than the molluscan fauna of the Errol Beds, the molluscs of the Powgavie Clay in themselves do not present much evidence of a pronounced climatic amelioration. The molluscan species present are indicative of shallow water conditions with sea temperatures considerably colder than at present (Graham in Paterson and others, 1981), as is consistent with the presence of dropstones. The change from Errol Beds to Powgavie Clay is, however, most apparent in the marine microfauna. Although E. clavatum, Acanthocythereis dunelmensis and Eucytheridea punctillata, all tolerant of cold water conditions, are dominant, these are accompanied by small numbers of species indicative of more temperate conditions, notably Elphidium williamsoni, E. albiumbilicatum, Buccella frigida, Protelphidium anglicum and Leptocythere sp. Of the arctic species characteristic of the Errol Beds K. glacialis is absent and C. montrosiense and R. mirabilis are only sporadically represented, and may be reworked from the Errol Beds. The relation of the Powgavie Clay to the Errol Beds is not clearly demonstrable in any known locality. In the Powgavie boreholes the Errol Beds are unfortunately absent, and the Powgavie Clay rests directly on fluvioglacial sand and gravel (Figure 20). The absence of any identifiable representative of the Errol Clay in the Powgavie boreholes has been explained as a result of mass movement of sediment on the late-Glacial sea floor before deposition of the Powgavie Clay (Paterson and others, 1981). Any investigation of the relation of the Powgavie Clay to the Errol Beds at the site of the IGS Burn-side Borehole was precluded by poor core recovery in the lower part of the borehole. In a borehole at Longforgan [NO 3205 2867], however, the Errol Beds occur in a sequence below sand containing a microfauna comparable with that of the Powgavie Clay (unpublished information). This sand is referable to Culfargie Beds which represent a coarse facies laid down at the same general period as the Powgavie Clay.

Away from the type locality, the Powgavie Clay has been identified with certainty only in the Burnside Borehole. It is not improbable, however, that the Powgavie Clay is also represented near Errol where a number of ostracod species, such as Cytheropteron latissimum, Elofsonella concinna, Hemicythere villosa, Hirschmannia viridis and Schlerochilus contortus, which are typical of the Clyde Beds, were listed, together with the fauna now known to occur in the Errol Beds, by Brady and others (1874) from an unknown locality, presumably one of the older claypits near Fala.

It is probable that by the time of the climatic amelioration which initiated the development of the marine fauna of the Powgavie Clay, the late-Glacial sea had fallen considerably in comparison with the relative sea levels existing during the deposition of the Errol Beds. The exceptional prominence of the Main Perth Shoreline has been attributed by Paterson (1974) to shoreline formation during a period of marked climatic amelioration when global ice-melt may have enabled eustatic sea level rise to balance local glacio-isostatic recovery. The coincidence of such an amelioration with the faunal change represented by the incoming of the Powgavie Clay (Clyde Beds) fauna and with the climatic transitions recorded at about 13 000 to 13 500 years BP (Ruddiman and McIntyre, 1973) and at the commencement of the Windermere Interstadial (Coope, 1977) seems not improbable. However, although the conclusion must be that the Powgavie Clay was laid down in the period following the formation of the Main Perth Shoreline, all known and presumed occurrences of the clay (including Fala and Inch-coonans) lie well below this shoreline.

Culfargie Beds

At the western end of the Firth of Tay, in lower Strathearn and in the lower Tay valley, silty clay and silt of the Errol Beds are overlain at the surface by a largely unfossiliferous deposit of sand with subordinate silt and gravel. The deposit was penetrated in the IGS Culfargie Borehole and was accordingly named the Culfargie Beds (Paterson and others, 1981, p. 14). It is considered that the Culfargie Beds pass seawards into the Powgavie Clay of the Carse of Gowrie area. The Culfargie Beds may be regarded as the foreset deposits of large prograding deltas descending the Earn and Tay valleys. The Powgavie Clay represents the contem-poraneous prodeltaic deposits.

In the Carse of Gowrie, sands referable to the Culfargie Beds occur above considerable thicknesses of Powgavie Clay in the Powgavie and Burnside IGS boreholes (Paterson and others, 1981), and here they contain a fauna closely resembling that of the Powgavie Clay. The base of the Culfargie Beds therefore appears to be markedly diachronous with sand deposition extending gradually farther eastwards in the post-Errol Beds period.

At Culfargie, the base of the sands cannot be related precisely to the climatic amelioration marked by the incoming of the Powgavie Clay fauna into the late-Glacial estuary, and it is possible that the earlier part of the Culfargie Beds here may have been laid down at the same time as the youngest parts of the Errol Beds in the Carse of Gowrie area. The advance into the late-Glacial estuary of the sands of the Culfargie Beds appears to have been foreshadowed by upward coarsening in sequences of the Errol Beds, as at Inchcoonans.

The Culfargie Beds appear at surface in the form of sand and silt terraces which are especially conspicuous on the south side of Lower Strathearn and in the Glencarse–Errol area of the Carse of Gowrie (Figure 17). The terraces which constitute the Lower Perth Shorelines (Cullingford in Gray and Lowe, 1977) mark successive stages in the reworking of the Culfargie Beds deltaic deposits in response to the general fall of relative sea level below the Main Perth Shoreline (Figure 15). Farther upstream, in the Lower Tay and Almond valleys north of Perth and in the area north of Dunning in Strathearn, sand and gravel terraces which overlie the Errol Beds may be regarded as proximal developments of the Culfargie Beds (Paterson and others, 1981). These deposits were however almost certainly formed above sea level and are fluvial rather than marine.

The erosion which accompanied and followed the deposition of the Culfargie Beds largely removed deltaic deposits of this period from the lower Tay valley below Perth. An indication of the level formerly attained by the delta-top in this area is however provided by an extensive, much dissected sand terrace at 19.6 m above OD near Glencarse House. This terrace, considerably higher than the Lower Perth Shorelines, is presumed to have originated from sediment carried down the lower Tay valley, and it is therefore inferred that the surface of the terrace formerly extended up that valley, towards Perth. However, in lower Strathearn, the Culfargie Beds are not found at heights greater than 16.7 m above OD, even in the embayment near Brickhall (Figure 17) where the deposits are likely to have been shielded from the full effects of later erosion. This level is appreciably lower than the original height of the delta surface at Glencarse House, particularly when allowance is made for subsequent isostatic tilt. It is therefore possible that at the time of formation of the Glencarse House terrace, the level of the delta-top actually descended up Strathearn from a point near Inchyra where the Tay delta emerged into the broad late-Glacial Tay–Earn estuary. This implies that as well as advancing seawards into the Glencarse area, the Tay delta also advanced westwards into lower Strathearn at this stage. If this is so, the contemporaneous Earn delta-front, advancing from the west, entered the late-Glacial sea west of Bridge of Earn. At later stages the Lower Perth Shorelines were cut into the deposits of the combined Tay–Earn delta with the production of the existing terraces.

The Culfargie Beds on the southern side of the Earn valley between Bridge of Earn and Abernethy were investigated by the Brickhall and Jamesfield boreholes (Paterson and others, 1981, p. 2). The first of these, sited on a terrace at a height of 15 to 16.5 m above OD, which Cullingford assigned to Lower Perth Shoreline LP-1, showed that the terrace surface was underlain by 1.96 m of silt and clay which rested on pebbly sand. Excavations [NO 140 163] for a pipeline about 1 km SE of the borehole site showed that the terrace deposit, which passed downward from silt to coarse sand, rested upon an uneven surface eroded in till and the basal red clay of the Errol Beds. A similar sequence, consisting of silt on grey sand, was proved by the Jamesfield Borehole, which was sited at a height of 11.4 m above OD on an extensive terrace, assigned by Cullingford (in Gray and Lowe, 1977, fig. 1, fragment 5) to Lower Perth Shoreline LP-3. In excavations [NO 178 163], about 2 km SW of the borehole site, the silts of the terrace deposit were seen to be closely laminated and penetrated by rusty tube-like markings, considered to be root-traces of Phragmites. The terrace deposit, which is therefore believed to have been laid down in intertidal conditions, rests on an uneven erosion surface, probably associated with a lower sea level than the overlying silts, which are thought to have been laid down during a subsequent transgression. It seems probable that each of the Lower Perth Shorelines relates to a minor transgression which interrupted the general late-Glacial fall of sea level (cf. Cullingford, 1971, p. 281).

The Lower Perth Shorelines in the area around Glencarse (Figure 17) were cut into the sandy and silty deposits of the formerly much more extensive Glencarse House terrace, and it appears that shorelines LP-1 and LP-2 (Cullingford in Gray and Lowe, 1977) were developed throughout the area of low ground north-west of the Errol 'island'. The later shorelines LP-3 and LP-4, however, were excluded from the area by the fall of sea level, shoreline LP-4 being largely concealed below the post-Glacial Carse Clay of the Kilspindie–Rait area (Cullingford in Gray and Lowe, 1977). Except near Glencarse, where grey, yellow-weathering, fine-grained, ripple-bedded sand was exposed in excavations along the Perth–Dundee trunk road, the material associated with the shorelines is mainly silt, probably overlying erosion surfaces cut into earlier deposits, as has been observed in the case of similar terraces in Strathearn.

North-west of Perth, the Huntingtower Borehole (Paterson and others, 1981, p. 2) showed Errol Beds overlain by sands considered to be Culfargie Beds at 18.5 m below OD whereas in the Almond Bridge Borehole (Paterson and others, 1981, fig. 2) and Borehole 26, only 1 km away, silts and clays of the highest faunal division of the Errol Beds occur up to at least 3.2 m above OD. The diachronous relationship between the Culfargie Beds and the Errol Beds is thus very marked here and it is suggested that, at Huntingtower, sedimentation was dominated by the early local entry of somewhat coarser material from the Almond valley, whereas at Almond Bridge the deposition of finer-grained sediment from the larger Tay valley continued for a longer period.

Erosion features cut during later part of late-glacial period

On the abandonment of the lowest of the Lower Perth Shorelines, the sea fell below the level of the present upper surface of the post-Glacial Carse Clay, which now largely conceals the deposits and features formed during the later part of the late-Glacial period. As the sea fell, an extensive, irregular erosion surface proved by drilling and probably formed by fluvial action rather than by marine planation, was eventually cut at between 3 m above and 5 m below OD into the Culfargie Beds in lower Strathearn ((Figure 19), section A, and Cullingford in Gray and Lowe, 1977). In the lower Tay valley, gravel-covered benches at or about Ordnance Datum were developed at sites near the new Friarton Bridge ((Figure 19), section B). Boreholes in the lower Almond valley between Huntingtower and Almond Bridge (Paterson and others, 1981, fig. 7, sec. 4) show an erosion surface at a height of about 5 m above OD.

Associated with these essentially planar erosion surfaces in lower Strathearn and in the lower Tay valley are channels, cut to much greater depths into the Errol Beds and Culfargie Beds. The channels are filled mainly with gravel and are considered to be of later date, formed as the sea level continued to fall. At Bridge of Earn ((Figure 19), section A), a channel is cut to a depth of more than 13 m below OD. South of Perth, at the site of the new Friarton Bridge, a channel descends to at least 29 m below OD, while in the Bogle Bridge Borehole [NO 099 257] put down north of the city by the Distillers Company, gravels, presumed to lie within a channel cut into laminated clays of the Errol Beds, descend to 19 m below OD.

In the Carse of Gowrie area, evidence of deep channelling, comparable with that in lower Strathearn and the lower Tay valley at Perth, was obtained in the Kingston IGS Borehole. Between levels of 4 m above and 17 m below OD, this borehole penetrated fine-grained sand, the Kingston Sand (Paterson and others, 1981), which overlies Errol Beds and therefore appears to occupy a channel incised through Culfargie Beds and Powgavie Clay.

A notable consequence of the fall of sea level during the late-Glacial period was the erosion of gullies in the now raised late-Glacial marine deposits in the Tay and Earn valleys. These are especially prominent in the Glencarse area, in lower Strathearn and in the lower Tay and Almond valleys. The gullying may have begun as soon as raised marine deposits were uplifted, and early deposits in Stratheden are affected. The main phase of erosion was undoubtedly after the deposition of the Culfargie Beds. These gullies do not appear to relate to streams at a higher level and frequently peter out before reaching the uphill limit of flat-topped deposits. It is suspected that they may have been cut by headward erosion in periglacial conditions.

The planar erosion surfaces in lower Strathearn and in the lower Tay valley probably relate to a sea level comparable with that marked by the Main Lateglacial Shoreline of the Forth valley. In the Carse of Gowrie this shoreline is probably represented by a thin bed of gravel, the Port Allen Gravel, which rests upon Errol Beds and Culfargie Beds in the Port Allen (Paterson and others, 1981, fig. 2) and Burnside (Figure 20) boreholes at heights of 3.2 and 4 m below OD respectively, and appears to be associated with an erosion surface cut into older deposits. The existence of shell debris in the gravel suggests that the erosion surface was cut by marine action.

The Main Lateglacial Shoreline has been considered to approximate to the lowest level reached by the sea in the late-Glacial period and to have been formed during the cold period of the Loch Lomond Stadial (Sissons, 1966). However, the depth of the channels in the Tay–Earn area appears to raise the possibility that the sea level may have fallen appreciably below the position of the Main Lateglacial Shoreline (McManus, 1972). Thus while the period of lowest sea level, that is, the sea level associated with the cutting of the deep channels, was almost certainly during the Loch Lomond Stadial, the Main Lateglacial Shoreline may be of a different age. It has been suggested (Paterson and others, 1981) that the shoreline may be referable to the period of a short-lived climatic amelioration dated at about 11 000 years BP and immediately preceding the Loch Lomond Stadial at Ardyne and Lochgilphead (Peacock and others, 1978).

It is possible, however, that the period of relative sea level stability necessary for the erosion of so prominent a feature may have extended over the later part of the temperate Windermere Interstadial, perhaps from about 12 000 to 11 000 years BP, while glacio-isostatic recovery was generally balanced by a eustatic rise.

Earn and Friarton gravels and earliest Carey beds

The sand and gravel infillings of the channels described above in lower Strathearn and in the lower Tay valley cannot, on existing evidence, be separated from the thinner spreads of coarse material on the flanking erosion benches. The coarse channel-fill and the thinner marginal deposits are therefore assigned in each valley to a single stratigraphical unit which is termed the Earn Gravel in Strathearn and the Friarton Gravel in the Tay valley. At Bridge of Earn, away from the channel, the Earn Gravel is little more than one metre thick ((Figure 19), section A). The Friarton Gravel, both thicker and coarser than its Strathearn equivalent ((Figure 19), section B), can be traced upstream from Friarton beneath the Carse Clay by means of boreholes (McManus, 1972; Cullingford in Gray and Lowe, 1977), and occurs directly below a thin surface layer of alluvium in the lower Almond valley, north-west of Perth (Paterson and others, 1981, fig. 7). It is presumed that in general these deposits were laid down, or at least that their formation was completed, as the sea rose from its lowest level, and it is probable that with further sea level rise, estuarine conditions began to extend up both the Earn and Tay valleys.

In lower Strathearn, the Earn Gravel is overlain by up to 8 m of sand and silt with thin bands of clay from which no fauna has yet been recovered ((Figure 19), section A). This deposit, named the Carey Beds (Paterson and others, 1981), is exposed below the post-Glacial Sub-Carse Peat at a number of localities, notably at a riverside section [NO 1747 1703] near Carey. It is probably of estuarine origin and appears to have an upper limit at about 7 m above OD. No representative of the Carey Beds has however yet been recognised in the lower Tay valley, where the Friarton Gravel is overlain by the Sub-Carse Peat or by still younger deposits ((Figure 19), section B).

Deposition of the Carey Beds to their upper limit took place some time prior to 9640 years BP, the basal date obtained from a peat (Callow and Hassall, 1970), which, near Carey, overlies a younger surface at 3.1 m above OD associated with a back-feature (Cullingford and others, 1980) cut into the Carey Beds. As the earliest Carey Beds post-date the Earn Gravel and therefore the lowest sea level associated with the Loch Lomond Stadial, they may be regarded as equivalent to the younger part of the silty clays of Unit 4 at Ardyne (Peacock and others, 1978), which are thought to have been deposited during a period of rising sea level ending after 10 245 years BP. The marine transgression to the highest levels attained by the Carey Beds therefore probably culminated between 10 245 and 9640 years BP. In the Forth valley contemporaneous deposits were laid down at levels up to the High Buried Beach (Sissons, 1966) which is thought to have co-existed with the ice of the Loch Lomond Stadial at its farthest limit near the Lake of Menteith.

The Carey Beds are represented in the Carse of Gowrie boreholes (Paterson and others, 1981, fig. 1) by up to 4.5 m of silty clays and sands with abundant Phragmites? root-traces, which lie between the Port Allen Gravel and the Sub-Carse Peat. In the Carse of Gowrie, as in lower Strathearn, the Carey Beds are not wholly of late-Glacial age, the younger sediments of this division being associated with marine features of post-Glacial age.

Younger Carey beds and Sub-Carse Peat

The rise of relative sea level, which culminated in the deposition of the older (late-Glacial) part of the Carey Beds up to the altitude of the High Buried Shoreline in lower Strathearn, was succeeded in the early part of the post-Glacial period by a general fall of relative sea level. This fall was probably interrupted by minor marine transgressions which led to the formation in lower Strathearn (Cullingford, 1971), and also perhaps in the Carse of Gowrie, of several terraces later buried by younger marine deposits. The terrace surfaces are probably underlain by thin transgression-deposits of silt and clay which collectively constitute the younger part of the Carey Beds and which were laid down between 10 000 and 8500 years BP.

Within the Carey Beds, which as a whole comprise the sediments between the Port Allen and Earn gravels and the Sub-Carse Peat and span the late-Glacial–post-Glacial boundary, there is little faunal evidence of the marked climatic amelioration known to have taken place after 10 000 years BP. In a borehole [NO 3205 2867] put down in 1979 at Longforgan, a microfauna from the Carey Beds is indicative of relatively cold, although by no means arctic, conditions (unpublished information).

As the marine terraces associated with the Carey Beds were successively abandoned by the sea, it became possible for their surfaces to be colonised by terrestrial vegetation. This resulted in the formation of a diachronous peat layer, known as the Sub-Carse Peat, which was eventually buried by the Carse Clay during the subsequent major rise of the sea to the Main Postglacial Shoreline. In the riverside exposure near Carey, one of the terraces, cut into pre-existing sediments of the Carey Beds, lies at a height of 3.1 m above OD. Its surface, probably underlain by a thin transgression-deposit, is overlain by peat from which basal radiocarbon dates of 9640 years BP. (Callow and Hassal, 1970) and 9524 years BP (Paterson and others, 1981, table 3) have been obtained. Such dates indicate, as pointed out by Cullingford (1971), that the Carey terrace corresponds to the Main Buried Shoreline of the Forth valley. A younger terrace, at a somewhat lower level near Innernethy [NO 189 178], is thought to represent the upper surface of a transgression-deposit (Cullingford and others, 1980). It is overlain by a peat with a basal date of 8505 years BP and is correlated with the Low Buried Beach of the Forth valley. The same terrace is probably present at the site of the Culfargie IGS Borehole (Paterson and others, 1981), where the Sub-Carse Peat rests on Carey Beds at about 2 m above OD.

In the Carse of Gowrie, the Sub-Carse Peat overlies older deposits over a wide range of altitudes (Figure 20) but the morphology of the buried terraces has not yet been investigated in detail. At the New Farm A IGS Borehole (Paterson and others, 1981), the surface of the Carey Beds, below a peat layer at 4.7 m above OD, possibly corresponds to the High Buried Shoreline. In the Port Allen, Powgavie and Burnside IGS boreholes (Paterson and others, 1981) and in the borehole put down at Longforgan, the Sub-Carse Peat overlies the Carey Beds at levels close to Ordnance Datum. Basal peat dates of 8320 and 8616 years BP suggest that the surface on which the peat rests corresponds to the Low Buried Shoreline. The altitude range of the peat base in these boreholes probably reflects the slope of the buried intertidal flat away from its back-feature.

The Sub-Carse Peat has been recorded at the edge of the modern channel in the Firth of Tay between Flisk Point [NO 312 227] and Ballinbreich [NO 273 202], and occurs below low water mark on the line of the Tay rail and road bridges at Dundee (McManus, 1972). Boreholes put down in 1961 at Dundee Head Post Office proved the Sub-Carse Peat at heights ranging from 2 to 3.5 m above OD. The peat occurs in lower Stratheden and in the Tentsmuir area and descends below present sea level in the Eden estuary (Chisholm, 1971). In all these more seaward areas no correlatives of the buried marine terraces of Strathearn have been identified, and any Carey Beds which may exist are not readily separable either from older deposits or, in the local absence of the Sub-Carse Peat, from younger deposits related to the subsequent transgression to the Main Post-glacial Shoreline.

Deposits of the Main Post-Glacial marine transgression and regression

The lowest sea level during the regression associated with the Carey Beds was probably reached about 8000 years BP (Cullingford and others, 1980). Thereafter a eustatic rise of sea level, at a rate which overtook glacio-isostatic uplift in central Scotland, caused a pronounced transgression and relative sea level rose to the Main Postglacial Shoreline. The marine deposits laid down during this transgression cover the earlier Carey Beds and Sub-Carse Peat, and occur in two principal forms. In the estuary areas of lower Strathearn and the Carse of Gowrie the deposits consist generally of intertidal silty clay, the Carse Clay, whereas in more exposed coastal areas the deposits are mainly of shelly sand and shingle.

In certain coastal areas however a two-fold sequence is apparent. Grey, planty clays which overlie the Sub-Carse Peat were laid down in lakes and marshes during the early stages of the transgression as the rising sea resulted in a parallel rise in the water table in adjacent low ground. As transgression continued, the coarser deposits laid down in intertidal and supratidal situations gradually encroached across the preexisting land surface and lay upon an erosion surface below which the earlier, finer-grained deposits of the transgression may be either preserved or eliminated (Chisholm, 1971).

On the exposed coast between Carnoustie and Arbroath, coarse beach deposits encroached landwards to the back feature of the Main Postglacial Shoreline and probably rest on a basal erosion surface cut into rock or till.

In the Tentsmuir and Buddon Ness areas the post-Glacial marine deposits consist over much of their extent of up to 12 m of sand which has an upper surface of generally subdued relief, masked to a considerable degree by the conspicuous topography associated with superimposed blown sand. In the more protected areas behind Tentsmuir, such as Moonzie and in lower Stratheden (Figure 16), the beach environment did not penetrate (Chisholm, 1971). The earlier deposits were nevertheless eroded, here, by tidal scour. Gravel, sand and silt were laid down in intertidal conditions above the resulting erosion surface. The flattish silty tracts so produced are reminiscent of the more extensive raised tidal flats which form the surface of the contemporaneous Carse Clay in the Carse of Gowrie and in lower Strathearn. Similar tidal flats occur in association with the highest levels attained by the transgression in a belt adjoining the back-feature of the Main Postglacial Shoreline in the St Michael's Wood area of Tentsmuir. Here however, there is a threefold sequence in which an upper layer of intertidal silt overlies the transgressive beach sands, and this change may be explained by the construction parallel to the shoreline and extending north of Cast of a sand bar behind which a lagoon formed (Figure 16). The situation was repeated farther east where a second elongate lagoon developed at a lower altitude as the sea level began to fall. North of the Tay, flat areas of silt occur close to the back feature west of Carnoustie, and a section in the Buddon Burn shows these deposits, with a basal coarse sand layer, resting on dark, planty clay not unlike the earliest transgressive deposits in north-east Fife.

The traces of former tidal channel systems have been detected on the carse-like flats at Tentsmuir and in lower Stratheden (Chisholm, 1971). In the Carse of Gowrie such channels probably determined the course of streams subsequently incised into the area's surface as sea level fell after the culmination of the transgression.

The transgression reached its highest level in the Tentsmuir area between 7605 and 5830 years BP (Chisholm, 1971) and in the Carse of Gowrie between 6600 and 6100 years BP (Morrison and others, 1981). Here and in lower Strathearn the Carse Clay, a grey silty clay with abundant rootlets, related to plant growth on the tidal flats during the transgression, was laid down to a maximum thickness of about 10 m. After the peak of the transgression the rate of eustatic sea level rise diminished, glacio-isostatic uplift reasserted its former dominance and relative sea level began to fall towards its present position. A number of marine terraces however were subsequently formed at levels below that of the Main Postglacial Shoreline as a result of fluctuations in the fall. These features were described under the name Lower Carse Shorelines by Cullingford (1971). Their surfaces are underlain by deposits resembling, though locally siltier than, the Carse Clay.

A terrace at 6 m OD in the Eden valley and another at about 4 m OD near Buddon Ness (Paterson, 1981) may correspond to certain of the Lower Carse shorelines (Figure 15) and to sea level fluctuations of this general period identified in NW England (Tooley, 1978). Near Inchyra (Figure 17) a radiometric date of 3425 ± 50, δ¹³C =–26‰ (NERC Radiocarbon Laboratory, East Kilbride) was obtained from wood near the base of a 2 m-thick terrace deposit. The surface of the terrace-deposit, at 6 m OD, relates to the Lower Carse Shoreline LC-1 (Cullingford in Gray and Lowe, 1977).

Kingston Sand and Buddon Sand

In the Kingston IGS Borehole the Carse Clay is directly underlain at a level of 4 m above OD by a thick deposit of sand, the Kingston Sand, which apparently occupies a deep channel incised through the Culfargie Beds and the Powgavie Clay into the Errol Beds. It is considered that the channel was cut during the period of low sea level during the Loch Lomond Stadial, when deep channels are believed to have been cut in lower Strathearn and the Tay valley at Perth. If this is not the case, a second period of drastically lowered sea level, in early post-Glacial times, is implied; this seems unlikely. The upper surface of the Kingston Sand is at a level which corresponds approximately, after allowance for isostatic tilt, with the highest levels reached by the Carey Beds in lower Strathearn. This might be taken as an indication that the Kingston Sand is of late-Glacial age, but the marine fauna contained by the sand does not support this conclusion. Samples from the Kingston Sand have yielded sparse but varied assemblages of foraminifera, molluscs and ostracods which are indicative of conditions distinctly more temperate than those suggested by the faunas in any of the late-Glacial deposits. A post-Glacial age for the bulk of the deposit appears likely. However, the bottom 5 m of the sand were not sampled and a late-Glacial age for this material cannot be excluded. The cold water indicator Elphidium clavatum, which is present in considerable numbers in both the Kingston Sand and the Carse Clay at Kingston, may have been derived from the late-Glacial deposits in which it is by far the commonest species.

The distribution and age of the Sub-Carse Peat demonstrate that the post-Glacial sea did not rise to the level of the top of the Kingston Sand before the time of the major transgression during which the Carse Clay was laid down. It follows that at least the upper part of Kingston Sand is equivalent in age to the lower part of the Carse Clay as developed elsewhere. This leaves the possibility that the lower part of the Kingston Sand contains equivalents of the younger (post-Glacial) part of the Carey Beds which comprises sediments laid down during the minor transgressions associated with the Main and Low buried shorelines.

The lateral extent of the sand body replacing the lower part of the Carse Clay at Kingston is known to be limited because an augerhole, put down at the seaward edge of the main Carse surface, 350 m SSE of the Kingston IGS Borehole, encountered only Carse Clay above the Sub-Carse Peat. It should be mentioned, however, that Cullingford (1971, p. 169) described a shelly sand layer within the Carse Clay north-west of Powgavie. The upper surface of this sand is at about 5 m above OD which is suggestive of correlation with the upper surface of the Kingston Sand.

A stratigraphical sequence similar to that revealed by the Kingston Borehole occurs near Buddon Ness, where the Buddon Sand, which contains a marine fauna resembling that from the Kingston Sand, rests upon Errol Beds (Paterson, 1981). The absence of any representative of the Powgavie Clay at Buddon Ness is probably a consequence of pre-Buddon Sand erosion. It is possible that the erosion surface associated with the Port Allen Gravel and with the Main Lateglacial Shoreline of the Forth valley (Sissons, 1974a) is represented in the Buddon Ness area by a bench cut into the Errol Beds at about — 8 m OD. Below this level the erosion surface, as at Kingston and elsewhere in the Tay–Earn area, descends to - 20 m, and may be associated with the low late-Glacial sea level postulated on p. 88. It is considered that both the Buddon Sand and Kingston Sand were laid down during post-Glacial times, in the course of the major marine transgression, during which the sea rose to the Main Postglacial Shoreline. That the sand-fill may have been laid down by tidal currents flowing into the Tay estuary is suggested by the evidence of large-scale foreset bedding detected in the Forth Beds in the Forth Approaches (Thomson, 1978, pl. 2a). The subhorizontal surface of the Buddon Sand at about 4 m above OD may mark a brief pause in the sea level rise, or possibly a short-lived regression.

Alluvium

Fluvial alluvium in the form of flood-plain deposits and terraced spreads is found along many of the streams and rivers, the most extensive tracts being in lower Strathearn, in the lower Tay valley, in the upper Tay valley between Birnam and Meikleour and in western Strathmore. In many parts of the district isolated patches of alluvium occupy hollows and mark the sites of former lakes, now silted up. In such places the deposits are generally of silt and mud, commonly peaty. The extensive spreads of alluvium in the Coupar Angus district are underlain in places by laminated clays, not unlike those described from Strathallan (Francis and others, 1970, p. 287). This material probably was laid down in late-Glacial times, when the drainage was re-establishing itself across an irregular, impeding surface of drift deposits, and transient lakes were formed (cf. (Plate 26)).

Along the rivers the alluvium is characteristically more variable in character, consisting of gravel, sand and silt in proportions changing from place to place. In the lower Earn valley alluvial deposits are usually silty. The floodplain of the River Tay at Perth is underlain by much silty material, but there is sand and gravel, and within the present channel of the River Tay gravel is carried downstream as far as Perth harbour. Further upstream the alluvium of the River Tay tends to be mainly of gravel and several terraces have been identified which relate to the post-glaciation downcutting of the valley. The alluvial terraces in the Birnam–Meikleour area are the product of the reworking of a valley fill which dates from the period of deglaciation. Large abandoned meanders are a feature of the surface of the alluvium here and in lower Strathearn.

Blown sand

Extensive areas of blown sand occur in coastal areas both north and south of the Firth of Tay. In the Tentsmuir area of north-east Fife, lines of low dunes trending N–S mark the positions of successive post-Glacial shorelines. Belts of higher more irregular dunes lie along the present coast. As at Pilmour Links (Chisholm in Forsyth and Chisholm, 1977, p. 251) south of the Eden estuary, 'blowout' dunes on Tentsmuir are orientated parallel to the prevailing wind direction which is from the south-west.

North of the Firth of Tay, blown sand occurs extensively on the ground between Monifieth, Carnoustie and Buddon Ness. The mapped boundary in this area is extremely generalised because of the nature of the deposits. There are large areas of relatively flat ground, exactly like the adjacent areas of post-Glacial marine deposits except that they carry numerous superimposed small irregular mounds of blown sand, usually less than 3 m high. Elsewhere there are elongate ridges, generally higher and trending ENE. A tendency for the ridges to form parabolic or 'hairpin' dunes closing towards the east-north-east suggests that the strongest winds blow from that direction. High dunes truncated by coastal erosion attain heights of over 25 m NNE of Buddon Ness. Lower sand ridges occur parallel to the coast east-south-east of the mouth of the Buddon Burn and south of the mouth of the Barry Burn near Carnoustie. Conspicuous dune topography extends to the east-north-east from Monifieth Links towards Barry and is a feature of the Carnoustie golf courses. North of Carnoustie, blown sand is superimposed on raised beach deposits at East Haven, south of Wormiehills and at Whiting Ness near Arbroath. MA

Chapter 10 Economic geology

Limestone

There is little limestone (or dolomite) in the district but there are a few very small disused quarries which were worked in the 19th century. The only bedded carbonates are the dolomitic limestones (' cementstones') in the mainly argillaceous Ballagan Formation of Lower Carboniferous age, which crops out around Dron, Newburgh and Errol. Boreholes in these areas showed that there are numerous beds of ferroan dolomite (Browne, 1980a, p. 10) but few exceed 40 cm in thickness and the ratio of mudstone to limestone was nowhere less than 5 to 1.

In the Kinnesswood Formation beds of nodular limestone ('cornstone') have been worked in quarries at Parkhill [NO 256 192] and Clunie Bleachfield [NO 221 176] near Newburgh, at West Dron [NO 129 161] and possibly also at Murie [NO 233 226] near Errol. The cornstone beds, which consist of calcite with subordinate dolomite (Browne, 1980a, p. 6), are 6 m thick in places but are generally less than 1 m thick. The beds are not thought to be laterally extensive.

Beds of nodular limestone occur also in a zone near the top of the Garvock Group of the Lower Devonian. This development, (the Stanley Limestone), has been worked on a small scale near Gallowhill [NO 1605 3545] and at Berryhills [NO 117 317] where there are the remains of limekilns. Individual beds of limestone are impersistent and do not exceed 40 cm in thickness, but the general zone can be recognised on a regional scale and may contain a total of more than 1.5 m of carbonate.

Building stone

The district was long an important centre for the production of building stone, paving-stones ('flagstones') and to some extent 'filestones' for roofing purposes, some of the quarries dating back more than four centuries (Mackie, 1980). The industry is now defunct but building stone was still being worked at Leoch Quarry [NO 359 361] in 1952 (Harry, 1952) and Ballindean Quarry [NO 254 292] was reopened for a time in 1969. In the heyday of the industry much material was exported to Norway, Germany and Australia and stone from Leoch was used in the construction of the Usher Hall, Edinburgh and the Glasgow Art Galleries.

The flat-laminated, fine-grained flagstones in the lacustrodeltaic members of the Dundee Formation were the most prized and justified the removal of many metres of overburden. The thinly bedded, fine-grained sandstones were also used for roofing tiles, as, for example, on old cottages at Kingoodie near Invergowrie. The more massive sandstones of the Dundee Formation and of the overlying Garvock and Strathmore groups were worked as building stone in numerous quarries throughout the district. Most of the sandstones are drab-coloured, but bright red sandstones of the Arbroath Sandstone were used in the construction of Arbroath Abbey. The grey sandstone used to build Dunkeld Abbey probably came from a quarry [NO 093 389] in the Teith Formation at Gelly Burn. The Scone Coronation Stone is probably of Garvock Group sandstone, as are many of the well known Pictish carved stones lodged in the Meigle Museum and in the National Museum of Antiquities, Edinburgh.

The reddish brown sandstones of the Upper Devonian Clashbenny Formation also have been worked through the centuries for building stone. The sandstone at Parkhill [NO 260 195] was used in the construction of Lindores Abbey. Clashbenny Quarry [NO 213 212], famous for fossil fish, was worked from the early 18th to the late 19th century, as were Pitkeathly [NO 114 160] and Quarry Hall [NO 112 183] quarries in the Bridge of Earn area. Quarries at Inchture [NO 276 294] and Ballindean [NO 254 292] were active during the last century, the latter being reopened during 1969 to provide stone for the reconstruction of the boundary wall of Rossie Priory. Sandstone of the Kinnesswood Formation was worked at Murie Quarry [NO 233 226].

Slate

The Birnam Slates of the Dalradian in the Dunkeld–Birnam area have been extensively quarried for roofing slates. The slates are mainly purple but grey and green varieties also occur. They were worked in a series of quarries at Birnam Hill [NO 037 404] and Newtyle Hill [NO 046 417] but the industry appears to have died out by about 1900 (Richey and Anderson, 1944). Limited amounts of slate spoil are still taken from one of the Birnam quarries for road bottoming.

Hardrock aggregate

Lower Devonian lavas and intrusions and late-Carboniferous quartz-dolerite and tholeiite dykes have all been exploited for hardrock aggregate (roadstone).

The Lower Devonian lavas are over 2400 m thick but there are relatively few major quarries along their outcrop. The most important resource consists of the exceptionally thick, mainly compact and uniform flows of hyperstheneandesite currently being worked at Ethiebeaton [NO 486 338] and Ardownie [NO 491 342] near Dundee and at Clatchard Craig [NO 244 177] near Newburgh. The lava flow at Clatchard Craig can be traced westwards to Glen Farg, where it has recently been worked at Castle Law [NO 083 050], and eastwards to Glenduckie Quarry [NO 281 191], Norman's Law [NO 305 202] and south of Coultra [NO 353 228].

The lavas of the Ochil and Sidlaw hills, the basic pyroxene-andesites and basalts, generally occur in flows less than 20 m thick, of which commonly more than half is slaggy or autobrecciated and of poor quality. Exceptionally thick flows with little basal autobrecciation are currently being worked at Collace [NO 208 316], Friarton [NO 118 213] and Brackmont [NO 431 224] quarries, and were formerly exploited in quarries at Burnside [NO 306 401], North Ballo [NO 246 354], Pepperknowes [NO 184 223], Fincraig [NO 231 169] and Forgandenny [NO 091 180]. There are numerous small quarries throughout the crop of the lavas (Ochil Volcanic Formation) where small quantities of stone have been extracted.

The Lower Devonian intrusions form two main groups, namely the basic porphyrites and the acid felsites. The basic porphyrites, which are closely related to the basic andesite lavas, are mainly fine- to medium-grained but in some cases contain zones of coarse-grained, more acidic material. They are important in the Dundee area where they occur as relatively thick and extensive sheet-like bodies intruded into the sediments and lavas of the area. Smaller irregular bodies and narrow dykes also occur. Basic porphyrite is being worked at Cunmont Quarry [NO 489 370] and was formerly worked at South Kingennie [NO 475 364], East Skichen [NO 522 417], Ballumbie [NO 440 349], Hilton of Knapp [NO 282 317], Longhaugh [NO 430 330] Craigie [NO 419 318] and Dundee Law [NO 392 313]. Considerable resources of these rocks are still available in the Dundee area.

The felsite intrusions are generally small but larger bodies occur at Forret Hill [NO 392 205] and Lucklaw Hill [NO 419 214]. There are considerable reserves at Lucklaw Quarry where the stone, classed as a rhyodacite, is worked as a 'redstone' aggregate. An orange-coloured porphyry has been worked at Ayton Quarry [NO 118 153] near Glen Farg, where it forms a small plug intruded into volcaniclastic sediments.

The late-Carboniferous quartz-dolerites (and tholeiites) form laterally extensive near-vertical dykes up to 35 m wide. They have been worked in numerous quarries, for example at Letham [NO 305 148], Wolfhill [NO 138 338], south of Huntingtower and Lamberkine [NO 058 208]. Where closely spaced horizontal joints occur, as at Monimail [NO 296 146] and Pitroddie [NO 204 253] quarries, the dolerite was formerly worked for pavement setts and kerbstones. There remain considerable resources of dolerite in the area and these are of particular importance if large-sized (1 m blocks) aggregate is required.

Mineral veins

The district is not noted for its mineral veins. Heddle (1901) recorded native copper (in prehnite) and malachite in the lavas exposed in the North Tunnel in Glen Farg [NO 162 144]. Farther south he also noted the occurrence of chalcocite, chrysocolla and cuprite with fluorspar. The presence of malachite in the North Tunnel area was confirmed in the recent survey. Heddle also recorded chalcopyrite in Newtyle Quarry [NO 285 405] and in the banks of the Water of May at Forgandenny. Copper ores were said to have been discovered in diggings for the foundations of Rossie Priory, Inchture [NO 285 308] according to Honey (1845) and in the glen [NO 270 310] north of Baledgarno (Adamson, 1793). In Clatchard Craig Quarry [NO 243 177] streaks of malachite occur in a mineralised fault breccia and on associated joints together with calcite, hematite, geothite, pyrite, kaolinite, barium clay mineral (after baryte?) and quartz. In places, lava fragments were observed to be coated with orbicular layers of calcite which had subsequently been fractured and recemented. Baryte was also recorded by Heddle (1901) near Forgandenny, at Ballindean and in agates at Balmerino. Baryte has been found recently in the road cutting on the southern approach to the Tay Road Bridge at Wester Friar-ton [NO 425 288] and in the felsite at Lucklaw Hill [NO 419 214] where streaks of malachite are also present. Galena has been recorded in the area only on Birnam Hill [NO 033 403] (Robertson, 1793, p. 358).

A wide variety of minerals has been recorded in the lavas of the Ochil Volcanic Formation either as films on joint surfaces or in vesicles within the lavas (amygdales). Amongst the minerals recorded are prehnite, datolite, analcime, laumontite, natrolite, fargite and saponite.

Gypsum in the form of white veins and pink nodules, and as poikilitic intergrowths, was found in some abundance in the strata of the Ballagan Formation (Lower Carboniferous) cut by the East Dron [NO 136 157] and Mains of Errol [NO 238 219] boreholes (Browne, 1980a, p. 10).

Sand and gravel

The sand and gravel resources of the area which lie within Fife Region have been reviewed by Browne (1977); those in Tayside Region by Paterson (1977). A more detailed assessment of the important deposits in the Newport on Tay area (National Grid Squares NO 42 and parts of NO 32 and 52) has recently been carried out by the Industrial Minerals Assessment Unit of the IGS (Laxton and Ross, 1981). Smaller deposits of fluvioglacial sand and gravel are currently being exploited at Forgandenny [NO 088 187], Cotside farm, Barry [NO 528 343], Loanleven [NO 055 257] and Luncarty [NO 086 295]. River gravels are being dredged from the River Tay below Perth [NO 120 226].

Brick and tile clays

The plastic marine clays of the late-Glacial Errol Beds have long been used for the manufacture of bricks, field drains and tiles. The clays are widely distributed in the lower Earn and Tay valleys and in the Carse of Gowrie area, both at the surface and beneath younger deposits (Paterson and others, 1981). Their physical properties have been discussed by Eyles and Anderson (1946). The greatest known thickness of the clays is 27 m, proved in a borehole at Dalreoch Bridge [NO 0037 1747], but deposits almost as thick are known from boreholes to lie beneath younger deposits in the neighbourhood of Almond Bridge [NO 0959 2666] and at Buddon Ness (Paterson, 1981). More usually the clays are from 3 to 6 m thick, as in the disused claypits at Scone [NO 118 283], Pitfour [NO 209 206] and Inchcoonans [NO 241 233], and in the most recently active claypit at Gallowflat [NO 213 207]. The clay was carried to Inchcoonans and made into pipes for field drains. According to Bonnell and Butterworth (in Eyles and Anderson, 1946), bricks made from the Errol Beds clays should, if fired at 1000°C, be satisfactory for use in conditions of normal exposure.

The grey, clayey silts and silty clays of the Carse Clay, which have a maximum thickness of about 10 m, are widely distributed in the lower Tay and Earn valleys, in the Carse of Gowrie and in Stratheden, where they form part of the post-Glacial raised beach assemblage. The clays have been worked on a limited scale at Forgandenny [NO 087 194], at Waterybutts [NO 237 255] and near Inchture [NO 270 298] and [NO 281 290] but little is known about their quality, even though the Waterybutts Pit was worked after 1945. The presence of ironstone concretions within the clay was said to be harmful. With judicious mixing of suitable additives, the quality of both this clay and the older late-Glacial marine clays could be improved to produce a higher quality product. Alumina could be extracted from some of these clays when the suitable technology is developed at the industrial scale.

Semi-precious stones

Many varieties of silica occur in the form of amygdales infilling vesicles (gas cavities) in the Lower Devonian lava flows. Several localities in the district such as Parbroath [NO 324 177], Dunbog [NO 280 180] and Balmerino [NO 360 250] are well-known to amateur gemmologists. The most common semi-precious stones are the variants of cryptocrystalline silica known as chalcedony. Both the concentrically laminated (agate) and parallel laminated (onyx) forms occur. Agate is the more common, showing red or grey colour banding. Other coloured forms of translucent chalcedony (normally used only for the grey or blue grey forms) which have been found are sard (yellow), carnelian (red) and chrysoprase (green). Jasper, the opaque red form of chalcedony, is quite common also.

Rock crystal also occurs, both in amygdales and as vein fillings, and amethyst and smoky quartz (cairngorm) were both found on Culteuchar Hill [NO 095 154] near Forgandenny.

Groundwater

Within the district, groundwater resources have been developed on a limited scale only (Jackson, 1967), although numerous springs and wells yield supplies adequate for farming or domestic purposes. Moderate supplies of water have been obtained by drilling from Lower Devonian sediments, mainly from sandstones but also from conglomerates. Yields of 100 000 g.p.d., (gallons per day) are not unusual, with a maximum recorded value of 530 000 g.p.d. in Dundee. However, a borehole [NO 0912 2571] drilled recently into the mudstones of the Cromlix Formation near Huntingtower was dry. Boreholes in the Lower Devonian lavas and volcanic conglomerates generally give only modest yields or are dry. The average yield seems to be about 1000 g.p.d. but a borehole at Myrecairnie [NO 371 178], near Cupar, produced 57 600 g.p.d. Yields up to 360 000 g.p.d., have been obtained from the Upper Devonian sediments, mainly sandstones, of the Clashbenny Formation in the Perth area and of the Burnside and Glenvale formations in Stratheden. A detailed study of the aquifer potential of the Upper Devonian in Stratheden, in particular the Knox Pulpit Formation, is given by Foster and others (1976). Few trial boreholes have been sunk into the Lower Carboniferous Ballagan Formation. Although the formation consists mainly of silty mudstone, a borehole near Taybank [NO 250 230] yielded 15 120 g.p.d., presumably from open joints or fissures.

Moderate to large supplies of groundwater have been drawn from shafts, wells and boreholes in the Quaternary deposits, mainly from fluvioglacial sand and gravel and from raised beach sediments. The yields vary from about 26 000 g.p.d. to a maximum reported value of 580 000 g.p.d., the best results being from the raised beach deposits.

Saline waters were formerly pumped from natural springs at Pitkeathly Wells [NO 117 164]. The yield of the wells is known for only one locality, where 360 g.p.d., were pumped; other wells are described as never being dry. The baths associated with the wells fell into disuse during the last century but the history of bathing extends back far beyond 1750.

A scheme to pipe large quantities of water from the River Earn to Glen Farg Reservoir [NO 105 110] to serve the water requirements of the Fife Region is under construction. Water is to be abstracted from the Earn, just upstream of the tidal limit, near Freeland, Forgandenny [NO 100 190]. MAEB

References

ADAMSON, J. 1793. The statistical account of Scotland, Vol. 9. (Edinburgh: Creech.)

ALLAN, D. A. 1928. The geology of the Highland Border from Glen Almond to Noranside. Trans. R. Soc. Edinburgh, Vol. 56, 57–87.

ALLEN, J. R. L. 1960. Cornstone. Geol. Mag., Vol. 97, 43–48.

ALLEN, J. R. L. 1974. Sedimentology of the Old Red Sandstone (Siluro-Devonian) in the Glee Hills area, Shropshire, England. Sediment. Geol., Vol. 12, 71–167.

ANDERSON, J. 1837. Organic remains in the Old Red Sandstone of Fife. Edinburgh New Philos. J., Vol. 23, 137–143.

ANDERSON, J. 1841. On the geology of Fifeshire. Trans. Hight Agric. Soc. Scotland, Vol. 7 (new series), 376–431.

ANDERSON, J. 1845. The (new) statistical account of Scotland. Vol. 9, 59–61. (Edinburgh, London: Blackwood).

ANDERSON, J. 1859. Dura Den, a monograph of the Yellow Sandstone and its remarkable fossil remains. (London: Constable.)

ANDERSON, J. G. C. 1947. The geology of the Highland Border, Stonehaven to Arran. Trans. R. Soc. Edinburgh, Vol. 61, 479–516.

ANDREWS, J. T. and DUGDALE, R. E. 1970. Age prediction of glacio-isostatic strandlines based on their gradients. Bull. Geol. Soc. Am., Vol. 81, 3769–3771.

ARMSTRONG, M. and PATERSON, I. B. 1970. The Lower Old Red Sandstone of the Strathmore Region. Rep. Inst. Geol. Sci., No. 70/12.

ATTRIDGE, J. 1956. A gigantic Holoptychius from Dura Den. Nature, Vol. 177, 232–233.

BALSILLIE, D. 1922. Notes on the dolerite intrusions of east Fife. Geol. Mag., Vol. 59, 442–452.

BOWES, D. R. and LEAKE, B. E. (Editors). 1978. Crustal evolution in northwestern Britain and adjacent regions. Spec. Issue Geol. J., No. 10.

BRADY, G. S., CROSSKEY, H. W. and ROBERTSON, D. 1874. A monograph on the Post-Tertiary entomostraca of Scotland. - Palaeontogr. Soc.

BRAITHWAITE, C. J. R. and JAWAD ALI, A. 1978. Heavy mineral distribution in the Old Red Sandstone of the northeastern Midland Valley of Scotland. Scott. J. Geol., Vol. 14, 273 -288.

BROWN, T. 1867. On the Arctic shell-clay of Elie and Errol, viewed in connection with our other glacial and more recent deposits. Trans. R. Soc. Edinburgh, Vol. 24, 617–633.

BROWNE, M. A. E. 1977. Sand and gravel resources of the Fife Region. Rep. Inst. Geol. Sci., No. 77/5.

BROWNE, M. A. E. 1980a. The Upper Devonian and Lower Carboniferous (Dinantian) of the Firth of Tay, Scotland. Rep. Inst. Geol. Sci., No. 80/9.

BROWNE, M. A. E. 1980b. Late-Devensian marine limits and the pattern of deglaciation of the Strathearn area, Tayside. Scott. J. Geol., Vol. 16, 221–229.

BROWNE, M. A. E. ARMSTRONG, M., PATERSON, I. B. and AITKEN, A. M. 1981. New evidence of late Devensian marine limits in east-central Scotland. Quat. Newsl., Vol. 34, 9–15.

CALLOW, W. J. and HASSALL, G. I. 1970. National Physical Laboratory radiocarbon measurements VII. Radiocarbon, Vol. 12, 181–186.

CAMERON, I. B. and STEPHENSON, D. 1985. British Regional Geology. The Midland Valley of Scotland. Third Edition. (London: HMSO.)

CAMPBELL, R. 1912. Cambrian, Downtonian and Lower Old Red Sandstone rocks near Stonehaven. Proc. Geol. Assoc., Vol. 23, 291–298.

CAMPBELL, R. 1913. The geology of south-eastern Kincardineshire. Trans. R. Soc. Edinburgh, Vol. 48, 923–960.

CHAMBERS, R. 1848. Ancient sea-margins. (Edinburgh: Chambers.)

CHISHOLM, J. I. 1966. An association of raised beaches with glacial deposits near Leuchars, Fife. Bull. Geol. Surv. G. B., No. 24, 163–174.

CHISHOLM, J. I. 1971. The stratigraphy of the post-Glacial marine transgressions in north-east Fife. Bull. Geol. Surv. G.B., Vol. 37, 91–107.

CHISHOLM, J. I. and DEAN, J. M. 1974. The Upper Old Red Sandstone of Fife and Kinross; a fluviatile sequence with evidence of marine incursion. Scott. J. Geol., Vol. 10, 1–30.

COOPE, G. R. 1977. Fossil coleopteran assemblages as sensitive indicators of climatic changes during the Devensian (last) cold stage. Philos. Trans. R. Soc. London, Vol. B280, 313–340.

CRAIG, G. Y. (Editor). 1965. The geology of Scotland. First edition. (Edinburgh: Oliver and Boyd.)

CULLINGFORD, R. A. 1971. Lateglacial and postglacial shoreline displacement in the Earn-Tay area and eastern Fife. PhD thesis (unpublished), University of Edinburgh.

CULLINGFORD, R. A. and SMITH, D. E. 1966. Late-glacial shorelines in eastern Fife. Trans. Inst. Br. Geogr., Vol. 39, 31–51.

CULLINGFORD, R. A. 1980. Late Devensian raised shorelines in Angus and Kincardineshire, Scotland. Boreas, Vol. 9, 21–38.

CULLINGFORD, R. A. CASELDINE, C. J. and GOTTS, P. E. 1980. Early Flandrian land and sea-level changes in Lower Strathearn. Nature, Vol. 284, 159–161.

DAVIDSON, C. F. 1932a. The arctic clay of Errol, Perthshire. Trans. Perthshire Soc. Nat. Sci., Vol. 9, 55–68.

DAVIDSON, C. F. 1932b. The geology of Moncreiffe Hill, Perthshire. Geol. Mag., Vol. 69, 452–464.

DAWSON, J. W. 1890. On certain Devonian plants from Scotland. Nature, Vol. 61, 537.

DE SOUZA, H. A. F. 1979. The geochronology of Scottish Carboniferous volcanism. PhD thesis (unpublished), University of Edinburgh.

DURHAM, J. 1877. The 'Karnes' in the neighbourhood of Newport, Fife, N.B. Geol. Mag., (2), Vol. 4, 8–13.

DURHAM, J. 1886. Volcanic rocks of the north-east of Fife. Q. J. Geol. Soc. London, Vol. 42, 418–434.

EYLES, V. A. and ANDERSON, J. G. C. 1946. Brick clays of north-east Scotland. Wartime Pamphlet Geol. Surv. G.B., No. 47.

FITCH, F. J., MILLER, J. A. and WILLIAMS, S. C. 1970. Isotopic ages of British Carboniferous rocks. C. R. 6e Congr. Int. Stratigr. Geol. Carbonif. (Sheffield 1967), Vol. 2, 771–789.

FLEMING, J. 1831. On the occurrence of scales of vertebrated animals in the Old Red Sandstone of Fifeshire. Edinburgh J. Nat. Geogr. Sci., Vol. 3, 81–86.

FORSYTH, I. H. and CHISHOLM, J. I. 1977. The geology of east Fife. Mem. Geol. Surv. G.B.

FOSTER, S. S. D., STIRLING, W. G. N. and PATERSON, I. B. 1976. Groundwater storage in Fife and Kinross–it’s potential as a regional resource. Rep. Inst. Geol. Sci., No. 76/9.

FRANCIS, E. H. 1982. Magma and sediment 1. Emplacement mechanism of late Carboniferous tholeiite sills in northern Britain. J. Geol. Soc. London, Vol. 139, 1–20.

FRANCIS, E. H. FORSYTH, I. H., READ, W. A. and ARMSTRONG, M. 1970. The geology of the Stirling district. Mem. Geol. Surv. G. B.

FRIEND, P. F. and WILLIAMS, B. P. J. 1978. International symposium on the Devonian System, (P.A.D.S. 78) September 1978. Palaeontol. Assoc.

GANDY, M. K. 1972. The petrology and geochemistry of the Lower Old Red Sandstone lavas of the Sidlaw Hills, Perthshire. PhD thesis (unpublished), University of Edinburgh.

GANDY, M. K. 1975. The petrology of the Lower Old Red Sandstone lavas of the eastern Sidlaw Hills, Perthshire, Scotland. J. Petrol., Vol. 16, 189–211.

GEIKIE, A. 1878. On the Old Red Sandstone of Western Europe. Trans. R. Soc. Edinburgh, Vol. 28, 345–452.

GEIKIE, A. 1897. The ancient volcanoes of Great Britain. 2 volumes. (London: MacMillan.)

GEIKIE, A. 1900. The geology of central and western Fife and Kinross-shire. Mem. Geol. Surv. G.B.

GEIKIE, A. 1902. The geology of Eastern Fife. Mem. Geol. Surv. G.B.

GEMMELL, A. M. D. (Editor). 1975. Quaternary studies in north east Scotland. (Aberdeen: Department of Geography, University of Aberdeen.)

GRAHAM, D. D. and WILKINSON, I. P. 1978. A detailed investigation of a late-Glacial faunal succession at Ardyne, Argyll, Scotland. Rep. Inst. Geol. Sci., No. 78/5.

GRAHAM, D. D. and GREGORY, D. M. 1981. A revision of C. F. Davidson's arctic fauna from Inchcoonans Claypit, Errol, held by the Museum and Art Gallery, Perth. Scott. J. Geol., Vol. 17, 215–222.

GRAY, J. M. and LOWE, J. J. (Editors). 1977. Studies in the Scottish lateglacial environment. (Oxford: Pergamon Press.)

GRIERSON, J. 1845. The new statistical account of Scotland, Vol. 10, 373–375. (Edinburgh, London: Blackwood.)

HARRIS, A. L. 1972. The Dalradian rocks at Dunkeld, Perthshire. Bull. Geol. Surv. G.B., No. 38, 1–10.

HARRIS, A. L. and FETTES, D. J. 1972. Stratigraphy and structure of Upper Dalradian rocks at the Highland Border. Scott. J. Geol., Vol. 8, 253–264.

HARRIS, A. L., SHACKLETON, R. M., WATSON, J., DOWNIE, C., HARLAND, W. B. and MOORBATH, S. (Editors). 1975. A correlation of the Precambrian rocks in the British Isles. Spec. Rep. Geol. Soc. London, No. 6.

HARRIS, A. L., BRADBURY, H. J. and MCGONIGAL, M. H. 1976. The evolution and transport of the Tay Nappe. Scott. J. Geol., Vol. 12, 103–113.

HARRIS, A. L., HOLLAND, C. H. and LEAKE, B. E. (Editors). 1979. The Caledonides of the British Isles  reviewed. Spec. Publ Geol. Soc. London, No. 8.

HARRY, W. T. 1952. The geology of the Dundee district from its quarries. Quarry Managers J., Vol. 35, 489–499.

HARRY, W. T. 1956. The Old Red Sandstone lavas of the western Sidlaw Hills, Perthshire. Geol. Mag., Vol. 93, 43–56.

HEDDLE, M. F. 1901. The mineralogy of Scotland. 2 volumes (Edinburgh: Douglas.)

HICKLING, G. 1908. The Old Red Sandstone of Forfarshire. Geol. Mag., Dec. 5, Vol. 5, 396–408.

HICKLING, G. 1912. On the geology and palaeontology of Forfarshire. Proc. Geol. Assoc., Vol. 23, 302–311.

HONEY, J. A. 1845. The new statistical account of Scotland, Vol. 10, 827–828. (Edinburgh, London: Blackwood.)

HORNE, J., JEHU, T. J., BOLTON, H., DON, A. W. R., FLETT, J. S., PEACH, B. N. and WOODWARD, A. S. 1915. The Upper Old Red Sandstone of Dura Den. Rep. Br. Assoc. Adv. Sci. (for 1914), 116–123.

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

JACKSON, N. P. D. 1967. Record of wells in the areas of Scottish one-inch geological sheets Kinross (40), North Berwick (41), Perth (48) and Arbroath (49). Water Supply Papers Geol. Surv. G.B., Well Catalogue Series.

JAMIESON, T. F. 1865. The history of the last geological changes in Scotland. Q. J. Geol. Soc. London, Vol. 21, 161–203.

JARVIK, E. 1950. On some osteolepiform Crossopterygians from the Upper Old Red Sandstone of Scotland. K. Svenska Vetensk. Akad. Handl., Vol. 2, Part 4, 1–35.

JAWAD ALI, A. and BRAITHWAITE, C. J. R. 1977. Penecontemporaneous weathering of the Old Red Sandstone of the Midland Valley of Scotland. Scott. J. Geol., Vol. 13 305–312.

JONES, S. J. (Editor). 1968. Dundee and district. Br. Assoc. Adv. Sci.

LAXTON, J. L. and Ross, D. L. 1981. The sand and gravel resources of the country around Newport-on-Tay, Fife Region. Miner. Assess. Rep. Inst. Geol. Sci., No. 89.

LEEDER, M. R. 1982. Upper Palaeozoic basins of the British Isles - Caledonide inheritance versus Hercynian plate margin processes. J. Geol. Soc. London, Vol. 139, 479–491.

LINTON, D. L. 1951. Watershed breaching by ice in Scotland. Trans. Inst. Br. Geogr., Vol. 15, 1–15.

MACDONALD, R., GOTTFRIED, D., FARRINGTON, M. J., BROWNE, F. W. and SKINNER, N. G. 1981. The geochemistry of a continental tholeiite suite: late Palaeozoic quartz-dolerite dykes of Scotland. Trans. R. Soc. Edinburgh: Earth Sci., Vol. 72, 57–74.

MACGREGOR, A. R. 1968. Fife and Angus geology. An excursion guide. (Edinburgh and London: Blackwood.)

MACKERROW, W. S., LAMBERT, R. ST. J. and CHAMBERLAIN, V. 1980. The Ordovician, Silurian and Devonian timescales. Earth Planet. Sci. Lett., Vol. 51, 1–8.

MACKIE, A. 1980. Sandstone quarrying in Angus–some thoughts on an old craft. Edinburgh Geologist, Vol. 8, 14–25. MACKIE, R. L. (Editor). 1939. Scientific survey of Dundee and district. Br. Assoc. Adv. Sci.

McMANUS, J. 1972. Estuarine development and sediment distribution with particular reference to the Tay. Proc. R. Soc. Edinburgh (B), Vol. 71, 97–113.

MACNAIR, P. 1908. The geology and scenery of the Grampians and the valley of Strathmore. Vol. 2. (Glasgow: MacLehose.) MELVILLE, L. 1939. The fair land of Gowrie. (Coupar Angus: Culross.)

MILES, R. S. 1968. The Old Red Sandstone Antiarchs of Scotland: Family Bothriolepidae. Palaeontogr. Soc.

MORRISON, J., SMITH, D. E., CULLINGFORD, R. A. and JONES, R. L. 1981. The culmination of the Main Postglacial Transgression in the Firth of Tay area, Scotland. Proc. Geol. Assoc., Vol. 92, 197–209.

MORTON, D. J. 1979. Palaeogeographical evolution of the Lower Old Red Sandstone basin in the western Midland Valley. Scott. J. Geol., Vol. 15, 97–116.

NEVES, R., GUEINN, K. J., CLAYTON, G., IOANNIDES, N. S., NEVILLE, R. S. W. and KRUSZEWSKA, K. 1973. Palynological correlations within the Lower Carboniferous of Scotland and northern England. Trans. R. Soc. Edinburgh, Vol. 69, 23–70.

OLDERSHAW, W. 1974. The Lochnagar granite ring complex, Aberdeenshire. Scott. J. Geol., Vol. 10, 297–309.

PATERSON, I. B. 1974. The supposed Perth Readvance in the Perth district. Scott. J. Geol., Vol. 10, 53–66.

PATERSON, I. B. 1977. Sand and gravel resources of the Tayside Region. Rep. Inst. Geol. Sci., No. 77/6.

PATERSON, I. B. 1981. The Quaternary geology of the Buddon Ness area of Tayside, Scotland. Rep. Inst. Geol. Sci., No. 81/1.

PATERSON, I. B. and HARRIS, A. L. 1969. Lower Old Red Sandstone ignimbrites from Dunkeld, Perthshire. Rep. Inst. Geol. Sci., No. 69/7.

PATERSON, I. B. ARMSTRONG, M. and BROWNE, M. A. E. 1981. Quaternary estuarine deposits in the Tay-Earn area, Scotland. Rep. Inst. Geol. Sci., No. 81/7.

PATON, A. W. and MILLAR, A. H. (Editors). 1912. Handbook and guide to Dundee and district. Br. Assoc. Adv. Sci.

PEACH, B. N., CLOUGH, C. T., HINXMAN, L. H., GRANT WILSON, J. S., CRAMPTON, C. B., MAUFE, H. B. and BAILEY, E. B. 1910. The geology of the neighbourhood of Edinburgh. 2nd Edition. Mem. Geol. Surv. G. B.

PEACOCK, J. D., GRAHAM, D. K. and WILKINSON, I. P. 1978. Late-Glacial and post-Glacial marine environments at Ardyne, Scotland, and their significance in the interpretation of the history of the Clyde area. Rep. Inst. Geol. Sci., No. 78/17.

POTTER, P. E. and PETTIJOHN, F. J. 1977. Paleocurrents and basin analysis. (Berlin: Springer.)

POWRIE, J. 1861. On the Old Red Sandstone rocks of Forfarshire. Q J. Geol. Soc. London, Vol. 17, 534–542.

POWRIE, J. 1862. The Old Red Sandstone of Fifeshire. Q J. Geol. Soc. London, Vol. 18, 427–437.

RAMOS, A. and FRIEND, P. F. 1982. Upper Old Red Sandstone sedimentation near the unconformity at Arbroath. Scott. J. Geol., Vol. 18, 297–315.

RAMSAY, D. M. 1964. Deformation of pebbles in Lower Old Red Sandstone conglomerates adjacent to the Highland Boundary Fault. Geol. Mag., Vol. 101, 228–248.

RICE, R. J. 1959. The glacial deposits of the Lunan and Brothock valleys in south-east Angus. Trans. Edinburgh Geol. Soc., Vol. 17, 241–259.

RICE, R. J. 1961. The glacial deposits at St Fort in north-east Fife: a reexamination. Trans. Edinburgh Geol. Soc., Vol. 18, 113–123.

RICE, R. J. 1962. The morphology of the Angus coastal lowlands. Scott. Geogr. Mag., Vol. 78, 5–14.

RICHARDSON, J. B. 1967. Some British Lower Devonian spore assemblages and their stratigraphical significance. Rev. Palaeobot. Palynol., Vol. 1, 111–129.

RICHARDSON, J. B. FORD, J. H. and PARKER, F. 1984. Miospores, correlation and age of some Scottish Lower Old Red Sandstone sediments from the Strathmore region (Fife and Angus). J. micropalaeontol. , Vol. 3, 109–124.

RICHEY, J. E. and ANDERSON, J. G. C. 1944. Scottish slates. Wartime Pamphlet Geol. Surv. G. B. , No. 40.

ROBERTSON, J. 1793. The statistical account of Scotland. Vol. 6. (Edinburgh: Creech.)

ROSSO, D. H. 1948. The Old Red Sandstone volcanic suite of Eastern Forfarshire. Trans. Edinburgh Geol. Soc., Vol. 14, 128–140.

RUDDIMAN, W. F. and MCINTYRE, A. 1973. Time-transgressive deglacial retreat of polar waters from the North Atlantic. Quat. Res., Vol. 3, 117–130.

RUSSELL, M. J. and SMYTHE, D. K. 1983. Origin of the Oslo Graben in relation to the Hercynian-Alleghenian Orogeny and lithospheric rifting in the North Atlantic. Tectonophysics, Vol. 94, 457–472.

SHAND, S. J. 1929. The classification of a glassy rock; the pitchstone of Wormit, Fifeshire. Geol. Mag., Vol. 66, 116–121.

SHACKLETON, R. M. 1958. Downward-facing structures of the Highland Border. Q. J. Geol. Soc. London, Vol. 113, 361–392.

SIMPSON, J. B. 1933. The late-Glacial readvance moraines of the Highland Border west of the River Tay. Trans. R. Soc. Edinburgh, Vol. 57, 633–646.

SISSONS, J. B. 1963. The Perth Readvance in central Scotland. Part 1. Scott. Geogr. Mag., Vol. 79, 151–163.

SISSONS, J. B. 1966. Relative sea-level changes between 10 300 and 8300 BP in part of the Carse of Stirling. Trans. Inst. Br. Geogr., Vol. 39, 19–29.

SISSONS, J. B. 1967. The evolution of Scotland's scenery. (Edinburgh: Oliver and Boyd.)

SISSONS, J. B. 1974a. Lateglacial marine erosion in Scotland. Boreas, Vol. 3, 41–48.

SISSONS, J. B. 1974b The Quaternary in Scotland: a review. Scott. J. Geol., Vol. 10, 311–337.

SISSONS, J. B. and SMITH, D. E. 1965. Raised shorelines associated with the Perth Readvance in the Forth valley and their relation to glacial isostasy. Trans. R. Soc. Edinburgh, Vol. 66, 143–168.

SISSONS, J. B. and CULLINGFORD, R. A. 1966. Late-glacial and post-glacial shorelines in south-east Scotland. Trans. Inst. Br. Geogr., Vol. 39, 9–18.

SISSONS, J. B. and BROOKS, C. L. 1971. Dating of early post-glacial land and sea-level changes in the western Forth valley. Nature, Vol. 234, 124–127.

STRINGER, P. 1957. Polyphase deformation in the Upper Dalradian rocks of the Southern Highlands of Scotland. PhD thesis (unpublished), University of Liverpool.

SUTHERLAND, D. S. (Editor). 1982. Igneous rocks of the British Isles. (Chichester: Wiley.)

TEALL, J. J. H. 1888. British petrography, with special reference to the igneous rocks. (London: Dulau.)

THIRLWALL, M. F. 1981. Implications for Caledonian plate tectonic models of chemical data from volcanic rocks of the British Old Red Sandstone. J. Geol. Soc. London, Vol. 138, 123–138.

THIRLWALL, M. F. 1983. Discussion on implications for Caledonian plate tectonic models of chemical data from volcanic rocks of the British Old Red Sandstone. J. Geol. Soc. London, Vol. 140, 315–318.

THOMSON, A. G. M. 1903. The position of the Old Red Sandstone in the geological succession. (Dundee: Leng.)

THOMSON, M. E. 1978. IGS studies of the geology of the Firth of Forth and its Approaches. Rep. Inst. Geol. Sci., No. 77/17.

THORPE, R. S. (Editor). 1982. Andesites. (Chichester: Wiley.)

TOOLEY, M. J. 1978. Sea-level changes in north-east England during the Flandrian Stage. (Oxford: Clarendon Press.)

WALKER, F. 1934. A preliminary account of the quartz-dolerite dykes of Perthshire. Trans. Proc. Perthshire Soc. Nat. Sci., Vol. 9, 109–117.

WALKER, F. 1935. The late Palaeozoic quartz-dolerites and tholeiites of Scotland. Mineral Mag., Vol. 24, 131–159.

WALKER, F. 1961. Tayside geology. (Dundee: Dundee Museum and Art Gallery.)

WALKER, F. 1963. The geology of Strathearn. (Dundee: Dundee Museum and Art Gallery.)

WALKER, F. 1965. The part played by tholeiitic magma in the Carbo-Permian volcanicity of central Scotland. Mineral. Mag., Vol. 34, 498–516.

WALKER, F. and IRVING, J. 1928. The igneous intrusions between St Andrews and Loch Leven. Trans. R. Soc. Edinburgh, Vol. 56, 1–17.

WATTISON, A. 1958. Note on the occurrence of Holoptychius nobilissimus Ag. in Dura Den, Fife. Trans. Edinburgh Geol Soc., Vol. 17, 179.

WESTOLL, T. S. 1951. The vertebrate-bearing strata of Scotland. Rep. 18th Sess. Int. Geol. Congr. G. B. 1948, Section 11, 5–21.

WHITE, E. J. 1963. Notes on Pteraspis mitchelli and its associated fauna. Trans. Geol. Soc. Edinburgh, Vol. 19, 306–322.

Appendix 1List of fossils recorded from the Devonian rocks of the district

In the following, the principal fossils found in the Devonian rocks of the district are listed. The chief fossil localities are arranged alphabetically within the major subdivisions of the Devonian strata, which are given in ascending sequence. Where the species is represented by specimens not housed in the collections of the British Geological Survey, Edinburgh, this is indicated; the abbreviations of the names of the depositories for these specimens are as follows:

• BM - British Museum (Natural History)
• D - Albert Museum, Dundee
• H - Hunterian Museum, Glasgow
• RSM - Royal Scottish Museum, Edinburgh

The list does not claim to be exhaustive. In particular, no attempt has been made to record the numerous localities, mainly within the Arbuthnott Group of the Lower Devonian, from which Parka sp. and unidentified plant fragments have been obtained.

Lower Devonian

Arbuthnott Group

• 1 Stream 365 m W43°S of Ardgarth [NO 2802 3690]
• 2 Balruddery Den 695 m W16°S of Mains of Fowlis [NO 3166 3223]
• 3 Balruddery Den 870 m W2°N of Mains of Fowlis [NO 3146 3243]
• 4 Balruddery Den 915 m W6°N of Mains of Fowlis [NO 3141 3250], possibly the original 'Balruddery Den' locality of 19th century collections.
• 5 Balruddery Den 1290 m W17°N of Mains of Fowlis [NO 3109 3272]
• 6 Clearie Burn 200 m upstream from junction with Boath Burn [NO 5364 3931]
• 7 Tributary of Corbie Burn 350 m ENE of Newton [NO 4582 4205]
• 8 Old Quarry 410 m E25°N of Corbiehill [NO 3434 2243]
• 9 Crombie Burn 340 m downstream from bridge at Crombie Mill [NO 5343 4006]
• 10 Dighty Water 220 m downstream from bridge at Linlathen [NO 4661 3392]
• 11 Dundee College of Education No. 13 Borehole [NO 4385 3142]
• 12 Duntrune Quarry 465 m S of Craighill [NO 4375 3028]
• 13 Elliot Water 90 m upstream from footbridge at Kelly Castle [NO 6062 4013]
• 14 Elliot Water 165 m upstream from footbridge at Kelly Castle [NO 6056 4032]
• 15 Spoil heap on bank of Fithie Burn 585 m S of Craighill [NO 4373 3515]
• 16 Fowlis Burn 620 m W10°N of church of Kirkton of Liff [NO 3270 3290]
• 17 Haining Burn 1085 m W15°S of Chamberwells [NO 3530 4200]
• 18 Haining Burn 1165 m W15°S of Chamberwells [NO 3528 4199]
• 19 Lochton Burn 110 m E of East Newton [NO 2703 3247]
• 20 Myreton Quarry 150 m NW of North March [NO 442 372]
• 21 Ninewells Hospital 450 m W39°N of Invergowrie House [NO 3597 3080]
• 22 Parkhill Old Quarry, probably the quarry 370 m E of Parkhill [NO 2495 1855]
• 23 Pitairlie Burn 1170 m NE of Cunmont [NO 4987 3738]
• 24 Pitairlie Burn 355 m SW of Denhead [NO 4874 3787]
• 25 Powrie Brae No. 7 Borehole [NO 4175 3499]
• 26 Rossie Burn 675 m E21°N of Rossie Priory [NO 2924 3107]
• 27 Rossie Burn 685 m E21°N of Rossie Priory [NO 2924 3108]
• 28 Sweet Burn 640 m NW of Whitehouse [NO 4269 3959]
• 29 Tealing Old Quarry 595 m N20°E of New Mains of Tealing [NO 4160 3976]
• 30 Westhall Quarry 200 m SE of Westhall Terrace [NO 446 367]
• 31 Whitehouse Den 895 m NW of Whitehouse [NO 4257 3982]
• 32 Wormit Bay, probably the exposure 730 m NW of Peacehill [NO 3851 2581]

Fossils

Plantae

• Nematophyllum caledonicum Lang 4 D
• N. forfarense (Kidston) 29 D cf. N. sp. 27
• Psilophyton sp. 12
• Zosterophyllum myretonianum Penhallow 20 RSM, 22 RSM, 30 H

Arthropoda

• Dictyocaris sp. 3cf., 13
• Eurypterid fragments 10, 28
• Kampecaris folarensis Page 32
• Pterygotus (Pterygotus) anglicus Agassiz 4 RSM, 5, 9, 11 cf., 12, 15 cf., 22, 29 RSM
• P. sp. 32

Pisces

• Acanthodians indet. 4, 7, 16, 18, 19, 21, 24, 25, 31
• Brachyacanthus scutiger Egerton 12, 29 RSM, 32
• B. sp. 4 RSM
• Cephalaspis pagei Lankaster 9, 12, 29
• C. powriei Lankaster (forma typica) 9, 29
• C. powriei var. brevicornis Lankaster 9
• C. sp. 4 RSM
• Climatius reticulatus Agassiz 2, 4 RSM, 12?, 28, 29
• C.? 27
• Eathacanthus grandis Powrie 12
• E. mucnicoli Powrie 12?
• E. sp. 27, 29 RSM
• Homocanthus sp. 4 RSM
• Zshnacanthus gracilis(Egerton) 4 RSM, 6, 7, 12, 14, 17, 28, 29 RSM, 32 I.? 29
• Mesacunthus mitchelli(Egerton) l?, 2?, 8?, 12, 17, 23?, 26?, 27?, 28, 29 RSM
• Parexus falcatus Powrie 5
• P. recurvus Agassiz 2, 4 RSM, 19, 29 RSM
• P. sp. 4 RSM, 8?, 17, 29

The fauna and flora, although extensive, do not appear to have any great stratigraphical significance. Waterston (in Craig, 1965, p. 282) describes the whole aspect of the fauna as 'Dittonian' and Westoll (in House and others, 1977, p. 73) suggests an age 'not older than the base of the Dittonian (sensu White)'. Richardson (1967; 1984) has suggested that the spores obtained indicate a Gedinnian age for the group.

Garvock Group

• Auchtertyre, possibly old quarries between 500 and 800 m SSE of Auchtertyre. [NO 286 404] Cephalaspis. sp. RSM and Pteraspis mitchelli Powrie RSM.
• White (1963, p. 320) could draw no stratigraphical conclusions concerning this fauna.

Strathmore Group

• Murthly, probably Gelly Burn Quarry, 230 m E18° S of Little Burnbane [NO 0938 3897] Psilophyton princeps Dawson

Upper Devonian

• 1 Exposures in old quarry and the banks of the Ceres Burn (Dura Den) between 740 m N66°E and 845 m N57°E of Dura Mains [NO 4162 1463] to [NO 4167 1478]
• 2 Old quarry 450 m WNW of Clashbenny [NO 2130 2118]
• 3 Culfargie IGS Borehole [NO 1675 1749]
• 4 Tributary of River Eden 650 m S14°E of Balass [NO 3935 1370]
• 5 Parkhill, probably the old quarries c. 1500 m NNW of Blinkbonny [NO 2605 1949]
• Fauna. Bothriolepis cristata Traquair 1 RSM, B. hydrophila (Agassiz) 1 RSM, B. sp. 2 RSM, Eusthenopteron? dalgliesiensis (Anderson) 1 RSM, Glytopomus kinnairdi (Huxley) 1 RSM, G. minor Agassiz 1 BM, Holoptychius flemingi Agassiz 1 BM, H. nobilissimus (Agassiz) 1 RSM, 2 RSM, 5 D, H. sp. 4. Phaneropleuron andersoni Huxley 1 RSM, Phyllolepis concentrica Agassiz 2 RSM, P. woodwardi Stensio 1 RSM, Fish scale in-det .3.
• Miles (1968, p. 12) suggested a general correlation between the strata at Clashbenny and Dura Den and equated them broadly with the Rosebrae Beds of the Elgin area and therefore with the upper part of the Upper Devonian.

Figures, plates and tables

Figures

(Figure 1) Generalised solid geology of the district

(Figure 2) Solid geology of the Birnam-Dunkeld area Dunkeld Grits

(Figure 3) Diagrammatic cross-section through an early fold of Dalradian sandstones and slates to demonstrate changing relationships between Slp and SI cleavage

(Figure 4) Distribution of principal Lower Devonian formations

(Figure 5) Relations of the Arbuthnott Group volcanic and sedimentary formations

(Figure 6) Characteristic sequences in the lacustro-deltaic rocks of the Dundee Formation

(Figure 7) Distribution of the Lower Devonian lavas and associated volcaniclastic deposits

(Figure 8) Distribution of the principal Lower Devonian minor intrusions in relation to the extrusive rocks

(Figure 9) Sedimentological and other features in the basal Carboniferous and older beds in boreholes (Browne, 1980a)

(Figure 10) Sedimentological and other features in the Ballagan Formation in boreholes (Browne, 1980a)

(Figure 11) Late-Carboniferous quartz-dolerite and tholeiite intrusions in the district

(Figure 12)Sketch-map showing the principal faults and folds in the district

(Figure 13) Quaternary marine deposits in the Tay—Earn area

(Figure 14) Distribution of Quaternary features and deposits, excluding till, in the district

(Figure 15) Shoreline-diagram for the district

(Figure 16) Quaternary deposits of the Stratheden and St Fort areas

(Figure 17) Quaternary deposits in lower Strathearn and the western part of the Carse of Gowrie

(Figure 18) Quaternary deposits in the lower Almond and Tay valleys

(Figure 19) Sections through Quaternary deposits in the Earn and Tay valleys Lines of section are shown on (Figure 17). Based on borehole information (Paterson and others, 1981)

(Figure 20) Sequences from boreholes through Quaternary deposits in the district

Plates

(Plate 1) Kinnoull Hill and lower Tay valley. The high ground is formed by Lower Devonian volcanic rocks. Superficial deposits forming the low ground conceal a deep, buried valley (D 1912)

(Plate 2) Structural features of the Dalradian rocks (photographs by Dr A. L. Harris)

(Plate 3) Flagstone quarry, flat-bedded, fine-grained sandstone, Dundee Formation, Dodd Hill, Sidlaw Hills (D 1209)

(Plate 4) Load-cast structures in sandstones and siltstones, Dundee Formation, Auchterhouse Quarry (D 1205)

(Plate 5) Primary current-lineation in flagstones, Dundee Formation, quarry on Auchterhouse Hill (MNS3911)

(Plate 6) Cross-bedded sandstones showing hardened, calcareous zones on upper parts of co-sets, Arbroath Sandstones, Whiting Ness (D 1036)

(Plate 8) Blocky, silty mudstones with thin channel-sandstone, Cromlix Formation, River Almond (D 3306)

(Plate 9) Lower Devonian fossils1 Parka decipiens Fleming, X 1, G.S.E.14007, old quarry 1025 m W30^°N of Balbeuchly 2 Cephalaspis powriei Lankaster, x 1, G.S.E.1171, Tealing Hill, Dundee 3 Mesacanthus mitchelli (Egerton), X 1, G.S.E.3909, Duntrune old quarry, Dundee All specimens held by BGS, Edinburgh

(Plate 10) Scarp formed by basic andesite lavas, Ochil Volcanic Formation, Black Hill, Sidlaw Hills (D 3300)

(Plate 11) Autobrecciated andesite lava, Ochil Volcanic Formation, Stannergate, Dundee (D 1021)

(Plate 12) Dip and scarp featuring on lavas of Ochil Volcanic Formation, Ochil Hills. Clatchard Craig Quarry in foreground is in the thick hypersthene-andesite flow of Norman's Law (D 2044)

(Plate 13) Froth-impressions and miniature sand-volcanoes in sediments infilling a lava-tunnel, Wester Keith, Sidlaw Hill (MNS3910)

(Plate 14) Intrusion of fine-grained, basic porphyrite showing basal contact with sandstones of Dundee Formation, Cunmont Quarry (D 1208)

(Plate 15) Unconformity of Upper Devonian conglomerate and sandstone on Lower Devonian Arbroath Sandstone, Whiting Ness, Arbroath (D 1034)

(Plate 16) Upper Devonian basal breccia with clasts of Arbroath Sandstone, Whiting Ness, Arbroath (D 2728)

(Plate 17) Conglomerate and sandstone at base of Upper Devonian sequence, Whiting Ness, Arbroath (D 2729)

(Plate 18) Upper Devonian fossils from Dura Den. 1 Goptopomus sp. and Ho/opochius sp., X 0.15, Royal Scottish Museum specimen 1966.39.8 2 Bothriolepis hydrophila (Agassiz), X 0.5, Royal Scottish Museum specimen 1936.38.1

(Plate 19) Late-Carboniferous quartz-dolerite dyke showing lateral displacement, Campsie Linn, River Tay (D 3311)

(Plate 20) Typical textures in quartz-dolerite and tholeiites from late-Carboniferous dykes. 1. Quartz-dolerite, Jack's Chair, Netherwood ((S68336), X 35, crossed-polars). Plagioclase laths partly embedded in subophitic augite (e.g. bottom left). Intersertal areas filled by micropegmatite intergrowths (centre and top right) and clear quartz (left centre). Skeletal titano-magnetite (top right) (PMS 389) 2.Tholeiite of Corsiehill type, Balruddery ((S52591), X 35, plane-polarised light). Plagioclase laths embedded in mesostasis of plagioclase microlites and reticulate ilmenite crystals in a turbid brown glass. Augite (left) has slight ophitic tendencies but is generally more granular than in the quartz-dolerites. Irregular pseudomorph after olivine (top right) (PMS390). 3. Tholeiite of Craigmakerran type, Carnbeddie, Williamston ((S56988), X 25, plane-polarised light). Plagioclase laths and rounded augite set in abundant glassy mesostasis containing microlites of plagioclase and ilmenite. Curved plagioclase laths enclose a spherical ocellus of glassy mesostasis, a characteristic feature of more glassy types of tholeiite (PMS388)

(Plate 21) Esker, Keptie Hills, Arbroath (D 1055)

(Plate 22) Moundy topography, fluvioglacial sand and gravel, Newton Hill, St Fort (D 2035)

(Plate 23) Fluvioglacial sand and gravel forming plateau, North Straiton, Fife (D 2034)

(Plate 24) Fluvioglacial sand and gravel, Barnhill, Perth (TS2041)

(Plate 25) Convoluted layer in fluvioglacial sand and gravel, pipeline trench, north of Meikleour (TS2039)

(Plate 26) Banded lacustrine silty clay, possibly varved, pipeline trench, south of Lily Loch (TS2040)

(Plate 27) Fossils from Inchcoonans Claypit (Errol Beds) 1 Phoca vittulina Linné, bone of, X 0.7, 1981.37.1 2 Portlandia arctica Gray, x 2, 1981.37.17 3 Thracia cf. serpentrionalis Jeffreys, X 2, 1981.37.18 4 Buccinum cyaneum Brugiere, X 2, 1981.37.2 5 Hiatella arctica (Linné), x 1.4, 1981.37.12 All specimens in the Davidson Collection, Museum and Art Gallery, Perth

(Plate 29) Post-Glacial fossils 1 Littorina littorialis (Linné), x 2, G.S.E.14027, Barry Buddon No. 1 Borehole, 8 m 2 Chlamys opercularis (Linné), (juvenile), X 2, G.S.E.14028, Barry Buddon No. 2 Borehole, 4m 3 Fabulina fabula (Gmelin) x 1.6, G.S.E.14029, as 2 above 4 Rissoa parva (da Costa), X 14, G.S.E.14030, as 2 above 5 Onoba semicostata (Montagu), x 13, G.S.E.14031, as 2 above 6 Macoma balthica (Linné), x 1.4, G.S.E.14032, as 1 above

Tables

(Table 1) Palaeocurrent directions as inferred from crossbedding azimuths in sandstones in the Garvock and Arbuthnott groups (corrected for tectonic dip)

(Table 2) Classification of Lower Devonian Lavas

Tables

(Table 1) Palaeocurrent direction as inferred from cross-bedding azimuths in sandstones in the Garvock and Arbuthnott groups (corrected for tectonic dip)