Geology of the Outer Hebrides: Memoir for 1:100 000 (solid edition) geological sheets, (Scotland)

By D J Fettes, J R Mendum, D I Smith and J V Watson

Bibliographical reference: Fettes, D J, Mendum, J R, Smith, D I, And Watson, J V. 1992. Geology of the Outer Hebrides. Memoir of the British Geological Survey, Sheets (solid edition) Lewis and Harris, Uist and Barra (Scotland)

British Geological Survey

Geology of the Outer Hebrides: Memoir for 1:100 000 (solid edition) geological sheets, Lewis and Harris, Uist and Barra (Scotland)

D J Fettes, J R Mendum, D I Smith And J V Watson

Contributors: Post-Lewisian minor intrusions N M S Rock; Post-Lewisian sedimentary rocks W Mykura

London: HMSO 1992. NERC copyright 1992 First published 1992. ISBN 0 11 884478 4. Printed in the UK for HMSO Dd 291137 C10 4/92

Authors:

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

Books

Maps

Preface

The Outer Hebrides cover a land area of 2900 km2, and consist largely of Lewisian Gneiss. The islands have never been systematically surveyed by the Geological Survey, but in the late 1960s large parts of the area were mapped by postgraduate students, many of them from Imperial College, London, and others from the universities of Birmingham and St Andrews. As there was no prospect of the Geological Survey undertaking a comprehensive mapping programme in the islands within the foreseeable future it was decided in 1970 that the Survey would collaborate with the late Professor J. V. Watson of Imperial College to complete the geological cover of the Outer Hebrides by using all available information supplemented by a rapid investigation of the unsurveyed ground. Doctors Fettes, Smith and Mendum and Messrs Henderson and Ross spent a total of six field seasons assessing the PhD theses and other published data as well as mapping the areas for which no data were available. This work resulted in the publication in 1981 of four geological maps at 1:100 000 (Lewis and Harris (North), Lewis and Harris (South), Uist and Barra (North) and Uist and Barra (South)). At the same time Dr J. D. Peacock of the Geological Survey investigated the Pleistocene and Recent deposits of the Outer Hebrides, and an account of these is published separately (Peacock, 1984).

The contributions to the memoir by the principal authors are as follows: D. J. Fettes: Chapters 1, 2, 4, 8, 9, 10 and 11; J. R. Mendum; Chapters 3, 7, 15 and 17 (the last jointly with with Dr W. Mykura); D. I. Smith: Chapters 6, 12 and 16 (the last jointly with N. M. S. Rock); J. V. Watson: Chapters 13 and 14. Dr R. Macdonald of Lancaster University provided help with various problems of petrochemistry.

Much of the material contained in this memoir was obtained from PhD theses and published papers. The authors are particularly grateful to the following, who have given permission for their unpublished work to be used in the production of both the maps and memoir: Drs K. H. Brodie, M. P. Coward, F. B. Davies, B. B. Dickinson, P. W. Francis, R. H. Graham, A. M. Hopgood, R. J. Horsley, R. J. Lisle, J. S. Myers, F. K. Palmer, R. H. Sibson, A. C. Skinner, S. R. Soldin and G. J. Witty; Misses E. Bishop, and M. B. Taft, and Mr K. Thamdrup.

Some chemical analyses were carried out by the staff of the Analytical Chemistry Research Group of BGS and its predecessor in IGS, and some of the isotopic age determinations were provided by Dr N. J. Snelling and Mrs M. Brook. Electron microprobe analyses were carried out by Mr R. I. Lawson and the photographs were taken by Messrs J. M. Pulsford and A. F. Christie.

F G Larminie, OBE Director British Geological Survey Keyworth Nottingham NG12 5GG March 1990

Summary of geology

The Outer Hebrides or Long Island form a 210 km long archipelago on the north western seaboard of Scotland. The islands take the full force of many Atlantic storms and display a great variety of scenery, ranging from the rugged cliffs and mountains of Harris to the gentle sandy machair of South Uist. They contain some of the most ancient rocks exposed in Europe, and once formed part of an extensive Precambrian North Atlantic craton, along with similar rocks in Greenland and eastern Canada. The islands are composed mainly of Archaean and Lower Proterozoic quartzofeldspathic gneisses, metasediments, banded basic and ultrabasic intrusions, and granites. Following Scourian deformation of the Archaean assemblage at 2800–2600 Ma, and intrusion of Late-Scourian intermediate and granitic bodies, the gneisses were intruded by a widespread suite of tholeiitic dykes at around 2400–2200 Ma. This Younger Basic suite of dykes is correlated with the Scourie dykes of mainland north-west Scotland. In southern South Harris a suite of gabbro, anorthosite, norite and diorite intrusions, also dated at about 2200 Ma, forms spectacular geology and scenery. This South Harris Igneous Complex has been intruded into metasediments which display granulite and upper amphibolite facies metamorphic assemblages.

The Laxfordian deformation and metamorphism at around 1700 Ma resulted in widespread reworking of the gneisses and basic dykes. The regional tectonic pattern is one of broad anti-forms, tight 'pinched-in' synforms and attenuated fold limbs. Studies of the degree of Laxfordian strain and metamorphic assemblages shown by the Younger Basics document detailed areal variations. Such variations, in part, reflect the preexisting Scourian tectonothermal pattern. Late-Laxfordian granites, mainly a series of sheets and veins, intrude the gneiss complex, notably in the Uig Hills and in South Harris.

A complex history of mylonitisation, thrusting and formation of pseudotachylite and 'cataclastic' gneiss is recorded in rocks within the Outer Hebrides Thrust zone, which runs down the eastern edge of the Long Island. The zone was initiated as a thrust belt late in the Laxfordian, and was subsequently reactivated in an extensional mode in Caledonian times. The Outer Hebrides acted as a horst block during the Phanerozoic with Devonian and Permo-Trias basins developing in the marginal sea areas, notably the Minch Basin and Flannan Trough. Permo-Carboniferous and Tertiary dykes are found throughout the Outer Hebrides, with PermoC arboniferous quartz dolerites and camptonite-monchiquites concentrated in Barra and South Uist and the Tertiary basaltic dykes more common in Harris and central Lewis. Differentiated basic-ultrabasic sills intrude Jurassic sediments in the Shiant Isles.

(Frontispiece) Castlebay and Heaval (383 m), Barra. The feature running upwards from the building on the right marks the position of a thrust within the Outer Hebrides Thrust zone (D 2936)

Chapter 1 Introduction

History of research

The first regional survey of the Outer Hebrides was by MacCulloch (1819), who recognised a monotonous series of hornblende gneisses relieved only by the red sandstones and conglomerates around Stornoway. Unlike the mainland Lewisian, the Outer Isles were not covered by the mapping of the Geological Survey in the late 19th century and early 20th century, and in consequence, although forming by far the largest area of Archaean crust exposed in Britain, the area remained comparatively unknown. This situation continued until Jehu and Craig published their series of papers between 1923–1934, in which they described the gneisses and their associated intrusions and outlined the full extent of the Outer Hebrides Thrust with its various crushed and broken rocks. Subsequently only intermittent work was carried out, and this was generally confined to relatively local or specialised problems. A more modern geological interpretation of the area was made by Dearnley (1962a), who suggested various correlatives with the better documented mainland Lewisian and erected some structural and petrological models. This was followed during the late sixties by a number of graduate theses, mainly from Imperial College, London, which gave rise to a large volume of data and provided detailed maps for almost a third of the area. In order to complete this survey of the islands a cooperative project was started between the Institute of Geological Sciences (now the British Geological Survey) and Imperial College. During the period of 1970-1976 the remaining area was surveyed (about two thirds by IGS and one third by IC); at the same time IGS integrated the existing maps, supplementing details as necessary. This memoir and the companion maps are the result of this cooperative project.

Because of the terrain and the nature of the geology it was felt that the production of 1:50 000 systematic maps would be a lengthy and probably unsatisfactory operation. In consequence two sets of maps have been produced at 1:100 000, one set showing lithology (published separately) the other structure. This account is not exhaustive, and many problems, particularly in the field of petrology, require further specialist study.

The sources of data for the compilation of the maps is given in (Figure 1). The use of material from unpublished thesis is particularly acknowledged.

Much of the work carried out in the 1960's on the Lewisian Complex of north-west Scotland and Precambrian rocks of southern Greenland was the subject of a conference held at the University of Keele in 1971. It included notable contributions on the geology of the Outer Hebrides which are reported in the conference proceedings (Park and Tarney, 1973). Many of these references are cited in this memoir. A follow-up meeting was organised at the University of Leicester in 1985 to review subsequent work on the Lewisian of north-west Scotland, including offshore results, and comparable Precambrian terrains in various parts of the world. Several of the papers given at this latter meeting provide an up-to-date summary of the Lewisian geology (Park and Tarney, 1987) including models of tectonic and metamorphic evolution and thrust formation.

Physical features

The Outer Hebrides, also referred to as the Outer Isles or Long Island, comprise the chain of islands off the north-west coast of Scotland, with the largest island of Lewis and Harris in the north and North Uist, Benbecula, South Uist and Barra strung out like a tail to the south. The island chain is about 210 km in length from the Butt of Lewis to Barra Head, and covers an area of about 2900 km2. It consists of hundreds of islands, many with highly indented outlines and deeply penetrating sea-lochs, so the coastline is highly intricate, with a length of thousands of kilometres. Four sets of outlying islands are normally included with the Outer Hebrides: the Flannan Isles and St Kilda (the latter not included in this memoir; see Harding et al., 1984) lying respectively some 30 km and 65 km west of Lewis; the northern isles of Sula Sgeir and North Rona, which are about 70 km north of the Butt of Lewis, and the Shiant Isles which lie eight kilometres south-east of Lewis in the North Minch.

The islands of the Outer Hebrides present a considerable variety of physical features ((Plate 1) and (Plate 2)). The northern half of Lewis is essentially flat with a thin fertile coastal strip and a monotonous peat-covered interior largely devoid of rock outcrop. Southwards the degree of exposure increases and the flat peat gives way to a complex mixture of lochs and rocky knolls as exemplified in the area north and south of Loch Erisort. This area rises westwards to the deeply incised and well exposed Uig Hills and southwards to the hills of Park and North Harris where Clisham (799 m) forms the highest top of the Outer Hebrides. The coastline around Lewis and North Harris is generally rocky with some notable cliff features. Perhaps the most spectacular of these lies on the west coast of Lewis south of Uig, where the Atlantic winds have kept many of the headlands free of lichen, so that there are some excellent 'three dimensional' exposures.

Across the narrow isthmus of Tarbert lies South Harris: an undulating rocky peneplain, graphically described by Heddle (1878, p.546) as the most barren part of the British Isles. Although both magnificently scenic, the two opposing coastlines are strongly contrasted. The west coast is formed by a series of sweeping beaches separated by small rocky headlands, whereas the east coast is made up of small rocky bays and promontories.

This contrast is also a marked feature of the Uists and Benbecula. On the west side is a gentle sandy coastline backed by a strip of fertile duneland or machair (see Ritchie, 1979). This gives way in the centre to a flat-lying maze of bog, rock and water, excellently seen from the summit of Rueval (125 m) in Benbecula. Running down the eastern seaboard is a spectacular spine of hills which reach their highest point at Beinn Mhor (608 m) in South Uist. The hills rise steeply on their western side and fall off eastwards to the cliff-lined east coast. The line of hills is broken by several sea-lochs, some of which penetrate more than halfway across the islands, thus providing excellent anchorages, for example Loch Boisdale, Loch Eynort and Loch Eport.

Traces of this tripartite division of machair, low lying rocky bog and hills/rocky coast can still be found in Barra, although the hill of Heaval (385 m) tends to dominate the island. Southwards from Barra lies the string of islands Vatersay, Sandray, Pabbay, Mingulay and Berneray, the last few being notably asymmetric, with high cliffs to the west sloping gently down to the east; the west coast cliffs of Mingulay and Berneray are over 200 m high.

Summary of the geology

With the exception of a small area around Stornoway, the islands of the Outer Hebrides are composed of an assemblage of rocks collectively known as Lewisian, the name being derived from the Isle of Lewis. These Lewisian rocks constitute, along with similar rocks on the north-west mainland, part of an ancient basement to the much younger Caledonian metamorphic rocks of the Scottish Highlands.

The main islands of the Outer Hebrides (the Long Island) lie on a Lewisian Ridge bounded to the east by the Minch Fault (Figure 2), which may have formed by reactivation of the Precambrian Outer Isles Thrust complex. The Minch Fault forms the western margin of the Mesozoic to late Palaeozoic sedimentary basins of the Minch which contain up to 4 km of fill. These sediments are extensively intruded by Tertiary basic sills which include the crinanite-teschenite sheets of the Shiant Isles, south-east of Lewis.

West of the Long Island Lewisian rocks form an extensive submarine platform, but north of 58°N the Flannan Trough, a shallow fault-bounded Permo-Triassic basin, occurs between Lewis and the Flannan Ridge on which Lewisian gneisses outcrop in the Flannan Isles. Westward of the Flannan Ridge are Palaeocene lavas and younger Tertiary sediments of the south-east side of the Rockall Trough.

Between the north end of Lewis and Sula Sgeir is the North Lewis Basin, which contains up to 5 km of Permo-Triassic and Liassic sediments. Sula Sgeir and North Rona, with their outcropping gneisses, lie on two of several small fault-bounded Lewisian inliers in the basin.

There are close similarities in rock type, age and geological evolution between the Lewisian of Scotland and the Precambrian rocks of Greenland. The Scottish rocks represent a tiny fragment of a large crustal block, the Laurentian Shield, which comprises much of Greenland and northern Canada; the western edge, carrying the Outer Hebrides and north-west Scotland, separated from the remainder on the opening of the North Atlantic during the Mesozoic.

Most of the rocks comprising the Lewisian were formed prior to 2800 Ma ago (Moorbath et al., 1975) and were repeatedly modified under conditions of high temperature and pressure deep within the earth's crust. The intensity of these periods of deformation and metamorphism has obscured the original nature and relationships of most of the rock units and many uncertainties regarding their geological history still remain. In general, however, the Lewisian Complex can be divided into the products of two broad tectonothermal events. The older Scourian event covers the time span from about 2900 to 2500 Ma and the younger Laxfordian event from 2500 to 1400 Ma. The separation of these two episodes is marked by a widespread suite of basic minor intrusions, the Scourie dykes, which were intruded about 2400 Ma ago. The dykes, therefore, cut across rocks and structures formed during the Scourian episodes and have themselves been affected by the later Laxfordian events. The Laxfordian terminated in a major thrust event which produced the Outer Hebrides Thrust Zone.

As can be seen from the accompanying map (Figure 2) the greater part of the Outer Hebrides is composed of a coarse-grained, banded or streaky gneiss, commonly with dark lenses composed almost entirely of hornblende. The gneiss on weathering tends to produce a subdued undulating landscape typified by the central part of Lewis. Within the gneiss in certain areas are rocks of a different character which are mainly of sedimentary origin (Coward et al., 1969). They occur on a variety of scales from well-defined belts extending over distances of up to 15 km down to small discontinuous lenses a few metres in length. Rock types recorded from these zones include calcareous and graphitic gneisses, quartzites and rare marbles, together with flaggy, biotite-rich garnetiferous gneisses. Intimately associated with many of these metasedimentary rocks are basic and ultrabasic rocks of igneous origin, some of which commonly show a well-developed banding which may mimic a pre-existing igneous layering.

All the rock groups described above were subjected to several phases of deformation and metamorphism during the Scourian event. Towards the close of this event two widespread, but not abundantly developed, suites of intrusions were emplaced. The first group comprises diorites, monzodiorites and microdiorites and the second group comprises potash-rich granites, monzonites and pegmatites. After the close of the Scourian event the Scourie dykes were intruded. Although subsequently modified during the Laxfordian event, the dykes can be recognised through the entire length of the Outer Hebrides from Barra Head to the Butt of Lewis, as well as on the mainland, and must have constituted part of a regional swarm of considerable extent. Their original orientation is difficult to determine in the Outer Hebrides but on the mainland they probably followed fissures running north-west to south-east. The dykes occur as distinctive dark brown or black, parallel-sided bands varying in width from a few centimetres to several tens of metres, and are too small and numerous to be shown on the 1:100 000 geological maps.

Subsequent to the intrusion of the Scourie dykes, but prior to the main Laxfordian deformation, an assemblage of igneous rocks of more substantial proportions was intruded at about 2250 Ma. Together they constitute the South Harris Igneous Complex with gabbro, anorthosite, norite and diorite as the main components. The only other meta-igneous body of comparable size and possibly age is the Corrodale Gneiss of South Uist (Coward, 1972). This is a banded gneiss of diorite composition which because of its massive nature forms Ben More and Hecla, the highest hills in the southern isles.

The present day large-scale structural pattern seen throughout the islands was established during the main Laxfordian event, the peak of which occurred at about 1700 Ma. The characteristic style of deformation is one of markedly asymmetric folds with broad rounded hinge zones and narrow attenuated limbs. The Scourie dykes act as a datum by which the intensity of the subsequent Laxfordian deformation may be judged. In the small areas of minimum Laxfordian strain Scourian features and Scourie dyke textures and even minerals may be preserved. In the larger areas of high Laxfordian strain the dykes have been brought into concordance with the gneiss foliation and the original minerals and textures have been reconstituted.

Towards the end of Laxfordian times, a suite of granites and pegmatites were intruded. Although they occur locally in the southern islands the granites reach their greatest development in south-west Lewis and South Harris (Myers, 1971) where several varieties of coarse grey and pink granite are present, all cut by veins of slightly younger, very coarse pink pegmatite. Relationships between the granites and the gneisses are complex, the granites forming a ramifying network of thick veins and irregular bodies in the gneiss. In general, however, a central north–south core in which granite is dominant is flanked to the east and north by a marginal belt of veined gneiss, the veins dying out fairly abruptly eastwards. The silica-rich composition of the granites and their hard and compact nature is one of the main reasons for the barren high landscape of the Uig Hills, the Forest of Harris and the central part of South Harris. Exposed to the full force of the Atlantic in west Lewis the granite-veined gneiss has given rise to one of the few areas of well-developed sea-stacks in the Hebrides.

After the emplacement of the granites, and probably before they had completely cooled down, there took place a phase of thrusting which imparted to many of the granite sheets, and less markedly to the gneiss, a planar fissility. These thrust effects become increasingly more prevalent eastwards, culminating in the Outer Hebrides Thrust Zone which commonly marks the eastern seaboard of the islands. Along this belt a fine-grained, dense black, flinty rock known as pseudotachylyte is widely developed. Pseudotachylyte is thought to have formed by brittle fracturing of the gneiss within the crust resulting in local melting by frictional sliding (Sibson, 1975). The hot molten material so formed, injected the adjacent gneiss and solidified to form an intricate vein network of black glass. In many places along the thrust belt this pseudotachylyte has welded together otherwise broken and friable gneiss making the rock more resistant to weathering and giving rise to prominent hills such as the Lees and Eaval on North Uist. Subsequently, at a much later time, the thrust zone was reactivated and the rocks were again thrust north-westwards with the formation of narrow belts of schist-like mylonite which are particularly well seen in south-east Lewis and along the east coast of South Uist. Later dip slip movement also occurred along the thrust zone. These stages are believed to be related to concurrent movements on the Moine Thrust of the mainland near the end of the Caledonian orogeny.

No evidence exists in the Outer Hebrides for sedimentary rocks of Torridonian or Cambro-Ordovician age such as occur extensively overlying the Lewisian of the mainland. If they were ever present they have since been eroded or removed by glacial action. The only sediments preserved on the main islands are the 'Stornoway Beds', a thick sequence of sandstones and conglomerates which occur as a down-faulted block around Stornoway on Lewis; they probably represent a marginal facies of the Mesozoic beds infilling the basin of the North Minch (Steel and Wilson, 1975). It is the relative ease with which these sediments have been eroded that has formed the good agricultural land around Broad Bay and led to the formation of the isthmus of Point. A few metres of Liassic shales and marls occur between the Tertiary sills of the Shiant Isles.

Igneous dykes of various ages cut the gneisses of the Outer Hebrides. The oldest, apart from the Lewisian suites, are quartz-dolerites of Permo-Carboniferous age (290 Ma) occurring as east–west dykes up to 35 m wide in a narrow swarm between north Barra and southern South Uist. These are cut by members of a slightly younger set, also east–west and belonging to the camptonite-monchiquite suite. They occur as thin dykes (average 1 m) throughout the Isles. Finally, extending from Skye across the Minch are thick dykes of Tertiary basalt and dolerite, especially in a northwest to south-east belt running from Loch Roag to Loch Shell on Lewis. Geophysical evidence indicates numerous basic sills intrusive into the Mesozoic sediments of the Minch basin; these outcrop as the differentiated crinanite-teschenite sill of the Shiant Isles and two sills at the mouth of Loch Maddy, one occurring west of the Minch fault on the mainland of North Uist.

Geological terminology

As part of the Lewisian complex of north-west Scotland the geological history of the Outer Hebrides has always been referred to the formalised system of the mainland. There are, however, important differences and it is necessary to define the usage of the terms in this account.

In their original discussion of the mainland rocks Sutton and Watson (1951) proposed the terms Scourian and Laxfordian to define the events separated by the emplacement of the Scourie Dyke suite. The Scourian was a period of high-grade metamorphism, commonly granulite facies and a rather irregular deformational style; the Laxfordian was a period of reworking with retrograde metamorphism (amphibolite facies) and a series of linear structures predominantly orientated NW–SE.

Subsequently Evans (1965) and Evans and Tarney (1964) defined a series of pre-Scourie Dyke Laxfordianoid shear zones with an accompanying retrograde metamorphism which they termed Inverian. Park (1970) in a review of Lewisian chronology suggested that since the main change in structural style occurred at the onset of the Inverian, the Scourian should be divided into the Badcallian (the early high-grade event) and the Inverian, each having similar status to the Laxfordian.

There are many inconsistencies in the use of these terms, some of which have been used in chronostratigraphic, lithostratigraphic and tectonostratigraphic senses by different authors. There is also a certain overlap in the usage of different terms. It is beyond the scope of this memoir to rationalise this situation. Here the terms have been used in the following way (Table 1). Scourian means the main tectonothermal event which culminated at around about 2700 Ma (Pidgeon and Aftalion, 1972; Moorbath et al., 1975). Its various phases are ill defined but it is generally responsible for the migmatisation and general 'gneissificiation' of the rocks. It is believed that there was no widespread granulite facies metamorphism at this time in the Outer Hebrides, but that the rocks recrystallised dominantly in the upper amphibolite facies. The assemblage of rocks which were subjected to these Scourian events are referred to by the informal term pre-Scourian complex. This complex contains a variety of lithologies with igneous and sedimentary parentage which may or may not cover an extensive period of Archaean history. Later than the migmatisation of the Scourian but prior to the intrusion of the Scourie Dyke suite are a variety of igneous intrusions with, at least locally, associated deformational events. The last of these intrusions comprise a suite of granites and pegmatites which have been dated at about 2600 Ma (see Chapter 6: Late-Scourian intrusions). The date given by these intrusions defines these deformational events as late Badcallian. Cutting this complex is a widespread but coherent suite of metadolerites and allied rocks which are termed the Younger Basics and which are believed to be equivalent to the Scourie Dyke suite (see below). The intrusion of some members of this suite appears to be associated with active shear zones, a tectonic setting similar to that described from the mainland by Park and Cresswell (1973); these shear zones are thus Inverian. It is therefore possible to define Badcallian, late-Badcallian and Inverian events in the Outer Hebrides and these terms have been used in specific discussion. However, there are several structural elements which cannot be satisfactorily ascribed to these events, so in general discussion the relatively less formal term late Scourian has been used.

A number of basic to intermediate dykes are believed to be of later age than the Younger Basics and are referred to as late-Laxfordian minor intrusions. These are believed to be largely of microdiorite, and the Younger Basics are distinguished from them by their almost total lack of biotite.

The Younger Basics and their host rocks were subjected to the polyphasal events of the Laxfordian episode, which was bracketed by the emplacement of the Scourie Dyke suite (about 2400 Ma, Lambert et al., 19706; Chapman, 1979) and the late Laxfordian granite and pegmatites at 1800–1700 Ma, van Breemen et al., 1971). The end of the Laxfordian is marked by granite intrusion and probably by the inception of the Outer Hebrides Thrust zone. No sediments have been recognised of post-Scourian and preLaxfordian age.

The problem of the Scourie dykes

It is obvious that the role of the Scourie Dyke suite is fundamental to the above scheme. The recognition of Scourie Dykes, particularly in a highly 'Laxfordianised' state is critical. If all the Younger Basics are of Scourie Dyke age then this recognition is comparatively easy. If on the other hand some of the Younger Basics are of intra-Laxfordian age, then recognition becomes virtually impossible and the use of the above scheme impractical. This possibility has been the focus of extensive debate in Lewisian literature (see Park, 1970). That some of the Younger Basics are of Scourie Dyke age is probably generally accepted. However, whilst many workers (e.g. Francis, 1973; Coward, 1973b; Davis et al., 1975) have followed Dearnley (1962a) in regarding the Younger Basics as a single suite identical to the Scourie Dyke suite, other workers (Bowes and Hopgood, 1975; Taft, 1978) have argued in favour of several ages of intrusion, some of which may be intra-Laxfordian.

It should be emphasised that this discussion is concerned only with the Younger Basics as defined above, that is metadolerites and allied rocks. Most workers would agree that the mafic microdiorites etc. defined above as the Late Laxfordian suite are of intra- or Late-Laxfordian age. Also, the Younger Basics can be distinguished in the field from the pre-Scourian Older Basic metadolerites by the Older Basic's generally higher degree of migmatisation (see Chapter 4: Older Basics and Ultrabasics).

Much of the argument on the Scourie Dyke problem rests on what is taken as a datum. The one group of workers (e.g. Hopgood and Bowes, 1972) believe that a distinct structural sequence can be discerned throughout the islands, and at any one locality individual folds can be placed within the sequence on the basis of characteristic geometry (style and orientation). If basic dykes are then seen to be variously affected by folds of different geometries, the folds and in consequence the dykes must be of different ages (Taft, 1978). The other group of workers, however, do not accept the concept of distinctive fold geometries in basement terrains that demonstrably exhibit a highly variable strain pattern (see discussion of Francis, 1973; Coward, 1973a) and assert, therefore, that the structural sequences do not provide a valid datum. Instead they believe, partly on the basis of correlation with the mainland Lewisian, that the Younger Basics are mainly of one age and as such provide a specific datum, so that structures that are cut by the dykes are Scourian, and those that affect the dykes are Laxfordian, following the original ideas of Sutton and Watson (1951).

Resolution of these opposing views is of fundamental importance and a considerable effort was made during the present survey to examine the critical evidence and to try to reach a conclusion. Essentially the problem may be considered as two questions:

Do the Younger Basics represent more than one age of intrusion?

1 There is an obvious variation in the strain state of the different deformational episodes across the area (e.g. Graham and Coward, 1973) which is reflected in fold tightness, style, axial plane orientation and related lineation and foliation development. Since in many places there is even a demonstrable variation in the geometry of parasitic folds around major folds, the view of unique sequences is difficult to substantiate and the arguments for multiple intrusion are seriously weakened (see also discussion of Francis, 1973; and discussion of Coward et al., 1969; Davies et al., 1975).

2 In the Outer Hebrides there are areas (e.g. Leanish, Barra; Ardivachar, South Uist; Garry a-siar, Benbecula) characterised by markedly cross-cutting Younger Basic intrusions. At the edges of these areas the dykes enter deformation zones where they become highly foliated and largely concordant. This transition is commonly abrupt and may take place over as little as 0.5 m (Coward, 1973a). Transitions into deformed zones have also been recorded from Fuday (Francis, 1973) and Pabbay (Graham, 1970).

Throughout the islands there is a clear variation in the degree of discordance or strain state of the Younger Basics. The current investigation has also shown a good correlation between the mineralogy of the Younger Basics and their deformational state. The critical point is that where within any one area the gneiss appears to have a consistent strain state, all the Younger Basics of similar size show corresponding degrees of deformation and recrystallisation. There is no indication of the variety of strain states or mineralogy that might be expected if the Younger Basics were of different ages in the structural sequence.

3. The cross-cutting dykes of Benbecula have been placed by Bowes and Hopgood (1975, table 1) between their F3 and F4 fold episodes. However, Coward (1973a) has demonstrated that the cross-cutting dykes may be traced into highly deformed f2(sensu Coward) zones. Folds within these f2 zones are tight to 'almost isoclinal', a fact at variance with the deformational style of Bowes and Hopgood's post-dyke F4–F6 folds which are described as 'open' or 'very open'.

We conclude that the Younger Basics are a largely coherent suite of metadolerites and allied rocks of regional significance, which exhibits a variable degree of deformation throughout the Outer Hebrides. (Plate 3) illustrates the local variations in the Younger Basic sheets at Howmore Quarry, South Uist. It cannot of course be proved that there are no metadolerites of different age, but if there are any, they are very few.

Is the Younger Basic suite of Scourie Dyke age?

1 The most obvious method of dating the suite is by geochronology. So far, however, only two published sets of dates are available. One from Cleitichean Beag in Lewis is a whole-rock K-Ar age of 2440 ± 60 Ma (Lambert et al., 1970b) close to the Rb-Sr whole rock age of 2390 ± 20 Ma for dykes from Assynt (Chapman, 1979). The other published source (Lambert et al., 1970a) gives whole-rock K-Ar dates of 1720, 1565 and 1500 Ma on dykes from South Uist and Benbecula: the authors refer these dates to the 'age of regional metamorphism and pegmatite intrusion'. Taft (1978) suggests that a comparison may be made between her third phase of basic igneous activity (that between her F3 and F4) and intrusive events in South Rona and the mainland dated at 1720 and 1740 Ma. However, such a comparison rests on the structural correlations which are open to doubt (see above), and Taft (1978) states that the absolute age of the dykes on Lewis has yet to be determined. It is hoped that new methods of geochronological dating (e.g. Nd-Sm systematics) may soon prove useful.

2 Dearnley (1962), in his original suggestion that the Younger Basics were equivalent to the Scourie Dykes, argued on the basis of similar chemistry and petrography, a view criticised by Hopgood (in discussion of Coward et al., 1969). We agree that a correspondence of tholeiitic chemistry cannot necessarily be taken to indicate similar ages. Equally, of course, the similarity does accord with a mainland and Outer Hebrides correlation.

3 Perhaps the most telling argument, however, is one of simple correlation. The Scourie Dykes form an extensive suite throughout the mainland Lewisian, and it is reasonable to assume that they would appear in the Outer Hebrides (although more or less affected by the Laxfordian episode). Only one suite of Younger Basics has been identified in the Outer Hebrides, and that suite is thus most likely to be equivalent to the Scourie Dykes (s.s.). This argument is particularly attractive in that the Scourie Dyke suite has been identified as part of a great swarm found throughout the North Atlantic Proterozoic continent (Escher et al., 1976a). It would, therefore, seem perverse to argue that the Younger Basics are not part of this suite.

We conclude that at least almost all Younger Basics are part of the Scourie Dyke suite and its equivalents in the North Atlantic swarm.

Miscellany

Place names

Spelling of place names in the Outer Hebrides varies considerably. This memoir conforms to those given in the Ordnance Survey Second Series 1:50 000 maps. National Grid references are given in the following forms: [NR 1234 5678] and [NR 123 567] Rarely the form [NR 12 56] is used.

Mineral abbreviations

The following abbreviations have been used in mineral assemblages in the text:

Normative minerals

The following abbreviations have been used in discussion of normative minerals

Chemical analyses

Full details of analytical methods and analysts are contained in BGS files held in Edinburgh.

Thin section numbers

Thin section numbers prefixed 'S' are part of the collection of the Scottish office of BGS. Those prefixed 'MC' are part of the collection made by M. P. Coward and deposited at Imperial College, London.

Use of the prefix meta-

All rocks comprising the Lewisian Gneiss have been more or less metamorphosed. However, in discussion of many igneous suites the prefix meta- has been dropped for convenience: e.g. Younger Basics instead of Younger Metabasics, and the tonalites (not metatonalites) of South Harris.

Radiometric age dates

All Rb-Sr ages have been recalculated using a 87Rb half-life of 4.88 X 1010 year.

Amphibole formulae

All amphibole formulae given have been calculated using a computer programme (Rock and Leake, 1984) which allocates total Fe between Fe3+ and Fe2+ to bring the sum [Si + Al + Fe3+ + Ti + Fe2+ + Mg + Mn] to 13.0, as recommended by Leake (1978).

Chapter 2 Lewisian Complex: introduction and quartzofeldspathic gneisses

Introduction

The term Lewisian Gneiss was suggested first by Murchison (1862) for the 'fundamental gneisses' of Scotland. It was not until the turn of the century, however, that the name came into general usage, when it was adopted by the Geological Survey as a composite term to cover all rocks of preTorridonian age. Murchison chose the name because of the preponderance of gneiss in the Isle of Lewis: indeed, with the exception of the Stornoway Beds and the Phanerozoic dyke swarms, the whole of the Long Island is composed of Lewisian rocks.

Within the Lewisian Gneiss complex of the Outer Hebrides several lithological groupings can be recognised, although in many areas deformation and metamorphism have blurred the distinctions between some of these groups and made them look similar. These groups form the basis of subdivision of the chapters on the Lewisian as follows:

Although the South Harris Complex contains elements of other groups (e.g. Late-Laxfordian granites and pegmatites) it has historically been treated as an entity both in research work and literature. We have therefore largely treated it as such in this memoir.

Lithology

Quartzofeldspathic gneiss is by far the most common rock type in the Outer Hebrides, forming the 'general country rock' of the area. The term is used to cover all the rocks not easily ascribed to any other lithological group and as such contains the highly metamorphosed and deformed derivatives of the other groups. To cover this rather amorphous grouping Dearnley (1962a) introduced the term 'grey gneiss'.

Although the gneiss is surprisingly uniform on a regional scale it is locally strongly heterogeneous. It consists of a grey medium- to coarse-grained rock which generally shows a rough banding defined by mafic concentrates. The composition varies from biotite gneiss to biotite-hornblende gneiss, with a variable but usually low amount of potash feldspar.

The gneisses may also exhibit lit-par-lit pegmatitic banding which is intense locally. In addition the gneisses generally contain a variety of basic and ultrabasic bands, lenses and clots, all of which may show some degree of assimilation and recrystallisation. Because of their long and involved history the gneisses show a wide variety of textures ranging from highly migmatised and homogenised rocks to fine-grained highly sheared banded gneisses. Further details of the banding and of the mesoscopic fabric are given in Chapter 3.

Lewis and Harris

Lewis and Harris consist of a mixture of biotite gneiss and biotite-hornblende gneiss. No attempt has been made during the present survey to delineate gneiss types. Jehu and Craig (1934) state that hornblende-biotite gneiss is prevalent in Lewis, and Stevens (1913) suggests that hornblende gneiss is dominant to the west of Stornoway. Beer (1952) reports hornblende gneiss as the prevalent rock type in northern Lewis, and suggests that hornblende-free biotite gneiss is relatively uncommon and restricted to localised outcrops where it characteristically weathers into small compact knolls. Pyroxene is rare in the gneisses.

Myers (1970a) and Soldin (1978) have distinguished and mapped out various gneiss types in Harris. Myers has three main types: biotite-banded gneiss, amphibolite-banded gneiss, and pegmatite-banded gneiss. Biotite-banded gneiss predominates in the west and south (Myers, 1970a, fig. 1). Soldin (1978, fig. 3.2) shows hornblende-biotite gneiss in a broad belt running westwards from Seaforth Island in the area south of Toddun [NB 210 030] and from Loch Trollamarig [NB 21 02] westwards to Tarbert. Jehu and Craig (1934) suggest that biotite gneiss is the main rock type in eastern North Harris. Although those gross regional differences probably exist it is generally demonstrable in any traverse that the gneisses have been derived from a complex sequence of granitic, granodioritic and dioritic parents (see below).

Migmatitic banding marked by lit-par-lit pegmatites, generally 2–5 cm thick, is developed throughout the region, although considerable local variation in intensity occurs. As shown below it records both Scourian and Laxfordian events. In Lewis the degree of migmatisation is relatively low over most of the region, increasing dramatically in the Uig Hills area. This is well seen on the roadside traverse westwards from Grimersta. At Loch Sgaire [NB 2014 2869] excellent exposures in a small roadside quarry show grey gneiss with some thin (0.2 cm) and widely spaced (c.30–50 cm) pegmatite bands cut by a basic dyke. Some of the pegmatitic lits show a slight discordance to the banding in the gneiss. Similar rock types can be seen in roadside cuttings westwards to Carishader [NB 100 330], a particularly good example being seen on the west side of Little Loch Roag at [NB 127 266]. Westwards from Carishader there is a marked increase in the degree of migmatisation as is well seen on the headlands at Ard More Mangersta [NB 004 333] and Aird Fenish [NA 993 294]. Migmatitic gneisses are also seen on the north side of Loch Resort and the south-west end of Loch Langavat at [NB 1530 1693].

In Harris migmatitic gneisses are particularly abundant in the south and west. They are equivalent to the pegmatite-banded gneisses of Myers (1970a) and the zone of granite gneiss of Soldin (1978). They are well seen south of Tarbert in various road cuttings, for example along the 'Golden Road' past Drineshader [NG 172 948] and south towards Geocrab [NG 116 910]. In contrast to those gneisses, the roadside exposures from Ardhasaig [NB 131 049] northwards to Ardvourlie [NB 188 105] show only thin and widely spaced pegmatitic lits.

Scourian migmatisation

This phase of migmatisation postdates the Older Basics; in many places the migmatisation invades and partially digests them to form basic agmatites. Agmatites of this type which have been relatively unaffected by later deformation can be seen at various localities (see Myers, 1971, fig. 1) in Harris. Elsewhere blebs and lenses of Older Basics can commonly be seen in the migmatitic gneiss, for example on the south facing cliffs of Aird Fenish. At the roadside quarry east of Loch Sgaire [NB 202 287] a basic dyke cuts a basic agmatite. The agmatite may be traced in a band, apparently representing a migmatised Older Basic sheet. This disruption of basic bodies or bands during the Scourian must have been in part tectonic. Extreme examples of these effects have led to the production of a rather unusual and distinctive gneiss type. The rock is a fairly uniform medium-grained leucogneiss with discrete amphibolite pods. Myers (1970a, fig. 20, plate 2A) refers to this as amphibolite-lens gneiss and regards it as formed from the disruption of basic bands and sheets during the Scourian with homogenisation of the gneiss during the Late Scourian. This gneiss can be seen at several localities (Myers, 1970a, fig. 1), some of the best localities being on the southern slopes of Lag Macgodrom [NB 06 09]. Lisle (1974, p.26) described occurrences of this lensoid gneiss' in north-west Lewis. Myers also suggests that the Late-Scourian migmatisation and recrystallisation leads to local homogenisation of the gneiss with the disappearance of the gneissose banding. For the biotite gneiss this leads to the production of a rock virtually indistinguishable from later granites.

The Scourian age of much of the migmatisation is easily demonstrated with numerous examples of Younger Basic dykes cutting migmatitic banding, for example at Enaclete [NB 121 275], Lundale [NB 1817 3262], Lurg a Mula [NB 1179 1595], Aribruaich [NB 2340 1832] and at the east end of Loch Erisort [NB 3855 2370]. In Harris some spectacular examples can be seen at [NB 055 083] by Loch Leosaid (see (Plate 8)), at [NB 019 096] south of Creagan Ruadha (see (Plate 9)), and on the west side of the Loch Chliostair dam road just beyond its junction with the Hushinish road (see also Myers, 1971, fig. 1).

Laxfordian migmatisation and pegmatites

The pre-dyke or Scourian migmatisation is quite separate from the extensive phases of recrystallisation or remobilisation of the migmatitic gneisses which occurred during the Laxfordian. At several localities near Flodabay [NG 09 88] granite bodies up to 20 m2 in area may represent an extreme degree of Laxfordian migmatisation. These granites typically form diffuse pods, lenses and veins within the gneiss.

This second phase of migmatisation is generally very variable and many areas are largely unaffected (see below). Since it results in part from a remobilisation of the earlier migmatitic lits it reaches its greatest development in areas of strong Scourian migmatisation. In its weakest form the Laxfordian migmatisation is evidenced by intense recrystallisation leading to pegmatite production in the sheared out or attenuated limbs of folds (e.g. the roadside north of Kinlochroag [NB 1400 2391], east of Loch Geshader [NB 122 323], on Bearasta Mor [NG 126 949], NE of Sgaladal [NB 030 220] and the west side of Loch Langavat at [NB 1530 1693]. Myers (1971b) describes folds of this type with axial surfaces east–west trending and general amplitudes of about 5 m. Pegmatitic patches also occur in the necks of disrupted or boudinaged basic dykes, and here a certain degree of recrystallisation and digestion of the dyke may also occur. The greatest effects are seen in south-west Lewis in the Uig Hills, and in western North Harris and South Harris, where recrystallisation may be so extensive that patches of rock of 0.5-1.0 m2 completely lose their gneissose fabric, forming a coarse granitoid rock. In the same general areas pegmatisation is extensive, and small pegmatites that have apparently been generated locally from remobilised gneiss penetrate the surrounding rock. There they cause or are associated with extensive recrystallisation and coarsening of the gneiss fabric. The junction between the gneiss and these pegmatitic patches is generally gradational and poorly defined. Larger cross-cutting and intrusive pegmatites may terminate in generally diffuse patches of coarsened gneiss. The pegmatisation and general coarsening of the gneiss produces a characteristically coarse and sugary appearance extensively seen in the northern and central parts of the Uig Hills, for example, at Lighe Mhor [NB 060 303], the head of Loch Raonasgail [NB 038 265], Tarain [NB 050 275], Loch Benisval [NB 100 200] and on the north side of Loch Tamanavay [NB 030 219]. Such gneiss has a poorly defined foliation which in parts seems to disappear completely on the weathered surface although on the fresh surface a rough compositional banding can generally be picked out. There is a definite relationship between extensive Laxfordian recrystallisation and homogenisation, and the presence of Late-Scourian pegmatitic lits. In Harris larger areas of apparently remobilised gneiss or, more particularly, remobilised Scourian pegmatitic lits do swamp early Laxfordian folds. There is no evidence that Laxfordian structures or fabrics are greatly destroyed by this Late-Laxfordian migmatisation. The minimum age for this recrystallisation is given by the relationship of the gneiss to the Late-Laxfordian granites. Within the Uig Hills and in west and south Harris many examples can be seen of discrete granites cutting recrystallised gneisses. Even where the gneiss has been extensively remobilised to form homogeneous granitoid patches the granites commonly have sharp contacts. Also, in the area west of the foot of Loch Suainaval, numerous examples can be seen of discrete xenoliths of coarse pegmatised gneiss contained in granitic sheets. The relationship of recrystallisation of the gneiss to the granites is further discussed in Chapter 11: Granite genesis.

Where the gneisses have not been subjected to extensive Late-Laxfordian recrystallisation, their appearance in the field is very much controlled by their degree of deformation. Where the rock is highly strained, the migmatitic Scourian banding is characteristically streaked out to present a fine- to medium-grained, finely banded rock. The pegmatitic lits are in many places disrupted to form a series of augen. Rocks of this type are seen on the road south from Stornoway [NB 415 341], on the road to Arnish [NB 40 32], in the Balallan–Arivruaich area and on the Tarbert–Ardvourlie road. One of the best examples of highly strained rock is seen at Tiumpan Head, where the gneiss forms a uniform finely banded rock, well exposed on the cliffs just west of the lighthouse. On the south cliffs of Aird Fenish a dramatic example can be seen of the juxtaposition of coarse migmatitic gneiss and a fine-grained, uniformly banded, highly strained variety.

In addition to these essentially ductile deformational effects, the gneisses of Lewis characteristically show some degree of late-stage cataclasis. Although these textures are generally greatest towards the thrust zone, nearly all the gneisses show, at least in thin section, evidence of late deformation. Peach and Home (1930) remarked on this widespread cataclasis and highlighted it as one of the differences between the gneisses of the Outer Hebrides and those of the North-west Highlands. Below the Outer Hebrides Thrust, late-stage deformation is locally intense within a belt running north-westwards across Lewis bounded on the south-west by Loch Roag and Loch Seaforth, and on the east by a line through Callanish and the head of Loch Erisort. Within this belt the gneisses commonly exhibit a platy texture with concordant bands of ultramylonite. Within the thrust zone the rocks typically show evidence of low-grade alteration. In South Harris this zone of alteration appears below the thrust. Here, the rocks have a characteristic rather soapy look, and are rich in epidote. These rocks and their probable relationship to the thrust are discussed in Chapter 15: Mylonites; Folding.

Uist and Benbecula

In this region, which is taken to include the islands in the Sound of Harris, there is a recognisably higher proportion of metasediments and characteristically banded Older Basics than in Lewis and Harris. In addition much of central North Uist and the islands in the Sound of Harris consist of fine-grained gneisses with metasedimentary affinities: these rocks are discussed separately below. Otherwise the gneisses are essentially the same as those in the northern isles, being a mixture of banded biotite gneisses and biotite-hornblende gneisses with a variable number of pink pegmatitic lits. There is little information on the relative distribution of the gneiss types: Jehu and Craig (1925) suggest that biotite gneiss is the most common rock-type west of the thrust in South Uist. On the north-west coast of North Uist, Jehu and Craig (1926) describe areas of very pink acid gneiss, which they suggest are probably later than the biotite-hornblende gneiss. Graham (1970) also distinguishes a 'rough' and a 'smooth' gneiss on the north-west coast, the smooth gneiss being described as 'unhanded, homogeneous and finer grained than "rough gneiss'. The 'smooth gneiss' would, therefore, appear to be the same as Jehu and Craig's acid gneiss. In this memoir this 'gneiss' is regarded as a Late-Scourian 'granite' and is discussed in the appropriate section below (p.45).

Coward et al. (1969) and Coward (1973b, fig.1) report the occurrence of pyroxene-bearing gneisses in the Loch Skiport area, which they attribute to an early granulite-facies metamorphism. Another small patch of pyroxene-bearing gneiss occurs in association with a large cross-cutting dyke on Beinn Bheag Dheas [NF 78 29] in central South Uist.

Dearnley (1962a) originally divided this area into northern and southern parts, the southern part being characterised by, among other things, a preponderance of migmatites analagous to those of the Uig Hills and West Harris. Dearnley (1962a, pl. ix) also showed the north-western part of North Uist as being highly migmatised. Dearnley and Dunning (1968) cite several localities, mainly on the west coast of South Uist, where the gneisses show evidence of strong Laxfordian migmatisation.

Scourian migmatisation

A distinction must be drawn, however, between Scourian and Laxfordian migmatisation. Scourian migmatisation is widespread throughout the region although, as in Lewis and Harris, of variable intensity. Migmatised gneisses evidenced by pink pegmatite lits are cut by Younger Basic dykes, for example, at Ludag [NF 7754 1468], Orosay [NF 7301 1737], Beinn Bheag [NF 8077 1684], Rarnish [NF 872 483] and Vallay [NF 7605 7649]. Migmatised Older Basic sheets or agmatites are also abundant, and can be clearly seen to be cut by later basic dykes at Veilish Point [NF 8150 7819], on the north side of Loch Druidibeg [NF 780 893], above Sloe Dubh on Loch Eynort [NF 8030 2891] and at Reineval [NF 7538 2569]. At Ardivachar Point [NF 738 460] Dearnley and Dunning (1968) describe a basic agmatite which is clearly older than the later cross-cutting unmigmatised Younger Basic dykes, and similar relationships can be seen to the north at Hornish Point [NF 759 473]. In addition to this extensive pre-dyke migmatisation, evidence of homogenisation of the gneiss can be seen at Gleann na Beiste [NF 9064 7599]. Here a Younger Basic dyke cross-cuts uniform gneiss that contains flattened lenses of amphibolite very similar to the amphibolite-lens gneiss described from the northern isles.

Laxfordian migmatisation

Post-dyke migmatisation and remobilisation of the gneisses was less widespread than in the Scourian phase, and as in Lewis and Harris it is most intense in areas of pre-dyke migmatisation. It is manifested by the general coarsening of the gneiss and a destruction of the gneissose banding. Coarsely crystallised bands have formed, some with disseminated hornblende, apparently as a result of a local recrystallisation of the gneiss. Elsewhere locally generated pegmatite stringers and veinlets intrude the surrounding gneiss. It is, however, around the Younger Basic dykes that the greatest degrees of remobilisation are observed, with numerous examples of the dykes being assimilated or recrystallised. For example, at Trollaskeir [NE 726 273] a basic band can be seen with a strongly recrystallised margin, remobilised gneiss penetrating the margin of the dyke in small stringers and veinlets and the adjacent gneiss becoming homogenised. The opposite margin of the dyke has remained curiously unaffected, with a sharp contact against the gneiss. At Beinn Bheag Tuath [NE 7853 3038] excellent exposures on flat-lying slabs show basic dykes cut by quartzofeldspathic veins. The gneiss shows a Scourian migmatitic banding, and the basic sheets have undergone blocky boudinage with the appearance of regularly oriented veins of remobilised gneiss between the blocks. In some places the gneiss shows little evidence of recrystallisation adjacent to the dykes; elsewhere bands of remobilised gneiss flank the line of basic blocks, giving the whole an appearance of an agmatite zone with regularly shaped and spaced blocks. Where remobilisation has been most intense, veins of remobilised gneiss may extend across from one basic sheet to its neighbour.

Examples of basic dykes being assimilated by remobilised gneiss are described by Kursten (1957), from the south face of Arnaval. It is important to emphasise the rapid variation in intensity of migmatisation. This is dramatically shown in the quarry at Howmore [NF 7662 3649] where basic sheets have been progressively invaded and recrystallised across the quarry face from north to south, the extreme assimilation of the basic sheet leading to the production of a coarse hornblende-bearing patch of remobilised gneiss. On the north side of Beinn Bheag Deas [NF 78 29], a dyke has been reworked along its length. At the south end the dyke is discrete and cross-cutting, and the margins are fine-grained with blocky joints, whereas, as one moves 200 m along strike to the north it becomes reworked and amphibolitised, with feldspathic stringers and lenses developed parallel to the axial plane cleavage.

The only area where this post-dyke remobilisation of the gneiss reaches an intensity comparable with that seen in Harris is in the area below the thrust between Hecla [NB 827 345] and Loch Skipport (Coward, 1973b, fig. 6A). In this area patches of gneiss appear to be 'transformed' into granite or pegmatite. This phenomenon is most intense in the area immediately north of Hecla, where Coward (1973b) reports that the gneiss has been wholly transformed into a homogeneous granitic mass with mafic clots. As in the northern isles this widespread migmatisation appears to be a fairly late event; Coward (1973b) associates it with his f4 folds, perhaps analogous to the east–west folds of Myers (1971) in Harris (see earlier in this chapter and Chapter 13: Later Laxfordian structures). Examples of remobilised gneiss in the sheared or attenuated limbs or folds can be seen on the west side of Kearsinish [NF 794 174].

Although there are regional zones of vertically foliated, uniformly striking gneiss the rock does not show the same intensity of deformation seen in similar zones in Lewis: the Scourian banding and pegmatitic lits are less broken and boudinaged. Indeed, except where it has been locally homogenised or remobilised during the Laxfordian the main banding seen in the gneiss can be considered as an essentially Scourian fabric.

Within the thrust zone the rocks are extremely deformed and affected by low-grade alteration. In northern South Uist this late alteration extends below the thrust and affects a large area of relatively undeformed rocks. These effects are discussed in Chapter 15: Mylonites, and Late-stage alteration.

Barra and the southern isles

Barra and the neighbouring isles exhibit a variety of gneiss types. On Barra thrust phenomena are confined to a relatively narrow zone, which more or less bisects the island. The rocks above and below this thrust zone are largely unaffected by thrust movements. Above the thrust zone pyroxene-bearing gneisses with a characteristically brown colouring are common on Flodday, Fuiay, Bruernish Headland, and in the area east of a line through Earsary and the head of Loch Obe. Below the thrust, pyroxene gneisses are common on Hellisay, Gighay, Greanamul and Garbh Lingay (see also Francis, 1973). Elsewhere the rock is a mixture of banded biotite gneiss and hornblende-biotite gneiss with a variable number of ultrabasic and basic bands, lenses and clots.

Scourian migmatisation

As in the northern islands the gneisses are commonly migmatitic, with abundant pegmatite lits. Gneiss of this type is excellently exposed at several localities, for example at Borve [NE 651 014], on the south side of Greian Head [NF 650 045], locally on Sgeir Liath [NF 65 03] and over most of the southern isles. The Scourian age of this banding can be demonstrated at several localities: for example, basic dykes can be seen cutting the pegmatitic banding on Fuday at [NF 732 076] and [NF 744 084] and on Rudha Charnain [NL 699 978]. It should be noted that since the east coast of Barra exhibits particularly low Laxfordian strain it is possible to recognise a number of Late-Scourian igneous phases (q.v.) which include a suite of microdioritic dykes and granitic veins, the granites being demonstrably later. The micro-diorite dykes cut the migmatitic pegmatite banding (e.g. on Fuday [NF 744 084] and [NF 741 080], demonstrating their age as post-Scourian migmatisation. Of course, there may well have been some remobilisation of the migmatitic lits during the phase of granite and pegmatite intrusion.

There are many examples of Older Basics being agmatised by the Scourian migmatisation, for example at Halaman Bay [NE 644 005] and Mingulay [NF 570 833]. Good examples of basic agmatites cut by the Late-Scourian dioritic dykes can be seen on Fuday [NF 741 091] and at Leanish [NL 704 987]. Agmatites cut by Younger Basics can be seen at Rudha Charnain [NL 699 978] where an agmatitic hornblende gneiss is cut by an anastomosing dyke.

Laxfordian migmatisation

Remobilisation of the migmatitic lits during the Laxfordian is common, reaching its strongest development on Barra in a zone running south-eastwards from Beinn nan Carnan [NE 683 982] to the coast (zone 2 of Francis, 1973, fig. 1). Within this zone basic dykes are partially agmatised and digested, as described above, remobilisation of the gneiss being commonly confined to areas adjacent to the dykes. One dramatic example is seen at [NL 689 976] where remobilised gneiss has broken through the nose of a large fold defined by a basic dyke, producing a patch of basic agmatite (Francis, 1973, pl. 2). Remobilisation of the highly migmatitic gneisses is seen at other localities, for example, on Borve Point [NE 651 014] where the pegmatite stringers, generated locally from the Scourian lits, cross-cut the gneissose banding. At the same locality remobilised gneiss has broken through a large (about 2 m) basic body producing a patch of agmatite 3–4 m2 in area.

Post-dyke migmatisation and remobilisation is also common on the other southern isles. On Pabbay the effects increase westwards with the amphibolites becoming strongly agmatised and swamped in pegmatite. Similar effects are also seen on the west side of Sandray, and on Mingulay some good examples of basic agmatites are seen on the shore at [NL 566 829]. On Mingulay Bowes and Hopgood (1969a) describe three generations of pegmatite development. The first is nearly parallel to the gneissose foliation and would appear to equate to the general phase of Scourian migmatisation described above. The second phase Bowes and Hopgood associate with asymmetric folds with north-east-trending axial planes: they regard this second phase as locally generated and linked to the earlier lits. They also relate this second phase to a recrystallisation of the gneisses. It would therefore appear that this phase of pegmatisation is equivalent to the general phase of remobilisation described above. The third set of pegmatites is late stage.

Note on the distant isles

The distant islands of North Rona, Sula Sgeir and the Flannans have not been surveyed during the present investigation. The following paragraph is included only as a reference.

On North Rona Nesbitt (1961) records two belts of acid gneiss flanking a central amphibolite body (see Chapter 4: Banded Basics of the northern isles). The general assemblage of the gneiss is given as qz-olig-kf-bi, with zircon as an accessory. Analysis of the gneiss by Bowes (in Nesbitt, 1961) is broadly similar to that of the gneiss on the Long Island ((Table 3) analyses 1 and 2). On Sula Sgeir Stewart (1933) describes the rock as a hornblende gneiss, containing 30 to 40 per cent hornblende with biotite and augite as the other mafic minerals. The only feldspars reported are orthoclase and microcline; quartz is usually absent. Superficially the rock appears to be an amphibolite, possibly poorly migmatised. On the Flannan Isles Stewart (1933) describes the rock as similar to that of Sula Sgeir with more biotite, andesine and iron oxides.

Petrography

It is clear from the above account that the components of the quartzofeldspathic gneisses range in composition from basic to acid. Most of the gneisses, however, lie near the centre of this range. Modal analyses of the main gneiss types (the samples were chosen to cover the range 65 to 71 per cent SiO2 in the analysed samples) are given in (Table 2). They show a range of quartz from 22 to 35 per cent and a range of 4 to 20 per cent in mafic (mainly hornblende and biotite) content. Potash feldspar is always less than 10 per cent and is not significantly higher in the biotite gneiss than in the hornblende-biotite gneiss. Potash feldspar is always subordinate to plagioclase even in the most acid varieties, plagioclase forming about 50 per cent of the rock. These analyses therefore define the gneisses as granodioritic to tonalitic (sensu Streckeisen, 1976) in composition (Figure 3).

In thin section the gneisses show a range of textures from granoblastic medium- to coarse-grained to schistose and highly recrystallised varieties ((Plate 4)c, d, e, f). Many of the gneisses show evidence of late-stage cataclasis resulting in straining and granulation of the main mineral phases ((Plate 4)a). The mafic minerals are generally concentrated in layers or schlieren but biotite may also be more uniformly disseminated.

Typically the gneisses contain qz-pl-kf-bi-(hbl)-(px) assemblages. The accessories may include apatite, zircon, epidote, allanite, sphene and iron ores with secondary white micas and chlorite. There is a complete gradation from hornblende-biotite gneiss to biotite gneiss, the biotite gneiss being the more leucocratic. Apart from chemical composition the mineral phases in the gneisses are controlled by their metamorphic and deformational state. The gneisses range from opx-cpx bearing assemblages to hbl-bi bearing assemblages, and are described as such below. It should be stressed that there is no implication that all the gneisses with lower metamorphic grade assemblages are retrograde products of original high-grade assemblages. Indeed, as is discussed later (p.23), it is believed that very few of the gneisses ever crystallised in the two-pyroxene facies.

The highest grade rocks contain pyroxene. They are particularly abundant in north-eastern Barra and the Loch Skipport area of South Uist but occur only rarely elsewhere.

These rocks generally exhibit granoblastic textures with individual crystals of about 1.0 mm ((Plate 4)e). The assemblage is cpx-(opx)-hbl-pl-qz, garnet being notably absent. The pyroxenes generally form small aggregates of equigranular crytals. The orthopyroxene, where present, is usually a pale pink hypersthene, commonly forming slightly corroded crystals. In the present investigation it has always been found to be subordinate to the pale green diopsidic clinopyroxene. Hornblende is generally brownish green and is present in some proportion in all the rocks. Rarely, it occurs as small equant grains with well-formed crystal faces in contact with the pyroxene, suggesting that it was part of an original equilibrum assemblage. More commonly the hornblende is retrogressive after pyroxene. Biotite may be present, forming deep-red crystal aggregates. The felspar shows no zoning and has a composition of mid-oligoclase to andesine. Quartz is generally restricted to small (c.0.2 mm) grains. Potash feldspar may be present as an accessory in small interstitial grains. One notable feature is the large modal proportion of opaques (magnetite and ilmenite) relative to other gneisses. Small round crystals of epidote and apatite may occur as accessories.

In lower-grade assemblages the aggregates of pyroxene are replaced by large (c.2.5 mm) crystals of hornblende. The small equant grains of hornblende may also be present. Potash feldspar is more common, generally as orthoclase, more rarely as microcline. Myrmekite is also relatively common. The plagioclase shows little change although a few crystals display marginal zoning to a more sodic variety. It is possible that the average anorthite content is lower than in the higher-grade assemblages as andesine appears to be less common. With the increase of modal hornblende at the expense of pyroxene the iron oxides virtually disappear, suggesting their incorporation into the amphibole. Those that remain invert in part to sphene, which appears as an accessory with apatite and epidote. At the same time quartz increases both in size and modal amount. The lowest-grade assemblages examined in this sequence contain pale green-brown biotite and large (c.6 mm) crystals of blue-green hornblende enclosing small subidiomorphic grains of quartz and plagioclase. The felspar and quartz crystals are generally smaller (c.2 mm) than the mafic minerals. They are similar to those of higher grade. The crystals of quartz and feldspar do not show any good crystal faces and the order of crystallisation is difficult to determine. Only potash feldspar, which tends to be interstitial, may be regarded as a late phase. In an extensive examination of hornblende-biotite and biotite gneisses in North Harris (Soldin, 1978) reports several mineral analyses that show the amphiboles to be in the range from ferro-edenitic hornblende to magnesian hastingsitic hornblende (sensu Leake, 1978). The plagioclase averages Or: Ab: An = 1: 72: 27. Soldin (1978) reports the presence of antiperthites and perthites; microcline he records as rare. Peach and Horne (1930, p.59) remarked on the relative abundance of allanite in the gneisses of Lewis in contrast to those of the north-east Scottish mainland. Beer (1952), however, records that the hornblende-biotite gneisses generally contain no allanite, that mineral being confined to the biotite gneisses, and even there it is much less common than in the intrusive Laxfordian granites (q.v.). Some of the biotite gneisses also contain a relatively large proportion of sphene; it is interesting to note that hornblende-biotite gneisses with similar amounts of TiO2 do not carry sphene, supporting the suggestion made above that at this grade the TiO2 is tied up in amphibole.

The degree of late-stage 'cataclasis' and low-grade recrystallisation has an important influence on the texture and paragenesis of the minerals ((Plate 4)a). The quartz crystals become strained and boundaries become sutured, and eventually the quartz recrystallises to a ribbon fabric or in extreme cases to a mylonite. Similarly, feldspars exhibit strain-induced perthite intergrowths and twinning, and eventual granulation and alteration to white mica and epidote. Myrmekite is common but as with microcline it is unclear if it is related to the effects of cataclasis. Similarly, it is unclear to what extent the occurrence of myrmekite and microcline is related to the late recrystallisation and deformation. One of the most noticeable features of this recrystallisation is the alteration of hornblende to epidote. The presence of small sphene granules in association with the epidote may indicate that it is a byproduct of this reaction. The alteration takes place very easily, and even when the degree of deformation is low amphibole may disappear completely. Where this has happened the epidote may form large (c.1.5 mm) subidiomorphic crystals or the amphibole may be represented by bi-ep-sph-carbonate-chl aggregates. The conversion of hornblende to epidote is independent of the nature of the hornblende, both the dark green and the blue-green varieties altering with equal facility. The alteration of hornblende to biotite, the biotite initially forming as thin slivers on the hornblende cleavage planes, may also be attributable to these late-stage effects since the biotite is seen to form from dark-green or blue-green amphibole, suggesting that it is not part of the 'normal' metamorphic sequence. Biotite itself appears in some thin sections to have altered to an aggregate of potash-feldspar and epidote. Late deformation also induces the exsolution of fine-grained iron oxide from the pyroxene and hornblende crystals, giving them a black dusty look. Iron oxides may also occur along strain induced 'cracks' in the feldspars.

The relationship of the various phases to the metamorphic history is discussed at the end of this chapter.

Geochemistry

The only previous work on the geochemistry of the gneisses of the Outer Hebrides is that by Tarney et al. (1972), Sheraton et al. (1973), Soldin (1976), and Bowes and Hopgood (1969a). Tarney et al. (1972), as part of a regional comparison of north-western Scotland and Greenland, characterised the chemistry of the Outer Hebrides gneisses and found that they showed a regional increase in K and Rb from the gneiss of the northern and eastern Lewis towards the south-west, that is, into the area dominated by intensely recrystallised Scourian migmatite and Late-Laxfordian granites. Soldin (1978), in an extensive study of part of northern Harris, produced analyses of biotite gneiss, hornblende-biotite gneiss and granitic rocks. Bowes and Hopgood's (1969a) work refers only to the island of Mingulay.

In the present survey a further 34 analyses of gneisses from Lewis and Harris have been carried out, and these, along with the results of Tarney et al. (1972) and Soldin (1978), give a coverage of 149 analyses in the northern isles. For comparative purposes a further 5 full analyses and partial analyses were made on representative gneisses from the Uists. The BGS analyses from Lewis and Harris were largely designed to give a representative sample from the west of the area. No attempt was made to collect the specimens on a grid basis.

In the areas of highly migmatised gneiss of West Harris and the Uig Hills, samples were collected to represent both the highly pegmatised 'sugary' gneiss and the remobilised gneiss, as well as rocks representing the least migmatised or recrystallised varieties, that is rocks without pink potassic lits and stringers. In sampling the banded gneisses or gneisses with migmatitic lits an attempt was made to collect the constituent parts in proportion to their appearance on the outcrop. A full list of the analyses is on file at the Edinburgh office of BGS. Average values are presented in (Table 3).

The full results show that the gneisses range from 62 to 75 per cent SiO2 and from 1 to 6 per cent K2O, the average analysis being broadly equivalent in composition, except for higher soda values, to a granodiorite/adamellite with a range from granite to tonalite (Le Maitre, 1976). The gneisses also have relatively high values of La and low Nb and Y compared with Taylor's (1966) averages for granitic and granodioritic rocks. Compared with Taylor's average crust the gneisses are relatively rich in Ba and La and poor in Y and Nb (values from Tarney et al., 1972). Ternary plots (normative Q-ab-or, an-ab-or) are shown on (Figure 23). Although there are regional variations it is obvious from (Table 3) that the rocks of the Outer Hebrides are similar to those of the Laxfordian areas of the mainland and also of Greenland (Tarney et al., 1972) and are significantly different from the Scourian granulites, being notably richer in K, Li and Rb. Soldin (1978) showed that in North Harris the biotite gneisses were on average significantly richer in SiO2, K2O and Rb but poorer in MgO, FeO, MnO, CaO and Zr than the hornblende-biotite gneisses. Soldin (1978) concluded that the differences were fundamental and did not represent any secondary metasomatic or metamorphic effect. The gneisses of the Uists are relatively more basic than those of Lewis and Harris, being richer in FeO, MgO and CaO but poorer in K2O and Rb (see AFM plot, (Table 5). x transitional rocks between basic and metasediment from (Table 5). garnetiferous metasediments from (Table 5), Sheraton et al. (1973), Palmer (1971) and Tarney and others (1972). 0:a average of four Langavat metasediments (Palmer, 1971; (Table 10)); b semipelites from the Langavat group (Table 10). +:D, average of Dalradian slates (Hickman, 1975, table 3); M average of Moine pelites (Butler, 1965, (Table 3)); G average granodiorite (Le Maitre, 1976). average of mainland Scourian gneisses, numbers refer to analyses in (Table 3). average of Outer Hebrides and mainland Laxfordian gneisses, numbers refer to analyses in (Table 3). Dashed lines define general field of metasediments (upper) and quartzofeldspathic gneisses (lower)" data-name="images/P936479.jpg">(Figure 6), p.32).

In order to test whether there are regional variations across Lewis and Harris (as suggested by Tarney et al., 1972) the ratios K/Ba, Ba/Rb, K/Rb, Ca/Y, Ba/Sr and Ca/Sr were selected as probably significant parameters and plotted. The results are presented in (Figure 4). For illustrative purposes arbitrary divisions have been made of low-, medium- and high-range values. It is obvious that the ratios Ba/Rb, K/Rb, Ca/Y, Ba/Sr and Ca/Sr fall from east to west and that the ratio K/Ba rises. It should be stressed that these changes are gradational and relate to an increase in the lighter elements from east to west. To determine if this change reflected a general change in lithological type or possibly a biased collection of rocks, the various ratios were plotted against the weight percent SiO2, the SiO2 content being regarded as a useful reflection of rock lithology. Initially the area was divided into four quarters along easting [NB 1100] and northing [NB 9200], and the regional variation of the elemental ratios against SiO2 examined. The north–south division was found to be not signifiCant, and only the east–west division was retained ((Figure 5)). Despite the fact that the division across the area was an arbitrary straight line it is obvious that there is a regional variation in the ratios for all values of SiO2. The average analyses and ratios for the two areas are given in columns 1, 2 in (Table 3). Examination of (Figure 4) and (Figure 5) also indicates that although there are east–west increases of Ba, K and Rb, the Rb content increases proportionally much more than the other elements. Since the east–west increase in these lighter elements is areally related to an increase in both Scourian migmatisation and Laxfordian remobilisation and recrystallisation, it is tempting to relate the chemical variations to the migmatisation. Within the 'zone of migmatisation' the K2O content may be taken as a general indicator of the degree of migmatisation: it is 5–6 per cent in migmatised rocks as opposed to 1-3 per cent in the adjacent unmigmatised rocks. The latter figure is comparable to that of gneisses outside the zone of migmatisation. In order to examine further the possibility that chemical variations are related to migmatisation, the elemental ratios were compared for gneisses with similar K2O content from within and outside the 'zone of migmatisation'. The results presented in (Table 4) clearly indicate that for all values of K2O the ratios K/Rb, Ba/Sr, Ca/Sr and Ba/Rb all fall, whilst K/Sr though variable on balance rises towards the 'zone of migmatisation'. These results, therefore, suggest that the east–west variation in gneiss chemistry may reflect a zonation unrelated to the degree of migmatisation. Soldin (1978) showed a correlation between modal mineralogy and various elemental ratios, K/Rb and Ba/Rb increasing with increasing hornblende content, whilst Rb/Sr decreased. This would seen to accord with Myers' (1970a, fig. 1) conclusion that biotite gneisses predominate in the east of North Harris. Soldin (1978) states that the difference in mineralogy reflects a fundamental difference in the rock types rather than a superimposed metasomatic effect.

In the Uists the results (column 3, (Table 3)) indicate that the gneisses there are even more depleted in the lighter elements than those of the eastern Lewis and Harris, having higher K/Rb, Ba/Rb, Ca/Sr and Ca/Y ratios and lower K/Sr, K/Ba, Rb/Sr and Ba/Sr ratios. In some respects these rocks have a chemistry intermediate between that of gneisses from Lewis and Harris and that of gneisses from the Scourian block of the Scottish mainland (Table 3).

These results are in accord with the ideas of McQuillin and and Watson (1973) who suggested that the Hebrides possessed a NE-SW lithological grain evidenced by a westward fall in the regional gravity anomaly map. The reasons for this regional variation are matters of considerable speculation. Two possibilities might be considered. Firstly, the general trend in the various elemental ratios in the gneisses from the Uists, through eastern Lewis and Harris to western Lewis and Harris is similar to changes recognised in Coll and Tiree by Drury (1974) and interpreted as due to the amphibolitisation of granulite-facies rocks. In this context it may be significant that the main areas of pyroxene-bearing gneiss in the Outer Hebrides are in eastern South Uist and east Barra (see below). Secondly, Tarney et al. (1972) suggested that mantle degassing during the Archaean might lead to a general upward diffusion of the lighter elements through the crust. Developing these arguments it is possible that the western half of Lewis and Harris may represent a higher crustal level than the east coast, or at least an area where the removal of lighter elements is less advanced. It is interesting to note in this context that Moorbath et al. (1975) state that the gneisses of South Uist, Benbecula and Barra have suffered extreme depletion in uranium relative to those of North Uist although they ascribe the depletion to Scourian metamorphism. The possibility that the differences may simply reflect a change in the original lithologies (see pp.12–14) seems less likely given the gradational nature of the variations.

This discussion is a preliminary statement of the problem. A possible solution must await a detailed investigation beyond the scope of the present survey. Chemical studies, in particular, will have to be conducted with great care in order to distinguish possible metasomatic effects from variation in original lithology, and will also need to take account of the possible preferential susceptibility of some lithologies to recrystallisation.

Metamorphism

As described above, the quartzofeldspathic gneiss shows a variety of mineral assemblages ranging from two-pyroxene granulite facies to the lower amphibolite/greenschist facies. At the present time by far the greater part of the gneisses are the amphibolite facies, the higher-grade rocks being restricted to a few localities mainly in Barra and South Uist. The question arises as to the age and extent of the granulitefacies metamorphism.

Dearnley (1962a) regarded Laxfordian crystallisation as falling into two distinct phases, an early granulite-facies metamorphism and a later retrogressive-amphibolite facies event. However this interpretation now seems unlikely (Coward et al., 1969; Moorbath et al., 1975) and the majority of workers probably regard Laxfordian recrystallisation to have taken place largely in the amphibolite facies (e.g. Myers, 1970a; Taft, 1978), with only very local recrystallisation in the granulite facies (Coward, 1973b; Myers, 1971). Although in some areas amphibolite facies recrystallisation appears to have continued beyond the deformational episodes, in others late deformational events are associated with upper greenschist conditions.

The main development of granulite-facies rocks is believed to be of Scourian age; the extent, however, is very uncertain. In Barra, Francis (1973) suggested that the boundary between his eastern and western gneiss, which roughly coincides with the present boundary between granulite- and amphibolite-facies rocks, was the junction between an infrastructure and a suprastructure. It seems more likely, however, to be the edge of extensive Laxfordian reworking, possibly controlled by pre-existing differences in the metamorphic grade, suggesting that the Scourian granulite facies did not extend much beyond its present outcrop.

The chemistry of the amphibolite-facies gneisses closely corresponds to that in the Laxfordian areas of the mainland, that is, they are relatively rich in lithophiles and thorium ('Carney et al., 1972), which might indicate that the rocks had not been subjected to granulite-facies metamorphism (Heier, 1973). It is worth noting in this context that in their chemistry the Uist gneisses are in part intermediate between the gneisses of Harris and Lewis and those of the Scourian block of the mainland, perhaps indicating a progressive depletion of lighter elements and increase in metamorphic grade southwards towards Barra and South Uist. Moorbath et al. (1975), working on lead isotopes show that the gneisses of South Uist and Barra, unlike those of North Uist, suffered extreme depletion in uranium during the Scourian; this may be taken to indicate more extensive granulite-facies metamorphism in the south with subsequent retrogression.

It is, therefore, concluded that during the Scourian the gneisses locally reached granulite facies particularly in the south and east, although the exact original extent of these high-grade rocks is unknown. The Laxfordian was dominated by amphibolite-facies recrystallisation, the granulite-facies rock that remains presumably resisted amphibolitisation because of the lack of available water.

Chapter 3 Lewisian Complex: metasedimentary and metavolcanic gneisses

Interbanded with the quartzofeldspathic gneiss is a series of rocks with distinctive mineralogies and appearance which have been ascribed to a group of metasedimentary and meta-volcanic gneisses. Many of these gneisses have distinctive chemical compositions (Table 5). This group reaches its greatest and most varied development in the Langavat and Leverburgh belts of Harris (discussed separately below) where graphite schists, marbles, garnet-kyanite gneisses and probable metavolcanic gneisses are well seen (Map 5). Metasediments within the Outer Hebrides, however, are more usually represented by rusty-weathering acid gneisses which tend to be more fine grained and finely banded than the quartzofeldspathic 'grey' gneiss. Characteristically they contain an abundance of biotite defining a good schistosity, and a high proportion of garnet. With the exception of some associated fine-grained basic and metavolcanic horizons (notably in the Langavat Belt of South Harris), amphibole, if present at all, occurs only in subordinate amounts to the other mafic minerals.

Paragneisses in the Hebrides were first noted by MacCulloch (1819), who recognised marbles in the Rodel area of South Harris. Subsequently Heddle (1878) recorded their mineralogy in some detail. Jehu and Craig (1927) defined the two major zones of paragneiss, the Leverburgh and Langavat belts flanking the South Harris Igneous Complex, and also recognised the high metamorphic grade of the impure marbles (diopside-forsterite) and pelitic gneisses (garnet-kyanite) of the Leverburgh Belt. They also gave a detailed description of the Langavat Belt rocks along the Bagh Steinigie [NG 019 938] to Borve [NG 035 950] coast section, noting the gradation into granitic gneiss at their north-east margin. Most subsequent workers studying the South Harris metasediments have added further details or differing interpretations. Dickinson (1974), in dealing with the metamorphism of the Leverburgh Belt paragneisses, is the only worker to concentrate solely on the metasediments.

The extent of the main metasedimentary belts is given by Coward et al. (1969), who reviewed their distribution throughout the Long Island. Further occurrences have been reported in Grimsay and eastern North Uist by Sibson (1977b) and in the Carloway area by Lisle (1974). During the present survey many new localities have been recorded mainly in the southern isles, notably in central and eastern Benbecula and in a broad belt across Uist from Baleshare to Lochmaddy. Isolated outcrops along the north coast of North Uist have also been found, linking previously recorded localities on the north-west coast of North Uist with those on the islands of Berneray and Pabbay in the Sound of Harris (Graham, 1970; Coward and others, 1969).

Although in South Harris the metasediments can be traced along strike for several kilometres, elsewhere they commonly occur as lenses a few metres long. Trails of such lenses may be locally followed along the gneiss foliation. On a larger scale, belts defined by an abundance of metasediment occurrences within the grey gneiss may be traced for considerable distances. One example, which seldom exceeds 0.5 km in width, can be followed for 20 km from west of Ben Tarbert in South Uist to the north coast of Benbecula. The metasediments characteristically show sharp contacts with the quartzofeldspathic gneisses across the strike but along the strike they have more gradational boundaries. These relationships and other criteria led Coward et al. (1969) to suggest that although most metasediments are non-migmatised they predate the Scourian deformation and migmatisation. Coward et al. cited an example of a Scourie Dyke cutting metasediments north of the Loch Skipport road at [NF 814 380]. They were also impressed by the close areal association of metasedimentary gneisses with large basic bodies, and suggested that the basic rocks had in some way preserved the distinctive textures and mineralogies of the metasediments (e.g. by hornfelsing). We have been unable in the present work, however, to substantiate this point. These relationships are further discussed in Chapter 4: Banded Basics of the Uists and Benbecula: Field aspects.

Lithology

The metasedimentary assemblage contains a variety of lithological types. Coward et al. (1969, table 1) make a useful distinction between metasediments sensu stricto and metasediments sensu lato. The first group contains rocks whose sedimentary origin is implied by their mineralogy or composition, and the second group contains rocks whose sedimentary origin is suggested by unusual textures seen in the field and/or by association with more distinctively sedimentary types. Most recorded occurrences of metasediments s.s. are of garnetiferous pelitic and semipelitic gneisses with subordinate amounts of psammites, quartzites, marbles and calcsilicates. Metasediments s.l. form parts of the Langavat Belt and the Ness assemblage in northern Lewis (Watson, 1969). These include flaggy quartzofeldspathic and striped hornblendic gneisses, and fine-grained basics, both of which may be volcanogenic. Texturally the metasediments s.l. grade into the ordinary quartzofeldspathic gneiss and it is difficult to estimate what proportion of these 'transitional' gneisses may belong to the metasedimentary assemblage. For example in North Uist, the belts of abundant metasediments s.s. contain a number of metasediments s.l. defined on their finer texture relative to the quartzofeldspathic gneisses. In addition there is a group of rocks of unusual composition, namely basic quartzofeldspathic gneiss with a high proportion of amphibole, and quartzitic gneiss with a relatively low proportion of mafics.

The occurrence of these rocks within and immediately adjacent to the belts of metasediments s.s. is sufficiently striking to define an association, and their general proportions are high enough to allow the whole zone to be separated on the main (1:100 000) map. The true proportion of metasediments within the 'grey gneiss' is masked by their extensive metamorphic history.

Langavat Belt

The metasediments of the Langavat Belt, lying on the northeast flank of the South Harris Igneous Complex, form a zone 0.8 to 1.4 km wide stretching from Loch Finsbay to Borve and Taransay. Elongate xenoliths of metasediments also occur in the adjacent diorite-tonalite at several localities, namely, in the Bagh Steinigie–Bleaval area (3.5 km long by 80 m wide), adjacent to Loch Langavat (500 m by 80 m), and in the Borsham area to the south-east (1.25 km by 70 m). The metasediments consist of regularly thin banded, fine-grained biotitic and hornblendic acid gneisses with pelite, quartzite and rare marble bands.

There has been little agreement between workers as to the distribution of lithologies within the Langavat Belt. Although Dearnley (1963) stated that the lithological sequence is repeated symmetrically by a near-isoclinal major fold, later workers (Myers, 1968; Palmer, 1971) have shown that there is no such repetition and that the minor folds of any one generation are generally consistent in vergence across the belt, thus implying that no major fold structure exists.

The subdivision into lithological units by Dearnley (1963) and subsequent divisions by Myers (1968) and Palmer (1971) are not used here in view of apparent lateral changes along the Langavat Belt, the gradational contacts of many of the units, and the general lack of agreement between different authors as to the lithology and distribution. The current survey has shown that the contact between metasediment units and the diorite of the South Harris Igneous Complex is at a slight angle in such a way that the quartzite-rich units that crop out in the Loch Eachkavat–Loch Langavat area are progressively cut out to the south. The garnetiferous biotitic metasediments that are well exposed just east of the summit of Sletteval [NG 052 917] may be traced intermittently south-south-eastward to Finsbay (see Map 5). The amphibolites within the belt apparently represent two modes of origin (see p.51). First, a major thick (c.150 m) amphibolite contains flattened feldspar aggregates in its central part, with locally (0.7 km west of Finsbay) marginal zones, 20–30 m wide, composed of laminated fine-grained amphibolite. The textures in the central portions are similar to those seen in some Younger Basics and are interpreted as metadoleritic or metagabbroic features. Second, in contrast, more finely banded amphibolites found notably in the more north-easterly parts of the belt consist dominantly of hornblende, oligoclase and quartz, interbanded with biotite acid-gneiss bands richer in quartz. These are interpreted as metasediments and/or metavolcanics.

The north-eastern boundary of the Langavat Belt is not clearly defined. On the coast section north of Borve Lodge [NG 0342 9495] there is a gradation north-eastwards from the regularly fine-banded amphibolite-rich gneisses of the Langavat belt into thin-banded acid gneisses with only minor amphibolites (locally discordant) over about 150 m. This change also corresponds with a marked increase in migmatisation, dominantly of Scourian age (Myers, personal communication), and with a Laxfordian overprint associated with the emplacement of the granite sheet complex. A similar gradational contact occurs in the Ardvey Peninsula immediately north-east of Finsbay Post Office, and on the Quidnish Peninsula. There finely interbanded hornblende schists and quartzofeldspathic gneisses, in parts down to millimetre scale, are folded into tight to isoclinal minor fold structures on Quidnish Peninsula at [NG 0954 8693]. However, both here and on the adjacent Ardvey Peninsula locally discordant amphibolite–gneiss contacts are present. The high degree of Scourian and Laxfordian deformation makes it difficult to assess the original nature of these units. They may represent sheeted basic dykes and/or metavolcanics. Pegmatite development in the acid gneisses was noted adjacent to basic bodies on both the Ardvey and Quidnish Peninsulas. The amphibolites which crop out on the south-east parts of these peninsulas show extensive alteration to epidote, actinolite and chlorite.

The coast section from Borve to Traigh Mhor and Bagh Steinigie provides a generally well-exposed cross-strike section across most of the Langavat Belt metasediments. It has been mapped and described in detail by several authors (e.g. Jehu and Craig, 1927, pp.476–478; Myers, 1968) and only a summary will be presented here.

The most north-easterly units are laminated and striped amphibolites and quartzose acid gneisses (about 340 m thick) which grade south-westwards into laminated amphibolites (25 m thick) locally with epidote, biotite and pyrite-rich bands. Dearnley (1963) records the mineralogy of such amphibolites as strained quartz, oligoclase (Ann), hornblende, minor biotite and accessory sphene. This is followed by 20 m of thin-banded acid gneisses characterised by the abundance of violet-tinged biotite, minor thin amphibolites, a marble band and abundant pyrite-rich bands. These beds crop out on the north-east side of Sta Bay [NF 0291 9496]. Many of these units may represent original metavolcanic horizons, particularly the regularly finely banded amphibolite/acid gneiss lithologies. On the south-west side of the bay is a thin band of marble adjacent to which lies 35 m of finely-banded amphibolite with diopside lenses up to 30 cm across containing abundant pyrite and epidote. The remaining 450 m consists of striped and thin-banded acid gneisses with amphibolites north-east of Traigh Mhor, and these are notable for the abundant tight to isoclinal minor folding and for the occurrence of several large pods of ultrabasic rocks (up to 55 m thick). These pods lie in a narrow zone which stretches south-east from Rubha Sgeir nan Sgarbh through Dun Borve and Loch na h-Uamha to Loch Meurach, 1.5 km north-west of Finsbay Post Office (see Chapter 4: Older Ultrabasic Complexes, Langavat Belt). Thin ultramafic bands are common within the metasediments adjacent to the large pods.

Additional calcsilicate lenses (up to 5 m wide and 15 m long) were recorded by Dearnley (1963) about 1 km southeast of Sta Bay where large diopsides are conspicuous in white banded rocks. Garnet- and epidote-bearing pelitic schists are associated with the marbles, particularly in the Bagh Steinigie inlier.

At Bagh Steinigie [NG 019 939] about 60–70 m of tightly folded fine-grained laminated amphibolites with subsidiary marbles, pelites and calcsilicate bands are well exposed, lying within the diorite-tonalite body. Of the four marble bands exposed the largest is 2.5 m thick and contains forsterite, commonly pseudomorphed by serpentine, clinohumite and diopside. Inland, exposures are sparse but high-grade pelitic rocks were noted by Myers (1968), usually within amphibolites. A similar association occurs at Borsham [NG 091 859] where an elongate xenolith occurs again in the diorite near to its north-eastern margin. Tremolite-garnet- and epidote-bearing acid gneisses are interbanded with fine-grained amphibolites and quartzose gneisses (see Dearnley, 1963).

The Bagh Steinigie assemblages preserve higher-grade (upper amphibolite) minerals than the assemblages in the more highly deformed metasediments of the main Langavat Belt.

At Aird Vanish on Taransay the Langavat metasediments are represented by garnetiferous semi pelitic schist which lies within acid gneisses typically containing the assemblage qz-pl-ep-(bi)-(hbl) (Myers, 1968, p.102). Myers attributed the nature of these acid gneisses to early gneissification (?Scourian) of the original metasediments and cited the presence of well-defined compositional banding as being a relict feature of their original metasedimentary origin.

Leverburgh Belt

This zone of metasediments, 1 to 1.7 km wide, lying on the southern flank of the South Harris Igneous Complex, stretches north-west from Vallay Island and Rodel through Leverburgh to Toe Head. A thinner zone extending from just north of Lingarabay [NG 070 850] west-north-west through Beinn Tharsuinn joins the main outcrop just northeast of Loch Steisevat (see Map 5). These paragneisses were metamorphosed under granulite-facies conditions but their original lithology is still very apparent. Garnet-bearing quartzose gneisses are dominant but marbles, graphiteschists, pelites and quartzites are all represented. In outcrop many of the rocks appear to be monotonous garnetiferous acid gneisses, but detailed petrography, notably by Dickinson (1974) and Dearnley (1963), has shown that the proportions and types of feldspars, aluminosilicates and ferromagnesian minerals are all very variable.

Dearnley (1963) subdivided the metasediments into three units; the Ben Obbe Quartzites, the Rodel Series and the Chaipaval Pelitic Series. In the Rodel–Leverburgh area both Dearnley (1963) and Witty (1975) envisage the Rodel Series as lying in the centre of a complex tight synform complementary to the Roineabhal Antiform to the north-east. However, to the north-west the Chaipaval Pelitic Series becomes dominant and north-west of Northton it is the sole representative of the metasediments. In view of the difficulties of correlation along the strike of the Leverburgh Belt and the gradational boundaries between the various rock types, only the lithologies are represented on Map 5; formal stratigraphic terms have not been used. Dickinson (1974, p.29) notes the lenticular nature of the various rock types on all scales and Dearnley (1963) enumerates in considerable detail several possible structural and stratigraphical complexities to account for the observed field relationships.

The metasediments have abundant calcsilicate and marble bands and these lie within two linear zones; one stretching north-west from St Clement's Church, Rodel, along Glen Rodel and Shranndabhal to Leverburgh and containing a series of forsterite-marble lenses up to 15 m long and 8 m thick; the second, narrower, zone extending from the Stuaidh peninsula, where Davidson (1943) records the marble as 49 m thick, to the western flank of Beinn an Toib (Ben Obbe). Graphite is a commonly recorded accessory in the calcsilicates and marbles, and both these zones also contain graphite-rich pelitic schists which are best seen in a small quarry 55 m east of St Clement's Church, on the Stuaidh Peninsula (S63101) at [NG 0424 8314] and just north-west of the Finsbay-Rodel road, 1 km north-east of Rodel. The graphite occurs in abundant rods up to 0.55 mm long, commonly together with pyrite and chalcopyrite. Within the Chaipaval Pelitic Series of Dearnley (1962b) calcsilicate lenses up to 3 cm wide and 1 m long with the assemblage dior-hbl-ap-qz were recorded by Dearnley in scapolite-bearing garnet-biotite paragneisses at the north-west end of the Chaipaval ridge [NF 961 938]. He suggested that the lenses represented thin impure dolomitic intercalations. Dearnley's analysis showed the rock to contain 2.11 per cent P2O5.

The garnet-quartz paragneisses are particularly well exposed in the 0.5 to 1 km-wide north-eastern marginal zone of the Leverburgh Belt around Sgriosan [NG 035 858] and north-east of Glen Coishletter. Basic dykes, pods and lenses, formerly gabbro and dolerite but now largely amphibolite are typically interbanded with the paragneisses. Garnet-hornblende-, garnet-biotite- and hypersthene-bearing quartz-rich gneisses are all found around Sgriosan (Dickinson, 1974).

In the Northton–Toe Head area perthitic microcline is a major component in massive, pale pink to fawn, acid gneisses (Dickinson, 1974, pp.44–48). These gneisses contain garnet-kyanite-biotite lenses up to 20 cm long in a quartz-perthite matrix which itself contains kyanites up to 2 cm long and subsidiary garnets up to 1 cm across.

The pelitic gneisses are best developed in the Rodel area and more extensively in the Northton–Toe Head region. They range from quartz- or feldspar-rich acid gneisses with kyanite/biotite and/or garnet, through to rocks in which kyanite attains 10 per cent of the mode. In the Chaipaval area certain bands crop out as prominent massive, pale blue-grey and red mottled gneisses. Over much of the area pelitic gneisses have a field appearance similar to the garnet-quartz gneisses. However, mauve-red garnets about 5 mm across and blades of bluish kyanite up to 7 mm long are abundant. Biotite is abundant only in the garnet-poor varieties. Sillimanite occurs as trails of tiny needles nucleated on biotite around plagioclase, microcline and quartz aggregates or marginal to garnet. On Bideinan garnetite (more than 75 per cent garnet) nodules with minor diopside, green hornblende, quartz and rutile (altered to sphene) occur in kyanite-gneisses. Dearnley (1963) has also recorded the assemblages gt-cum-mt and gt-mt-bi with minor quartz from iron-rich metasediments near Northton.

Cordierite-bearing gneisses are uncommon but have been described by Davidson (1943) from the areas adjacent to Rodel crofts and Rodel Hotel, and by Dearnley (1959) from the Stuaidh Peninsula. The cordierite is only rarely fresh, generally being altered to fine aggregates of yellowish pinite.

As with the garnet-quartz gneisses adjacent to the main gabbro sheets, pelitic gneisses adjacent to the norite become strongly schistose and the acid gneisses mylonitic, particularly in the Beinn an Toib–Carminish area. Dickinson (1974) notes that in these rocks perthite is converted to microcline, garnet to biotite, and sillimanite is abundant. Palmer (1971) also records plagioclase replacement by microcline and presents chemical analyses of sheared and unsheared pelites (see the description of the geochemistry at the end of this chapter).

The metasediments of the Lingarabay-Beinn Tharsuinn strip are dominantly garnetiferous quartzose gneisses similar to those adjacent to the south-west side of the anorthosite-gabbro complex. Dearnley (1963) describes the occurrence of the assemblage qz-pl(olig/and)-gt-sil-bi-spinel from this zone. The rocks are particularly highly deformed and Witty (1975, p.186) envisages the zone as forming the core of an isoclinal synform.

Sound of Harris metasediments

This zone was included in the general account of metasediments given by Coward et al. (1969). Quartzites and garnet-bearing semipelitic gneisses were recorded near Alarip Bay on the north coast of Pabbay by Graham (1970, p.54), where they lie adjacent to a large basic body. In fact a narrow strip of semipelitic gneiss surrounds this basic body (cf. South Harris), suggesting that the metasediments may have been preserved by hornfelsing. The steeply dipping rocks of central south Pabbay, north-east Berneray and on much of Killegray are thin-banded, fine-grained quartzbiotite and hornblendic acid gneisses similar to those of the Langavat Belt, according to Graham (1970, p.57).

There is no agreement between authors as to the original nature of most of the acid gneisses on Ensay. Dearnley (1963) and Witty (1975) assume that all the rocks are meta-sediments, Witty (1975) correlating them with Langavat Belt rocks. Jehu and Craig (1927) and Graham (1970), who studied the island in detail, recognise only minor zones of metasediments within dominantly acid gneisses. This distinction is important when the regional structure of the area is considered (Chapter 14: The Sound of Harris and southern Harris).

The gneisses on Ensay are typified by the assemblage qz-pl(olig)-bi-ep-hbl but range to more quartz-rich varieties. Finely banded hornblendic acid gneisses similar to those found on the north-east side of the Langavat Belt are also common.

North Rona metasediments

In North Rona Nesbitt (1961) records garnet-sillimanite-bearing rocks in a 'transition zone' between banded amphibolite and acid gneiss (cf. Dougal, 1928). These are represented on the 1:100 000 lithology map (Map 1) as metasediments but they may be more closely allied to the 'hybrid' rocks seen at Loch Bee in South Uist (see Chapter 4: Banded Basics of the Uists and Benbecula: Petrography), a suggestion supported by the analysis given by Bowes (in Nesbitt, 1961).

Petrography

Areas other than South Harris

The fine-grained biotite-bearing metasediments include a wide range of compositions. Typical assemblages encountered in the present investigation are:

qz-pl qz-pl-cum-bi
qz-pl-bi qz-pl-hbl-cum-gt-bi
qz-pl-bi-ep qz-pl-ant-bi
qz-pl-bi-gt qz-pl-bi-ep-ky-kf
qz-pl-hbl-bi-gt qz-pl-bi-sil-gt-(kf)-(ms)

Accessories include opaque ores, sphene, apatite, epidote, zircon and allanite. In addition to these mineralogies Coward and others (1969) report orthopyroxene from metasediments in the Loch Skipport area of South Uist, and staurolite from the Loch Bee metasediments. Calcareous rocks seem to be confined to the Langavat and Leverburgh belts.

The rocks generally show a fine-grained banding, and the various assemblages are interbanded on a centimetre scale. Textures are generally schistose, although quartz and feldspar grains may define an equigranular matrix. Alteration and cataclasis due to late-stage deformation are locally present. The plagioclase is mainly andesine although locally oligoclase and albite occur. They are generally fresh but may show some alteration and secondary twinning. Potash feldspar (mainly as microcline) is largely confined to rocks with aluminosilicate polymorphs, suggesting a relatively less basic composition. Locally it may revert to muscovite. Biotite is generally 'foxy' red-brown. Garnet is poikilitic and commonly heavily corroded, being rimmed by plagioclase. Coward et al. (1969, p.394) suggest that the pink garnet from biotite-bearing pelitic gneiss is close to pure almandine with a small content of manganese. A distinctive feature is the common occurrence of amphiboles of the cummingtonite-anthophyllite series. Hornblende and cummingtonite may occur together. Anthophyllite generally occurs as euhedral crystals in pelites, quartzites and quartzofeldspathic gneisses. Epidote also occurs in well-shaped crystals but it is unclear if it is 'primary' or secondary. Chlorite and secondary biotite may be present as alteration products. One unusual assemblage encountered during the present investigation in Benbecula includes a quartz-feldspar rock with sulphide-rich bands perhaps equivalent to the pyrite-rich rock reported by Coward et al. (1969, table II) from the north-west coast of North Uist. Other exotic rock types apparently transitional between metasediments and basics will be described in Chapter 4: Banded Basics of the Uists and Benbecula: Petrography. The mineral assemblages indicate an upper-amphibolite/granulite facies grade of metamorphism. It is difficult to relate mineral growth to any specific deformation phase or sequence of crystallisation, although Coward et al. (1969) suggest that the main crystallisation occurred during their F2 (dL2 of this memoir) event.

Langavat Belt

The mineral assemblages of the Langavat Belt meta-sediments are summarised in (Table 6). The most common accessories are sphene and apatite. Textures within the metasediments are generally even grained and granular, being largely produced by recrystallisation under amphibolitefacies conditions. Average grain size is less than in the Leverburgh Belt rocks, although feldspar porphyroclasts and other relict minerals are preserved even in the most strongly deformed parts of the belt.

A typical laminated biotite-rich metasediment from the north-east side of Sta Bay ((S63072), [NG 0294 9496]), consists of elongate aggregates of quartz, oligoclase, foxy-brown biotite and subsidiary potash feldspar. Grain size varies with individual bands from 0.2 to 1 mm, averaging about 0.3 mm. Strain shadows are common particularly in the larger quartz grains and pyrite is notably abundant in certain bands. Ilmenite locally overgrows rutile. A nearby laminated amphibolite ((S63073), [NG 0288 9494]) is composed of aggregates of elongate green hornblende, plagioclase, quartz and relict clinopyroxene, with an average grain size of 0.2 mm (rarely grains reach 1.2 mm). Such units may represent original basic volcanic horizons but the degree of deformation and recrystallisation masks earlier features. Coarser-grained amphibolite bands are also present, and these contain more equant grains with triple junctions commonly at about 120°. Zoned plagioclase is present with grain compositions ranging from high andesine to oligoclase. Minor amounts of late-stage calcite are present.

Myers (1968) has recorded the presence of grunerite/cummingtonite and anthophyllite in Langavat Belt rocks. He also describes staurolite enclosing earlier fabrics defined by iron oxides and quartz, in turn enclosed by garnet. The garnets locally enclose curved inclusion trails which are continuous with the matrix foliation but some are also elongated parallel to the strong lineation. Subsequently garnet broke down to biotite and hornblende at the same time as cordierite and fibrolite grew (Coward et al., 1969). These minerals are typical of upper amphibolite grade rocks and also occur in metasediments in North Uist and Benbecula.

Near the south-west end of Loch Langavat at [NG 0504 8844], and within 30 m of the highly sheared diorite contact, pods and lenticular bands of hornblendic material up to 35 cm wide occur in flaggy quartzofeldspathic gneiss with thin diorite sills. Although in the field they look similar to calcsilicates, in thin section (S64543) the assemblage opxcpx-green hbl-bi-sph-pl-cc is found. This is most probably a hybridised metagabbro-metasediment assemblage and suggests that locally granulite-facies conditions were attained in parts of the Langavat Belt.

Leverburgh Belt

The mineral assemblages found in the various Leverburgh Belt lithologies are taken mainly from the work of Dickinson (1974) and given in (Table 7).

Thin-section examination of the impure marbles and associated calcsilicates show the calcite to be generally clouded (average grain size is 0.5 mm) in contrast to the dolomite (grains up to 5 mm in diameter), which is clear and invariably twinned (Dickinson, 1974). Forsterite (Fo97) commonly forms small grains 0.5 to 4 mm across, and diopside grains up to 5 mm in diameter are abundant. The phlogopite (1 to 2 mm long) may be bent or kinked. Colourless spinel is common. In the calcsilicates diopside and pargasite are abundant and crystals may reach 3 cm in length. Thin calcsilicate lenses occur in Glen Coishletter, where Dearnley (1963) records the assemblage hbl-pl-cc-scp-chl in a rock containing magnetite and pyrite lenses.

In thin section the garnet-quartz paragneisses show a very variable grain size (0.5 to 5 mm) with striping defined by such variations. Towards the margins of the Leverburgh Belt quartz becomes more sutured and shows abundant undulatory extinction and subgrain development. Dearnley (1963) has recorded instances where lenticles of deformed quartz-feldspar mosaics are surrounded by sinuous highly elongate grains. Compositional lamellae in the gneisses are defined by variable amounts of minor fine-grained biotite. Plagioclase (up to 1 mm across) and more rarely microcline are notable constituents in some thin sections and myrmekite and antiperthite are common minor constituents. Pale lilac (pink in thin section) garnets commonly up to about 6 mm across are abundant in many quartzose gneisses. Minor green hornblende and/or muscovite may also be present. Hypersthene is uncommon but was noted around Sgriosan by Dickinson (1974), where it occurs typically in grains 1 to 4 mm across that show schiller effects. It is generally retrograded to a spongy network of biotite-quartz-magnetite. Scapolite is common in the garnet-hornblende gneiss but not in garnet-biotite gneiss. It forms grains up to 2 mm across in quartz-plagioclase aggregates.

Where the quartzose gneisses lie adjacent to the gabbros they are more highly foliated, with a reduced grain size and commonly with ribbon quartz textures. Also, Dickinson (1974) records that near these contacts garnet gneisses are replaced by hornblende-biotite gneisses.

In thin section pelitic gneiss from the Chaipaval peninsula ((S63087) [NF 9812 9124]) shows kyanite plates (3 to 7 mm long) which are generally bent, kinked and in parts twinned. Small grains of kyanite are also present. Garnets show marginal embayment by quartz and are commonly zoned. They may also be shattered and partially replaced by biotite or kyanite. Dickinson (1974) analysed garnets (Table 8) from pelitic gneiss from Traigh na 'h Uidhe near Northton [NF 983 905], and showed them to be notably different from those in the meta-igneous rocks. They are particularly iron-rich but because no whole-rock analyses are available to define the bulk chemistry it is difficult to assess whether or not these garnets are typical of the metasediments. In contrast Dearnley (1963) used optical methods to estimate garnet compositions and found them to lie within the range of those from the gabbros (ND = 1.785-1.795). Garson (1971) also recorded garnet from kyanite-bearing pelitic gneiss as have ND = 1.783 and a unit cell of 11.53Å, indicating a considerable pyrope component. Biotite shows pale yellow to deep foxy-brown pleochroism and forms sheaves defining the foliation; it also occurs as irregular laths, more randomly oriented. Quartz occurs as small grains (less than 2 mm) in aggregates associated with twinned plagioclase; both minerals show evidence of fine-grained recrystallisation. Sillimanite is uncommon in the Rodel pelites but widespread in the Toe Head–Northton area. It invariably forms matted trails of tiny acicular crystals commonly nucleated on biotite. In the Kyles House area sillimanite is very abundant, forming tabular plates, and even overgrowing garnet ((S63095), [NF 9948 8849]). Perthitic microcline is a minor component of pelites in the Rodel–Leverburgh zone where it is rimmed by microcline. Muscovite is an important constituent of the pelites in the Bideinan–Aird an't Sruith area. It replaces sillimanite, itself a late-developed mineral in these rocks, and overgrows biotite and more rarely kyanite and quartz-feldspar aggregates. Its presence coincides with zones of coarser grain-size (the average diameter of quartz grains is 1.5 mm). Late-Laxfordian pegmatite pods are also found near Kyles House, indicating that this area has apparently been strongly affected by late-stage potash metasomatism.

In the perthite-rich gneisses of the Northton Toe Head area the perthites are generally cracked and turbid, averaging 5 mm across but commonly up to 1 cm. They invariably show marginal recrystallisation to microcline. Sillimanite is also locally present.

Geochemistry

Many analyses of metasediments are available for the Langavat and Leverburgh belts, although few have been published. Geochemical data on the more widespread biotite-rich acid gneiss are however sparse. In this memoir a full analysis of a pelitic gneiss from Rarnish, Benbecula, and a partial analysis from Rueval, also on Benbecula, are presented (Table 5). The results are similar to the values given by Tarney and others (1972) and Palmer (1971).

Skinner (1970) presented major and trace-element data for twenty-three metasediments from the Leverburgh Belt and four from the Langavat Belt (see Palmer, 1971, pp.200 - 204). Witty (1975) gave eight additional analyses from quartzose psammites (locally garnetiferous) of the Leverburgh Belt but these were taken from xenoliths in, and beds adjacent to, the metagabbros. In the course of a study of rare earth element (REE) content of Moine and Lewisian metasediments (Mendum and Plant, personal communication) six samples of Langavat and Leverburgh meta-sediments were analysed (for trace elements and REES). Representative average analyses from these data are included in (Table 9) and (Table 10). In (Table 9) the consistency of the analyses is perhaps surprising considering the variety of rock types sampled, with the average value (Sheraton et al., 1973) being very similar to that for pelitic gneisses given by Palmer (1971). Although varied rock types are seen in the field it appears from the geochemistry that the metamorphism has exaggerated small compositional differences in the original sediments. Differences in Mg/Fe and Fe3+ /Fe2+ ratios, and the presence of excess K2O may all play a part in determining the final mineral assemblage.

Metasediments from the Leverburgh Belt show significant depletion in K, Rb, Cs, Th and U, and enrichment in Ca, Sr, Ba and Zr. In general, alkalis are low and CaO + MgO values high. K/Rb ratios average 435 for 21 metasediments whereas for typical Dalradian shales/psammites the ratio lies in the range 200-250 (Hickman, 1975). These features are compatible with granulite-facies metamorphism of an original shale-sandstone sequence (see Heier, 1973). Mica breakdown caused either by increased pressure-temperature conditions and/or a change in the fluid composition (H2O to CO2) results in release of K and Rb into the fluid phase.

Palmer (1971) noted that the pelites show high total Fe, Mg, SiO2 and Al2O3 contents, and are similar to published shale and Dalradian slate average values. When plotted on an AFM diagram (Table 5). x transitional rocks between basic and metasediment from (Table 5). garnetiferous metasediments from (Table 5), Sheraton et al. (1973), Palmer (1971) and Tarney and others (1972). 0:a average of four Langavat metasediments (Palmer, 1971; (Table 10)); b semipelites from the Langavat group (Table 10). +:D, average of Dalradian slates (Hickman, 1975, table 3); M average of Moine pelites (Butler, 1965, (Table 3)); G average granodiorite (Le Maitre, 1976). average of mainland Scourian gneisses, numbers refer to analyses in (Table 3). average of Outer Hebrides and mainland Laxfordian gneisses, numbers refer to analyses in (Table 3). Dashed lines define general field of metasediments (upper) and quartzofeldspathic gneisses (lower)" data-name="images/P936479.jpg">(Figure 6) along with values for garnetiferous metasediments from the Lewisian of the Scottish mainland and Greenland (Tarney et al., 1972), all values are seen to correlate. These concentrations are separable from the average quartzofeldspathic gneiss values. The rocks are confirmed as metasediments, their SiO2 content being too high relative to the Fe and Mg content for 'normal' igneous parentage. A CaO/Y plot (Table 3). x metasediments from analyses 1 and 2, (Table 5); and from Tarney and others (1972, table 1, analyses 5, 8, 11, 12 and table 2, analysis 9)." data-name="images/P936480.jpg">(Figure 7) also serves to discriminate between metasediments and quartzofeldspathic gneisses. These results relate mainly to pelitic and semipelitic gneisses and for the psammitic units other discriminants may be needed (e.g. TiO2/SiO2 plots–see Tarney, 1976). It does appear that many of the pelitic rocks have remained relatively closed systems during their extensive history of deformation and metamorphism.

Palmer (1971), however, used Skinner's (1970) data to show that the sheared pelites of the Leverburgh belt, typically adjacent to the meta-igneous contacts, have a markedly changed geochemistry and mineralogy (see column 4, (Table 9)). The recorded increase in K2O and Rb and decrease in MgO and FeO correlates well with the observed replacement of garnet by biotite, and plagioclase by microclinc.

Only four complete major oxide analyses are available from the Langavat Belt metasediments (Palmer, 1971, pp.203–204) and their average value is close to the average gneiss composition (Table 5). x transitional rocks between basic and metasediment from (Table 5). garnetiferous metasediments from (Table 5), Sheraton et al. (1973), Palmer (1971) and Tarney and others (1972). 0:a average of four Langavat metasediments (Palmer, 1971; (Table 10)); b semipelites from the Langavat group (Table 10). +:D, average of Dalradian slates (Hickman, 1975, table 3); M average of Moine pelites (Butler, 1965, (Table 3)); G average granodiorite (Le Maitre, 1976). average of mainland Scourian gneisses, numbers refer to analyses in (Table 3). average of Outer Hebrides and mainland Laxfordian gneisses, numbers refer to analyses in (Table 3). Dashed lines define general field of metasediments (upper) and quartzofeldspathic gneisses (lower)" data-name="images/P936479.jpg">(Figure 6). The average K/Rb ratio is 167, considerably lower than that of Leverburgh Belt rocks; this probably reflects the multiple amphibolite-grade metamorphic effects within these rocks. Both Myers (1968) and Palmer (1971) have argued that the 'grey' gneiss of the Hebrides is derived from such metasediments, but as the pelites of the Leverburgh Belt have remained relatively closed systems with respect to major elements, it is possible that the limited number of analyses merely reflects variable metasediment compositions. The problem of gneiss formation is discussed below (Chapter 5) and we conclude that they are largely of igneous origin.

Rare earth element (REE) data show a spread from a 'normal' metasediment Lewisian Gneiss curve with enriched light REE (LREE) values and low values of heavy REE (HREE), down to LREE depleted specimens (Figure 8). This suggets that no partial melting has occurred in the Leverburgh Belt paragneisses despite their high metamorphic grade. Langavat Belt rocks show small but significant positive Eu anomalies, and the notably quartzose psammite collected near to the diorite contact [NG 0404 9179] shows very marked LREE depletion. This may be due either to its highly quartzose nature or to localised partial melting near to the diorite. Subsequent amphibolite-grade metamorphism and/ or shearing apparently failed to restore original LREE levels.

No analyses of the metavolcanic rocks are currently available.

Chapter 4 Lewisian Complex: older basics and ultrabasics

This chapter covers all basic and ultrabasic rocks which have suffered some degree of Scourian migmatisation and/or deformation. Much has been written about the range of basic and ultrabasic gneisses in the Scourian of the Scottish mainland (Peach et al., 1907; O'Hara, 1961b, 1965; Bowes et al., 1964, etc). Within the gneisses of the Hebrides, however, the basic and ultrabasic fractions occur for the greater part as rather diffuse lenses and discontinuous sheets. The following groups have been identified:

Early Basics of Barra

Banded Basics of Uists and Benbecula

Banded Basics of the Northern isles

Older Ultrabasic complexes.

Early basics of Barra

On the east coast of Barra, around Leanish and Earsary, Laxfordian deformation is particularly low. This allows a number of pre-Younger intrusive phases to be recognised, some of which have been described by Hopgood (1971) and Francis (1969). Francis (1969) recognised an early/ultrabasic suite, an 'early dyke suite' and a suite of granites and pegmatites. The last two groups are discussed in Chapter 6. The basic/ultrabasic rocks form a series of discontinuous outcrops. They are locally pyroxene-bearing, and Francis (1969, p.24) describes them as having sharp contacts with the gneiss, possibly cross-cutting the foliation. The bodies post-date the Scourian gneiss-forming period but are veined by pegmatitic material and have been subjected to part of the Scourian deformation. K-Ar dates on hornblendes from those bodies have given an age of 2585 Ma (Francis et al., 1971).

Banded basics of the Uists and Benbecula

Field aspects

Within the southern isles there are various older basic rocks which characteristically exhibit a mineralogical layering or banding. The large basic bodies that form a belt of small hills across the northern part of North Uist locally exhibit slight layering, but they are believed to be Younger Basics and are discussed in Chapter 10. There are also other basic bodies of uncertain age and affinities that exhibit a crude layering.

One group of older banded basics stands out as distinctive. Such rocks are found mainly in South Uist and Benbecula, locally associated with metasediments. They crop out in long linear belts within the gneiss foliation. These basic rocks have a characteristic marked compositional layering which ranges from ultrabasic through basic to felsic, the sequence being commonly repeated several times across the width of the body.

In general the bodies are no wider than several tens of metres, but some can be traced intermittently for many kilometres. One such belt runs from Loch Stulaval, round the south and east side of Arnaval, northwards across Loch Eynort and north-westwards to the Stoneybridge road junction. A more spectacular belt runs from the north side of Loch Druidibeg across Loch Bee and the South Ford to Liniclate, northwards to the Market Stance quarry in Benbecula, and then eastwards to form part of Rueval Hill in central Benbecula. Excellent examples are also seen 500 m north-east of Rubh' Aird-mhicheil (Plate 5) at [NF 734 338], at Borve Point, Benbecula [NF 765 497], on the shore west of Balivanich [NF 758 552] (see Coward, 1973a), and at Orosay [NF 726 173] in South Uist. Where recognised, these basic rocks are always concordant with the gneiss foliation and show a uniform layering traceable for considerable distances. Locally, however, they are agmatised, migmatised and strongly deformed. In two such places they are apparently cross-cut by later, less deformed basic, dykes, namely, at Sloe Dubh in Loch Eynort [NF 803 289] (Coward, personal communication) and on the south-west slopes of Reineval, South Uist [NF 754 256]. In the Reineval exposure the dykes do not cut the agmatised banded basic directly but are found in the adjacent gneiss. The migmatised basic rock forms a sheet more than 30 m thick found in isolated crops over several hundred metres, whereas the cross-cutting dykes are only 10-20 cm thick and occur in gneiss in the vicinity of the large basic body. If the cross-cutting dykes are Younger Basics (of Scourie Dyke age) in both of the above occurrences, then the agmatisation of the banded basics is to be dated as Scourian and the basics themselves as pre-Scourian. At Balivanich and Borve Point the banded basic masses are flanked by gneisses with many cross-cutting Younger Basics although none were found within the masses themselves. Previous work includes mention by Jehu and Craig (1925) of the 'banded gneisses of Arnaval' in which they describe the rapid alteration of felsic and mafic layers. Kursten (1957) also refers to the Arnaval gneiss and correlates it with the similar rocks of central Benbecula. Kursten draws attention to the fact that the Arnaval gneiss is in part assimilated by the quartzofeldspathic gneiss and concludes that the banded basics must be older. He does not, however, allow for the possibility that the gneiss was remobilised later and then intruded and digested the basics bodies.

In western North Uist between Bayhead and the coast near Tigharry, a narrow zone of subvertical rocks trending northwest and averaging 1 km wide is found (see Graham, 1970, for details of coastal section). Within this zone are two major amphibolite bands and several minor ones, and meta-sedimentary gneiss and granitic gneiss lenses. Although the basics are strongly attenuated and contain a particularly penetrative fabric and lineation, small-scale compositional banding, locally defined by garnetiferous bands, is widely seen. Pyrite and minor chalcopyrite are prominent in parts. On the basis of these features the basics are correlated with the banded basics of Benbecula and South Uist.

Coward et al. (1969) were impressed by the relationship of metasediments with banded basics and suggested that the metasediments might be preserved in the aureoles of the basics. As many of the newly recorded metasedimentary localities do not occur near basic bodies this view can no longer be entirely supported. However, the general association of the two lithologies in long linear belts is very striking. A similar relationship is described from the Loch Laxford area of the mainland (Davies, 1974b; Davies and Watson, 1977), and from the Fiskeneaesset region of Greenland (Windley et al., 1973; Myers, 1975a, 1976). In these localities the association has been attributed to the emplacement of basic magma into the supracrustal host rocks. This close association of banded basic rock and metasediment is well seen at Rueval in Benbecula and at Loch Bee in South Uist. At Loch Bee there is a series of rather exotic lithologies, mainly garnet-quartz and coarse garnet-biotite rocks, which may well represent lithologies transitional between the basic rock and the metasediment. Although the association with the metasediments and the ex istence of possible transitional types might suggest a possible volcanogenic or tuffaceous parentage for the banded basics, their strong resemblance to layered cumulates (cf. Davies, 1974b) argues for an intrusive origin (cf. Coward et al., 1969, p.403), and points to the transitional types as possible hybrids or contaminated metasediments, although the possibility that there has been metamorphic diffusion cannot be excluded. Because of the apparent pre-Scourian age of the basics and their association with metasediments it is very tempting to correlate the banded basics with the layered complexes of the Greenland Archaean (Windley et al., 1973; Myers, 1975a) as well as those of pre-Scourian age on the Scottish mainland (Davies and Watson, 1977). Care, however, must be exercised in extending correlations to the mainland because many of the layered complexes described from the Scourian (Bowes et al., 1964; Sills et al., 1981) do not exhibit such well-defined layering as the Outer Hebrides examples (see below).

Petrography

The layering is essentially defined by the varying proportions of hornblende (± clinopyroxene), garnet and plagioclase (f quartz) but the style and scale of the layering are variable. Layers are from 1 to 15 cm in thickness with rapid, sharp alternations of ultrabasic and felsic elements or with the amount of feldspar gradually increasing. Variation from ultrabasic to felsic occurs in units of 15–20 cm within larger units of 1–2 m, which themselves also show a gradual change to bulk felsic assemblages. The best example of this phenomenon is seen at Rubh' Aird-mhicheil [NF 734 338]. Here a basic body about 250 m wide shows vertical layering striking uniformly NW–SE. The mass becomes increasingly felsic to the north-east, the eastermost 50–70 m being a banded anorthosite. Within the mass, however, there are several complete units of 1–2 m varying from a dense ultramafic base to a felsic top with the same sense of variation as the overall body. These units themselves showing marked finer-scale layering. In general the ultramafic and anorthosite portions of banded basics are very subordinate to the basic (metadoleritic) section, the layering there being largely defined by variation in the garnet proportion.

One particular banded basic east of Arnaval [NF 792 258], first described by Coward (1969), has a gradation from a metagabbroic facies to a hornblende-clinopyroxene ultramafic facies. This is in close association (although no contact can be seen) with a knob of harzburgite which shows far less sign of alteration or metamorphism. The exact relationship is unknown.

The textures and mineralogies of the basics are completely metamorphic. The texture is usually subequigranular with polygonal grains. There is no evidence of former cumulate textures (cf. Davies, 1974b, p.282). In the ultramafic portions the usual assemblage is dominantly hornblende with variable amounts of clinopyroxene and garnet, and subordinate amounts of plagioclase. The same assemblage essentially persists throughout the various facies of the layered unit, the plagioclase increasing with respect to the mafics to such an extent that the anorthosite contains only subordinate amounts of mafics. Clinopyroxene may be a major constituent, especially in the more massive bodies. It may be partially altered to hornblende but in general seems to be in an equilibrium assemblage. The hornblende is usually greenish to greenish brown although it is retrograded to blue-green varieties and rarely to biotite. Analyses of amphiboles from the banded basic at Loch Bee (Appendix 1) give an average composition of ferrohornblende (sensu Leake, 1978). Analyses of garnet from a hbl-gt-cpx-pl-qz assemblage in Market Stance Quarry in Benbecula gave Also An3 Gr23 Py10 Sp4 with no significant variation from rim to core. The bulk rock chemistry is given in (Table 5), analysis 8. In a lower-grade rock from Loch Ollay with a hbl-gr-pl-qz assemblage two garnet analyses gave compositions of Al51 An4 Gr26 Py14 Sp4 and Al57 An3 Gr22 Py14 Sp4. The garnet is generally poikilitic and commonly rimmed by a corona of plagioclase. In some of the most ultrabasic rocks plagioclase is confined to these coronas, and quartz appears in the ground mass as a byproduct of the garnet breakdown. In some rocks the garnet is replaced by a polygonal mosaic of plagioclase (cf. Chapter 10: Mineralogy). The plagioclase in the highest-grade rocks (i.e. clinopyroxene-bearing) is labradorite, partially retrograded to oligoclase/andesine in the lower grades. Sphene, iron ores, zircon and apatite are common accessories. Carbonate occurs as a late-stage prod- uct. Sulphides are in relatively abundant, one sample from the Market Stance quarry containing 5 per cent pyrrhotite. In the same quarry a lens (80 X 35 X 15 mm) of molybdenite was recorded. No other examples of molybdenite were found and analyses of basic rocks and neighbouring meta-sediments (analyses 2, 8, 9, (Table 5)) show values of Mo of less than 5 ppm.

No attempt has been made to relate the growth of minerals to deformational episodes although the low-grade retrogression and introduction of carbonate may well be associated with the late brittle deformation along the Outer Hebrides Thrust zone. The age of the pyroxene is difficult to assess but since it occurs in rocks with pronounced Laxfordian fabrics it may be assumed to have at least recrystallised during this period, although its age may well be Scourian.

The 'transitional' rocks of Loch Bee show the following assemblages, from the most basic to the more leucocratic.

  1. qz-pl-hbl-gt-bi (S62089)
  2. qz-pl-bi-gt-hbl (S62090)
  3. qz-pl-gt-bi (S62091))

Assemblage 2 contains a much higher proportion of biotite over hornblende than assemblage 1. Assemblage 3 contains accessory zircon and secondary muscovite and carbonate, as well as large quartz crystals which may be late stage. The textures are completely metamorphic. These assemblages are considered to be intermediate between the meta-sediments and the basic rocks.

Geochemistry

No systematic investigation of the chemistry has been attempted and only a few analyses are recorded for descriptive purposes (Table 5). Despite the possibility that the bodies represent cumulates, the analyses define them as Q-normative iron-enriched dolerites very similar to the Younger Basics. Indeed the average analysis of the banded basics is almost identical to that of the Younger Basics (Table 24). The suite is comparable with modern tholeiite suites (Carmichael et al., 1974, tables 9–8 to 9–11), although with lower Ba and K. This is in marked contrast to the layered complexes in the Fiskenaesset region of Greenland, which belong to a talc-alkaline and largely o/-normative suite (Windley et al., 1973; Myers, 1975b). This is reflected in the few mineral analyses available; the amphibole from the Loch Bee (South Uist) banded basic is a ferrohornblende with Mg/FeO + MgO (mg) = 0.40 (20.00 per cent FeO), whereas that from a Younger Basic near Caltinish, South Uist is a ferroan pargasitic hornblende with mg = 0.46 (18.60 per cent FeO) (Appendix 1). These differ from amphiboles in the Fiskenaesset complex which are mainly magnesiohornblende and tschermakitic hornblende with mg = 0.5 (Myers and Platt, 1977). However, many of the early layered complexes of mainland Scotland are also tholeiitic (Sills et al., 1981). The possible hybrid rocks at Loch Bee have chemical characteristics transitional between the metasediments and the banded basics (Table 5). This is illustrated on the AFM plot (Table 5). x transitional rocks between basic and metasediment from (Table 5). garnetiferous metasediments from (Table 5), Sheraton et al. (1973), Palmer (1971) and Tarney and others (1972). 0:a average of four Langavat metasediments (Palmer, 1971; (Table 10)); b semipelites from the Langavat group (Table 10). +:D, average of Dalradian slates (Hickman, 1975, table 3); M average of Moine pelites (Butler, 1965, (Table 3)); G average granodiorite (Le Maitre, 1976). average of mainland Scourian gneisses, numbers refer to analyses in (Table 3). average of Outer Hebrides and mainland Laxfordian gneisses, numbers refer to analyses in (Table 3). Dashed lines define general field of metasediments (upper) and quartzofeldspathic gneisses (lower)" data-name="images/P936479.jpg">(Figure 6). It should be noted that analysis 3 (Table 5) has 77.1 per cent SiO2; the sample analysed shows large grains of quartz, which may be secondary.

Conclusions

The banded basics are something of an enigma. The field evidence strongly supports their correlation with other preScourian (or Archaean) layered complexes. The geochemistry, however, is ambiguous and would allow a correlation with the Younger Basics. The mineral assemblages and fabrics would be quite consistent with this view, but such an interpretation would carry with it the implication that the rare dykes that cross-cut banded basics in South Uist are considerably later than the Younger Basics. However, the strong association of the banded basics with the meta-sediments, their confinement to long linear belts within the gneiss foliation, and the fact that elsewhere the layered anorthosite sequences are of Archaean age, all argue in favour of a pre-Scourian age.

Banded basics of the Northern Isles

Zones of layered basics like those described from the southern islands have not been identified in the north. Several localities, however, may contain rocks of a similar age. Just south-east of the mouth of Loch Leosaid [NB 055 083] in North Harris there is a large amphibolitic band heavily veined by quartzofeldspathic material, the whole being cut by a discrete parallel-sided Younger Basic. The amphibolite band can be traced across the outcrop for several metres. Associated with the amphibolite is a series of disrupted blocks of anorthosite and leucogabbro. The blocks range up to 0.5 m and are separated by the quartzofeldspathic veins. There seems little doubt that the basic and anorthositic rocks represent a layered igneous body of early Scourian age subsequently heavily veined and agmatised before the end of the Scourian. The cross-cutting Younger Basic dyke shows very low deformation. It is a matter of speculation whether similar bodies exist elsewhere but masked by the higher degrees of Laxfordian deformation.

In the ground lying south-west of the Carloway river valley in west Lewis, Lisle (1974) has identified belts of metasediments. Associated with these is a series of old amphibolites carrying garnets and rare pyroxene (Lisle, 1974, p.29). Lisle (1974) reports that they exhibit an episode of veining not seen in the Younger Basics. It is tempting to correlate this general association of metasediment and basic with the banded basics of the Uists.

On the island of North Rona Nesbitt (1961) has recorded an amphibolite body that occupies much of the central part of the island. She describes the rock as banded, with hornblende- and plagioclase-rich bands, locally garnetiferous. The island was not visited during the present investigation, but the description suggests that the amphibolite may be equivalent to the banded basics of the Uists. Analyses of the amphibolite presented by Bowes (in Nesbitt, 1961) show its composition to be broadly similar to that of the layered basics. Nesbitt describes garnet-sillimanite gneiss that she classes as part of a general transition from the amphibolite into the neighbouring acid gneisses. Although these are shown on the main map as metasediments, they may be equivalent to some of the mixed or 'hybrid' rocks seen at Loch Bee (p.36).

Mention may also be made at this point under this heading of the anorthositic rocks of north-east Lewis. These were first described by Dougal (1928) and by Jehu and Craig (1934, p.8.55) and were mapped in detail by Watson (1969, fig. 2). The rocks are strongly tectonised and appear to predate the early deformation of the area. The anorthosites contain a number of thin amphibolitic bands and are associated with metasediments. It is therefore tempting to correlate the sequence with the metabasic/metasedimentary association of the southern islands.

Jehu and Craig (1934, p.862) report an unusual feldspar rock from near the mouth of Loch Claidh [NB 267 016]. Here, there is a small exposure of altered feldspar-epidotesphene rock, which they termed albite rock. It is probable that this rock is altered anorthosite, possibly that reported by Dougal (1928) from the 'Park district'.

Older ultrabasic complexes

Throughout the islands there are a number of ultrabasic masses. Both Jehu and Craig (1934, p.57) and Soldin (1978) recognised that these masses fall into two distinct age groups: an earlier highly altered group and a later less altered group. The earlier group, here referred to as the Older Ultrabasics, is thought to be Scourian or pre-Scourian. The later group is thought to be post-Scourian and is thus described in Chapter 9: Ultrabasic Complexes: Northern isles. The view that the Older Ultrabasics are of Scourian age is supported by Soldin (1978, p.161, fig. 4.10), who describes a fresh metadolerite containing a simple amphibolite assemblage that cuts an altered ultrabasic mass south of Tarbert. The metadolerite is thought to be a member of the Younger Basic Suite, so the alteration of the ultrabasic body must be Scourian.

The Older Ultrabasics are characteristically altered, commonly in concentric zones of assemblages bearing successively actinolite, talc and carbonate. They may be equivalent to some of the ultrabasic masses described from the mainland Scourian (Bowes et al., 1964; O'Hara, 1961, 1965).

Northern isles

Ultrabasic rocks mostly occur as isolated lensoid masses, but in places the masses line up in tracts along the gneiss foliation. Ultrabasic rocks have been reported from the island of Scarp by Jehu and Craig (1934, p.857), who described a body as consisting of a fibrous amphibole and actinolite. At Eilean Ghlas on Scalpay a sill-like mass is reported to contain veins of steatite and fibrous chrysotile (Jehu and Craig, 1927, p.478). Soldin (1978) describes a number of altered ultrabasic bodies in the area south of Tarbert. Most of them consist of tremolite-, anthophyllite-, actinolite- and talc-bearing assemblages, but one of the larger masses retains a relic core of olivine and orthopyroxene; within this core, banding on a scale of 20–50 cm may represent a primary layering. In field appearance and mode of alteration these bodies strongly resemble ultrabasic layered assemblages which predate Badcallian metamorphism in the mainland outcrop. Soldin (1978, p.161) gives the mineral compositions as olivine (Fo78) and orthopyroxene (En80) and presents an analysis of the rock ((Table 23), column 10 see p.78). Ultrabasic bodies described by Lisle (1974) from the area of Loch Laxavat and Dalbeg area are probably also pre-Badcallian.

Langavat Belt

One of the greatest developments of small ultrabasic bodies is in the Langavat Belt (q.v.: Chapter 3) of South Harris, where they form a series of isolated lenses ranging from a few metres wide up to 900 m long and 100 m wide between Rudha Sgeir nan Sgarbh and Loch Meurach (Map 5). The lenses are elongate with their axes parallel to the regional gneissic banding but the train of bodies regionally transgresses the compositional banding of the metasediments. These ultramafic pods, which weather to prominent buff-coloured rocky hillocks, have been noted by Heddle (1878), Jehu and Craig (1927), Dearnley (1963), Livingstone (1963) and Myers (1968). Livingstone (1976a) notes that seven major lenses occur and no trace of contact metamorphism could be found at the few exposed contacts. The lenses are typically composed of a central zone of olivine-tremoliteserpentine rock which grades outwards through an anthophyllite-rich ultramafic rock into chlorite schist or chloriteactinolite schist. Forsterite (Fo) values for relict olivine range from Fo89 in dunites to Fo85.5 in the rocks with ol-tremserp assemblages. Accessory minerals include pyrite, chalcopyrite, carbonate minerals, talc, chlorite and phlogopite.

In the gneisses marginal to the lenses are elongate ultramafic lenses and bands (now generally hornblendite), commonly only 5 to 10 cm thick and concordant with the gneissic banding. Cummingtonite is also found in these adjacent gneisses.

Two main exceptions to the general pattern were noted by Livingstone (1976a). These are a mica-peridotite lens (ol = Fo90.5, opx = En87) near Loch Dubh Sletteval [NG 055 898] and a partially layered ultramafic sequence at Scara Ruadh [NG 055 889]. The Scara Ruadh body has layers dipping about 70° towards the west, and consists of serpentinite layers 10 cm to 2.5 m thick with rare relict olivine (Fo845) and chlorite/tremolite layers up to 60 cm thick. Finer (about 3 cm wide) banding was also recorded but this was more irregular in nature. Livingstone (1976a) also analysed the layers for major and trace elements and suggested that the layered body was originally a layered alpine-type peridotite (harzburgite), subsequently modified by locally extensive metamorphic and metasomatic processes. These changes included the addition of CaO, H2O and SiO2 and the formation of zones of Cr- and Fe-rich opaque phases.

Myers (1968) has suggested that there is a uniform zonation of rock types from centre to margin of the lenses, with serpentine-tremolite rocks of the centre enclosed by ultramafics rich in enstatite and anthophyllite and by actinolitechlorite rocks, and rimmed by hornblendites. Livingstone (1963) analysed amphiboles from these ultramafic lenses and found them to range between pargasite, pargasitic hornblende and magnesio-hastingsite. Myers attributed this pattern to isoclinal folding of an original layered sequence in such a way that the present distribution reflects a fold interference pattern. However, the enveloping metasediments (Chapter 3: Langavat Belt) show no evidence of any similar fold structures on the scale necessary to create the observed ultramafic lenses, and the minor tight-to-isoclinal folds are shown by Myers (1968) and Palmer (1971) to verge consistently north. The explanation we prefer is that the present outcrops represent a Scourian layered ultramafic body that has become boudinaged, altered and deformed as a result of both Scourian and Laxfordian tectonic and metamorphic events.

Sound of Harris

Altered ultrabasic bodies are found on the islands of Berneray and Pabbay. They have been likened to the ultramafics of the Langavat belt in Harris (Graham, 1970, p.5.5). Graham (1970) reports the rocks as varying from almost pure dunite in the core to act-ant-talc-carbonate assemblages arranged in zones of increasing retrogression outwards. Livingstone (1965) describes the principal rock types as amphibole-bearing orthopyroxene peridotites and harzburgites, and records a vein of sagvandite (bronzitemagnesite rock) cutting across the intrusion.

Chapter 5 Lewisian Complex: the nature of the early gneisses

The Outer Hebrides comprise two fundamental (preScourian) rock assemblages: the quartzofeldspathic 'grey gneiss' and the group of associated metasedimentary, metavolcanic and banded metabasic rocks. These two assemblages are ubiquitous throughout the Archaean of the North Atlantic. As discussed in Chapter 3 (Lithology), the grey gneiss almost certainly contains deformed and migmatised metasedimentary and metabasic rocks; nevertheless the greater bulk of the gneiss consists of the characteristic and relatively uniform hornblende-biotitegneiss and biotite gneiss.

Because of their distinctive geochemistry and highly variable nature the parentage of the metasedimentary and metavolcanic rocks is almost certainly part of a supracrustal sequence. What is less certain is the original nature of the quartzofeldspathic gneiss and its relationship to the metasediments. Dearnley (1962a) and Palmer (1971) raised the possibility that the great bulk of the gneiss is migmatised metasediment, and Palmer (1971) cited the chemistry of the Langavat Belt (intermediate between that of the Leverburgh Belt and the gneiss) as evidence of a possible stage in the migmatisation process. Coward et al. (1969) considered this possibility, but favoured the idea that the metasedimentary and metavolcanic assemblage formed a supracrustal unit on a gneiss basement (see also Watson and Lisle, 1973). Unfortunately the several periods of metamorphism and deformation which the gneisses have undergone have completely obscured any field evidence that could possibly bear on the problem. The only clues remaining are the geochemistry of the rocks and the nature of possible correlatives in the less deformed areas of Greenland.

As described above in Chapter 2 (p.19) the geochemistry defines the bulk of the quartzofeldspathic gneiss as the equivalent of a calcalkaline suite ranging from granodiorite to adamellite. Chemically such a suite could be derived from an igneous, sedimentary or mixed parentage. However, the remarkable uniformity of the gneiss chemistry within the North Atlantic area argues against either a sedimentary source (cf. the varied sequences in the Leverburgh belt, which could have been sedimentary) or a mixed source, and points to an igneous source. This view is supported by Tarney (1976) who showed on plots of TiO2 against SiO2 that the gneisses have much stronger compositional affinities to igneous than to sedimentary rocks. Moorbath et al. (1975) found an initial 87Sr/86Sr ratio of 0.7014 (±) 0.0007 for the gneisses, and suggest from this and from data on Pb isotopic ratios that the parents of the grey gneiss were derived from upper mantle sources, not more than 100-200 Ma prior to the peak of Scourian metamorphism, and this they date at c.2650 Ma. Chemically therefore the evidence suggests a derivation from igneous rocks intruded at c.2800 Ma.

Myers (1976, 1978), working on the Archaean gneisses of West Greenland, describes the formation of quartzofeldspathic gneisses (Graedefjord gneiss) by the progressive deformation of a series of tonalite and granodiorite igneous intrusions. The 'gneiss' was emplaced as a series of subcordant sheets and vein networks into an already deformed series of metasediments, metavolcanics and layered gabbro and anorthosite sheets, these layered sheets being broken up into blocks by the intrusive 'gneiss.' The 'gneiss' now forms the dominant rock group of the area. The Graedefjord gneiss is itself cut by a complex sequence of later tonalites, diorites, granites and basic dykes (Myers, 1976, table 1). This whole sequence is variably deformed and subjected to amphibolitefacies or granulite-facies metamorphism. Myers (1978) suggests on the basis of various isotopic age dates that the rocks were formed, deformed and metamorphosed in the span 2800-3000 Ma. It is easy to see parallels between this area and the Outer Hebrides, and arguments for a common history are strong.

We conclude, therefore, that the bulk of the grey gneiss of the Hebrides is a complex sequence of tonalite, granodiorite and granite intrusions, aged of c.2800–2900 Ma, intruded into (or in part extruded onto) a sequence of rnetasedimentary and metabasic rocks, the metabasics therefore being the oldest rocks of the area. Myers (1978) has several illustrations of the progressive deformation of the various rock units of the Graedefjord gneiss demonstrating that with high degrees of deformation and migmatisation there is a marked convergence of appearance between the various rock types. Because of this it is impossible even to guess at the full range of original lithologies 'hidden' in the grey gneiss of the Hebrides. In general biotite gneiss may be ascribed to monzogranitic parents, the hornblende-biotite-gneiss to granodioritic and tonalitic parents, and the basic units possibly to minor intrusives, although lithological differences may in part have been enhanced by metamorphic segregation.

The regional variations in geochemistry described in Chapter 2 (Geochemistry) may relate to broad changes in the type of igneous parent; for example, the higher K/Rb ratios in western Lewis may indicate a higher preponderance of granite although the transitional nature of the variations would argue against this.

This explanation implies that during the late Archaean there was a widespread intrusion of granodiorite of apparent upper mantle origin throughout the North Atlantic area. What this may mean in terms of geological models is wholly a matter for speculation (Tarney, 1976) and is beyond the scope of this memoir. It also implies that the gneisses of the Outer Hebrides may not have undergone the more extensive history proposed for the mainland Scourian (Davies, 1975).

Chapter 6 Lewisian Complex: Late-Scourian intrusions

The rocks described in this section are a group of intrusions, mainly intermediate and acid, which occur sporadically throughout the Outer Hebrides. Locally they are sufficiently large or numerous to be shown on the 1:100 000 map sheets, but in the main they are small bodies which can be classed as minor intrusions. They can be defined as rocks of igneous origin that postdate the Scourian metamorphism and migmatisation but are older than the Younger Basic suite. Ideally, therefore, one should be able to identify them by seeing them both transgress the gneissose banding of the country rock and being cut by members of the Younger Basic suite. Such a combination of circumstances is rare, to be seen only in the small areas that have largely escaped the effects of subsequent deformation and metamorphism. The best exposed and documented area of this type is southeastern Barra, where the full range of intrusion types is present, and age-relationships within the suite can be unambiguously demonstrated. It was described by Hopgood (1964) and by Francis et al. (1971). Francis et al. established the following sequence of intrusions:

Only the early dyke suite, granites and early pegmatites are dealt with in this chapter; the ultrabasics have been covered in Chapter 4 and the Younger Basics and Laxfordian pegmatites are dealt with in Chapters 10 and 11 respectively.

A five-point Rb/Sr isochron from three early pegmatites gave an age of 2465 ± 53 Ma, subsequently recalculated to 2610 ± 50 Ma (Moorbath et al., 1975).

A major problem with the Late-Scourian intrusions is that although they are undoubtedly of igneous origin, most are deformed and to some extent recrystallised, and their origins are obscured. It must be emphasised therefore that classification and nomenclature is based upon their present mineralogy, which may not correspond with that of the original igneous rock.

Metadiorites and microdiorites

Field aspects of minor intrusions

Cross-cutting pale grey sheets and dykes range in thickness from about 1 m down to a few centimetres. They are best displayed on the very well-exposed coastal section of southeast Barra between Leanish [NL 700 986] and Earsary [NL 705 998]. The dykes are normally rectilinear and unfolded, but commonly exhibit an internal fabric oblique to the dyke walls and defined by flattened and/or folded mafic schlieren and/or aligned biotites (Plate 6). This fabric is cut by virtually undeformed members of the Younger Basic suite and by veins of pink 'granite' and early pegmatites (Plate 7). On the basis of cross-cutting relationships at Leanish and variations in composition, the dykes of this suite have been subdivided into three sets (Francis, 1973). The oldest are basic and the youngest acid. None exceeds 25 cm in thickness and Francis (1969) has recorded fewer than six of each. The third set, intermediate both in age and in chemistry, are very much more abundant. North-eastwards along the coast from Leanish, there are many dykes of the intermediate set that cut the gneiss foliation discordantly. Intermediate dykes also cut the gneiss above the main (Outer Hebrides) thrust on the islands of Fuiay and Flodday. No intermediate minor intrusions have been positively identified to the west of the main thrust on Barra itself. On the island of Fuday in the Sound of Barra, however, a swarm of intermediate dykes is present and the relationships between these and the Younger Basic suite are particularly well displayed. On the south-east side of the island both the intermediate and the Younger Basic dykes are markedly discordant to the migmatitic gneiss foliation and individual intrusions of each can be continuously traced for up to 80 m. Close to Rudha Carraig-chrom, at the eastern extremity of Fuday, a discordant Younger Basic dyke, 1 m thick, can be followed for 75 m. On the waterwashed slabs just above the high-tide mark, the dyke truncates a 30 cm intermediate sheet and its pronounced internal fabric. The fabric resembles that seen at Leanish but is more intensely developed and displaces the margins of the sheet to form a series of similar folds with an axial-plane fabric continuous with that in the gneiss (Plate 8). No fabric is visible in the Younger Basic dyke. Nearby a 5–30 cm intermediate sheet is folded along with the gneiss, and these folds are cut by a 10–30 cm pink pegmatite, typical of those in the area, which in turn is cut by a Younger Basic dyke 10 cm thick (Plate 9).

The islands to the south of Barra have not been surveyed in detail but on Flodday [NL 612 924] at least one thin, pale grey, concordant intermediate sheet has been found. Farther south, on the east side of Mingulay [NL 569 826], several thin sheets lying within the gneiss banding are petrographically almost identical to the intermediate dykes at Leanish and probably belong to the same suite. Northwards from the Sound of Barra, small, Late-Scourian intermediate dykes have been found at only two localities. The first lies on the south coast of South Uist about 1 km west of Ludag [NF 760 139]. There, two or three dykelets, none thicker than 20 cm, cut the migmatitic gneiss foliation at a high angle and are more deformed than members of the Younger Basic suite nearby. There are similar relationships near Garry a-siar on the west coast of Benbecula [NF 759 529]. A narrow, slightly transgressive intermediate sheet is strongly folded along with the gneiss, while a few metres away Younger Basic dykes are markedly discordant and relatively undeformed.

Field aspects of major intrusions

The phase of igneous activity which produced the intermediate suite described above also gave rise to intrusions large enough to be shown at the 1:100 000 scale (Maps 1 and 2). They are best developed in Barra, above the major thrust, in a zone about 1 km wide to the east of Castlebay, and they can also be found at the same structural level on the islands of Vatersay and Sandray to the south. An uninterrupted cross-section through this zone is exposed between Orosay and Rudha Charnain on the south coast of Barra. The dominant rock type is a homogeneous, foliated, pale grey metadiorite occurring in massive sheets up to several hundred metres thick inclined eastwards at 50-60°. These are interbanded at the west end of the section with normal quartzofeldspathic gneiss cut by veins of pseudotachylite. Towards the eastern limit of the zone, coarse migmatitic and themselves contain leucocratic lits giving the rock a migmatitic aspect. Throughout the section, members of the Younger Basic suite are present, and some of them cut the fabric in the country-rock gneisses and in the metadiorites.

Smaller, irregular sheets of coarse-grained grey foliated metadiorite are found at Earsary on the east coast of Barra [NL 705 998], but the dominant type of intrusion in that area is pink 'granite' (see below).

A thick sheet of metadiorite is present in the south-east part of the island of Flodday [NF 755 023], again a short distance above the thrust. Nearby, below the thrust, the entire island of Lamalum [NF 729 033] and the east side of the Black Islands [NF 728 021], are composed of typical foliated massive metadiorite, locally very coarse and hornblende-rich, the total width of the outcrop being about 250 m.

Northwards from the Sound of Barra no intermediate intrusions of any significant size are encountered until the north-west coast of Lewis is reached. At the mouths of the Borve River [NB 409 573] and Dell River [NB 488 624] two mappable intrusions are exposed (Davies et al., 1975). The Borve body is a rather homogeneous smooth-weathering the coast for a distance of 2 km. The inland margin is strongly deformed, but the interior portion is weakly foliated and carries conspicuous oval white feldspar spots 0.5 to 1 cm in length. The smaller exposure at the Dell River, although more strongly foliated, is very similar to that at Borve; projecting the foliation trends in the adjacent country rock gneiss suggests that the two bodies are in structural continuity. The correlation is strengthened by the close association of .metagabbroic and metasedimentary rocks with both sheets. Petrographically these metadiorites bear a strong resemblance to the Rudha Charnain rocks and are therefore tentatively allotted the same age.

Petrography of the minor intrusions

As stated earlier, the intermediate dykes on Barra have been subdivided by Francis (1973) into three types. The oldest are simply mafic varieties of the main microdiorite suite, rich in greenish brown hornblende, biotite and clinopyroxene. Two generations of pyroxene are present; one occurs as relict cores in hornblende and may represent original igneous pyroxene, while the other forms small (0.5 mm) rounded clear diopsidic grains ((S58118)).

The youngest set of intermediate dykes, of which few are to be found, are leucocratic, fine-grained microgranodiorites. They are rich in oligoclase and carry subordinate potash feldspar in the form of strained perthite and microcline. Quartz is interstitial. The only mafic mineral is a dark brown biotite which defines a weak fabric.

The main set of dykes are microdiorites with equigranular textures and an average grain-size of less than 1 mm. Plagioclase (An28) is always dominant, constituting 46 -60 per cent of the rock (Table 11). It occurs as simply twinned crystals, subhedral to anhedral, forming a polygonal texture. Rarely it is elongated parallel to a mica fabric. Deep-green hornblende, subhedral to anhedral, is normally the dominant mafic mineral. It varies in length from 0.5 to 1 mm, the long axes of the larger crystals lying in the plane of the biotite fabric. Where mafic schlieren are present they are composed of aggregates of polygonal hornblende grains. Biotite is always present as foxy-red or dark brown flakes, 0.75-3 mm in length, which are usually aligned to give the rock its pervasive schistosity. A typical feature of the microdiorites is the high percentage of accessory apatite, 2.5 per cent being recorded in one specimen. The other main accessory mineral is an opaque ore commonly rimmed by hematite. Allanite and zircon are rare accessories. Potash feldspar is seldom present; the modal analysis (Table 11) that recorded 11.4 per cent (S57801) is quite exceptional. Quartz is similarly rare other than in rocks containing potash feldspar where it occurs as strained interstitial grains.

The microdiorites west of the Outer Hebrides thrust at Mingulay, Flodday, Ludag and Garry a-siar, though similar to those on Barra, show evidence of retrogression. The hornblende is pale bluish green, and many crystals contain small blebs of quartz; potash feldspar and quartz are more common; sphene is present as narrow rims to some of the opaque minerals; and some of the accessory allanite is rimmed by epidote.

Petrography of major intrusions

The metadiorites are in most respects identical to the microdiorites apart from their coarser grain-size: 1.5–2 mm. A few carry strained antiperthitic potash feldspar and elongate strained quartz aggregates, which suggests a possible transition towards some of the augen 'granites' described below. Modal analyses of specimens from Earsary (S59979), (S57805) and Rudha Charnain (S58101) show them to be diorites or quartz monzodiorites (Table 11). Biotite is usually more abundant than dark green hornblende, and commonly grows across the amphibole to give the rock a penetrative schistosity. Pyroxenes, both hypersthene and clinopyroxene, are rare, and tend to occur as small altered rounded grains.

Granites

Under this heading is included a variable group of 'granitic-looking' rocks which occur principally in Barra and North Uist. Variations in original rock type, in the intensity of deformation and in metamorphism subsequent to their formation have tended to modify some of these intrusions radically. Igneous textures have apparently been largely obliterated and mineral assemblages considerably altered. Some intrusions that lack good age-criteria and are concordant, may be difficult to distinguish from the host country-rock, and their classification as Late-Scourian is based upon slender petrographical evidence.

Field aspects, Barra

Two completely dissimilar types of granitic rocks can be distinguished on Barra and the adjacent islands. The first is confined to the area of low-intensity Laxfordian deformation between Leanish and Earsary on the east coast where intrusive relationships are still preserved. Numerous coarse massive pink sheets, up to about 10 m in thickness, with diffuse margins and weak planar fabrics, invade the country-rock gneiss (Francis, 1973). These intrusions were described as granites by Francis (1973) and their Late-Scourian age is confirmed by the fact that they carry inclusions of microdiorite but are themselves clearly cut by undeformed members of the Younger Basic suite. Similar rocks are found above the main thrust on Vatersay.

The second group occurs within the main outcrop of metadiorite on the south coast between Orosay and Rudha Charnain. There, coarse pink granitic sheets, concordant and transgressive, with a pronounced augen foliation, locally cut both the quartzofeldspathic gneiss and the metadiorite. The pink augen average 1 cm in length and exceptionally reach 3 cm. Here again the augen granite is cut by members of the Younger Basic dyke suite. The pink augen granites ap pear to become more deformed when traced inland and are well exposed as concordant gneissose granites interbanded with grey foliated metadiorite immediately above the major thrust to the west of Heaval. Below the thrust on the island of Lamalum a 50 m band of strongly schistose micro-augen granite occurs within the dominant foliated metadiorite.

Field aspects, North Uist

On the north-west coast of North Uist Jehu and Craig (1926) noted areas of very acid gneiss which they thought to be younger than the associated hornblende gneiss and hornblende-biotite gneiss. In the same area Graham (1970) recognised two lithological gneiss types— 'rough' and 'smooth'. The smooth gneiss he described as unbanded and homogeneous with augen of pink microcline, and suggested that it represented a pre-Laxfordian porphyritic granite. It forms concordant bands folded by at least two sets of minor folds. Graham (1970) also discovered a sheet of deformed granite 0.75 km thick on the island of Berneray in the Sound of Harris, which is grey and coarse grained with a strong fabric defined by aligned biotites. A few deformed members of the Younger Basic suite are found within the granite and Graham (1970) concluded that it was of Scourian age. As a result of the present survey, a large number of additional granitic bodies have been found in North Uist. The largest is a coarse-grained white lenticular body about 3 km long near the north-west coast [NF 725 680]. Most are much smaller, and range from pale grey massive medium-grained rocks with white augen (perhaps original phenocrysts) to intensely flattened laminated grey-pink rocks with strongly developed rodding. Some of these are very similar to Graham's 'smooth' gneiss but others, e.g. on Vallay [NF 772 767], are less deformed and contain rectangular pink potash feldspars up to 5 cm in length. Numerous elongate lenses of granitic gneiss have been mapped in the Newton Ferry/Beinn Bhreac [NF 904 770] area. Most are massive and pink-coloured, locally with feldspar augen, and are apparently cut by amphibolites of the Younger Basic suite. Areas of unbanded grey and pink rodded granitic gneiss with augen are also common on the small islands on the south side of the Sound of Harris. Several Younger Basic sheets occur within the granitic gneiss in this area; one dyke 2 m thick appears to cut the planar fabric in the gneiss [NF 981 752].

Taken together, most of the granites assumed to be of Late-Scourian age in North Uist lie in a zone 2-4 km wide which mimics the overall tectonic pattern and suggests that most of the intrusions lay at roughly the same structural level prior to the onset of Late-Laxfordian deformation (see Map 2).

Field aspects, Lewis

A few bands of granite gneiss, possibly of Scourian age, have been distinguished on the north-west coast of Lewis [NB 370 540] and [NB 435 595], and on the Eye Peninsula [NB 535 365]. These were mapped out largely because they are pink, leucocratic, and smooth-weathering, in contrast to the normal grey gneiss country-rock. No intrusive relationships have been detected. The sheets are usually strongly foliated, and some appear to be tightly folded along with amphibolites that probably belong to the Younger Basics suite. They seem to have been affected by all stages of Laxfordian deformation, and on this basis, together with their resemblance to the 'smooth' gneiss of North Uist, they are included here as Late-Scourian granites.

Petrography, East Barra

In the present re-examination of the area of exceptionally low Laxfordian deformation between Leanish and Earsary, one of the most unexpected results has been the discovery that the Late-Scourian granites of Francis (1973) are in fact monzonites ((S57804), (S59977), (Table 11)). The monzonites form a uniform suite of even-grained pinkish rocks with equigranular textures and weak planar fabrics. The dominant constituent (up to 50 per cent of the rock) is pink potash feldspar occurring as anhedral strained string perthite. Some of the crystals are lenticular, the long axis averaging 1.5 mm. A few exceptionally large aggregates 2.5 cm long may represent original phenocrysts. Plagioclase is polygonal albite oligoclase, occasionally containing vermicular quartz. Quartz seldom exceeds 5 per cent, and is interstitial and internally highly strained. The total mafic content is usually about 10 per cent. Foxy-red 'clean' biotite, up to 1.5 mm long, is always in excess of 'dirty' relict green hornblende which shows retrogression to biotite aggregates and quartz. The clean biotite is the mineral (along with the slightly lenticular potash feldspar) which gives these rocks their fabric, and it commonly cuts across hornblende. The opaque iron oxide (2–3 per cent) is in large homogeneous crystals of ?magnetite or ilmenite. Apatite and zircon are accessory.

Petrography, South Barra

A second granitic type occurs within the area above the major thrust to the east of Castlebay. Rocks of this group are typically coarse pale grey to salmon-pink granites with a pervasive augen foliation. The pink augen consist of lenticles of potash-feldspar aggregates, usually a mixture of highly strained microcline microperthite grains and strained stringperthite subgrains. The clear augen are mainly quartz in all stages of breakdown into strained lenticular aggregates up to 5 mm long. Plagioclase is subordinate and commonly myrmekitic. Biotite, in part showing late-stage kinking, occurs in aligned laths and is the sole mafic mineral. Some of this deformation may be related to movements on the underlying thrust. Accessories are iron ore, apatite (very common), epidote, zircon and allanite.

Petrography, North Uist

Unequivocal age relationships for the early Uist granites have not been found. It is not known with certainty that the rocks mapped as granite are all of Late-Scourian age or are genetically related. The petrographic factors which, lend weight to their being included in this section are that most of them are rich in pink microcline, which distinguishes them from the country-rock grey gneiss; they show a gradation from typical rocks with granitic textures to granoblastic gneisses; and they are unlike the Laxfordian granites.

These North Uist granitic rocks are difficult to describe in a systematic way. Variations in mineralogy, degree of recrystallisation, and late-stage retrogression all overlap to some degree. As a starting point, however, two types may be distinguished.

An example of one type is the large lenticular body on the west side of the island near Hanglam [NF 725 680]. This is locally gneissose but virtually undeformed internally, and it preserves a coarse granitic texture with an average grain size of 2 mm. It consists dominantly of anhedral clear oligoclase, slightly strained quartz, and interstitial microcline. The chief mafic mineral is green hornblende up to 3 mm in length, overgrown by bright brown biotite, the biotite showing local retrogression to chlorite and epidote. Apatite, allanite, zircon, and a little iron-oxide (S67673) are accessory. On the classification of Streckeisen (1976) this rock, with less than 2 per cent potash feldspar and about 10 per cent of quartz, is a quartz diorite. Total mafics are approximately 10 per cent, with hornblende in excess of biotite.

An example of the other type is a concordant band of porphyritic granite at least 50 m thick, on the north side of Vallay [NF 772 767], which, although partly foliated, preserves a relict igneous texture (S61331). The pink megacrysts consist of strained microcline and occur as rectangular tablets and lenticular augen up to 5 cm long but averaging 2 cm. The groundmass is inequigranular and consists of equal amounts of microcline and turbid plagioclase up to 1.5 mm across, and over 20 per cent of highly strained quartz. Both the plagioclase and the quartz are locally recrystallised to strained subgrains 0.5 mm in diameter. Dark green hornblende contains quartz blebs and is crossed by dark brown biotite crystals up to 0.75 mm in length. Much of the biotite shows partial alteration to chlorite and is associated with epidote. Accessories are allanite (some rimmed by epidote), apatite, sphene (one crystal is 0.75 mm long) and zircon. This rock is a hornblende-biotite granite. There are several almost identical bands at the west end of Vallay, which exhibit varying degrees of deformation and recrystallisation. The end product is a homogeneous pale coloured gneissose granite with a pervasive biotite-epidote fabric. The average grain size in these bands is 1 mm, with the quartz tending to form highly strained elongate ribbons. Some replacement of clean polygonal plagioclase by muscovite and epidote is evident. Hornblende is absent.

On the islands of Hulmetray [NF 983 750] and Vaccasay [NF 976 755] in the Sound of Harris, bands of strongly lineated white augen granitic gneiss are present ((S61346), (S61348)) which may represent highly tectonised versions of the quartz-diorite type described above. They have a well-developed augen texture with an average grain-size of 0.5 mm. Plagioclase (An30) is dominant, microcline is interstitial, and quartz occurs either as elongate lenticles of strained subgrains, or simple rounded less strained crystals. The fabric is defined by the quartz lenticles and by aligned dark green hornblende and dark brown biotite. Epidote is associated with the biotite, and apatite, allanite, zircon and opaque iron oxide are accessories. The amount of quartz in these rocks is sufficient for them to be classified as tonalites. The south end of the nearby island of Tahay [NF 964 750] is composed of rodded augen gneiss which could be a highly deformed version of the porphyritic granite found on Uist. The rock is intensely foliated and lineated, with pink augen 0.5 cm wide and 2 cm long. A thin section (S62182) cut parallel to the long axes of the augen shows alternating layers, about 1 mm thick, of elongate slightly strained quartz lenticles (6 mm) and granoblastic-textured aggregates of microcline and subordinate plagioclase. Triple junctions of 120° are common in the feldspar layers where the average crystal size is 0.2 mm. The plagioclase is commonly myrmekitic and partially overgrown by muscovite. The layers are often separated by discontinuous rows of brown biotite and rare hornblende of single crystal width. Apatite is the only notable accessory. The granite gneiss on Sarstay [NF 974 759] is white and displays many of the same textural features. The main difference is that the feldspar layers in the Sarstay gneiss are of polygonal-textured plagioclase (average size 1 mm) with very subordinate interstitial microcline. Hornblende is more common, in bands 1 mm wide, and is overgrown by biotite and epidote. Apatite, zircon and some opaque iron ore are accessory (S62178). This rock is probably an intensely deformed and recrystallised quartz diorite.

Pegmatites

Transgressive pegmatite veins are found throughout the Outer Isles but there are no reliable criteria to distinguish the various suites in the field. According to Cunningham (1981) allanite is a common accessory in the Laxfordian pegmatites but is absent from the Late-Scourian veins in which monazite may be found. By definition, the Late-Scourian pegmatites are older than the Younger Basic suite but it is only in areas of low Laxfordian deformation and metamorphism that clear-cut age relationships have been preserved. Thus at Leanish in Barra [NL 703 986] a 1 m dyke of the Younger Basic suite cuts a 0.3 m pink pegmatite which itself cuts a microdiorite dyke 0.1 m thick (Plate 7). Nearby [NL 703 987] a dark grey Younger Basic dyke 8-12 m thick cuts a 3.5 m pink pegmatite with a prominent white quartz core and large clots of magnetite. This pegmatite contributed to a Rb-Sr isochron age of 2610 ± 53 Ma (Francis et al., 1971). A small Laxfordian pegmatite (0.3 m) cross-cuts the dyke 20 m to the north, along the outcrop of the basic dyke. In addition, on the island of Fuday a fine-grained black Younger Basic dyke, [NF 743 084], 5 cm thick, cuts a planar pink pegmatite vein, 10 cm thick, which in turn cuts a folded microdiorite dyke.

At Ardivachar [NF 738 457], South Uist, a well-documented area of low Laxfordian finite strain, Dearnley and Dunning (1968) described veins and pods of gneissose pegmatite which locally cut the banding in the country-rock gneiss and are in turn cut by Younger Basic dykes. The pods consist of orthoclase, microcline microperthite, albite and quartz, with subordinate biotite and accessory opaque oxides and metamict rare-earth minerals. The larger alkali feldspar crystals locally show a graphic texture and incipient granulitisation. They argued that the pegmatite is 'pre-Scourian' in age but an indifferent isochron gave an age of 2560 ± 80 Ma (Lambert et al., 1970a). Several interpretations of this date have been advanced, but its near coincidence with the Barra date suggests to us that the pegmatite belongs to the Late-Scourian suite.

Dearnley and Dunning (1968, table 1) also record Late-Scourian pegmatites at Garry a-siar [NF 756 531] on the west coast of Benbecula. These cut the main early foliation in the gneiss. Detailed mapping of the area during the present survey revealed one locality where a pegmatite, unequivocally of Late-Scourian age, can be seen. Here a 0.2 m pegmatite is cut by a 3.5 m amphibolite of the Younger Basic suite.

No examples of Late-Scourian pegmatites have been recorded from North Uist, Harris or Lewis. On the island of Scarp, however, Myers (1970a) describes and figures small veins of Scourian granite and pegmatite which cut across the gneiss banding and are cut by thin amphibolites, probably of the Younger Basic suite [NA 989 138].

Geochemistry

Microdiorites

Four analyses of microdiorites are presented in (Table 5) analyses 12 -15, two from above the thrust and two below. They range from Q normative to ne normative with SiO2 45 - 56 per cent; FeO total + MgO 10 -16 per cent; CaO 4 - 8 per cent; and alkalis 5 - 8 per cent. Chemically the rocks are equivalent to the trachyandesite - trachybasalt series (Carmichael et el., 1974, table 2 -3; Le Maitre, 1976). The AFM plot ((Table 23), analyses 1-5. A ultrabasic bodies, (Table 23), analyses 6-8. layered basics, (Table 5), analyses 7-11. O anorthosite from layered complex, (Table 5), analysis 6. x Late-Scourian 'microdiorites', (Table 5), analyses 12-15. * Late-Scourian 'granites', (Table 5), analyses 16-17. range of members of mainland Scourie Dyke suite from Tarney (1973). Younger Basic metadolerites (Outer Hebrides: this work). 'Cleitichean Beag' basic from Dearnley (1963, table 7, analysis 4). 0 noritic rocks from Soldin (1978, figure 4.11); 'average' of Maaruig ultrabasic from Soldin (1978, figure 4.11)." data-name="images/P936484.jpg">(Figure 11), p.80) shows them to have a typical alkaline trend (Carmichael et al., 1974, fig. 10-7). They may be considered as broadly chemically equivalent to the Late-Caledonian microdiorite suite of the West Highlands although they show a greater iron enrichment (Fettes and McDonald, 1978, fig. 9) with an Fe:Mg value of about 2.5 as opposed to 1. There is no significant difference between those above and those below the thrust. This might suggest that the minor petrographical differences described above may reflect essentially isochemically variations with metamorphic grade.

Metadiorites

Only one sample of metadiorite has been analysed- (S57805) ((Table 5), number 16). The rock is Q normative with SiO2 56 per cent and NaO + K2O 7.8 per cent. Chemically it shows no significant difference from the microdiorites.

Granites

One typical pink granite from the east side of Barra has been analysed ((S57804), (Table 5), number 17). It is Q normative with a SiO2 value of 60 per cent and NaO + K2O totalling 10.6 per cent. The granite is chemically similar to the metadiorite but has a much higher potash content (6.34 per cent), which would define it as syenitic (Le Maitre, 1976). This confirms the petrographical evidence that these rocks on Barra are not true granites. Unfortunately the data are too sparse to assess accurately whether the high potash is a reflection of primary chemistry or due to secondary effects.

Pegmatites

No published analyses are available. Cunningham (1981) describes the Scourian pegmatites as having significant chemical differences from the Laxfordian pegmatites (see Chapter 11).

Conclusions

1 A variety of igneous rocks were intruded into the Outer Isles Lewisian at c.2600 Ma.

2 The sequence of intrusion has been established by numerous cross-cutting relationships, as follows (from younger to older):

Potash pegmatites

Potash granites (monzonites) and granites Diorites, monzodiorites and microdiorites

3 The original nature of the intrusions is best seen in areas of low subsequent tectonism, e.g. the east coast of Barra and the north coast of North Uist. Where deformation is high the pegmatites and granites are represented by varieties of unbanded homogeneous augen granitic gneiss. The most strongly deformed and recrystallised granites are found on the islands in the Sound of Harris where they occur as 'pencil' gneisses in the core of a large DL3 Laxfordian fold (see (Table 27)).

4 Definite evidence exists at Leanish (Barra), Fuday (Sound of Barra), Ludag (South Uist) and Garry a-siar (Benbecula) that a phase of deformation and recrystallisation took place in the interval between the intrusion of the microdiorite dykes and the Younger Basic suite, and probably also predating the suite of pegmatites. This was interpreted as a Laxfordian event (i.e. post-Younger Basic suite) by Francis (1973, pp.188 - 189) which in some way failed to affect the Younger Basic suite to any extent. However, Davies et al. (1975, p.49), dealing with North Lewis, show a sequence of events as follows:

Emplacement of Scourie Dyke suite
Scourian Formation of marginal lineaments
Scourian Penetrative deformation and metamorphism
Scourian Emplacement of dioritic and other intrusions

The dioritic intrusions have been correlated with the Barra metadiorite–microdiorite suite (p.41), which implies that the deformation and metamorphism of these Lewis intrusions is of Late-Scourian age. Whether or not this phase can be correlated with the Inverian on the mainland (Evans and Lambert, 1974), is dealt with in Chapter 1.

5 On the island of Coll (Inner Hebrides), Drury (1972) tentatively correlated quartzose pegmatites, emplaced at the beginning of his Inverian, with those cut by the Younger Basic dykes at Leanish on Barra. On the adjacent island of Tiree, Westbrook (1972) identified early pegmatites which he equated with the Late-Scourian pegmatites mapped by Dearnley and Dunning (1968) at Garry a-siar, Benbecula.

6 On the Scottish mainland near Stoer, Sills and Windley (1982) described six thin discordant amphibolite dykes which were intruded, metamorphosed, and folded prior to the intrusion of the Scourie Dykes. Although chemically unlike the microdiorites these amphibolites appear to be about the same age and are classed by Sills and Windley as Late-Scourian or early Inverian.

7 A direct correlation seems to exist between the Late-Scourian potash pegmatites of the Outer Hebrides and those on the Scottish mainland. The mineralogy and isotopic dates are equivalent (Evans and Lambert, 1974, p.142). The Hebridean evidence further suggests that some of the bands of microcline-rich granitic gneiss of the mainland, e.g. at Loch Kirkaig (Evans and Lambert, 1974, p.131) and at Assynt (Sheraton et al., 1973), may have been originally porphyritic granites rather than pegmatites. Davies and Watson (1977, p.129) cite examples of microcline-rich gneiss and pegmatite cross-cut by amphibolites that they interpret as metadolorites of the Scourie Dyke suite. The potassic gneisses may well have been granites and could belong to the Late-Scourian suite.

8 Farther afield in west Greenland, the Qorqut biotite granite which contains equal amounts of microcline, plagioclase and quartz has been dated at 2520 ± 90 Ma (Moorbath and Pankhurst, 1976). The scale difference (the WI-gut intrusion is 50 km long and up to 18 km thick) is immense, but its age would place it in the Late-Scourian category. Bridgwater et al. (1973a, table 1, p.496) suggest an equivalence of an intrusive granite-pegmatite event in Labrador, Greenland and Scotland at 2400-2600 Ma, thus by implication suggesting that this granitic event may mark the boundary between the Archaean and the Proterozoic.

Chapter 7 Lewisian Complex: the South Harris Igneous Complex

In this chapter we describe the igneous geology of that part of South Harris which lies south-west of a line between Borve [NG 035 950] and Rubha Quidnish [NG 103 865]. Exposure is extremely good over much of the area, particularly on the north-east side of Loch Langavat and south-east of Roineabhal; ice scouring has removed almost all the drift cover in these parts. Locally extensive peat obscures a large area of outcrop in Glen Horsaclett and east of Loch Langavat. There is also a lot of scree on Roineabhal.

The various component units of the Igneous Complex intrude a series of metasediments which comprise the Langavat, Leverburgh and Ensay belts, These metasediments are described in Chapter 3. The Igneous Complex outcrops over 46 km2 forming 64 per cent of exposed rock in the south-west part of South Harris. Although the more obvious geological features were recorded by the early geologists who visited the Outer Hebrides there has subsequently been a great deal of detailed work recorded in PhD theses on the area during the period 1963–1978. Much of the following section is based on these accounts. Map 5 shows the detailed distribution of rock types in south-western South Harris.

Previous work

MacCulloch (1819) distinguished the Roineabhal area from the surrounding gneisses and noted the garnet metagabbros and marbles of the Rodel area. Heddle (1878) again drew attention to the garnet metagabbro, which he termed 'eclogite', and to the 'serpentine' prominences in the Langavat area.

Jehu and Craig (1927) made a detailed appraisal of the nature and distribution of rock types of South Harris, They distinguished a 'gabbro-diorite' (here termed metadiorite) body adjacent to the meta-anorthosite and described granulite-facies mineralogies from within the body near Mula [NG 024 899]. They also described zones of abundant pegmatites marginal to and within the Langavat Belt meta-sediments which extend north-westward on to Taransay.

Subsequently Davidson (1943) carried out detailed petrographic work and fieldwork in the Rodel–Roineabhal area. He distinguished ultrabasic bodies in the metasediments and large metagabbro sheets, one adjacent to the anorthosite and the second in the Renish Point–Strond area marginal to the Sound of Harris (here termed a metanorite). He considered that all the meta-igneous rocks were part of a single highly differentiated intrusive complex. He also distinguished two metamorphic events: a high-grade garnet-forming event (in the meta-igneous rocks), and a later retrograde event.

Dearnley (1959, 1963) produced a detailed synthesis of the whole region including chemical and modal analyses. His map, which is at a scale of 1:25 000, remains substantially unaltered by the subsequent work. He proposed that the 'Igneous Complex' and the widespread Scourie Dyke suite (which is notably absent in south-west South Harris) were both related to a tholeiitic parent magma. He also proposed that the two major metamorphic events were Laxfordian in age. The 'metagabbros' (metanorites) bordering the Sound of Harris were termed pyroxene-granulites, and in them he recognised relict textures interpreted as pre-Laxfordian in age.

Since Dearnley's work, theses have generally concentrated on particular aspects or lithologies of South Harris, e.g. the Laxfordian granite complex of Harris and related gneisses (Myers, 1968), a comparison of the gneisses of South Harris and the Supertoq region, east Greenland (Palmer, 1971), ultramafic rocks (Livingstone, 1963), meta-anorthosite and metagabbros (Witty, 1975), and metadiorite and metanorite (Horsley, 1978).

General lithology

The South Harris Igneous Complex is composed of four major distinct meta-igneous lithologies which in decreasing order of age are: metagabbro, meta-anorthosite, metanorite and metadiorite. A large elongate body of metatonalite occurs within the metadiorite. (As elsewhere in this memoir, the prefix meta- is mostly omitted in the rest of the text of this chapter.) The mineralogical and geochemical features of the igneous bodies show them to be separate but related bodies of a highly differentiated calcalkaline suite.

The metamorphic effects are more fully described at the end of this chapter (under the headings Metamorphism and Synthesis). In brief, the oldest and highest-grade mineral assemblages in the igneous rocks are products of granulitefacies metamorphism which probably occurred soon after intrusion and continued for a considerable time. Lower-grade (amphibolite- to greenschist-facies) metamorphic events occurred during the Laxfordian but are restricted in many of the igneous bodies to marginal shear zones and are in general structurally controlled. The metamorphic pattern is one of progressive decrease in pressure and temperature from c,2000 Ma (Cliff et al., 1983) to the Late-Laxfordian pegmatite intrusions at c.1500 Ma.

Magnetic data (Westbrook, 1974) suggest that the basic igneous bodies, notably the gabbros, extend offshore for about 24 km to the north-west of South Harris, maintaining the approximate width of outcrop seen on land. Westbrook (1974) suggests that gabbros underlie the diorites at a depth of about 7 km. As the outcrop of the anorthosite body widens to the south-east it is taken to extend off the eastern shore; however, effects of the thrust zone and Minch Fault in that area must greatly modify the original contact relationships.

Ultramafic and mafic dykes and pods intrude the meta-sediments and larger igneous bodies.

The relative ages of the four major rock types and their relationship to the metasediments are shown by well-documented field relationships. The metasediments occur as foliated xenoliths in the various igneous bodies. Locally they have internal fold structures. Dearnley (1963), for example, recorded 'foliated and contorted' quartzose paragneiss xenoliths in the anorthosite north of Rubha Sguta [NG 062 835] and noted that gabbro sheets intrude the Leverburgh metasediments. He also recorded clear evidence that the diorite postdates the gabbros from Sletteval Quarry [NG 059 855] where angular blocks of gabbro lie within hybridised diorite below the base of the pegmatite sheet. Gabbro also forms abundant xenoliths in the diorite, particularly in the Bleaval area. Horsley (1978) interpreted these as relict screens left as the diorite magma rose through them. Gabbro and amphibolite (?mafic gabbro) xenoliths are also present in the norite body although deformation here is greater (Horsley, 1978; Dearnley, 1963). Gabbro dykes indistinguishable compositionally from the lenticular gabbro bodies are found cutting across the early fabric of the three main igneous bodies. The relative ages of the anorthosite and diorite are not directly determinable. Horsley (1978), however, noted an anorthosite xenolith in norite on the shore section near Strond [NG 027 843]. The implied order of intrusion is thus gabbros, anorthosite, norite, diorite and tonalite.

Major intrusive bodies

Metagabbros

The gabbros constitute about 15 per cent of the exposed igneous complex, and typically form sheets which are abundant marginal to the anorthosite, norite and diorite. They attain their maximum development (350 m thick) in largely unexposed ground between Loch na Moracha and Bhoiseabhal (Beinn Tharsuinn region [NG 035 876]) and on the north-east side of Chaipaval Peninsula (e.g. around Leomadal [NF 978 934]). They also occur as large lenticular inclusions up to 3 km long and 0.5 km wide (e.g. Bleaval [NG 030 915]) in the diorite, and smaller pods and dykes are abundant throughout South Harris. Within the Langavat Belt there is a thick amphibolite which runs from Borvemore [NG 027 943] through Loch Langavat to Finsbay (see Chapter 3: Langavat Belt). It trends parallel to the diorite contact, and the present mapping has shown that it progressively transgresses the supracrustal lithological units (Map 5). It locally has a relict coarse-grained possibly igneous texture and shows compositional banding. Its margins are generally fine-grained laminated amphibolite. It is interpreted as a retrograded and highly deformed gabbro sheet related to the marginal gabbros which are abundant in South Harris (see also Palmer, 1971). Other smaller amphibolite sheets and pods in the Langavat Belt may also be representative of the gabbro suite. Dearnley (1963, p.282) has shown that the trace-element geochemistry of these amphibolites is very similar to the higher-grade gabbros to the south-west. Myers (1968, fig. 20) shows illustrations of two exposures where highly deformed amphibolites in the Langavat Belt still have discordant relationships with the metasedimentary banding (Borve, and the Loch na h'Uamha area).

The gabbros typically contain a metamorphic assemblage composed of garnet, clinopyroxene and plagioclase (An40-45) Hornblende and orthopyroxene may occur and accessory quartz, rutile, ilmenite, sphene and apatite are typically found. Secondary hornblende, biotite and scapolite are locally developed (Dearnley, 1963). The garnetiferous variety is coarsely crystalline with garnets up to 1 cm across. Where the gabbros lie within the granulite-facies diorite on Bleaval, [NG 030 915] Dearnley (1963) records the assemblage cpx-gt-and with minor feldspar replacement by scapolite. Accessories are rutile, sphene, quartz and ilmenite. The rare orthopyroxene-bearing metagabbros were noted by Davidson (1943) near Rodel and on Roineabhal and Beinn Tharsuinn by Dearnley (1963) and Witty (1975).

Both brown-green and blue-green hornblende occur in the metagabbros, the former a product of granulite-facies metamorphism on possible primary igneous hornblende, and the latter a product of low-grade retrogression mainly from clinopyroxene. The brown-green hornblende is a magnesio-hastingsite (sensu Leake, 1978) and the blue-green hornblende a magnesio-hornblende. Dearnley (1968) found that the blue-green hornblende was typical of epidoteamphibolite-grade metamorphism.

Garnet compositions from six localities are also given by Dearnley (1963); these range from Al44Py42Gr13 to Al70Py13Gr15.

Witty (1975) has studied the gabbro-rim of the anorthosite in detail and the following account is largely based on his work. The rim varies from 120 to 350 m thick and within it are three main rock types: banded ultramafics and mafic gabbros occurring as pods within quartz-gabbros; the quartz gabbros themselves, which make up 55–60 per cent of the rim; and garnet-biotite-quartz paragneisses which occur as strips and lenses up to 270 m long and 18 m wide. The paragneisses are partially assimilated by anorthosite where they lie along the contact with both the anorthosite and gabbro, but junctions between paragneisses and quartz-gabbro are sharp.

Banded ultramafites and mafic gabbros

These rock types show compositional banding or layering in the central parts of lenses, but are amphibolitised and foliated around their margins. There is a complex inter-banding of gabbro, paragneisses and amphibolitic ultramafite to mafic gabbro. Paragneiss inclusions also occur within the mafic gabbro pods, either oriented parallel to the compositional banding, or near the margins, parallel to the deformation fabric in the amphibolite.

Although the observed features and mineralogy of the rocks are largely a product of granulite-facies metamorphism, some relict primary igneous textures remain. The ultramafic rocks are typically clinopyroxenites, garnetites and garnet hornblendites and the mafic gabbros contain the assemblage cpx-gt-pl-hbl. Because of the high metamorphic grade, the rocks appear to be more basic than would have been indicated by their original mineralogy. This results from feldspar consumption at the site of new amphibole and garnet growth. The original bytownite is recrystallised and partially consumed but remains apparently compositionally unchanged by the granulite-facies metamorphism. The banding, described in considerable detail by Witty (1975), occurs as both large- and small-scale rhythmic structures. These units are best seen on Ha-cleit [NG 032 873] and north of Abhainn Easean Chais around [NG 029 866], where they show that the sequence is inverted. The units, here shown as right way up, may be summarised as follows:

The major mineral assemblages in these units are as follows:

Ultramafites–gt-cpx, cpx-hbl, gt-cpx-hbl-il-mt, gt-cpx-opx-(hbl); epidote, sphene and symplectite (plagioclase-hornblende intergrowth after garnet) are also common.

Mafic gabbros–pl-gt-cpx-hbl-il-mt, pl-(amph); epidote and symplectite are again common, and apatite, rutile, sphene and rarely biotite are also found.

As in the gabbro outside the rim complex the hornblende in these rocks is a magnesio-hastingsite (sensu Leake, 1978). It commonly encloses clinopyroxene and garnet in a poikilitic texture. Although its composition is metamorphic (Table 12) it forms fine igneous-like banding and larger 'meshwork' textures, and much of it was probably initially of primary igneous origin. Clinopyroxene lies in the augitesalite range (see (Table 12) for representative analysis) and the rare orthopyroxene has a composition of En65. Orthopyroxene is absent from the mafic gabbros, reflecting their low total Fe content (5-12 per cent, rarely up to 14 per cent). Garnet is weakly zoned and dominantly of almandinepyrope composition (Table 12), showing a range from Al33Py44 to Al42Py28. The grossular component ranges from 9 to 29 per cent but the compositional range does not simply relate to rock type. Where calcic clinopyroxene is abundant the grossular content is generally low, and the almandinepyrope ratio apparently increases as the Fe/Fe + Mg ratio increases, but more slowly.

Plagioclase feldspar always exhibits lamellar twinning, and has a composition ranging from An85 at the base of mafic gabbro units to An46-58 at the top. Weak zoning occurs and grain boundaries are smooth and curved with 120° triple junctions common. H2O values are extremely low in the calcic plagioclase (less than 0.03 per cent) but rise to 0.12 per cent in andesine. Locally plagioclase is altered to scapolite.

Epidote is common and may be found forming en-echelon planar veins or irregular vein networks.

Quartz gabbros

Witty (1975) records these as more uniform rocks, locally massive but more generally with a penetrative fabric. Included quartzose paragneiss lenses are particularly abundant. The rocks range from mafic gabbro with up to 4 per cent modal quartz and about 50 per cent plagioclase (An53-56) through to quartz gabbros and melagabbros with 4-12 per cent modal quartz, 15- 50 per cent plagioclase (An14-30) and minor microcline. Mineral assemblages include pl-cpz-opx-hbl-qz-bi-il, pl-cpx-hbl-gt-qz-bi-il and mic-pl-hbl-bi-il. Apatite is a common accessory and symplectite (hornblende-feldspar after garnet) is ubiquitous. Mineral compositions are similar to those of the mafic gabbroultramafite sequence, both being metamorphic. Orthopyroxene (En62) was only found in two slides by Witty (1975). The oligoclase shows faint albite twinning, undulatory extinction and bent twin lamellae. Recrystallisation has occurred widely, and biotite has grown at the expense of hornblende during Laxfordian retrogression to the lower amphibolite facies.

Gabbro bands are also abundant in the outer (i.e. lower) parts of the anorthosite body (see Meta-anorthosite section, p.53). Palmer (1971) has recorded discordant mafic gabbro dykes cutting the larger intrusions (see also Dearnley, 1973; Witty, 1975) but sharing a common chemistry and petrology with the larger gabbro bodies. These are further discussed in the section on minor intrusive bodies (Mafic gabbro dykes and sheets p.65).

Geochemistry

Witty (1975) analysed 54 gabbros from the rim sequence of the anorthosite and 9 mafic gabbros from the Beinn Tharsuinn sheet. Both Dearnley (1963) and Palmer (1971) analysed three gabbros. Average values for the ultramafic to mafic gabbro sequence, the quartz gabbros, the mafic gabbros of Dearnley and Palmer and discordant mafic gabbro dykes in the anorthosite (Witty, 1975) are given in (Table 13). The geochemistry of the mafic gabbro-ultramafite sequence corresponds closely with that of layered basic intrusives, such as the mafic olivine ferrogabbros of Skaergaard (Wager and Brown, 1968). Variations in the Fe/FeO + MgO ratio in the mafic gabbro-ultramafite sequence, for instance, are typical of differentiation trends found in gravity-stratified basic intrusions. This suggests that the early metamorphism (granulite facies) has been largely isochemical with element equilibration restricted to a few millimetres.

The ultramafic-banded gabbros are highly undersaturated with silica: they are ol- and ne-normative (Table 13). CaO values are high (up to 45 per cent normative anorthite) and yet the Sr value (88 ppm) is extremely low. Both Cr and Ni are also anomalously low considering the basic and Fe-rich nature of the rocks. The normative and observed compositional values of feldspar for these rocks are very similar, unlike those in the anorthosite.

The quartz gabbros show higher total alkalis and SiO2, and lower CaO and Al2O3 contents than the mafic gabbros. (Figure 9) (after Witty, 1975, p.206) shows the compositional fields of the two gabbro types on the AFM plot together with nine quartzose paragneiss specimens. The quartz gabbros are most probably derived by assimilation of the paragneiss by the parental mafic gabbro. The more quartz-rich bands remain as relict lenses.

It is unfortunate that so few trace-element data are available for the South Harris rocks, for trace-element variations are well documented for the layered basics of the Skaergaard intrusion (Wager and Brown, 1968). The high V values (600 and 250 ppm) quoted by Dearnley (1963) for the metagabbros from Ha-cleit and Bleaval suggest that the gabbros at present exposed represent a magma that was 70-80 per cent solidified, which is in accord with the low Sr, Ni and Cr values.

By correlating variations of different elements (e.g. CaO, SiO2, FeO, MnO and TiO2) in the mafic gabbro-ultramafite sequence, and using Pearce variation diagrams, Witty (1975) has deduced that the fractionating phases in the original magma were Ca-rich clinopyroxene (En35Fs12Wo48), ilmenite (for ultramafics) and plagioclase (An85). Analyses of ten clinopyroxenes from the metagabbros show an average composition of En40Fs14Wo44, very close to the theoretical value (cf. anorthosite).

Meta-anorthosite

The anorthosite forms a well-exposed wedge-shaped mass outcropping over about 5.2 km2 around Roineabhal (460 m) in southern South Harris (see Map 5). The following sections are largely taken from Witty's detailed account (1975). The anorthosite forms the core of a major fold, tight to isoclinal and antiformal to sidewards-closing; its axis is sub-vertical or plunges steeply to the north-west. The overall geometry and nature of the anorthosite show that beds become younger as we go towards its centre. Exposure at present allows mapping of a planar cross-section of the intrusion. Although the original mineralogy has been largely modified by early granulite-facies metamorphism many 'igneous' features and geochemical trends can still be recognised. Within the anorthosite there are numerous gabbro bands and schlieren, and the body may be divided into three.

Lower zone of banded anorthosite

The lower zone is 70 m thick where attenuated on the fold limb but 600 m wide in the hinge area. It is made up of fourteen gravity-stratified gabbro-anorthosite units individually 3 to 40 m thick, which show that the sequence youngs towards the centre of the anti-form and is inverted. The banding is reflected by the geochemistry, which shows an overall increase in normative feldspar, an increase in the FeO/FeO + MgO ratio following an iron enrichment trend, and an increase in the differentiation index inwards from the inverted base. Clinopyroxene is virtually absent from much of this lower zone, increasing to approximately 10 per cent in the gabbros of the uppermost units. In the lower four-fifths of the zone mafic gabbros (hbcpx-gt-pl-il) form layers 1 -30 m thick in the gabbros and gabbroic anorthosites. They are less frequent upwards. Plagioclase composition ranges from An37-39 (white plagioclase) in the mafic gabbros to An44-48 in the gabbros and An64-68 (pink plagioclase) in the anorthosite 'tops'.

Middle zone of banded anorthosite

This zone is well exposed on the north-western flanks of Roineabhal. It consists of rhythmically banded 'anorthosite' and gabbro with eight major gabbro bands, each 55-120 cm wide, recorded in addition to numerous smaller ones (each less than 30 cm). Clinopyroxene is abundant. The thicker gabbros are generally coarse grained with patches of hbl-white pl-cpx and gtcpx-pink pl assemblages. In the anorthosite (sensu stricto), bands of circular or oval clots of gabbro up to 30 cm across are present (e.g. 230 m east-north-east of Roineabhal summit) in zones 8 to 20 m below the base of the gabbro bands. Feldspars range in composition from An48-52 in the gabbros to An64-69 in the gabbroic anorthosites and An69-72 in the anorthosites.

Upper zone of schlieren 'anorthosite'

This division is best seen on the south-east slope of Roineabhal. Gabbroic anorthosites comprise 65-70 per cent of the rocks, but bands of massive pure anorthosite are also abundant. The gabbro schlieren range from diffuse to well defined, and range from near spherical forms 20 -30 cm across to oblate ellipsoids (axial ratio up to 7:1). Dearnley (1963) records that the schlieren are of near-uniform size on a local scale. Mafic minerals are largely absent from the adjacent anorthosite. There are also large patches of garnetiferous anorthosite up to 100 m across.

Petrography

Thin section examination of the 'anorthosite' within the core of the major fold shows coarse-grained equilibrium textures characteristic of granulite-facies assemblages with bent twin lamellae and only minor undulose extinction. Assemblages include gt-hbl-cpx-pl-(qz)-(il), hb-pl-(qz)-(il), hbl-cpx-pl-qz, and gt-pl-qz.

Orthopyroxene was found by Witty (1975) in only one locality. Plagioclase is the dominant mineral; it varies in colour in the field from pink (1 per cent FeO on average) to white (0.1 per cent FeO on average). The pink plagioclase contains tiny exsolved hematite rods. Compositional zoning is rare in both varieties, and where present the variation never exceeds 5 per cent An from the more calcic centre to the rim.

In the fold limbs Laxfordian effects strongly overprint the earlier granulite-facies mineralogy. There, plagioclase recrystallisation and myrmekite formation at grain boundaries are common features, together with grain fracturing and bent twin lamellae. Scapolite is abundant; Davidson (1943) gives optical data implying a composition of about Me60 (60 per cent meionite). Witty (1975) cites a single electron-probe analysis of Me65 (the co-existing feldspar is An56Ab40Or2). These values are close to those obtained from the highest-grade Younger Basics (Chapter 10: Feldspar mineralogy).

Anorthosites from shear zones show medium-grained polygonal textures. Retrogression to the amphibolite facies caused garnet to break down to a hornblende-plagioclase symplectite and clinopyroxene to a pale green hornblende. Where shearing is locally intense, clinozoisite, zoisite, chlorite, biotite, calcite and epidote may be present (Witty, 1975). Retrogression was noted by Witty to relate particularly to Late-Laxfordian shear zones; both retrogression and shearing increasing in intensity south-eastwards from Roineabhal.

At the south-eastern end of the anorthosite body saussuritisation related to movements along the Outer Hebrides Thrust Zone has affected most rock types. Below the major thrust there is only minor fracturing, in contrast to the rocks above. Irregular masses of rock remain unaffected within this altered zone, and the transition from saussuritised to unaltered rock is less than 60 cm wide. In saussuritisation plagioclase changes to fine-grained zoisite, paragonite and quartz. Mafic minerals are altered to ep-clz-chl-(zo)-(trem).

The degree of alteration increases south-eastwards with virtually no original feldspar remaining at the thrust zone. The grain size is further reduced in the crushed gneisses above the thrust, and paragonite and quartz are more abundant. Late-Laxfordian alkali pegmatites intruded into the rocks prior to thrusting have been deformed into thin bands and lenses. A chemical consequence of saussuritisation is a loss of alkalis, presumably by leaching, with a resultant excess of Al2O3 and SiO2 (see (Table 15)).

Mineral composition

Clinopyroxenes from the anorthosite show a range from calcic augite/salite in the gabbro and anorthosite units of the middle zone to diopside/salite in the upper zone. The composition of the clinopyroxenes was shown by Witty (1975, pp.280–282) to be typically metamorphic and characteristic of the granulite facies (representative compositions are given in (Table 14)). The Fe/Fe + Mg ratio for clinopyroxene increases broadly with that for the host rock although it is always less than the host rock ratio. Orthopyroxene crystals in the mafic gabbro layers are uncommon; they have low Al2O3 (2.27 per cent) and high FeO + MgO (45 per cent) values, which places them too in the metamorphic field (Bhattacharyya, 1971). Garnets from the anorthosite are more calcic than those from the norite and diorite (Table 14). Average values are Al45Py31Gr24 but.up to Gr33 is present in some garnets. Garnets co-existing with clinopyroxene have a grossular content roughly inversely proportional to the amount of clinopyroxene in the rock. Zonation is common, and Witty records a decrease in pyrope content towards the rim (end-member variations are Py24.8 to Py14.6 and Py35.8 to Py25), although the grossular content remains constant across the crystal. The Fe/Fe + Mg ratio for the garnets varies inversely with that of the host rock, being much higher in iron-poor rocks and decreasing as the rock becomes more iron-rich. The ratios for garnet and host rock are about equal at a value of 0.62.

Amphibole analyses (nine specimens analysed by Witty, 1975) fall into distinct groupings which range from ferroan pargasite and ferroan pargasitic hornblende (sensu Leake, 1978) in the mafic gabbro bands, through to edenite in the gabbro horizons of the lower zone anorthosite. During Laxfordian amphibolite-grade metamorphism retrograde amphiboles were formed by marginal alteration of clinopyroxenes in the gabbros of the lower zone. These amphiboles have much lower total alkali and lower Al values; they are magnesio-hornblendes (sensu Leake, 1978). Representative analyses are given in (Table 14).

Witty (1975) used co-existing pyroxene pairs from the orthopyroxene-bearing mafic gabbro and the banded ultramafic gabbro units of the rim metagabbros to estimate palaeotemperatures. The resultant values of 1135°C to 1315°C (± 70°C) were obtained using the method of Wood and Banno (1973). These apparently reflect igneous equilibration temperatures rather than metamorphic values.

Geochemistry

Witty (1975) presented 25 XRF analyses of the anorthosite and a further 25 are available in an unpublished report for the Highlands and Islands Development Board by Robertson Research Ltd (McKenzie and others, 1973). In addition Palmer (1971) gives nine XRF analyses and Davidson (1943) two wet chemical analyses.

Anorthosite geochemistry largely reflects the dominance of plagioclase feldspar in the rocks with high Al2O3 values (24-34 per cent). Mafic gabbros, gabbros and anorthositic gabbros are generally ol-normative, and gabbroic anorthosites and anorthosites are Q-normative. This is in agreement with the presence of modal quartz. Palmer (1971) suggests that many of the rocks are ne-normative, but discrepancies in SiO2 totals between Witty and Palmer perhaps throw doubt on the validity of Palmer's results. The gravity-stratified units show variations in the FeO/FeO + MgO ratio which still preserve typically igneous differentiation trends. The ratio increases from the base of the unit (0.538) up into the mafic gabbro (0.553), then after a subsequent decrease (0.505), increases to the anorthosite top of the unit (0.654). The ratios also show an overall average increase from about 0.55 (± 0.1) in the gabbroic basal 'anorthosite' to about 0.85 ( ± 0.2) in the uppermost exposed anorthosites (Witty, 1975, p.204).

The South Harris anorthosite complex is notably calcic by comparison with other anorthosites (e.g. in the Adirondacks, and in Norway), and the mineralogy of its lower gabbros has been markedly changed by the granulite-facies metamorphism. The growth of garnet and change in composition of other mafic minerals, all at the expense of feldspar, have served to make the rocks appear more mafic than the normative chemistry implies. Up to 50 per cent normative feldspar may not be reflected in the mode (Witty, 1975, p.216).

Representative analyses of the different members of the anorthosite complex are given in (Table 15) together with the average composition. The small amount of trace-element data precludes detailed comparison with establised trends for other layered basic bodies. Witty's (1974) values relate dominantly to the more mafic parts of the anorthosite but Dearnley's (1963) and Palmer's (1971) analyses (twelve in total) are in good agreement. Cr values of 110 ppm from Witty are probably too high, and Palmer's (1971) value of 22 ppm is probably more representative. High Ni values (average 40 ppm) reflect the original chemistry of the anorthosite which here is best expressed in the norm. Rb values (2 ppm) are generally very low, although Witty records values up to 54 ppm, which may be caused by introduction of Rb locally along shear zones. Ba values of 100 ppm are low and follow from the low K2O value of the anorthosite since Ba substitutes for K. Both elements show preference for the residual liquid in fractionating igneous bodies. Sr values (average 266 ppm) are generally low, ranging from about 130 ppm in the mafic gabbros to over 400 ppm in the upper-zone anorthosite. Considering the dominance of calcic plagioclase in these rocks the Sr values are surprisingly low. Smith (1974, p.82) has shown that the Sr content of plagioclase decreases markedly with increasing An values because Ca discriminates against Sr (Sr content is a maximum at about An35-40).

In the anorthosites Na2O correlates with SiO2, Al2O3, FeO, MnO and MgO and there is a good correlation of FeO with MgO (cf. mafic gabbros and ultramafics rimming the 'anorthosite'). The mafic gabbro bands in the lower-zone anorthosite show affinities with the quartz gabbros and mafic gabbros of the rim, but the gabbros within the anorthosite show more affinity with the anorthosite (s.s.). Pearce variation diagrams were used by Witty (1975) to show that for these rocks with about 80 per cent normative plagioclase, fractionation of plagioclase with a composition of An71 (average modal composition An69.9) dominated crystallisation (cf. gabbros of the rim complex). Minor clinopyroxene with a Fe/Mg:Si ratio of 0.5 was determined as the only fractionating mafic phase (average measured ratio 0.606). The fractionating clinopyroxene composition is calculated by Witty (1975) to be En22Fs23Wo50 as opposed to the average analysis (five values) of En38Fs23Wo49. Witty used the detailed chemistry of a complete gravity-stratified unit (unit 3 in lower-zone anorthosite) to show that from an initial gabbroic-anorthosite magma both clinopyroxene cumulates (mafic gabbros) and plagioclase cumulates would result, with the remaining residual liquid corresponding to basal gabbroic anorthosite (subsequently gabbro). Bulk-rock compositions for the different members of unit 3 are exactly as predicted by this model.

Metanorite

This igneous body crops out along much of the south-west coast of South Harris; its margins are strongly deformed by many shear zones largely of Laxfordian age. Geophysical data and outcrops on several islands suggest that much of the Sound of Harris is underlain by the norite body. Where unaffected by later amphibolitisation it is a uniform dark grey basic orthogneiss with a typical 'glassy' charnockitic appearance. The norite crops out particularly well on the Renish Peninsula [NG 042 826], on the south-west side of Chaipaval (Toe Head Peninsula), at Strond [NG 026 827] and in north-east Ensay.

Although Jehu and Craig (1927) mentioned igneous rocks associated with sediments in the belt bordering the Sound of Harris, Dearnley (1963) regarded the contacts as thrusts and termed the meta-igneous body a pyroxene-granulite. He suggested that it predated the main gabbro-anorthositediorite suite as the internal foliation is apparently cut by basic dykes (mafic gabbro) belonging to that suite. The internal foliation is a poorly defined alignment of pyroxenes which is destroyed near the margins by the amphibolitefacies shearing. However, the contact relationships between this body and its surrounding metasediments resemble those of the other South Harris meta-igneous bodies. This factor allied to a compatible geochemistry strongly suggests that the body is a deformed and metamorphosed norite belonging to the South Harris Complex. This interpretation was initially proposed by Graham (1970) and Palmer (1971), and subsequently supported by Horsley (1978). A sequence of subvertical 3 to 5 m thick metasedimentary bands and a gabbro sheet, 100 m thick near Strond, but locally only 2 m near Kyles House, border the norite (Map 5). On the Isle of Ensay a zone of schistose amphibolite lies adjacent to highly retrograded and sheared norite. Retrogression to amphibolite and foliation development are also well seen at the northeast margin of the norite.

Both norite and diorite are intruded by dykes and veins (up to 10 m thick) of gabbro and pyroxenite. Both also contain gabbro/amphibolite xenoliths, and, as previously stated, Horsley (1978) noted an anorthosite xenolith in the norite near Strond [NG 029 842]. Metasediment xenoliths are also locally found in the norite (e.g. Rubh'an Teampuill [NF 970 913]).

The norite is considerably deformed internally, reflecting its more uniform response to the strain that in the diorite resulted in discrete shear zones, especially at its margins. The norite is not uniform in composition and these variations are reflected in its mineralogy. Almandine-pyrope garnets (up to 1 cm across) are abundant in the Strond-Renish Point section but absent elsewhere. In Ensay Palmer (1971) reports that the norite is rich in large orthopyroxene crystals.

The typical high-grade metamorphic mineralogy of the norite is as follows (Horsley, 1978): pl-cpx-opx-(hbl)-mt, and pl-cpx-opx-gt-(hb1)-mt. When retrograded to amphibolite facies (Laxfordian effects) the assemblage is: pl-hbl-qz-(bi)- (cum)-(ep)-(sph). This is similar to that found in the diorite but the proportion of hornblende to plagioclase is greater.

On Saghay More [NF 998 867] in the Sound of Harris typical outcrops of norite show a crude foliation on the outcrop scale but are homogeneous in hand specimen. The rocks consist of about 50 per cent plagioclase, generally perthitic or antiperthitic andesine (An43-50), with highly pleochroic pink-green orthopyroxene and pale green clinopyroxene in approximately equal amounts. Accessory magnetite, apatite and rutile are present, and quartz and garnet locally occur in small quantities (Dearnley, 1963). There is an inverse relationship between garnet and orthopyroxene in the norite. 'Primary' olive-green hornblende is also common, showing exsolved ore along the cleavage planes. Throughout the norite, twin lamellae in plagioclase are bent and broken, and zones of crushed feldspar and pyroxene are prevalent. Feldspar recrystallisation resulting from deformation has also been reported by Palmer (1971) and Dearnley (1963).

As in the diorite, retrogression is patchy but is generally more intense towards the margins of the intrusion. The nature of the retrogression detailed by Palmer (1971) is similar to that in the diorite. Clinopyroxene alters to hornblende, orthopyroxene to biotite, and 'primary' hornblende to a sieved pale green hornblende with no ore inclusions which in turn may be rimmed by small, pale brown biotites. Relict clinopyroxene typically persists after all the orthopyroxene has disappeared. In shear zones plagioclase is totally recrystallised and pale green actinolite is present together with hornblende and biotite. Tiny biotites define a new fabric, showing that conditions did not favour extensive grain growth.

Mineral composition

The chemistry of the minerals in the norite is similar to that of the minerals in the diorite. (Table 16) shows representative analyses of the three main metamorphic minerals (after Horsley, 1978). The clinopyroxenes of the norite plot on or adjacent to the salite–augite boundary and have slightly more CaO and SiO2 than those in the diorite. The orthopyroxenes of the norite have Fe/Fe + Mg ratios of 0.37 to 0.40, slightly lower than those of the diorite, but all fall in the central part of the hypersthene field. Analyses of garnet from the norite are not available, but optical determinations (ND = 1.785, sp. gr. = 4.01) by Dearnley (1963) indicate a composition of Al60Py34. Plagioclase compositions range from An28 to An33 with up to 2 per cent Or.

Geochemistry

A considerable number of whole-rock and partial analyses of specimens from the norite are available. Horsley (1978) presents thirteen analyses (eight from granulite-facies norites), Palmer (1971) a further nine and Dearnley (1963) one complete analysis and four alkali- and trace-element determinations. (Table 17) gives representative single and average analyses. Although the norites are geochemically similar to the diorites they nevertheless show the following differences:

  1. Slightly lower SiO2 and Al2O3 values and higher total alkalis (Na2O + K2O).
  2. Significantly lower FeO/FeO + MgO values (0.47 to 0.60) than the diorite (0.56 to 0.68).
  3. Differing normative feldspar proportions and compositions (57 per cent and An37 for the norite and 67 per cent and An48 for the diorite).
  4. Higher Ni, Cu and Cr values in line with the more mafic nature of the norite.

The geochemistry of the norite is similar to that of a basaltic andesite. Unlike the diorite which is invariably Q-normative the norite varies between from Q- and ol-normative. Using Pearce variation diagrams Horsley (1978) calculated that in the original fractionating magma the feldspar phase was An66-68 and that clinopyroxene and orthopyroxene were additional phases.

Horsley (1978) also gives abundant trace-element data for the norite, and Palmer (1971) and Dearnley (1963) provide supporting data. Sr, Ba, Zr and Y are notably high in the norite compared to the other South Harris intrusives. Ni values are high, lying in the same range as those for the gabbro suite, which is more mafic, and Cr values are even greater. Rb values are low, a feature of all component bodies of the South Harris igneous suite. Lower crustal rocks with a low K2O content (about 0.5 per cent) have K/Rb ratios of about 625, whereas such ratios for rocks in the South Harris Complex range from 846 for the norite to 2270 for the diorites (Horsley, 1978). Horsley interprets this as a primary feature with Rb being discriminated against by hornblende, plagioclase and potash-feldspar. Rb tends to be concentrated in the late liquid fractions as a result. In the granulite-facies norite and diorite K/Rb ratios correlate well with most major oxides, giving typical differentiation curves. This correlation does not hold for rocks amphibolitised during the Laxfordian, suggesting that there is widespread movement of K and/or Rb during high-grade hydrous metamorphism.

Horsley (1978), Witty (1975) and Palmer (1971) suggested that though the norite is an integral part of the South Harris Complex it was the product of a different parent magma from the one from which the diorite and anorthosite bodies were differentiated.

Dearnley (1963), Jehu and Craig (1925) and Kursten (1957) all noted the similarity between the norite in the Sound of Harris and the Corodale Gneiss of South Uist (Chapter 8). Both bodies are cut by mafic gabbro 'dykes' (garnet-clinopyroxene amphibolites), and show relict igneous features and granulite-facies mineralogies. The single analysis obtained during this work (see Chapter 8) suggests that the Corodale Gneiss has geochemical features intermediate between the norite and diorite but with higher total alkalis.

Metadiorite

The diorite is the largest single meta-igneous body exposed in South Harris (about 30 km2). The term diorite is adopted in this report in preference to tonalite or gabbro-diorite (Jehu and Craig, 1927; Dearnley, 1963) in view of the detailed mineralogical and geochemical work carried out on the body by Horsley (1978). Typically it is a light to dark grey massive rock with sporadic garnets. Locally the diorite is more gabbroic (e.g. Scarista foreshore [NG 014 939]), and pyroxene is the abundant phase (Palmer, 1971). Aligned pyroxenes may form a crude banding which Horsley (1978) and Palmer (1971) have interpreted as a dominantly igneous feature. The banding generally strikes north-east - southwest and dips about 75° to the north-west. This banding is best seen on Maodal [NF 998 905] and on the Bolaval-Scarasta ridge [NG 011 915].

The earliest textures seen in the diorite are coarsely granoblastic, with quartz and andesine forming straight grain boundaries and 120° triple junctions. Clinopyroxene forms large subhedral pale green grains with irregular birefringence and exsolution lamellae, whereas orthopyroxene forms smaller anhedra] grains, strongly pleochroic pink to pale green, and rimmed by secondary biotite (Palmer, 1971). Olive-green hornblende is also present as large euhedral grains, locally with exsolved ore along its cleavage. Large irregular garnets are common. These mineralogical and textural features are products of the early granulite-facies metamorphism which most authors suggest overlapped with the period of igneous intrusion (Horsley, 1978; Palmer, 1971; but cf. Dearnley, 1963).

In the more extensively deformed parts of the diorite, quartz and feldspar form a sutured intergrowth with undulose extinction, deformation bands and bent feldspar twins. Clinopyroxene has broken down to pale bluish green hornblende, and the olive-green hornblende has also changed its composition. The quartz and plagioclase have been progressively strained and recrystallised to medium-grained mosaics of unstrained grains ('mortar texture').

Dearnley (1963) divided the diorite into four discrete metamorphic zones ranging from granulite/upper-amphibolite grade down to greenschist grade. The high-grade zone in the Gleann Horsacleit-Bolaval region he characterised by the assemblage di-hy-and(An50)-qz. Dearnley attributes minor garnets in Zone II diorite on the Mas Garbh ridge [NF 044 878] to slight contamination with nearby metagabbro. No mention was made of those garnets more widely distributed in the south-west part of the diorite (see Map 5). The grade becomes progressively lower to the northeast as the diorite margin is approached until in the strongly foliated marginal zone the typical assemblage is hb-ep-bi-olig(An30)-qz. These mineralogical changes are attributed by Dearnley to Laxfordian deformation and attendant retrogression, the effects of which are generally manifested as shear zones. Work by Palmer (1971) and Horsley (1978) showed, however, that such simple divisions are untenable. Dearnley is correct that high grade (gt-cpx) assemblages do lie mainly adjacent to the southwest margin of the metadiorite and opx-cpx assemblages are abundant in the Bolaval- Scarasta region (Map 5). Although there is a general increase in deformation and retrogression to the north-east such effects are best described as patchy and localised. The following assemblages are typical of the granulite-facies assemblages in the diorite:

pl-cpx-opx-(hbl)-(mt)-(qz), pl-cpx-(gt)-(hbl)-mt-qz.

The changing mineralogy on retrogression may be summarised as follows:

p61 Insert unnumbered diagram

A typical amphibolite-grade assemblage is hbl-pl-qz-bi(cum)-(ep). Pyroxene and garnet change to hornblende-quartz and hornblende–plagioclase (symplectite) intergrowths respectively. The proportions of the various minerals changes with retrogression as follows (Horsley, 1978):

Quartz 0.1% to 3.82%

Feldspar 66.5% (An42-46) to 55.3% (An34-41)

Hypersthene 17.0% to 1.9%

Hornblende 2.9% to 25.8%

Dearnley (1963) records values as low an An35 for plagioclase in the sheared marginal zone adjacent to the Langavat Belt. Deep red garnets vary in abundance but are up to 4 cm across; compositions lie in the almandine-pyrope field.

Numerous small basic and ultrabasic dykes and bodies intrude the diorite (see the section on Minor intrusive bodies p.62). The gabbro inclusions and large lensoid masses have already been referred to (see the section on Metagabbros earlier in this chapter). The intrusion is also cut by quartz-plagioclase and hornblende-plagioclase pegmatite veins. The abundant gabbro xenoliths are surrounded by contaminated garnet-rich diorite (in parts with a secondary hornblende-plagioclase rim), and clinopyroxene-rich diorite. Adjacent to the gabbro-diorite contact between Loch na Moracha and Greabhal [NG 004 891] such effects extend for up to 20 m into the diorite (Dearnley, 1963). Within many of the larger gabbro xenoliths diorite streaks are common. The presence of amphibolite xenoliths in the north-east part of the diorite reflects the more highly deformed and retrograde nature of this part of the intrusion.

Mineral composition

Horsley (1978) determined 24 mineral compositions from the diorite using the electron probe. Five representative values are given in (Table 18).

The clinopyroxenes are clearly metamorphic in origin, with a higher Wo (CaSiO3) component than is generally found in dioritic igneous bodies. They lie mainly in the calcic augite field near the augite/salite boundary. The orthopyroxenes are also of metamorphic origin (see Fleet, 1974), being hypersthenes with Fe/Mg ratios of 0.4 to 0.45. The Wo component is uniformly low (0.9 to 1.3 per cent) compared with igneous orthopyroxenes. The hornblendes lie in the ferroan pargasite and magnesian hastingsite fields of Leake (1978). The high TiO2 values (1.97 to 2.12 per cent) are typical of the hornblende-granulite-facies (Raase, 1974). The octahedral AlVI component is uniformly low (0.48 to 0.54), and Horsley (1978) suggests that this may be a pyroxene/hornblende partitioning effect. A single specimen of garnet shows compositional zoning from Al55.Py24.1 at the core to Al57 , Py21.1 at the rim. Use of Wood and Banno's (1973) geothermometer for co-existing garnet and clinopyroxene gives equilibration temperatures of 697°C for the rim (Horsley, 1978). The Fe/Mg ratios for individual mafic minerals in the diorite decrease from garnet through hornblende and orthopyroxene to clinopyroxene. Hornblende and orthopyroxene are transposed from the theoretical order. Plagioclases lie in the range An37-46 for the granulitefacies diorite dropping to Ar134-41 for amphibolite-grade rocks. The orthoclase component is low (1.5 to 2.3 per cent). Zoning is rare and is found when the feldspar margin is partially retrogressed; e.g. core An39 margin An35 (Horsley, 1978).

Geochemistry

Palmer (1971) presented nine XRF analyses of the diorite, Dearnley (1963) two wet chemical analyses, and Horsley (1978) 56 XRF analyses. These analyses show a high degree of uniformity and are similar to published values for diorites (Le Maitre, 1976). Horsley (1978) analysed 16 specimens from granulite-facies rocks, 19 from the granulite/amphibolite transition zone and 21 from amphibolite-facies rocks. Averages for these three types and for Palmer's (1971) analyses are listed in (Table 19). Horsley (1978) noted some consistent differences between his analyses and those of Palmer (1971), notably a 1 to 2 per cent discrepancy in Al2O3 values. Geochemically the diorite is similar to a calcalkaline high-alumina andesite/basaltic andesite, having a normative mafic index of 30 to 40 (Horsley, 1978). (Figure 10) shows the compositions and trends of the diorite analyses in comparison with the other units of the South Harris Complex. Nickel (20–30 ppm) and chromium contents are low. Dearnley (1973) gives about 30 ppm for Cr, and Palmer (1971) gives a range from 15-195 ppm with an average of 63 ppm. The low Ni and Cr values are a product of early enrichment of Ni and Cr in the ultramafics and progressive depletion on differentiation. With amphibolitisation there is a significant increase in Ba (following an increase in K2O ), Rb, Zr and In granulite-facies diorites the Sr/Ba ratio decreases regularly (1.16 to 0.55) and the Rb/Sr ratio increases with fractionation mainly because Sr decreases. The K/Rb ratio correlates well with most major oxides in the granulite-facies rocks but gives a flat trend in amphibolitised diorites. In the granulite-facies rocks inter-element correlation is highly significant for 31 per cent of the matrix values, particularly between felsic elements and major oxides. Mafic elements and major oxides show very low correlation coefficients. The amphibolite-facies rocks do not give good differentiation trends except for the more immobile elements: Al2O3, CaO and Zr. Horsley (1978) has documented in detail the igneous processes likely to lead to the observed geochemical trends and concluded that igneous trends are preserved in the granulite-facies diorites but are strongly modified in amphibolite-grade rocks. The igneous trends show that plagioclase (An63-67) fractionation dominated the crystallisation sequence although an overall decrease in FeO, MnO and MgO with differentiation suggests that minor clinopyroxene fractionation has also occurred. The subsequent granulite-facies metamorphism resulted in a redistribution of CaO from plagioclase into orthopyroxene, clinopyroxene and garnet.

Metatonalite

Using field and geochemical criteria Horsley (1978) has distinguished an elongate body of tonalite within the diorite and adjacent to the metasediments of Bagh Steinigie [NG 019 938]. The exposed body is 9 km long and 100 to 150 m wide (Map 5). The rock is generally deformed and retrograded, and contains more free quartz than the diorite. Rarely microcline is present Horsley attributes the recorded presence of calcite to migration of abundant CO2 and some fluid from the adjacent marble-rich metasediments. Calcic scapolite replaces feldspar in this unit (Dearnley, 1963) and alteration to sericite is abundant.

Geochemically the tonalite shows features typical of a highly differentiated product of a diorite magma. Na2O values reach 5.34 per cent (but K2O is only 0.92 per cent) and Sr and Ba average 1356 ppm and 900 ppm respectively. Rb values are very low, however, suggesting that Rb depletion may be a primary feature of the whole South Harris Igneous Complex. FeO/FeO + MgO values reach 0.65, showing some Fe fractionation relative to the diorite. Horsley (1978) attributes the origin of the tonalite to filter pressing of the residual liquid of the diorite.

Minor intrusive bodies

The detailed work carried out over the past 20 years in South Harris, notably by Dearnley (1963), Livingstone (1963), Witty (1975), and Horsley (1978), has shown that minor in trusions, particularly those of a mafic and ultramafic nature, are abundant. Several 'suites' of intrusions, such as the mafic gabbros, are particularly widespread. Others such as the anorthosite/ultramafic breccias are necessarily restricted to or associated with particular igneous bodies. Early ultramafic lenses form a linear zone within the Langavat Belt metasediments. These are described in detail at the end of Chapter 4. The ultramafic rocks in the Leverburgh Belt, however, are thought to relate to the South Harris Igneous Complex, and are described in the following section.

Ultramafic rocks of the Leverburgh Belt

These rocks are best developed in the vicinity of Rodel, where they occur in numerous small lenticular bodies up to 150 m long and 45 m wide. Davidson (1943) has described their petrography in some detail. He termed them edenite saxonite and edenite peridotite. Subsequently Livingstone (1963, 1976b) provided detailed whole-rock and mineral analyses and discussed their paragenesis. He distinguished two main rock types: spinel-amphibole lherzolite and garnet-amphibole lherzolite.

In thin section Livingstone (1976b) shows the spinel lherzolite to consist of pleochroic bronzite, pargasitic hornblende (cf. Davidson, 1943), clinopyroxene and partially serpentinised olivine (Fo84). Magnetite and hercynite are common accessories. Textural studies led Palmer (1971) to suggest that the spinel may be a product of olivine resorption, the mineral assemblage resulting from the incomplete reaction

ol + pl H2O opx + cpx + spinel + amph. (reversible reaction)

The garnet-rich lherzolite on the east side of Loch Rodel contains orthopyroxene, amphibole (cummingtonite), olivine (Fo84), kelyphite (plagioclase + amphibole), spinel, magnetite and rare small garnets 0.5 mm across (Livingstone, 1967). Subsequently Livingstone (1976b) found that the garnet composition (Py55) and Cr levels (1000 and 100 ppm in two garnets analysed) were appropriate for crust-generated garnet peridotites in gneissic terrains. He showed that the bulk composition of the garnet-amphibole lherzolite was indistinguishable from the spinel-bearing variety.

Although previous authors have commented on the lack of clear contact relationships between the ultramafics and adjacent rocks, during the course of this survey a massive discordant ultramafic dyke was found at the head of Loch Rodel [NG 0425 8345]. The following assemblage was recorded in thin section (S73098): ol-opx-cpx-hercynite-mt-serp. The hercynite (green Fe2+-Al spinel) is notably abundant. Although this mineralogy does not match that of the specimens cited by Livingstone (1976b) the body is undoubtedly of igneous origin. This new data supports the hypothesis that these ultramafic bodies are genetically part of the South Harris igneous suite. Livingstone (1963) notes that moderate totals of Na2O + K2O, and high CaO and Al2O values also suggest such a relationship, with the ultramafics representing accumulative admixtures of olivine, pyroxene and plagioclase of the igneous suite. Normative calculations show olivine to lie in the range Fo86-93 and plagioclase in the labradorite/bytownite field. This also explains the observed gradation in some outcrops between gabbro and ultramafite.

Livingstone (1976b) suggests that the garnet lherzolite may result from locally developed tectonic overpressure or, more probably, that the rocks were metamorphosed under conditions close to the boundary between the spinel lherzolite and garnet lherzolite fields.

Hornblendites

Apart from the hornblendite lenses and bands associated with the Langavat ultramafites, dykes and pods of hornblendite are also found widely in the Leverburgh metasediments. Palmer (1971, pp.225–226) explains their presence as a function of metamorphic grade, for they pass locally into gabbro. They are composed almost entirely of pargasitic hornblende. Large hornblende pods, now saussuritised, also intrude the anorthosite complex where it lies within the thrust zone, and a similar but altered body occurs in the rim gabbros immediately south-west of Sletteval Quarry [NG 059 855].

Thin (less than 20 cm) hornblendite dykes locally with weak internal fabrics subparallel to the dyke margins, trend approximately east–west and cut the compositional banding of the anorthosite (Witty, 1975). As well as the predominant amphibole (dark green pargasite), minor biotite, plagioclase, microcline, quartz and epidote are present. Rare large clinopyroxene grains are also noted. Witty (1975, pp.175–177) has also recorded in detail the lenticular hornblendite bodies in the thrust part of the anorthosite. He ascribes their distribution to intrusion during an initial thrusting episode and related saussuritisation. He records that anorthosite inclusions in the hornblendite are saussuritised, but the hornblendite itself is unaffected. BGS work in Lewis on the thrust zone rocks has shown that hornblende-rich basics are commonly resistant to thrust effects and have been little affected by the late retrogression. In view of this it seems probable that the hornblende pods have formed a locus for thrust development, rather than having formed as a result of the thrusting as suggested by Witty (1975, p.177).

Mafic gabbro dykes and sheets

Intrusions of mafic gabbro are abundant throughout South Harris, particularly within the meta-igneous bodies and the Leverburgh Belt. They are the oldest intrusives in the anorthosite; in the lower zone, although locally discordant they generally form regionally concordant dyke-like bodies (Dearnley, 1973). They are typically 1 to 2 m wide but in places attain 7 m. More rarely they form net-veins. In the upper parts of the anorthosite the dykes tend to be more discordant; in the Lingara Bay area [NG 062 847] large discordant gabbroic pods (up to 170 by 30 m) were noted by Witty (1975). Davidson (1934), Palmer (1971) and Dickinson (1974) have all described gabbroic dykes from the Rodel district that are commonly hypersthene-bearing. They range in size from a few centimetres up to 50 m wide, and locally show gradation to more ultramafic members. They may also be retrograded to varieties richer in biotite and chlorite.

Graham (1970) noted that amphibolite dykes are common within the norite and diorite; in both intrusions they may be folded, with development of a new axial-plane foliation. The dykes may be boudinaged, but more generally they merely lie oblique to the 'igneous' foliation of the host rock. These differences were attributed by Graham (1970) to various differences in competence between host and amphibolite, but they may also reflect variations in the intensity of later deformation or the presence of several generations of mafic dykes. The presence of gabbroic dykes cutting the diorite which in turn contains gabbroic xenoliths, e.g. on Greabhal [NG 004 891], shows that basic minor intrusives both predate and postdate the major intrusive bodies.

Where mafic gabbros occur as sheets, for example adjacent to Loch Thorsagearraidh [NG 047 842], Witty (1975, p.133) has recorded the presence in them of internal compositional banding commonly showing a considerable angular discordance to the sheet margins. Their contacts with meta-sediments are marked by interbanding of the two rock types. The internal banding in the sheets is caused by the presence or absence of plagioclase feldspar; within the thick mafic gabbro bands elongate clots or thin lenses of coarse-grained garnetclinopyroxene aggregates occur.

The gabbros generally have a cpx-amph-pl-il-(gt) assemblage with garnet locally absent. Hypersthene-bearing norite dykes are also present. The plagioclase is andesine but ranges from An37 to An49, dependent on the location and degree of retrogression. Quartz is a common accessory mineral. Dearnley (1963, 1973) and Witty (1975) have described members of this suite and found that they are mineralogically and chemically indistinguishable from both the Scourie Dyke (Younger Basic) suite and the basic layers of the banded gabbros of the lower-zone anorthosite. It is also significant that the present study has failed to separate chemically the Older and Younger Basic suites (Chapters 4 and 10).

Hornblende-plagioclase pegmatites

A swarm of hornblende-plagioclase pegmatites 1 to 90 cm thick was noted by Davidson (1943) to cut across the norite in the Strond/Renish Point area. Witty (1975, p.132) noted that the pegmatites cut across the basic and pyroxenite dykes. Similar pegmatites are also seen in the Bàgh Steinigie area and north of Chaipaval on the Toe Head Peninsula where they cut the diorite. They consist of coarse-grained plagioclase (andesine) and amphibole, the amphibole forming prismatic crystals up to 6 cm long, commonly lying perpendicular to the pegmatite contacts (Palmer, 1971, p.241). Minor quartz and apatite are also present, and Palmer (1971) noted a relict orthopyroxene in one specimen. The pegmatites show no evidence of folding; local recrystallisation and bent twin lamellae in feldspar show that there has been only minor internal deformation. The margins of the pegmatites are commonly more basic than the interior, and the adjacent rock may be depleted in plagioclase. Palmer (1971, p.241) noted small displacements across the pegmatites in the norite but suggested that these pegmatites were formed by segregation from the host rock, for they are much thinner where they cut ultramafic dykes. This feature may, however, be a product of the differing ductilities of the two igneous rocks (norite and ultramafite), for such features are common in mineral veins cutting mixed lithologies (e.g. in the North Pennines).

Pyroxenites

Websterite (pyroxenite) has been described from a small quarry 270 m NNW of Rodel Hotel and from a pit by the footbridge to Rodel Schoolhouse, (Livingstone, 1963; Davidson, 1943; Witty, 1975). It is composed of orthopyroxene and clinopyroxene with abundant magnetite and brown spinel, along with minor amounts of amphibole (pale yellow-green actinolitic hornblende-Witty, 1975, p.137), biotite and serpentinised olivine. The pyroxenites merge laterally through black clinopyroxene amphibolites, in part garnetiferous, to plagioclase amphibolite. Amphibole and garnet-amphibole pyroxenite form isolated lenses between Rodel and Leverburgh (e.g. 460 m south-east of Leverburgh Post Office, and on the east shore of Loch Cisteabhat [NG 027 857]). The typical assemblage is clinopyroxene (salite), garnet, amphibole (pargasite) and orthopyroxene (En69-77), with garnet abundant and olivine and spinel absent. Apatite and oligoclase/andesine are common accessories. Livingstone (1963) has described small-scale banding in these rocks defined by alternation of garnet-clinopyroxene layers 4 mm thick with dominantly amphibole layers.

Similar locally podded garnetiferous websterite bodies up to 5 m thick are exposed on the coast between Traigh an Teampuill and Traigh na Cleavag. A representative thin section (S63089) [NF 9756 9120]) from a massive dyke-like body shows a mosaic of orthopyroxene and clinopyroxene with locally abundant garnet and minor biotite, ilmenite, quartz and rare plagioclase.

Witty (1975, p.139) has described three large dykes of websterite north-east of Rudh'an Teampuill where they trend parallel to the regional gneiss foliation. Similar dykes were noted south-west of Northton. Clinopyroxene: orthopyroxene ratios are approximately 5:1, and the rocks contain accessory plagioclase and ilmenite. The abundance of garnet, biotite and quartz in the dykes adjacent to meta-sedimentary rafts (up to 15 m long) and near dyke margins was attributed to assimilation of the country rocks. Witty (1975) also noted locally abundant sulphide phases in the dykes, forming up to 15% of the exposed rock. Dendritic pyrrhotite intergrown with minor chalcopyrite is abundant, and pyrite and vein chalcocite are also present.

These rocks are identified mineralogically with the thick pyroxenite dykes (up to 10 m across) and the related thin interconnected vein network near Strond, at [NG 029 838], which cut the norite. Palmer (1971, p.228) first described these intrusions. Subsequent descriptions have shown that the dykes are preferentially orientated near parallel to the early (north-south) foliation in the norite. Dearnley (1963) has discussed narrow websterite veins cutting the norite at the north end of the beach near Ensay House on Ensay.

These rocks are composed of granoblastic intergrowths of clinopyroxene and orthopyroxene, with the clinopyroxene making up 60-79 per cent of the rock. Minor secondary hornblende is present. The observed breakdown of euhedral clinopyroxenes to smaller polygonal aggregates suggests that there has been widespread recrystallisation (Palmer, 1971, p.228). Minor plagioclase and/or biotite have also been reported (Dearnley, 1963), and ilmenite and apatite are accessories.

Related ultramafic bodies were also mapped by Horsley (1978) in many parts of the diorite body, where they are dominantly garnetiferous and range from garnet-pyroxene hornblendites to garnet pyroxenites and hornblende pyroxenites, through to metagabbros. On Toe Head near Geo Mor [NF 960 947] Palmer (1971, p.226) has recorded net-veining of metasediments and early metagabbro dykes by thin garnetiferous metagabbro dykes. These probably relate to the adjacent diorite to the east.

(Table 20) lists the average analyses by Horsley (1978) and Palmer (1971) for the intrusive ultramafic bodies in the norite and diorite. The composition of the ultramafic rocks in the diorites and norite is similar to that of the banded ultramafic component of the anorthosite complex. Garnet pyroxene hornblendites are ne-normative (up to 8 per cent ne) and the pyroxenites and hornblende-bearing pyroxenites are hy-normative. The Q-normative analyses of Palmer (1971) may reflect some error in the major oxide values. Horsley (1978) showed that on a TiO2:Zr/P2O5 plot the dykes cross the alkaline basalt/tholeiite boundary, and on a Ti-Zr plot they fall in the low-K tholeiite field. There is a gradual decrease in alkalis with increasing SiO2 (cf. the increase in alkalis with SiO2 for the norite and diorite.). The geochemistry of the ultramafites is a product of clinopyroxene and minor plagioclase (high calcic) differentiation. Horsley (1978) has used Pearce variation diagrams to calculate the clinopyroxene composition as Wo42En24Fs16. He concludes that these ultramafites are tholeiitic showing extreme Fe enrichment and notes that within the two major igneous bodies the ultramafic intrusions have differing geochemical characteristics. Fe/Mg ratios for ultramafites in the diorites range from 0.44 to 0.67 whereas those in the norites range from 0.33 to 0.53. Although the ultramafites have a major element compositional range which overlaps that of the early ultramafic rocks in the Langavat Belt, the Cr and Ni values are distinctly different.

Anorthosite-ultramafite 'breccias'

Witty (1975, pp.100 -115) has described at length the occurrence of lenses of breccia in the marginal zones of the anorthosite complex, notably along the quartz gabbro and 'anorthosite' contact. He recognised two types, which he designated types 'A' and 'B'.

Type 'A' breccias occur mainly along the south-west flank of the anorthosite complex, notably north-east of Abhainn Easean Chais, north of Loch Cachlaidh and on the southeast slopes of Beinn na h-Aire (see Map 5). The 'breccia' consists of disorientated angular foliated and banded 'anorthosite' blocks with narrow interconnecting ultramafite veins. Small intrusive bodies of massive ultramafite are commonly associated with the breccias. Witty (1975) stated that marker horizons in the basal 'anorthosite' sequence can be traced through the breccia zones with only minor lateral displacements. The ultramafites are dominantly composed of pale green augite with minor scapolite and symplectite. Iron-oxide rods are found in scapolite and as exsolutions in the clinopyroxene. Garnet (almandine-pyrope) occurs in diffuse patches and streaks as well as in nodules, where it forms 30–70 per cent of the mode. Rare olive-green ferroan pargasite plates occur in the nodules.

The best development of type 'A' breccias is north-east of Abhainn Easean Chais [NG 034 864] adjacent to a large lenticular massive ultramafite body which itself contains elongate xenoliths (up to 30 cm long) composed of foliated and banded 'anorthosite' and quartz gabbro. Garnetite and garnetiferous ultramafite nodules are also present in the ultramafite and Witty (1975, p.105) interprets these as cognate xenoliths. The breccia is characterised by highly disrupted large angular blocks (10 m3 to 100 m3) and smaller more rounded blocks of 'anorthosite'. The intervening veins and irregular patches of ultramafite (as well as rare garnetite nodules) tend to dip vertically. In places the anorthosite blocks lose their fine foliation, becoming discontinuous and indistinct; the end product is massive anorthosite with included irregular ultramafite lumps. Adjacent to these zones narrow discordant anorthosite veins and stringers cut the breccia and ultramafite.

Type 'B' breccias crop out 450 m west and 300 m southeast of Ha-cleit summit [NG 032 873]. Small xenoliths of foliated and banded 'anorthosite', quartz gabbro and ultramafite occur in a massive anorthosite matrix. 'Stoping' features were observed in several localities (e.g. 400 m west-north-west of Ha-cleit summit). Ultramafic xenoliths are preferentially rounded and some show epidote-rich reaction rims 5-10 mm thick. Large rafts of type 'A' breccia occur within the type 'B' lenses.

Although the breccias appear random, Witty (1975, pp.101–115) has found that the larger blocks show a preferred orientation. In the type 'B' breccias west of Ha-cleit this foliation trends at 035°, almost perpendicular to the anorthosite/gabbro contact and markedly different from the foliation trend in the adjacent rocks, which is 105°. In the type 'A' breccias north-east of Abhainn Easean Chais there are two distinct maxima, one trending at 123°, parallel to the strong regional foliation, and the other at 065°. There appears to be no satisfactory explanation for the 065° preferred orientation except that it may result from rotational effects during a late-deformation event.

Witty (1975, p.109) suggests that type 'A' breccias were formed by explosive gas action fracturing the basal units of the 'anorthosite' followed rapidly by general upwelling of gabbroic magma. This highly basic material was subsequently transformed by the m2 metamorphism to a more ultramafic mineralogy (see Metamorphism p.72, and Synthesis pp.73 -74 this chapter and the section on Metagabbros, pp.51 - 53 this chapter). This would explain the contrast between angular large blocks and more rounded smaller blocks in the breccias. At a locality 640 m south-west of Hacleit summit the ultramafite matrix appears to be vesiculated; this may be a secondary weathering effect. The presence of scapolite in the ultramafite suggests that CO2 may have been a significant constituent of the gas phase, unless the scapolite was formed entirely during the later m2 metamorphism (see Metamorphism; Synthesis at the end of this chapter). Type '13' breccias are thought by Witty (1975) to result from anorthosite injection, local partial melting and accompanying fluid flow in the same areas where the type 'A' breccias are developed. This process may have happened during a later phase than that in which the type 'A' breccias were formed.

It should be noted that the breccias were intruded during the DL1 deformation but prior to the tight folding of the anorthosite complex (Witty, 1975). Some of the 'breccias' textures illustrated by Witty superficially resemble those described by Windley et al. (1973) as 'cumulate' textures in the Archaean layered basic body at Fiskenaesset in West Greenland.

Anorthositic gabbro dykes and trondhjemite pegmatites

The anorthositic gabbro dykes are generally parallel-sided bodies no more than 25 cm wide and traceable for up to 36 m (Witty, 1975, p.115). They are restricted to the anorthosite complex, particularly the marginal banded ultramafites and mafic gabbros, and are more abundant at its north-west end. The dykes cut the compositional banding and early foliation and are themselves cut by early Laxfordian shear zones. They have a central zone enrichment in mafic minerals, which form clots typically oriented oblique to the dyke margins. The dykes may reflect the bulk composition of the magma that differentiated to form the anorthositic complex (Witty, 1975, p.115). They are composed of pink andesine crystals (An32-38) 15 to 20 mm across, coarsely intergrown with quartz, and diffuse elongate aggregates of pale green amphibole and remnant clinopyroxene. Garnet occurs locally, notably in the dykes cutting ultramafites and mafic gabbros. The feldspars show bent twin lamellae and undulose extinction, and recrystallisation textures are common.

Dearnley (1963) describes similar dykes and veins from anorthosite in the Rodel Park area with plagioclase (An25-46) forming 80 per cent of the rock. He termed these dykes trondhjemitic pegmatites. They are partly altered to zoisite for they lie within the zone of saussuritisation. In some places these pegmatites and anorthositic gabbro dykes are basic, and elsewhere they are intermediate to acid, matching the host rock in this respect; Witty (1975, p.117) suggests that they were, therefore, formed by segregation from the surrounding rocks.

Late potash-rich basic dykes

Thin potash-rich basic dykes, less than 50 cm wide, with K2O values of 5.5 to 7.5 per cent, cut obliquely across the early foliation in the diorite north of Scarista near Sgeir Liath [NG 013 939]. Palmer (1971, p.242) records twelve such dykes and notes that they trend about north - south and dip steeply to the east. The dykes cross-cut the hornblende-plagioclase pegmatites in the diorite but Myers (1968, p.94) notes that they are cut by a late series of deformed thin pegmatite veins. Similar dykes, commonly folded, are also found at the south-east end of Traigh an Teampuill [NF 973 913], where they cut the mylonitic foliation in the acid gneisses at the margin of the Leverburgh Belt. The dykes typically have a granular appearance; locally they have a zone of marginal foliation. In some thin dykes this foliation is present throughout the dyke, the fabric being subparallel to its length.

The dykes consist of large twinned clinopyroxene laths in a granular intergrowth of microcline, deep-brown poikilitic biotite and subsidiary albite. Apatite is common and iron ores, sphene and quartz occur as accessory minerals. The following assemblages were recorded from two samples: the first (S63084), from near Sgeir Liath at [NG 0133 9388] consists of cpx-bi-hbl-mic-qz-ab-ap; and the second, (S63120), from Rubh'an Teampuill [NF 9731 9131], cpx-kf-qz-ab/olig-ms-scp-ap-chl-sph-mt. Dark green hornblende (mainly after clinopyroxene) and chlorite increase in abundance with the degree of foliation. Extreme retrogression gives a hbl-ep-mic assemblage with minor biotite, quartz and sphene (Palmer, 1971).

The lath shape, twinning and texture of the pyroxenes are commented upon by Palmer (1971) and a subsequent analysis (Table 21) by Horsley (1978) confirms his suggestion that the pyroxene is of 'igneous' origin. The potash feldspar and biotite are assumed by Myers (1968) to be primary minerals, and he terms the rocks metalamprophyres. Palmer (1971) ascribes all the potash minerals to metasomatism such as is found marginal to the Leverburgh Belt. Horsley (1978), however, points out that the adjacent rocks are undeformed and show no metasomatic effects. He presents three full analyses of typical dykes (Table 22) and notes that they lie on an extension of the calcalkaline trend on the AFM plot ((Figure 10)). Although their K/Rb variation is typical of unaltered igneous rocks; it does not accord with the pegmatitic-hydrothermal trend of Shaw (1968). Horsley suggests that the dykes represent a late-stage of island arc magmatism and terms them shoshonites.

Pegmatites

Pegmatites are particularly abundant in the zone of migmatite and granite-gneiss around the north-east margin of the Langavat Belt. Although some smaller bodies may be of Scourian age the majority of the larger lenses are related to the Late-Laxfordian granite sheets that become abundant immediately to the north and east of the Langavat Belt.

Small pegmatite veins and lenses up to 7 cm thick are found in many parts of the Leverburgh metasediments. Dickinson (1974, pp.64–65) illustrates one such locality east of Rodel at [NG 053 831]. The veins and lenses are usually concordant with the regional foliation; in the marginal parts of the Leverburgh Belt they occur as pink streaks up to 2 cm wide (e.g. at the east end of Rubh' an Tempuill [NF 975 913]). The pegmatites are typically composed of quartz, perthitic potash feldspar and minor biotite and plagioclase. The present size of quartz and potash-feldspar aggregates, about 1.5 cm across, and their minor discordance in less deformed zones led Dickinson (1974, p.66) to suggest that these were pegmatitic segregations formed during the earliest metamorphic event.

The Laxfordian pegmatites have their greatest development on Taransay, where they form about 25 per cent of the exposed ground on the Aird Vanish Peninsula and up to 75 per cent just east of Traigh a' Siar [NG 010 010]. The pegmatites (cf. Myers, 1971) form elongate lenses typically 3–4m across, with their long axes subparallel to the regional banding. Jehu and Craig (1927, pp.481–483) showed these rocks to be composed of microcline (rarely orthoclase), perthitic albite-rich plagioclase (up to An33), quartz and small biotite-flakes. Magnetite occurs in irregular masses up to 3 cm across.

On mainland South Harris von Knorring and Dearnley (1960) described the mineralogy of a particularly well-exposed pod at Loch a' Sgurr [NG 0708 8652] near Finsbay. This consists of albitic plagioclase, microcline, quartz and minor biotite. Magnetite segregations up to 15 cm long are associated with a green-yellow mica, uraninite, thorite, and monazite. Zircon is a common accessory, and also present are the following: beryl, spessartine, gahnite, columbite, pyrite, tourmaline, apatite, thorogummite uranophane, kasolite, pyrochlore and allanite.

Dearnley (1963) subdivides the South Harris pegmatites generally into dominantly biotite-magnetite-bearing and dominantly muscovite-bearing varieties, the former being typical of the Langavat Belt and adjacent igneous rocks (Roineabhal–Beinn Tharsuinn area) and the latter of the Leverburgh belt. In places there are also pegmatites that lack significant amounts of mica or opaque minerals. Von Knorring and Dearnley (1960) relate the abundance of magnetite in the Langavat pegmatites to the paucity of biotite. They also note the scarcity of tourmaline in the South Harris pegmatites.

In South Harris two large pegmatites at Sletteval [NG 061 856] and Chaipaval [NF 976 920] (Map 5) were worked on a small scale for potash feldspar during the 1939–45 war. They were included in an economic appraisal of the Western Isles carried out for the Highlands and Islands Development Board by Robertson Research Co. Ltd (McKenzie et al., 1973). Parts of the following account are taken from this report.

The Sletteval pegmatite consists of two veins respectively 20 and 6 m thick. An irregular vein joins the two main veins at their north-eastern end. The veins dip about 65° to the west-north-west, and are composed of microcline (rare orthoclase), microperthitic plagioclase, quartz and biotite.

Quartz-feldspar graphic intergrowths are common. Magnetite crystals up to 2 cm across are common, and garnet, apatite and tourmaline also occur. Von Knorring and Dearnley (1959, 1960) noted in addition allanite, zircon, thorite and thorogummite, most of which were included in biotite. Three small quarries were worked in the pegmatite between 1940 and 1945.

The Chaipaval pegmatite is 2 to 3 km north-west of North-ton. It ranges from 6 to 25 m in width; at its north-eastern end it dips 70° to the south-east, but in the south-west it dips 40° to the north. It is represented at the coast of the Sound of Harris by irregular veining. The constituent minerals are quartz, plagioclase, microcline, muscovite, and minor biotite and magnetite. In the vicinity of the quarry, quartz and potash-feldspar crystals attain 45 cm in length in central lenticular zones (up to 2 m wide). Muscovite locally forms 'books' up to 20 cm across and 3 mm thick. Graphic intergrowths of quartz and feldspar are common, particularly adjacent to the central coarse-grained lenses. Garnet (Sp54.5, Al41.2) is a common minor component, and von Knorring and Dearnley (1959, 1960) noted beryl, columbite, gahnite, zircon, monazite, allanite, tourmaline, pyrite, uraninite and thorite in concentrates from the Chaipaval pegmatite. Tourmaline is present, but nowhere abundant.

Other large lenticular pegmatite bodies (3–20 m wide) have been noted in the vicinity of Kyles House [NF 998 881] associated with the local potash metasomatism. Large bodies also crop out on Beinn Tharsuinn and Bhoiseabhal [NG 0433 8720]. The Beinn Tharsuinn lenses dip gently northwards and are up to 12 m thick and 100 m long. They are dominantly quartz-feldspar pegmatites but biotite, garnet and magnetite are also found. The Bhoiseabhal bodies lie just south-east of the summit and are 6-12 m wide. They contain intergrown smoky quartz, tourmaline and garnet. Pegmatites up to several metres wide and locally containing feldspars up to 30 cm long and 20 cm across were reported by Jehu and Craig (unpublished field notes) from north-west Killegray in the Sound of Harris. Numerous bodies also occur in west and south-west Ensay in elongate lenses approximately parallel to the regional strike.

Cunningham (1981) has worked on the geochemistry of large granite-pegmatite bodies in the Lewisian of north-west Scotland and the Outer Hebrides. He found that the South Harris muscovite-bearing pegmatites with the assemblage pl(An10)-qz-ms-gt-kf formed in conditions of lower temperature and lower fO2 than the potassic pegmatites which contain a qz-kf-bi-mt assemblage. Biotite from Laxfordian pegmatites has a higher Fe/Fe + Mg ratio than biotite from Scourian pegmatites, which implies that the Laxfordian pegmatites formed at a lower temperature than the Scourian pegmatites. Rb, Sr and Ba determinations (ppm) for feldspars from pegmatites in South Harris are listed below (Cunningham, 1981; whence the specimen numbers are taken).

Rb Sr Ba
kf pl kf pl kf pl
GC 6 2486 833 26.1 7.0 145 8.1
GC 23 1568 112 33.4 56.3 101 23.6

The pegmatites of South Harris have been dated at c.1100 Ma from K-Ar mineral ages (Kursten, 1957) and at between 1496 and 1662 Ma by Giletti et al. (1961) using Rb-Sr and K-Ar methods on biotite and muscovite. 238U/206P analyses on uraninite from the Chaipaval pegmatite (Dearnley in Giletti et al., 1961; Bowie, 1962) give an age of 1490 ± 35 Ma. Further U-Pb ages on monazite and uraninite from the Loch a'Sgurr pegmatite and from the Chaipaval body give 1565± 100 Ma (Bowie, 1964). The pegmatites were emplaced in the time period 1700 -1500 Ma with the lower-temperature pegmatite of Chaipaval (with a mt-ab-gt assemblage) dated around 1500 Ma.

Structure

The overall structure and tectonic relationships of the South Harris rocks are covered in Chapters 13 and 14, but it is appropriate here to discuss in more detail the history of the igneous complex and the contrasts between the Leverburgh and Langavat belts. The terminology for the structural events is that set out in (Table 27) (at the beginning of Chapter 13). Most of the Langavat Belt rocks have been affected by Scourian amphibolite-grade metamorphism (Myers, 1968, p.151) and show the effects of repeated folding and high finite strain. In the Leverburgh Belt, the marble-pelite divisions are tightly folded and there are deformed migmatitic fabrics in some of the acid gneisses (Dickinson, 1974), but the rocks generally lack the widespread strong planar fabric of the Langavat Belt. Graham (1980) has assigned the early deformation in the Leverburgh Belt rocks to a Scourian event, for the granulitefacies mineralogy appears to overprint the fabrics. The igneous bodies contain a primary foliation which has subsequently been locally deformed by numerous shear zones; these are particularly well developed adjacent to the margins of the bodies, and there they have regional significance. If the Nd-Sm isotopic age of 2180 ± 60 Ma (Cliff et al., 1983) obtained from the anorthosite is the maximum age of the South Harris Igneous Complex then all structures in the igneous rocks must be post-Scourian in age. It is probable that a large fault or series of faults trending north-west lies offshore in the Sound of Harris.

Pre-DL1 event–metasediments

The inclusion of highly deformed metasediments within only slightly deformed diorite (notably in the Bagh SteinigieBleaval area) suggests that a major gneiss-forming episode occurred prior to igneous intrusion. The presence of finely foliated amphibolite associated with these sediments on the south-east flank of Bleaval suggests that this was an amphibolite-grade event. Individual Scourian deformation phases, that can be identified for example in Barra, cannot be distinguished readily in South Harris. This major development of foliation prior to intrusion of the Igneous Complex is attributed to dS2 (see (Table 1)). Within the Langavat Belt itself no early fabrics can be recognised because the component metasediments and metavolcanics have subsequently been much more intensely deformed than the diorite.

Primary foliation/banding in the igneous bodies (dL1)

As noted in the previous sections a primary foliation commonly defined by the alignment and concentration of mafic minerals has been recorded in many igneous bodies on South Harris. Horsley (1978) noted an alignment of pyroxene in parts of the diorite (this chapter: Metadiorite); in a few areas such as Sgeir Liath near Scarista this fabric is absent. In the central parts of the diorite it strikes north-east and dips steeply (about 75°) to the north-west. Similarly, in the norite Horsley (1978) noted parallel alignments of pyroxene glomeroporphyroblasts.

Within gabbro and norite pods in the metasediments of the Northton- Toe Head region, Graham (1970) reports a steep internal fabric trending south to south-east, markedly oblique to the boundaries of the pod and to the regional foliation. A thick boudinaged garnet-metagabbro band (up to 50 m wide) north-west of the norite contact in this area contains a strong internal fabric, steeply inclined and striking east - west. We correlate this fabric with that observed by Witty (1975, pp.148 - 149) in the anorthosite complex, particularly in the lower-zone rocks and marginal mafic gabbros. Here it is largely subparallel to the compositional banding. Witty (pp.150 -159) describes a complex series of structures associated with the fabric. He recognises an early migmatisation and related 'plastic' deformation with local development of small-scale tight folds. Ultramafic/mafic gabbro dykes (this chapter: Anorthosite-ultramafite 'breccias') have been preferentially intruded parallel to the d L1 fabric and appear both to predate and postdate the main phase of fabric development. The earlier dykes are locally folded and foliated, and show segregation of felsic components; in contrast the later dykes are commonly discordant on a small scale, with net-vein relationships. A later minor phase (also attributed to dL1) of small-scale parallel subvertical shears trending east - east-north-east locally disrupts the earlier structures but predates the later dyke phase. Quartzofeldspathic lensoid segregations are locally extensively developed subparallel to the d L1 fabric and to the banding in the anorthosite complex, and also within the later thin shears. The degree of dL1folding and 'contortion' was found by Witty (1975, p.149) to correlate with the amount of migmatisation. Witty (p.155) illustrates syn-dL1 garnets overgrowing tight to isoclinal microfolds in the gabbroic complex. In other areas granulite-facies metamorphism resulted in an anhydrous assemblage and the early deformation is absent; the rocks therefore were resistant to subsequent deformation. Witty (1975) assigns the boudinage and amphibolite-grade marginal foliation of the banded gabbros to the du event, but during the present survey these features were found to be more in accord with the dL2 structures elsewhere in South Harris. Horsley (1978) and Witty (1975) both suggest that the intrusion of the igneous bodies was coeval with the dL1 event.

DL2 event

This deformation phase is correlated with the main penetrative Laxfordian deformation found in many parts of the Outer Hebrides (see Coward et al., 1970). It has resulted in a wide variety of structures in South Harris dependent largely on the rock types affected and the attitude of major lithological contacts. The igneous bodies show prominent marginal and subsidiary internal shear zones. The norite body has apparently been deformed to an ovoid shape with either thick shear zones (up to 100 m wide) or a series of small zones developed along its south-west and north-east margins. The sense of shear across these zones changes from dominantly dextral in the Northton–Toe Head area to dominantly sinistral in the Renish Point area. The larger diorite/tonalite intrusion now forms a sigmoidal body with a major amphibolite-grade shear zone at its north-east margin. Shear zones generally have a consistently dextral sense of movement throughout this body (Graham and Coward, 1973; Graham, 1980); some of these zones may be a product of the later dL3 event. The internal deformation of the diorite is less than in the norite. In the anorthosite complex the central zones generally show little evidence of dL2 deformation except for a locally strong linear fabric marked by spindle-shaped mafic clusters, but the margins are strongly foliated and retrograded with extensive boudinage. Witty (1975, p.163) showed that there is a relationship between the observed thickness of banded gabbro units (1 cm to 7 cm thick) and the angle between the internal compositional banding and the margin of the boudin (subparallel to 60°). Although there were probably differences in the original thickness of the banded gabbros the observed variations are largely due to tectonic attenuation (Witty, 1975, p.163; Palmer, 1971, pp.212–213). The banded gabbro lenses all have an outer zone of amphibolitised material 1–4 m wide which grades outwards into a thicker highly foliated amphibolite sheath.

The anorthosite complex itself is folded into a major tight antiform, the axis of which plunges 80° to the north-west at its south-eastern end but vertically in the area around Loch Steisevat and Loch na Moracha. It is hence both respectively steeply downward-facing, and sidewards- or north-west-facing along its axial trace. Although the dL1foliation is folded around the hinge (Witty, 1975, p.158) the granulitefacies assemblages and early textures are largely unaltered. Witty (pp.160–165) has attributed the boudinage of the marginal banded gabbros and attendant development of foliation to the dL1 event with subsequent dL2 overprinting. However, the weak axial-plane biotite foliation noted in the hinge of the Roineabhal Antiform appears continuous with the extremely strong fabric on the limbs. This strongly suggests that the fold is dominantly a dL2 structure although it does not totally preclude initial open-fold development during dL1, as suggested by Witty. The limbs appear to have acted as a locus for marginal shearing both during the dL2 event and subsequently.

West of Northton the perthitic and other metasedimentary gneisses show a strongly attenuated (mylonite) banding along their junction with the norite. This deformation apparently relates to shearing along the margins of the norite and is taken here to be a dL2 effect. Within these gneisses discordant potash-rich basic dykes are folded, showing that their intrusion predated the development of the dL2 foliation. This does not preclude the development of earlier shear zones coeval with intrusion of the major units of the South Harris Igneous complex.

dL3 event

This event is documented by Coward et al. (1970) (whose nomenclature is broadly followed in this memoir). They define a regional-scale series of broad antiforms and tightly pinched synforms, and show that the South Harris Complex lies in a broadly synformal zone with a complementary major antiform occurring in the Sound of Harris (see Map 3). It is difficult to determine whether dL3 effects in South Harris are limited to the tightening up and modification of pre-existing structures, perhaps with further development of localised shear zones, or whether the Roineabhal antiform and related synforms are largely dL3 structures. Witty (1975, pp.147–149) and Horsley (1978) appear to have confused the nomenclature of Coward et al. (1970) in that they recognise only one major Laxfordian tectonic event (dL3). Horsley (1978) assigns the major shear zones and amphibolite-facies metamorphism to the dL3 event, omitting any mention of the penetrative dL2 event. We suggest here that the effect of the dL3 event was to tighten up pre-existing structures, possibly causing further attenuation on the limbs of the major dL2 folds and the formation of numerous, predominantly dextral, north-north-west–trending shear zones in the igneous bodies. Horsley (1978) records a series of such steeply dipping zones, trending north north-west, along the south-west margin of the diorite, where they are marked by the development of extremely fine-grained amphibolite. These zones are less abundant in the central parts of the diorite; there they dip steeply eastwards and may be up to 2 m wide. In the Gleann Uachdrach/Sgeir Leomadal [NF 985 939] zone, a lenticular mylonitic diorite zone trending 148° and up to 4 km long has been recorded by Horsley (1978). Witty (1975, pp.166–169) has carefully documented similarly trending shear zones in the anorthosite and has shown that they extend for up to 500 m laterally and range in width from a few centimetres to 22 m. He records anticlockwise rotation of up to 20° of material lying between major dL3 shears (e.g. a pod of gabbro/ultramafite at Abhainn Easean Chais on the south-west flank of the anorthosite complex). This effect can be traced through from the anorthosite and into the paragneiss (Witty, 1975, p.167).

dL4 and later events

Witty (1975, pp.167–170) records two distinct but broadly related sets of shears within and adjacent to the anorthosite complex: major east-trending sinistral shear zones up to 4.5 km long with displacements of up to 140 m, and minor sets of shears trending north-east to east-north-east which locally cross-cut the first set. The second set of shears is visible on aerial photographs, where extensive lineaments can be seen. Only minor displacements (both sinistral and dextral) can be observed. Dearnley (1963) noted that near-vertical shear zones trending approximately east–west are related to the large Late-Laxfordian alkali pegmatites. Many authors have noted that the Sletteval Pegmatite is locally sheared with a dextral displacement, and Myers (1968) also noted shearing and pseudotachylite development in a Late-Laxfordian pegmatite on the north-west shore of Loch Langavat [NG 0425 9085]. Some of the pegmatites have been intruded along pre-existing shears; they exhibit slight peripheral granulation owing to continued movement (Witty, 1975) and their feldspar orientation appears to reflect growth in a stress field. This dL4 or dL5 event is therefore of almost certain Late-Laxfordian age, and Dearnley (1963) has commented on the similarity of trend between these shears and those found in the Loch Broom-Kylesku area on the Scottish mainland, which are pre-Torridonian in age.

The thrusting and shearing associated with the Outer Hebrides Thrust Zone are discussed elsewhere (see Chapter 15); the effects in South Harris are restricted to the south-eastern part of the anorthosite complex and to the prominent south-eastern peninsulas (e.g. Renish Point, Sron Ghaoithe). Witty (1975, p.170-177) provides a very detailed description of the thrust 'anorthosite'. It should be noted that pseudotachylite was a fairly widespread product of shearing in the granulite-facies gneisses, provided no water was introduced. Its occurrence on South Harris is not specific to the thrust zone.

A considerable number of olivine-free tholeiitic basalt dykes were emplaced in South Harris prior to the main Tertiary faulting. This faulting resulted in a series of prominent lineaments trending north-north-east and between northwest and north-north-west (Witty, 1975). In the southeastern part of the island the fault/dyke lineaments are commonly eroded to form deep gullies.

Olivine-dolerite and basalt dykes largely postdate the Tertiary movements but are associated with the faulting. Their trends are seen on Map 1. Witty (1975) has argued that the highly vesicular nature of the earlier Tertiary dykes suggests emplacement at crustal depths of less than 100 m, implying minimal uplift since Middle Eocene times. The faults displace features of the Outer Hebrides Thrust Zone and are similar in trend to the pronounced shears which cut the Tertiary basalts of Skye (Dearnley, 1963).

Metamorphism

The mineral assemblages which characterise the different rock types in the South Harris Complex have been listed and discussed earlier in this chapter. Here, we assemble the evidence to give a coherent picture of the metamorphic conditions in Scourian and Laxfordian times and to suggest possible values of pressure and temperature for particular events. Specific metamorphic interpretations are also discussed. Dickinson and Watson (1976) described the evolution of the P, T path of the Leverburgh area in relation to other parts of the Lewisian Complex. Their general thesis of continued but periodic uplift is endorsed here but modified in the light of more recent work (Horsley, 1978; Cliff et al., 1983).

The metasediments show no assemblages that can be definitely dated as prior to the intrusion of the major igneous bodies, although, as stated previously, the igneous bodies do contain metasedimentary rafts with a strong tectonic striping, typified by amphibolite-grade assemblages. Also it seems probable that the South Harris metasediments can be correlated with those of the Scourie area of north-west Scotland, and hence they have probably undergone amphibolite- to granulite-grade metamorphism (m1) during the Scourian. In zones of locally strong Scourian deforma tion, e.g. pelitic or marble-rich zones, water partial pressure (PH2O) levels were almost certainly greater than in the low-deformation areas where progressive dehydration reactions occurred with little 'tectonic interference'. Some migmatisation probably predates the intrusions but critical field relationships are generally absent. In the Langavat Belt Myers (1968, p.103) has recorded concordant quartz and quartzofeldspathic segregations parallel to the banding; these are in turn cut by quartz-oligoclase veins and sweat-outs which are folded by early Laxfordian structures. In the Leverburgh Belt pre-Laxfordian metamorphism culminated in the following critical assemblages:

These assemblages result from the superimposition of m1 and m2; m2 is an event apparently synchronous with the intrusion of the igneous suite at about 2200–2000 Ma (Cliff et al., 1983). The intrusive rocks of the Igneous Complex show evidence of recrystallisation and local equilibration to granulite-facies conditions (yet igneous geochemical trends are remarkably unaffected) during this period. Critical assemblages in the diorite, norite, anorthosite are:

In parts of the norite and anorthosite gt-opx-cpx-bearing assemblages are found. These are typical of the hornblende-granulite facies, and imply pressures of 1000–1300 MPa100 Megapascals (MPa) = 1 Kilobar (Kb) and temperatures of 825 ± 25°C. The gt-cpx-qz assemblages in the diorite and norite are characteristic of higher pressures than the opx-cpx-qz-bearing rocks, although their occurrence could also be a product of compositional variation. Witty (1975) accepted the latter explanation, but Horsley (1978) has shown that the garnet-bearing assemblages are dominantly in the south-west parts of the diorite body (see Map 5) and appear to be restricted largely to the originally low structural levels (the sequence now being inverted). This would suggest that higher pressures were attained in the garnetiferous parts of the diorite.

Hypersthene-plagioclase intergrowths form reaction rims around garnet in the diorite and norite, and Dickinson (1974) has enumerated other minor assemblages (m2a) compatible with pressures of 800 ± 100 MPa and temperatures of 760± 40°C and with PH2O less than 0.3 of the total pressure (PTotal). Dickinson (1974) rated these metamorphic assemblages as a later separate phase than m2, but because of their restricted occurrence it is more likely they reflect rapid crustal uplift (Horsley, 1978). These modified m2 assemblages (m2a) are only found in the originally deeper levels of the complex with no mineralogical evidence for modified reactions in the higher levels. The m2a assemblages imply a pressure decrease of 400 MPa (equivalent to about 10 km nearer the surface), and a change in geothermal gradient from 22°C/km at the peak of m2 metamorphism to about 38°C/km during m2a. Cliff et al. (1983) show that this event was over by 1840 Ma.

The Laxfordian metamorphic assemblages imply an overall progressive decrease in temperature and pressure with time, but retrogressive metamorphic assemblages are confined to specific zones, mainly in metasediments, because the associated deformation pattern is so inhomogeneous. The massive igneous bodies and metasediments which show virtually anhydrous granulite-facies metamorphism are not susceptible to retrogression. The main retrogressive event (m3) was coeval with the main Laxfordian deformation and is typified by the following retrogressive assemblages:

The m3 event forms the dominant assemblages in the Langavat Belt and on the sheared north-west side of the Igneous Complex. Sillimanite is locally common in pelitic units of the Leverburgh Belt, and Horsley (1978) records m3 partial melting in shear zones in the diorite. This implies that if PH2O = PTotal then the temperature was 675°C and pressure was 800 MPa. Myers (1968) records sil-st-co-gt-and cum-grn-bearing assemblages from pelitic rocks in the Langavat Belt implying pressures of between 450 and 700 MPa (crustal depth respectively 13–21 km) and temperatures of 660 ± 30°C . These assemblages are polymetamorphic, containing inherited components from the m1 and m2 events; Myers (1968, pp.103–105) himself assigns the mineralogy to two discrete episodes. Implied geothermal gradients for the m3 event lie in the range 31-48°C/km. This event is dated on the mainland and in the Outer Hebrides (Moorbath and Park, 1972; Bowes, 1978) at c.1700 Ma

Localised potash-rich metasomatism occurred in the Northton–Leverburgh coastal zone either during the m3 event or during subsequent emplacement of pegmatites, particularly where the rocks are strongly folded/deformed and in the marginal shear zones of the norite. This results in the replacement of feldspar by muscovite, notably around Kyles House, where fibrolite is widely developed. Dearnley (1963) has also recorded that limestones show minor scapolitisation. The metasomatism has also led to the extensive development of biotite in the sheared norites/amphibolites.

Although Dickinson (1974) records the local development of chlorite after biotite, and minor calcite, there was no widespread low-grade retrogression. Pressure and temperature during this late phase are estimated at about 300 MPa and 500°C (Horsley, 1978); in the upper greenschist facies. Final pressure and temperature patterns in the area may be reconstructed from the mineralogy of the Late-Laxfordian pegmatites. These are dated between 1703 Ma and 1533 Ma (Giletti et al., 1961), and details of their mineralogy are given earlier in this chapter: see pp.68–70, Pegmatites.

Synthesis

In view of the large amount of information on South Harris, we present an interpretation of the Complex and adjacent metasediments, and discuss the broader implications. This is based largely on the major references cited throughout this section.

The metasediments represent an original sequence, probably marine, of sandstones and shales with minor bands of black shales and limestones. In the Langavat Belt basic volcanics and/or tuffs appear to have been locally abundant. Before the tectonic inversion of the South Harris Complex the original structural sequence was probably:

The metasediments were strongly deformed and metamorphosed under amphibolite-grade conditions (c.2800 Ma), probably whilst being tectonically 'transported' (perhaps subducted or faulted) to lower crustal levels. Subsequently, at about 2200–2000 Ma (Cliff et al., 1983), at these deep levels the sequence was extensively intruded by a range of ultramafic to anorthosite plutons and a network of dykes and smaller pods. The sequence of intrusive rocks shows a change from tholeiitic to calcalkline to alkaline basalt similar to the sequence in island-arc volcanism (Jakeg and White, 1969; 1972) and in suites from orogenic belts on continental margins. Although Dearnley (1963) postulated that the sequence represented a differentiated tholeiitic suite, later work by Palmer (1971), Witty (1973) and Horsley (1978) has shown conclusively that calcalkaline trends characterise the major igneous bodies. Hence, the layered gabbros and related ultramafics, which are extensively developed, belong to an olivine-tholeiite suite. The Langavat ultramafic pods may represent a refractory olivine-orthopyroxene residuum produced by fractional crystallisation of an original olivine-basalt magma to give the high-alumina intrusive bodies (diorite-norite-anorthosite). However, Livingstone (1976a) concluded they represent alpine-type peridotites and dunites emplaced into oceanic sediments during plate-tectonic processes. Horsley (1978) points out that this should result in about three times as much ultramafite material as high-alumina basalt. Hence, if the Langavat ultramafic complex and South Harris Igneous Complex are derived from the same parent, Livingstone's explanation would seem unlikely. The South Harris Igneous Complex shows a regional discordance to the metasediment banding whereas the ultramafic pods trend subparallel, implying that the Complex and the metasediments are unrelated in time. The c.2200 Ma age for the igneous suite also supports this implication. The minor ultramafic dykes within the South Harris Igneous Complex have affinities with olivine-tholeiites and alkali basalts; each major intrusive body has its distinct suite. These rocks are strongly hy- or ne-normative and Horsley (1978) suggests that they are derived largely by clinopyroxene fractionation from tholeiitic magma. It is clear that the Igneous Complex could be derived from one or more olivine-tholeiite parent magmas with differing amounts (or rates) of uprise and fractionation at lower crustal levels. The diorite and norite bodies have geochemical similarities which strongly suggest a single parent magma (Horsley, 1978). The minor later shoshonite dykes may represent about 5 per cent of partial melting of the upper mantle. The total sequence (tholeiite-shoshonite) is compatible with magma generation at progressively greater mantle depths.

The range of rocks now exposed appears to represent material originally formed at widely differing crustal levels, ranging from deposition at the surface down to the diorite body which probably crystallised at depths near to 35 km (Horsley, 1978) and the shoshonites which may have crystallised even deeper. The subsequent granulite-facies metamorphism which affected most of the rocks is a natural consequence of equilibration at the pressures and temperatures prevailing at these basal crustal levels. For the higher-grade assemblages (gt-cpx and opx-cpx in diorite and norite) they were around 1300 MPa (38 km depth) and 825 ± 25°C (Horsley, 1978; Dickinson and Watson, 1976). The growth of cpx ± pl in garnet-bearing norite and diorite indicates pressures around 800 MPa and temperatures of 760 ± 40°C (PH2O < 0.3 PTotal). Rb-Sr isotopic data (12 points) of Chapman (in Horsley, 1978) give an age of 2250 ± 160 Ma attributed to the granulite-facies assemblages in the diorite. The range of values is very small and by varying the points chosen, values of up to 2800 Ma may be obtained. However recent Sm-Nd isotopic measurements (Cliff et al., 1983) have been interpreted to show that the granulite-facies metamorphism of the South Harris Complex probably peaked around 2000 Ma and is not older than 2180 ± 60 Ma. The data give no indication that any of the igneous bodies are appreciably older than this. The data may be interpreted as giving an older age limit (Inverian?) for the deep-level granulite-facies metamorphism in South Harris (see also Dickinson, 1974; Horsley, 1978).

Subsequent Laxfordian deformation inverted much of the South Harris Complex, possibly by regional shearing and/or recumbent folding. Tight folding followed, mainly on axes plunging NNW to north-west and on north-west-trending axial planes (dipping steeply north-east or near vertical), and Myers (1968) records a corresponding amphibolite-grade metamorphic episode. In the igneous rocks mainly dextral north-west-trending shear zones developed extensively. The intensity of metamorphism progressively declined during this period (see Dickinson and Watson, 1976), and after abundant Laxfordian granite and granodiorite sheets had been injected, at around 1800-1600 Ma accompanied by increased heat flow, the area rose rapidly to near its present level. This uprise was accompanied by thrusting and pseudotachylite formation in the south-east parts of South Harris (see Chapter 15: Outer Hebrides Thrust Zone).

The recognition in the Karakorum range of north Pakistan of a Cretaceous island-arc system (Kohistan Arc), now exposed down to former upper-mantle levels, has interesting parallels with South Harris (Coward et al., 1982). A large layered norite-gabbro-ultramafite body, which recrystallised under granulite-facies conditions and intruded amphibolites, metagabbros (representing oceanic crust) and granulite-facies metasediments, is exposed. Diorite and granodiorite plutons of calcalkaline affinity occur at intermediate levels and flysch, pillow lavas and tuffs at shallow levels. Garson and Livingstone (1973) have suggested that South Harris may be an overthrust island-arc–ocean-crust sequence but place the main movement in the Laxfordian.

Before the South Harris Complex was folded it was perhaps about 60 km long. If it and the anorthosite-metasediment assemblage at Ness in North Lewis (McQuillin and Watson, 1973; Watson, 1969) were parts of the same complex, then the total outcrop length would have been more than 160 km. The recent Nd-Sm isotopic age date (2180 ± 60 Ma, Cliff and others, 1983) on the South Harris Igneous Complex rule this out if the Ness assemblage is of Scourian age. Modern arc systems are commonly in excess of 1000 km but available data from other Archaean areas (e.g. Glikson, 1976) suggest that smaller crustal plates may have been present in the Archaean. The implied crustal depth (35 km or more) for the m2 metamorphism and intrusion of the igneous suite in South Harris suggest that crustal thickness was of the same order as at present or possibly even greater.

Chapter 8 Lewisian Complex: Corodale Gneiss

The Corodale Gneiss, so termed by Coward (1972), is a distinctive, crudely banded clinopyroxene-orthopyroxenegarnet-bearing granulite-textured metadiorite, which forms the prominent ridge of hills on the eastern side of South Uist. First described by Jehu and Craig (1925) as the Crushed Gneiss, the Corodale Gneiss was subsequently described by Kursten (1957), Dearnley (1962a) and Coward (1972). The gneiss is everywhere separated from the 'grey gneiss' by tectonic contacts, and indeed its western contact forms the main feature of the Outer Hebrides Thrust Zone in South Uist. Its southern contact is difficult to trace in the field, lying in the complex tectonic melange north of Ru Melvick, but the boundary must approximate to a line from Hartabreck [NF 830 151] south-eastwards to Camas an Lochain [NF 833 135]. Corodale Gneiss forms the headlands of Rubha na h-Ordaig and Meall an lasgaich, is found on the side of Hartavagh [NF 830 158], and forms the high ground of Rudha Meall na Hoe [NF 826 171]. Across Loch Boisdale the outcrop widens and to the north Corodale Gneiss forms the distinctive summits of Triuirebheinn, Stulaval, Beinn Mhor and Hecla. In the north the western contact of the gneiss swings offshore at the mouth of Loch Skipport [NF 854 381]; it passes the island of Ornish and continues up through Caolas Luirsay. Luirsay Dubh [NF 866 406] and Luirsay Glas are thus the northernmost outcrops of Corodale Gneiss. To the east the gneiss is bounded by a belt of phyllonite (the Usinish mylonite of Jehu and Craig (1925)), which is best seen on the east coast of South Uist at Rubha Rossel, Usinish and Rubha Bolum, and on Stuley Island.

The western contact is a remarkably well-defined feature. Below the thrust plane the grey gneiss is virtually unaffected by cataclasis, while above it the Corodale Gneiss has been rendered down to a pseudotachylite mélange, the amount of pseudotachylite decreasing upwards. This melange forms a massive rock which stands out as an abrupt dark cliff face above the grey gneiss platform, a feature particularly well seen at Rudha Meall na Hoe [NF 826 171] and on Triuirebheinn [NF 807 206]. Above the thrust the degree of deformation and development of cataclastic fabrics is highly variable and it is probable that the Corodale Gneiss is affected by a series of minor internal thrusts. The sequence of deformational events is described by Coward (1972).

The Corodale Gneiss shows a well-developed banding on the scale of a few centimetres upwards to several metres. The banding is defined by the variation from ultramafic assemblages to plagioclase-rich assemblages. Coward (1972) suggests that in addition to this finer banding there is a much larger-scale compositional variation; most of the ultramafic rocks occur in the structurally lower levels, whereas the higher levels are marked by a preponderance of felsic rocks. Both Coward (1972) and Dearnley (1962a) report that the banding is cut by dyke-like basic masses which they correlate with the Younger Basics. Dearnley (1962a, pl. vii) figures an excellent example of a cross-cutting basic dyke. Some apparent cross-cutting relationships have, however, to be treated with scepticism. Late deformation has disrupted the banding and juxtaposed ultramafic layers and the banded sequence in an apparent cross-cutting relationship; films of pseudotachylite at the boundaries suggest that the contacts are not primary.

Petrography

The differing mineralogy of the rock is due in part to original banding and in part to subsequent retrogression and deformation. A generalised assemblage can be given as opx-cpx-gt-hbl-pl with accessory biotite, quartz, ilmenite, apatite and scapolite.

The 'original' assemblage was apparently opx-cpx-(gt)-pl, the amount of plagioclase ranging from zero to 50 per cent. The plagioclase is generally andesine/labradorite. Hornblende is always present, apparently secondary after pyroxene. Jehu and Craig (1925, p.733) mention a scapolitebearing pyroxene granulite from Hartavagh [NF 830 158] which from its description is probably Corodale gneiss. Scapolite has also been identified in the present investigation (e.g. (S62231)). Texturally the rock is equigranular, although it is everywhere affected to some extent by deformation, which locally has tended to give the rock a rather augen-like appearance. Incipient mylonitisation (low-grade recrystallisation) is confined to planes wrapping around the larger grains. Elsewhere the minerals show more brittle forms of deformation with little incipient recrystallisation.

Geochemistry

Very little chemical work has been published on the Corodale Gneiss. One full and one partial analysis are presented in (Table 23) and the full analysis plotted on an AFM diagram (Figure 10). The analyses show the gneiss to be ol-normative and broadly equivalent to a diorite (Le Maitre, 1976), although richer in soda and relatively low in potash. If the rock is of igneous origin some of these features may be due to later metamorphic depletion.

The age of the Corodale Gneiss

Kursten (1957), Dearnley (1962a) and Coward (1972) all regard the Corodale Gneiss as a meta-igneous body. Coward cites the banding and the general trend to more felsic assemblages at higher levels as evidence of magmatic layering. Coward (1972) and Dearnley (1962a) suggest that it can be correlated with the South Harris Complex. Both regard the South Harris Complex as post-Scourian, and imply that the granulite-facies assemblages seen in the Corodale Gneiss are of early Laxfordian age. (However, Coward would argue that pyroxene crystallised in regional upper-amphibolitefacies conditions because in these rocks the water partial pressure was relatively low. If the Corodale Gneiss was relatively dry, it would also follow that Laxfordian deformational events would probably be evidenced by shear zones and fracturing with concomitant pseudotachylite generation, as is seen in South Harris).

Both the South Harris Igneous Complex and the Corodale Gneiss have granulite-facies assemblages and igneous-like layering, both show similar deformational fabrics and have similar chemistry, all of which support a correlation between the South Harris Igneous Complex and the Corodale Gneiss. However, the recent date of 2180 t 60 Ma (Cliff et al., 1983) on the South Harris Complex which is taken (Chapter 7, p.74) to be close to the age of intrusion places the Complex as younger than the Younger Basics. If the dykes cutting the Corodale Gneiss are taken as Younger Basics then the correlation with the South Harris Complex cannot hold. If the dykes are regarded as equivalent to the lategabbro and ultramafic intrusions seen in the South Harris Complex then the correlation can still be made.

Chapter 9 Lewisian Complex: Ultrabasic Complexes

Scattered throughout the islands is a series of distinctive ultrabasic and layered ultrabasic/basic bodies distinguished from the Older Ultrabasics (Chapter 4) by their less altered state. They are apparently intrusive into the gneiss and characteristically form small knolls or knobs, with a yellowish-brown weathered crust. The rocks are mainly peridotites; locally dunites and pyroxenites are found. In some of the layered bodies the assemblages range from ultrabasic rock to leucogabbro. The rocks are relatively fresh.

Northern Isles

Ultrabasics are particularly common in North Harris and in south and central Lewis. They have been described by, among others, Dougal (1928), Jehu and Craig (1927, 1934), Myers (1971), Lisle (1974) and Soldin (1978). In addition, short descriptions and analyses of two bodies are given by Guppy (1956). The ultrabasic rocks generally occur as isolated knobs, some of which are aligned to form trails. At one locality on the south side of Loch Shell opposite Eishken [NB 304 092], a narrow belt of ultrabasic 'knobs' and debris can be traced at intervals for almost 2 km. Whether this represents a true dyke-like intrusion or a glacial debris trail is uncertain. In central Lewis a zone (3 km wide and about 10 km long) of particularly abundant ultrabasics can be defined (Map 1). Individual bodies range from a few square metres to 100 m by 50 m.

The rocks are generally peridotites with local variations to dunites. Jehu and Craig (1934, p.856) report an almost pure dunite from Brandarsaig [NB 088 059] in North Harris. The rocks are generally fresh; locally there is some marginal alteration. The contacts of the bodies are mostly not exposed, but they can usually be defined within a few metres. This suggests that if an altered zone is present it is thin. In addition to the marginal alteration, belts in which the rock is somewhat serpentinised traverse individual bodies. In general the ultrabasic rocks are free of deformational fabrics.

In thin section the rocks are seen to consist of olivine, showing various degree of serpentinisation and large plates of orthopyroxene which may include the olivine. The pyroxene plates are up to 3–5 cm long. Hornblende is invariably present, either as small equigranular grains or as large almost porphyroblastic crystals; both varieties are almost certainly secondary. Phlogopite is also common, and to a lesser extent clinopyroxene. Magnetite, chromite and chrome spinel are accessories.

One of the largest and best-described bodies is that at Maaruig in North Harris [NB 202 062]. The mass lying on the north side of Maaruig Bay covers about 0.3 km2. It was extensively studied by Soldin (1978), who concluded that it was a layered cumulate complex. Soldin (p.157) gave the main assemblages from base to top as ol-crm-amph, cpx-ol-crm-amph, cpx-ol-amph-pl. From the distribution of these assemblages Soldin (fig. 4.23) concluded that the mass is arranged in a central anticline flanked by two synclines. He found it difficult to match modal variation to the fine layering, but a cryptic variation was defined by the mineral chemistry. The mineralogy given by Soldin (pp.160 and 193) shows olivine ranging upwards from Fo82 to Fo88, orthopyroxene ranging from En87 to En90 and the plagioclase at the top as An56, with accessory biotite, spinel and chromite. In some specimens the orthopyroxene was partly altered to amphibole. From various theoretical considerations Soldin (p.209) suggested that the mass was formed by the co-precipitation of olivine (Fo88) and orthopyroxene (En84Fs10 Wo6). Because these theoretical compositions for the precipitating phases are close to the actual values of the present minerals, he concluded (p.209) that the olivine and orthopyroxene may be primary, but the amphibole is wholly secondary, for it does not appear in the theoretical assemblages.

It is interesting to note that the Maaruig body, although layered, is essentially ultramafic throughout. This feature seems to be a characteristic of many of the bodies examined, namely that they appear to be isolated in the grey gneiss and not associated with any form of basic rock. One possible exception is found at Loch Ionsil [NF 15 36] on Bernera. There Lisle (1974) has recorded a strip of ultrabasic rock lying next to an olivine-gabbro member of the Younger Basic suite (a 'Cleitichean Beag' type: see Chapter 10). However, the contacts of the two rock types are not exposed and their true relationship cannot be defined.

Southern Isles

Ultrabasic rocks are found at scattered localities in the southern islands, mainly in North Uist and the northern part of South Uist. The main occurrences are, at Craig Haston [NF 743 668], Creag Loisgte [NF 823 447], east of Ben Tarbert [NF 813 393] and at East Gerinish [NF 822 447] in South Uist. The mass at Creag Loisgte is largely a pyroxenite with large (3–4 mm) crystals of pyroxene partially retrograded to an aggregate of fine (c.0.2 mm) equigranular crystals of amphibole. On its south side the mass is in contact with a fine-grained basic which contains an equigranular cpx-gt-hbl-pl-opq assemblage, although whether this represents a later intrusion or a facies of the ultrabasic is unclear. The mass on the east side of Ben Tarbert is also a pyroxenite, more actinolite-rich at the margins than within. It contains opx-cpx-hbl assemblages, with the orthopyroxene crystals ranging up to 5–6 mm. One of the most interesting bodies is at East Gerinish. It consists of a largely undeformed, layered cumulate sequence ranging from ultramafic at the base to leucogabbro. Deformation is confined to shear zones. Generalised assemblages through the body can be given as:

Within the leucogabbro various fine- and coarse-grained facies can be identified. The fine-grained facies is probably due to deformation. The generalised assemblages are:

The minerals mostly have replacement textures, although some of the large (c.10 mm) crystals of clinopyroxene may be primary. The body has suffered various degrees of retrogression and alteration, involving changes from orthopyroxene to clinopyroxene, clinopyroxene to hornblende, anthophyllite to hornblende, hornblende to biotite, and anthophyllite perhaps to talc; plagioclase also shows evidence of alteration.

Quartz and ore occur in some specimens as accessories. Where sheared the minerals may show complete retrogression to hornblende or hornblende-plagioclase assemblages. Chemically the East Gerinish complex shows a strong calcalkaline trend ((Table 23); (Table 23), analyses 1-5. A ultrabasic bodies, (Table 23), analyses 6-8. layered basics, (Table 5), analyses 7-11. O anorthosite from layered complex, (Table 5), analysis 6. x Late-Scourian 'microdiorites', (Table 5), analyses 12-15. * Late-Scourian 'granites', (Table 5), analyses 16-17. range of members of mainland Scourie Dyke suite from Tarney (1973). Younger Basic metadolerites (Outer Hebrides: this work). 'Cleitichean Beag' basic from Dearnley (1963, table 7, analysis 4). 0 noritic rocks from Soldin (1978, figure 4.11); 'average' of Maaruig ultrabasic from Soldin (1978, figure 4.11)." data-name="images/P936484.jpg">(Figure 11)).

The age of the ultrabasic rocks

The age of the ultrabasic rocks is uncertain. They postdate the Scourian metamorphism, and it is therefore tempting to associate them with the Younger Basic suite. Certainly Myers (1970b) and Lisle (1974) regard them, along with rocks of the 'Cleitichean Beag' type (olivine gabbro, picrite, norite), as the most basic members of the Younger Basic suite and as such equivalent to the bronzite picrites etc. of the mainland (Tarney, 1973). Soldin (1978) also regards the Maaruig ultrabasic as closely related to a suite of noritic rocks that might be considered as a more basic facies of the Younger Basic suite. However, in an extensive examination of the chemistry of Younger Basic amphibolites, and of the ultrabasics and norites, he concludes that the differences are so great that although they may be related they cannot be regarded as part of a coherent chemical suite. Support for a general relation between the Younger Basic suite and the ultrabasics is given by the orientation of the zone of abundant ultrabasics in central Lewis, for it is broadly parallel to the zone of abundant basic rocks of 'Cleitichean Beag' type in southern Bernera, as well as the zones of abundant metadolerites of the Younger Basic suite (Watson, 1968, fig. 3) in northern Bernera. One difficulty in solving the nature of any relationship is the lack of visible contacts between rocks of the two groups. Although Jehu and Craig (1925, p.627) report an ultrabasic body (ol-cpx-opx-spinel) from Loch Stulaval in South Uist which is cut by a small dyke with chilled edges, this does not preclude the two rocks belonging to the same general magmatic episode.

In South Uist, Coward (1969, p.324) regards the East Gerinish body as a relatively undeformed equivalent of the banded basic and associated ultrabasic (harzburgite) which lies east of Arnaval (Chapter 4: Banded Basics of the Uists and Benbecula: Petrology). However, the chemical characteristics of the East Gerinish body (it is calcalkaline) seem to distinguish it from both the banded basics and the Younger Basic suite, which are tholeiitic.

It is, therefore, impossible at this stage to be certain of the age of the unaltered ultrabasics, although on balance we think that most are probably of Younger Basic age.

Whatever their age, the mode of emplacement of the ultrabasics is unusual; in general they occur as isolated plugs, and even where, as some do, they form trails, they are generally too widely separated to be boudins of a tectonically disrupted dyke. The bodies therefore have been emplaced as lensoid masses. It is possible that they are gravitationally transported cumulate restites from a fractionating magma.

Chapter 10 Lewisian Complex: Younger Basics

The Younger Basics are defined as a suite of basic rocks whose intrusion postdates the main phases of Scourian migmatisation and deformation. In areas where Laxfordian deformation is low or moderate it is relatively easy to distinguish Younger from Older Basics because the Younger Basics are discordant to the Scourian gneiss fabric. In areas of high deformation where all the basic rocks are generally concordant it is still possible to tell them apart where the Older Basics show a higher degree of migmatisation than the Younger Basics. Only in a relatively few areas, where Laxfordian remobilisation and migmatisation have been particularly strong, affecting all the basic rocks, is it impossible to tell them apart.

It was Dearnley (1963) who first suggested on chemical as well as structural grounds that the Younger Basic rocks of the Hebrides were equivalent to the Scourie Dykes of the Scottish mainland. Subsequent workers have generally accepted that correlation, at least in part, although some have argued (e.g. Taft, 1978; Hopgood and Bowes, 1972) that several of the Younger Basics are of intra-Laxfordian age. This point is fundamental to the interpretation of much of the geology and has already been argued (Chapter 1); we concluded that almost all the Younger Basic intrusions are pre-Laxfordian in age. It should be emphasised, however, that there was a considerable time span from the cessation of Scourian migmatisation and deformation (at about 2450 Ma) to the first major movements of the Laxfordian (at about 1800 Ma), this period covering the Inverian. The Younger Basics could thus have been intruded during this interval (cf. Evans and Tarney, 1964), and in consequence there may well be a number of magmatic lineages. The terms 'Scourie Dykes' and 'Scourie Dyke age' should therefore be read with this in mind. Nevertheless, most of the dykes do appear to form a coherent suite which is part of the great North Atlantic swarm (Escher et al., 1976a). This swarm, which trends between east and. south-east, runs from Scotland across Greenland to Baffin Island affecting 250 000 km2 of continental crust (Escher et al., 1976a).

On the Scottish mainland Tarney (1973) divides the suite into several groups: bronzite picrites and olivine gabbros, early dolerites, the Loch Croach peridotite, noritic dolerites, and late tholeiitic dykes. In the Hebrides there are a few noritic and picritic bodies, but as their relative age is uncertain they have been lumped together under the descriptive title of 'Cleitichean Beag' dykes, after one of the largest and best described members of the group, which is at Cleitichean Beag [NB 271 367], near Breasclete in Lewis. The ultrabasic layered dunites and peridotites described elsewhere (Chapter 9) could, perhaps, be assigned to the same suite. The remainder of the pre-Laxfordian Younger Basic rocks are classified under the heading of metadolerites.

Field aspects

The Younger Basics have a wide variety of appearance in the field ((Plate 10), (Plate 11), (Plate 12), (Plate 13)), ranging from large coarse-grained lensoid bodies and discrete parallel-sided cross-cutting dykes, to bodies that have been almost completely digested in remobilised gneiss (see p.14 Chapter 2: Uist and Benbecula; Laxfordian migmatisation). Almost all, however, are concordant to subconcordant bodies, averaging 1-3 m in thickness and generally discontinuous along strike. Individual members are locally traceable for several hundred metres as series of pods and lenses.

'Cleitichean Beag' dykes

The 'Cleitichean Beag' dykes have been recognised in North Harris and Lewis (see also Chapter 9), where they lie predominantly in two east–west belts, the southern one running from the mouth of Loch Seaforth westwards to Hushinish, and the northern one from Cleitichean Beag to Great Bernera (Map 1). Within these belts, however, occurrences are not numerous and individual bodies may be separated by several kilometres. Indeed Jehu and Craig (1934, p.852), who originally described the group, recorded just 17 localities, and only a few additional dykes have been noted subsequently (Myers, 1971; Lisle, 1976; Soldin, 1978; and the present investigation). It should be emphasised that the 'Cleitichean Beag' dykes were initially distinguished from the normal metadolerites on their field appearance, being generally coarser and more massive. It is not easy, however, to distinguish between 'Cleitichean Beag' dykes and some of the large coarser-grained metadolerites, for example in North Uist and the Park area of Lewis. Indeed some rocks which have been considerably recrystallised differ only in chemistry. In consequence it is possible that some 'Cleitichean Beag' dykes have been mapped as metadolerites (metabasics on Maps 1 and 2).

The 'Cleitichean Beag' dykes generally form large, rather craggy features of considerable extent, the type dyke itself covering about 0.5 km2. They seldom show marked elongation; where present the elongation is normally east–west although Cleitichean Beag itself is elongated NW–SE. The contacts with the gneiss are generally not exposed but they appear to be mainly concordant, without, as far as can be seen, apophyses or branching dykelets. The rock has a grey to brown, rather pitted weathered crust and a coarse texture, and is composed of large crystals of pyroxene and plagioclase. It is generally undeformed and the coarse-grained textures persist up to the margins, except where there are retrograded margins with amphibole and biotite prominent. Retrogression also accompanies localised internal shear zones.

The relative ages of the individual dykes are uncertain. They are thought, by the present authors, to be broadly contemporaneous because they tend to concentrate in zones. Their relationship to the metadolerites is also obscure, although one excellent outcrop from near Lundale [NB 181 326] in Lewis shows a 1–2 m metadolerite cutting a large noritic body. Elsewhere the masses are cut by fine-grained dykes (e.g. at Cleitichean Beag (Lambert et al., 1970b)), although it is not known if these are metadolerite.

Jehu and Craig (1934, p.852) subdivide the group into norites and olivine norites, the norites being the more abundant. As examples of the norites they describe two localities in North Harris: the more southerly is on the shores of Loch Seaforth near Trilleachan-beag [NB 206 085], with a reported area of 0.4 by 1.2 km; the other, the larger, is known as the 'Ardvourlie dyke' and can be traced from the shore south of Creag Chaise [NB 202 094] to the head of the Langavat valley [NB 144 105], a distance of 6 km with an average width of 100–150 m. Soldin (1978) presents a detailed map of the eastern end of this dyke, where patches of norite are surrounded by amphibolite which he interprets as the retrograded equivalent of the norite. Myers (personal communication) suggests, however, that the norite may well be cutting the amphibolite.

Metadolerites

The metadolerites constitute by far the greater part of the Younger Basic suite. They include a large variety of basic rocks ranging from discrete cross-cutting dykes to highly deformed and recrystallised amphibolites. They are found extensively throughout the islands, although in terms of crustal extension Watson and Lisle (1973) estimate that they represent only about 5 per cent. The dykes tend to concentrate in zones (Watson, 1968; Myers, 1970b; Map 1) which in places are oblique to the trend of the overall gneiss foliation even where individual members of the zone are largely concordant. Elsewhere (Chapter 13: The Younger Basic suite and first Laxfordian events) we argue that the original strike of the dykes was probably between east and south-east, thus conforming with those of mainland Scotland and Greenland (Escher et al., 1976a).

The metadolerites range in size from large bodies to small stringers and pods. Some of the largest and least deformed examples define a distinctive zone across the north of North Uist, where they form many rocky knolls around Vallay Strand as well as the summits of Crogary Mor and numerous other small hills. By far the largest is the one centred around Carra-crom [NF 734 735], a roughly lensoid mass measuring 3 by 1 km. There are also similar large masses in the Park area of Lewis. In general, however, the bodies seldom exceed 0.2 km2. They tend to be very coarse-grained massive rocks with an equigranular mosaic of pyroxene, garnet, hornblende and plagioclase. Localised shearing causes retrogression and the development of deformational fabrics. The margins of these large bodies are generally obscured but where seen are commonly foliated amphibolites. Similar-sized bodies also occur in Lewis between Loch Grunavat and Little Loch Roag, but there they are commonly foliated and largely amphibolitic.

Many of the large bodies exhibit a crude layering generally arising from variations in felsic content, although in places it is also due to variations in garnet content or even to the degree of alteration of garnet to plagioclase. This is well seen in the body at Alioter [NF 890 728], in some of the bodies around Beinn Mhor, North Uist [NF 898 762] and in the large body south of Caiteshal in south-east Lewis [NB 243 035]. The banding may survive even when the body has been largely recrystallised, for example on Roineabhal in Lewis [NB 233 212], and on a smaller scale in a small body near Loch Claidh [NB 261 033]. It seems possible that the banding might represent a modified form of igneous layering; if it does, these large bodies were probably sills as opposed to the smaller bodies, which have a dyke-like form.

Most of the metadolerites are relatively small, seldom exceeding a few square metres. In general they are now concordant or subconcordant; a few are markedly cross-cutting. Individuals can seldom be followed for any distances; a few have been traced for up to 2 km. The cores of the larger dykes are similar to the large bodies described above, commonly having an equigranular and rather mottled appearance characteristically with conspicuous garnet, the texture becoming foliated and retrograded towards the margins. Some of the dykes have a coarse-grained ophitic, almost gabbroic appearance, although the 'igneous' minerals have been pseudomorphed by fine-grained aggregates. In the area east of Little Loch Roag, for example on Aird Orasay [NB 129 315], and near Loch Crocach [NB 152 306] some of the small basic bodies are crossed by a series of small (about 5 cm) locally branching veins, which are characteristically garnet-rich with subordinate amounts of plagioclase. It is probable that these veins represent some form of late igneous veining.

All the dykes show some evidence of Laxfordian deformation; the amount varies greatly. With increasing deformation the equigranular rock is progressively foliated and sheared until it becomes a 'simple' hornblende-schist. Shear zones, many spectacular, have developed within the basics, and examples from the mass at Caisteal Odhair [NF 726 765] in North Uist were figured by Ramsay and Graham (1970). At the margins of the shear zone garnets can be seen rimmed by plagioclase; further into the zone they are completely pseudomorphed, and finally the feldspar pseudo-morphs themselves have been extensively strained and flattened.

The extensive 'Laxfordianisation' of the dykes makes it generally difficult to assess what their contacts with the gneiss country rock were like originally. In areas where the Laxfordian strain is low it is possible to make some general observations. The dykes have sharp, clean-cutting contacts and the fabric of the gneiss is unchanged; there is no evidence of melting or hornfelsing in the gneiss, nor of chilling in the dykes. The dykes characteristically have small apophyses or branching dykelets, for example at Ardivachar [NF 740 462] and Garry-a-siar [NF 756 532] (Dearnley and Dunning, 1968; see also Watson and Lisle, 1973), thus suggesting a comparison with many of the dykes described from Greenland by Escher et al. (1976a), who took these features to indicate synkinematic intrusion. It is probable that some of the dykes were intruded into a series of active shear zones, which in places followed earlier ductile shears, but elsewhere were caused by the dykes own magmatic pressure. The propagation direction was largely controlled by the regional stress system, which had controlled regional structures both before and after dyke intrusion. Watson and Lisle (1973) suggest that this was so for the whole North Atlantic swarm, a view supported by work on the Scottish mainland. For example, Tarney (1973), and Park and Cresswell (1973), present strong evidence to show that the emplacement of the Scourie Dyke suite was controlled by Inverian shear zones in the Assynt and Torridon areas of north-west Scotland respectively. J. S. Myers (personal communication) has recently applied these concepts to the Outer Isles. He cites a dyke that crops out on the south side of Aird Fenish at [NA 9920 2926]. Here a complex folded Scourian gneiss sequence is crossed by a shear zone 400 m long and trending NE–SW, within which the gneiss foliation is intensely streaked out and recrystallised. Within the shear zone and close to its margin is a 1 m metadolerite which is slightly cross-cutting and contains only a weakly developed fabric (Plate 13). Elsewhere, examples of relatively undeformed gneiss are seen. According to Escher et al. (1976a) and Park and Cresswell (1973) the metabasic dykes in Greenland and the Scottish mainland become markedly thicker and discordant as they leave the shear zones. There was, therefore, a marked structural control over the mode of intrusion of these dykes. In the Outer Isles, however, it seems that most dykes do not lie in pre-existing shear zones and originally may well have been highly discordant. This is based on the observation that where Laxfordian strain is low the dykes are strongly discordant. Care must, however, be taken to try to assess the degree of Laxfordian strain on other criteria than the structural state of the Younger Basics. It could be argued, for example, that Laxfordian deformation was concentrated in the areas where the dykes are largely concordant, that is in the areas of postScourian movement. Much has been written about the response of the dykes to the phases of Laxfordian deformation, leading to their reorientation into concordant attitudes (Hopgood, 1965; Francis, 1973; Coward, 1973a; Myers, 1970b). This reorientation can take place over a few metres and several examples have been figured by Coward (1973a). Care must be exercised in deciding between genuine reorientation and the possible variation of a dyke entering a shear zone. Francis (1973) suggests that the early mineralogy of the dykes may well have affected their response to Laxfordian deformation, dykes with granulite facies cores being more competent than those with amphibolite facies assemblages. He shows several examples of structures produced in dykes of variable competence relative to gneisses.

Petrography

Tleitichean Beag' dykes

These rocks can be treated as a series of norites and picrites (cf. Jehu and Craig, 1934, p.852). The picrites, as exemplified by the Cleitichean Beag intrusion itself, have been described by Jehu and Craig (1934, p.854) and Dearnley (1963, pp.273–274). The rock has an assemblage opx-cpxol-hbl-pl-opq with minor amounts of biotite, spinel, magnetite and epidote. Dearnley (1963, table 7) gives the following mode for his analysed specimen: plagioclase (An60) 16.3 per cent, orthopyroxene 44.2 per cent, clinopyroxene 17.2 per cent, olivine 15 per cent, hornblende 5 per cent, and biotite 2 per cent. In hand specimen the rock has a relict igneous texture with large plates of orthopyroxene, fresh olivine and clouded plagioclase with a characteristic green colour. Dearnley (1963) describes the orthopyroxene as enstatite bronzite, with a rim of clear secondary orthopyroxene. At the junction of olivine (fosterite) and plagioclase grains, coronas of clinopyroxene have appeared, some of which have completely replaced the olivine. Garnet is absent from these coronas. Secondary clinopyroxene and labradorite form small equigranular grains at the margins of the orthopyroxene. Biotite is at least in part secondary. Petrographically these rocks are similar to the olivine gabbros or bronzite picrites described by Tarney (1973) from the Scottish mainland.

The noritic rocks at Ardvourlie have been described by Jehu and Craig (1934, p.853) and Soldin (1978), and the following description is largely based on their work.

The typical assemblage is opx-pl-hbl-(gt)-opq. Orthopyroxene (En72-76) and plagioclase (An53-66) are the dominant minerals with orthopyroxene forming 50 per cent of the mode and plagioclase 35 per cent. The orthopyroxene is partly altered to small grains of secondary clinopyroxene and hornblende. Plagioclase is generally twinned, and has antiperthitic intergrowths. Garnet is present in places at the boundary between plagioclase and orthopyroxene. Biotite appears to be largely secondary, and is more common with increasing recrystallisation and retrogression. Accessories include ilmenite (less than 2 per cent of the mode), magnetite, pyrrhotite, epidote and some rutile. Soldin (1978) suggests that the noritic assemblage of orthopyroxene and plagioclase passes into the surrounding amphibolite, the orthopyroxene having first altered to garnet and pargasitic amphibole.

The norite sheet near Lundale has large corroded plates of orthopyroxene set in a matrix of equigranular clinopyroxene, orthopyroxene, pale amphibole, plagioclase (oligoclase/andesine) and flakes of brown biotite.

These rocks are petrographically similar to the noritic dolerites on the mainland of north-west Scotland (Tarney, 1973).

Metadolerites

The petrography of the metadolerites has been extensively described (e.g. by Jehu and Craig, 1923; Dearnley and Dunning, 1968; Soldin, 1978). The main assemblages are listed by Dearnley (1962a, 1963). The mineralogy and textures of the metadolerites are highly variable, depending on the intensity of deformation and metamorphism which the rock has suffered. The assemblages range from relict igneous through granulite-facies to lower-amphibolite-facies. The textures range from primarily ophitic and subophitic to equigranular and schistose (Plate 14). Within any dyke there may be various assemblages and textures. Granulite-facies assemblages, where present, tend to be restricted to the cores of bodies which have amphibolitic and foliated margins, most small bodies are amphibolitic throughout. Each phase of Laxfordian deformation has tended to overprint the dyke to some extent with new fabrics and further amphibolitisation and recrystallisation. This leads to a complex sequence of metastable assemblages relating to different deformational phases within the dykes. For descriptive purposes, however, five assemblages can be considered, viz:

1 Relict igneous: Although none of the dykes shows completely unaltered igneous assemblages, several contain large plates of partly altered clinopyroxene and plagioclase with an ophitic texture. More rarely orthopyroxene may be present, apparently as a primary mineral. The pyroxenes show various degrees of alteration to finer-grained equigranular aggregates, clinopyroxene going to a complex of new clinopyroxene and hornblende (e.g. (S58589)). In addition the plagioclase has recrystallised to smaller, less calcic grains and garnet has appeared at plagioclase/clinopyroxene boundaries (e.g. (S58584)). Orthopyroxene, in small grains, is less common (e.g. (S61511)). In all the rocks of this type that were examined, hornblende seemed to be a secondary phase after pyroxene; however, some may well be primary (see below).

2 Granulite facies: With complete recrystallisation of the igneous texture the rock is composed of finer-grained equigranular aggregates with typical granulite-facies assemblages: opx-cpx-pl-hbl-opq; gt-cpx-pl-hbl-opq; and opx-gt-cpx-pl-hbl-opq. The plagioclase is generally mid-andesine. The hornblende is brownish and largely secondary after pyroxene. Quartz is as an important accessory. Scapolite is rare. Examples of these assemblages are contained in (S58053), (S58095) and (S58146).

3 Subgranulite facies: With increasing retrogression hornblende is the dominant phase. Orthopyroxene quickly disappears and garnet becomes rimmed by plagioclase (e.g. S 1962); the general assemblage can be given as: hbl-cpx-gtpl-qz-opq, which is clearly unstable. It is probable that most of the rocks were tending towards a stable assemblage of hbl-cpx-pl-opq-(sph)-qz, although in the less calcic varieties, which were apparently rarer, the stable assemblage is hbl-gtpl-opq-(sph)-qz. In the less calcic varieties the garnets may grow to 10–20 mm. The plagioclase is generally oligoclase/ andesine and ilmenite is mantled by sphene (e.g. (S58777)). Examples of these rocks are (S57157) and (S57154).

4 Amphibolite facies: These rocks have a simple assemblage: hbl-pl-sph-qz. The hornblende by this stage has developed into large crystals, although small relic cores of clinopyroxene are still to be seen in places. Quartz was now an important phase. Garnet persisted in the more basic layers, but it has generally been completely pseudomorphed by plagioclase.

5 Lower amphibolite facies: In these rocks biotite is an additional phase, although generally minor, (e.g. (S57133)) to give hbl-pl-qz-bi-sph. In some rocks epidote, scapolite and blue-green hornblende are also present.

In all of these assemblages various textures may be found. Relict ophitic and subophitic textures are found in granulite-facies rocks (S60059), and persist into some group 5 assemblages (S57133), in which concentrates of mafics and plagioclase have pseudomorphed the original igneous textures. Such relict ophitic textures persist even to the margin of some metadolerites. The degree of schistosity is not uniform, and highly schistose textures occur locally in groups 3–5. In general, the degree of schistosity development and the amount of hornblende are related, both commonly increasing towards the dyke margins.

Relating the various assemblages to deformational episodes is problematical. Where there are penetrative fabrics, it is possible in some areas, but it is difficult with the equigranular textures. For example, generally it is not feasible to assess the age of equigranular hornblende, since the amphibolitisation of the granulite-facies assemblages and textures may have been a continuous process spanning several deformational phases.

The regional degree of amphibolitisation of the dykes can be usefully expressed by reference to the assemblages present in the cores of the larger dykes. A plot of such assemblages is given in (Figure 12). This shows that the areas of lowest grade broadly coincide with the areas of highest Laxfordian deformation, although in western North Harris low-grade assemblages cover an area of only moderate Laxfordian deformation (cf. Myers, 1970b, fig. 1, and Myers, 1971, fig. 7). The possible significance of this will be discussed later (p.96).

Mineralogy

Amphibole

Analyses of an amphibole from a cross-cutting metadolerite dyke in North Uist are presented in Appendix 1, (Table 1). The dyke has a good granoblastic texture with equidimensional crystals. The analyses place the amphibole in the ferroan pargasitic hornblende field. Four amphiboles from Younger Basics in North Harris (Soldin, 1978), although somewhat more edenitic, have similar compositions.

Garnet

Analyses of garnets from five metadolerites are presented in Appendix 1, (Table 2). In addition to garnet, the rocks all contain clinopyroxene and various amounts of orthopyroxene and hornblende. The textures are generally granoblastic, although in one specimen (S62282) large plates of clinopyroxene show partial alteration to a mosaic of small equidimensional grains of clinopyroxene, and where the clinopyroxene is in contact with large crystals of plagioclase a reaction rim composed of a mosaic of small garnets has developed. The analyses show that the composition of the garnets varies very little, an average composition being Al60An2Gr18Py16Sp4. This analysis is very similar to that given by Turner (1968, p.328) as a composition typical of basic rocks at the amphibolite-granulite facies transition. The relatively high percentage of the grossular component is undoubtedly due to the breakdown of bytownite. One crystal (S62266) shows a slight increase in Fe/Mg from core to margin but we do not regard this as being of major significance. These analyses are very similar to those presented by Soldin (1978) from metadolerites in North Harris. Soldin (1978) shows that garnet in the norite is more magnesian than in the metadolerites, with Al: Py at 4:1, in which Al: Py = 7.1.

Pyroxene

Analyses of six orthopyroxene and eight clinopyroxene crystals are presented in Appendix 1, (Table 3). All are from cross-cutting metadolerites with good granoblastic or partly altered igneous textures. One large orthopyroxene (S62282) and two clinopyroxenes (S61964) are probably igneous crystals because they are only partly recrystallised to a granoblastic mosaic. The pyroxene compositions are plotted in (Figure 13) along with the results from Soldin (1978). The clinopyroxenes form a remarkably close group within the salite field. The two large crystals that are thought to be igneous have identical compositions to those in the granoblastic-textured metadolerites. The orthopyroxenes plot mainly in the ferrohypersthene field with the exception of crystals from one sample (S62282), these are believed to be igneous and are more magnesian. The different composition may indicate a slight rise in Fe/Mg during recrystallisation.

Scapolite

Forty-seven analyses of scapolite to within error limits of ± 2% were obtained from 9 specimens of Younger Basic rocks. The specimens were taken from central and south Lewis, and North and South Uist after petrographic indentification of scapolite. Representative analyses are given in Appendix 1, (Table 4). Scapolite composition is notably calcic, CaO values ranging from 9.8 per cent to 17.7 per cent. They lie mainly in the mizzonite field (Me41-72) although those with the lower meionite (Me) values fall in the dipyre field (see Shaw, 1960). Kwak (1977) related the meionite content of the scapolite to the grade of metamorphism for a series of calcsilicates and subsidiary metadolerites and scapolite schists in South Australia. He found the rocks of the amphibolite facies to be characterised by meionite values between Me50 and Me73, apparently independent of bulk rock composition except where the original rock was an evaporite. He noted that scapolites were sulphur-free even when adjacent to sulphide veins.

The widespread occurrence of scapolite in Hebridean metadolerites is most probably a consequence of their intrusion into the gneisses under pressure-temperature conditions of the upper amphibolite/granulite facies. The fluid phase in the gneisses under such conditions would be predominantly CO2, which would promote scapolite formation (see Newton and Goldsmith, 1975). The basic rocks themselves may also be a source of CO7 (see Touret, 1971). Hence when the ig neous plagioclase equilibrated to local conditions, scapolite was formed (e.g. (S71853)). The low SO3 contents (all <0.05 per cent except for one analysis from (S60059) which gave 0.25 per cent SO3) preclude a primary igneous origin for the scapolite. Provided that the dykes did not become hydrated as pressure and temperature fell following their intrusion (giving epidote/clinopyroxene), the plagioclase-scapolite mineralogy would not be superseded, but would equilibrate to progressively lower metamorphic conditions (e.g. (S61319), (S60011)). The excess CaO resulting from this re-equilibration results in sphene formation and an increase in the amount of modal hornblende.

Specimen (S62250) from Loch Ollay in South Uist has an unusual assemblage: it contains coarse-grained clinopyroxene, green hornblende and plagioclase (An45-60) with large calcic scapolites (Me61-73) and a high sulphur component (SO3 = 3.2 per cent to 3.96 per cent). This is compatible with a deep crustal uncontaminated basic intrusion (see Goldsmith, 1976), but field observations show that it is a migmatised, boudinaged, felsic-veined amphibolite. It is significant that nearby metadolerites contain the assemblage opx-gt-cpx, showing that they recrystallised at or about the amphibolite/granulite facies boundary. (S62250) may in fact be an Older Basic rock deformed and migmatised during the Scourian event, particularly as the locality is close to Rudh' Aird-mhicheil where a thick, banded basic body crops out (Chapter 4: Banded Basics of the Uists and Benbecula).

Feldspar

Analyses of feldspar were obtained from the same nine metadolerites that had been selected for scapolite analysis. The eighty-five analyses lie within total error limits of ± 1.2% . These analyses, in conjunction with those of scapolite, were undertaken to try to determine the pressure-temperature conditions during and subsequent to Younger Basic intrusion.

In only one specimen (S60110) was potash feldspar found, the remainder containing plagioclase with an overall compositional range of An27 to An85. Orthoclase content of the plagioclase ranges from 0 to 1 per cent (except (S62220)). Zoned feldspars are not abundant and where observed have a core with a high An content (bytownite-labradorite) and a marginal zone richer in Ab (andesine-oligoclase). Feldspar zoning is thought to result here from partial crystal equilibration to progressively lower pressure-temperature conditions. Only specimens with relict textures that might be igneous ((S61853), (S60059), (S60087)) show An values in the bytownite (An70-90) range. These values are typical of the upper part of the feldspar compositional field for basalts (Brown, 1967) which lies between An50 and An85. Even in the Younger Basic samples with bytownite feldspar, however, there have been textural changes, commonly as a result of downgrading, e.g. in (S61853), where bytownite grains (An74-78) lie adjacent to andesine-labradorite grains (An49-52).

In the remainder of the specimens, which exhibit metamorphic textures, feldspar compositions range from An28 to An50 with concentrations around An30 and An48. The higher values are typical of the recrystallised, equigranular metadolerites which still retain clinopyroxene and orthopyroxene ((S60011), (S60079)). In general An values became progressively lower as the metadolerites were increasingly amphibolitised. Rarely (e.g. (S60110) from the centre of a cross-cutting basic dyke 3 to 4 m wide at Rarnish, Benbecula) feldspar lies in the range Al28-32 and yet relict clinopyroxene and minor orthopyroxene and garnet remain, suggesting that plagioclase feldspar may have equilibrated to lower pressure-temperature conditions before the pyroxene was fully retrograded.

Specimen (S62220) is unusual in that the rock contains large biotites, apparently of primary origin, in a coarse-textured cpx-gt-pl-hbl-bi rock. The felspar has a surprisingly low An content (An27-48, mainly in range An27-32 and contains 0.8 to 2.8 per cent (averaging about 2 per cent) orthoclase component. In view of the texture and mineralogy, we suggest that feldspar compositions in this specimen may have originally been more sodic than in the other basics.

In specimen (S60110) potash feldspar (5 analyses), containing between 0.5 and 1.12 per cent BaO and 0.38 to 1.2 per cent Na2O, occurs as anhedral grains and as inclusions within the plagioclase. The compositions suggest that the observed potash-feldspar veining and associated migmatisation in the sample have resulted from a widespread K2O enrichment.

Geothermometry

The granoblastic textures present in many of the metadolerites suggest that the co-existing garnets and clinopyroxenes have equilibrated against each other; the garnets are small and unzoned in these rocks. It is therefore possible to use the methods proposed by Raheim and Green (1973) to obtain estimates of the temperatures at which the assemblages equilibrated. Three specimens from the Uists were examined ((S62087), (S62266) and (S62282), Appendix 1) and four from Harris (data from Soldin, 1978), including one from the 'norite' at Ardvourlie. Assuming a pressure of 400–700 MPa the specimens indicate an equilibration temperature of 620-640°C (note that variation of pressure outside these limits does not greatly alter the temperature (Raheim and Green, 1973, fig. 11)). Although it would be unwise to draw too firm conclusions from these figures certain tentative observations can be made.

1 There is no significant difference between the values for Harris and those for the Uists, suggesting that the dykes were intruded at approximately the same structural levels in both areas, assuming similar geothermal gradients. This is in marked contrast to the central region of the mainland, where O'Hara (1977) suggests 450°C as the temperature of metamorphism at the margins of Scourie dykes. This temperature difference is consistent with textures and mineral assemblages in the two areas and supports the model for emplacement of the dykes discussed above.

2 Temperatures of 620–640°C at 400-700 MPa are consistent with the upper amphibolite facies (see Turner, 1968, fig. 8–6) and support the model proposed above, in which the dykes were intruded into gneisses which were recrystallising under such conditions; the dykes contain garnet- and clinopyroxene-bearing assemblages which have equilibrated, and they are thought to have done so because the partial pressure of water relative to total pressure was low. Soldin (1978, tables 5–8) used a variety of methods to estimate the pressure-temperature conditions of the metadolerites and suggests an average value of 700°C at 600–700 MPa.

Shaw (1960, pp.279–280) plotted the meionite content of scapolite against the anorthite (An) content of co-existing plagioclase to try to define a correlation between the two minerals but found no clear trend. His specimens, however, came from diverse sources and the data presented here do show more coherence. (Figure 14) shows a plot of meionite values for scapolite plotted against An values for plagioclase for nine Younger Basics. The resulting trends appear to relate to metamorphic conditions during and subsequent to Younger Basic intrusion.

Many of the specimens show internally consistent mineral compositions implying that plagioclase and scapolite were in equilibrium at least over the distance of a thin section (e.g. (S61319), (S60011), (S60110)). The more calcic scapolites are in equilibrium with plagioclase in the An44-61 range. Within this An range meionite values for scapolite decrease rapidly as An values fall. Within the lower meionite range of scapolite values (less than 50 per cent), the An content of plagioclase declines rapidly and uniformly while meionite values fall slightly. By analogy with coexisting ferromagnesian minerals (garnet, hornblende, pyroxene) which also show retrogression, this trend apparently reflects a decrease of metamorphic grade within the amphibolite facies.

Specimen (S62220) comes from the potash-rich basic body immediately south-east of Skealtraval [NF 8557 7018] in North Uist. This, as already discussed (see Feldspars, above), has anomalously low plagioclase An values. Its meionite values are compatible with a higher metamorphic grade than implied by its An values.

In some metadolerites, scapolites were formed adjacent to plagioclase in the range An50-60. This most probably took place during the period when the igneous material equilibrated at the ambient granulite-amphibolite facies conditions prior to the more widespread regional amphibolite-grade retrogression. In some basic bodies, notably those with well-documented variations of plagioclase An content both between and within grains, retrogression also resulted in a patchy decrease in the meionite contents. In other rocks, the retrogression was penetrative, and scapolite compositions are more uniform, as are An values commonly. The slope of the graph of lower An values against meionite values on (Figure 14), which presumably reflects this equilibrium and the decline in pressure and temperature within the amphibolite facies, gives the approximate relationship: Meionite % = 39.5 + 0.3 (Anorthite %–20) —within the range An30-60 and neglecting results from (S62220).

Geochemistry

The general chemical character of the Scourie Dyke and equivalent suites has already been described by several workers: on the mainland by Burns (1966), O'Hara (1961a) and Tarney (1973) among others, and in the Outer Isles by Dearnley (1963) and Soldin (1978) working respectively in South and North Harris. During the present investigation 32 analyses of Younger Basics were obtained, extending the coverage in the Outer Isles (Figure 15). Average analyses are given in (Table 23) (analyses 14, 17). Much work still requires to be done, however, before the petrogenesis and crystallisation history of the suite are fully understood.

Interpretation of the chemistry of the Younger Basics in terms of modern analogues must be carried out with caution.

Because of the great age of the Younger Basics, the discriminatory methods which link modern magma types to specific tectonic settings cannot be assumed to hold and have not been used here. Another potential problem concerns the mobility of ions during metamorphism; fortunately, as will be shown later, the metamorphism appears to have been essentially isochemical (cf. Burns, 1966; Park and Cresswell, 1973; Soldin, 1978).

Only a few analyses of 'Cleitichean Beag' dykes are available. Dearnley (1963) presented an analysis of the Cleitichean Beag dyke which shows it to be ol-normative with a Mg/Fe ratio similar to the picrites and 'amphibolitised peridotite' of the mainland (Tarney, 1973, table 1). The sheet at Lundale is Q-normative with 60 per cent normative hypersthene. The analysis is broadly similar to the norite analysis of Tarney (1973, table 2) although with slightly higher Mg/Fe and Cr. The Ardvourlie norite' of Soldin (1978) is ol-normative, and its chemistry is slightly more basic than that given by Tarney (1973, table 2) but with similar Mg/Fe and Cr. Compared with Le Maitre's (1976) average norite, these norites are Mg-rich. The AFM plot (Table 23), analyses 1-5. A ultrabasic bodies, (Table 23), analyses 6-8. layered basics, (Table 5), analyses 7-11. O anorthosite from layered complex, (Table 5), analysis 6. x Late-Scourian 'microdiorites', (Table 5), analyses 12-15. * Late-Scourian 'granites', (Table 5), analyses 16-17. range of members of mainland Scourie Dyke suite from Tarney (1973). Younger Basic metadolerites (Outer Hebrides: this work). 'Cleitichean Beag' basic from Dearnley (1963, table 7, analysis 4). 0 noritic rocks from Soldin (1978, figure 4.11); 'average' of Maaruig ultrabasic from Soldin (1978, figure 4.11)." data-name="images/P936484.jpg">(Figure 11) shows that the dykes are more basic than the metadolerites, but lie on the same high-iron-enrichment trend. The Cleitichean Beag dykes are chemically similar to the norites and picrites of the mainland.

In the average analyses of the metadolerites presented in (Table 23), analyses 14-17 define the suite as typical Q-normative to ol-normative continental tholeiites with relatively low Al2O3 and high FeO + MgO contents (Carmichael et al., 1974, table 9–10; Manson, 1967, table IV; Prinz, 1967, table II; Gottfried et al., 1977, table 4). The AFM plot (Table 23), analyses 1-5. A ultrabasic bodies, (Table 23), analyses 6-8. layered basics, (Table 5), analyses 7-11. O anorthosite from layered complex, (Table 5), analysis 6. x Late-Scourian 'microdiorites', (Table 5), analyses 12-15. * Late-Scourian 'granites', (Table 5), analyses 16-17. range of members of mainland Scourie Dyke suite from Tarney (1973). Younger Basic metadolerites (Outer Hebrides: this work). 'Cleitichean Beag' basic from Dearnley (1963, table 7, analysis 4). 0 noritic rocks from Soldin (1978, figure 4.11); 'average' of Maaruig ultrabasic from Soldin (1978, figure 4.11)." data-name="images/P936484.jpg">(Figure 11) shows a high-iron-enrichment trend characteristic of tholeiite suites. Variation plots of selected elements against MgO are presented in (Figure 16). MgO was chosen as an index of fractionation (see Wright, 1974) because the Thornton-Tuttle differentiation index is affected by alkali mobility (cf. Soldin, 1978). On the plots most elements show coherent trends comparable with modern tholeiites (e.g. Wright and Okamura, 1977, fig. 15). The scatter of points in some diagrams is probably due, in part, to the existence of several magmatic lineages. This is well shown for TiO2, total FeO, P2O5, SiO2, Y and Zr, where most points define a standard ol-normative to Q-normative trend, but two ol-normative points ((S58793) and (S59777)) lie on a separate, possibly slightly undersaturated, trend. Some of the scatter must, however, be attributable to subsequent metasomatism and leaching. This is most pronounced on the alkali and Ba plots, although it should be emphasised that the general igneous trend is still recognisable.

Compared to the 'Cleitichean Beag' dykes the metadolerites have lower MgO, Cr and Ni but higher total FeO, alkalis, Rb, Sr, Y and Zr. Soldin (1978) finds broadly similar contrasts between his norite at Ardvourlie and the surrounding amphibolite, which is chemically identical to the regional suite of amphibolitised basic rocks. Soldin (1978) argues that these regionally distributed amphibolites could not be retrogressed norites as the chemical changes are too improbable; but he also asserts that the amphibolite surrounding the norite is a retrogressive equivalent of the norite. We consider that J. S. Myers' view (personal communication) that the norite cuts the amphibolite is more likely and we accept it in this memoir.

To examine the possibility of mobility of elements during metamorphism, we first divided the suite into five groups based on their present mafic mineralogy: 1 opx-cpx-(gt)- (hb1); 2 cpx-gt-hbl; 3 cpx-hbl; 4 hbl; and 5 hbl-bi. The distribution of the various groups on the MgO variation plots was found to be largely random. Likewise, for most elements the trends defined by assemblage 5 are just as coherent as those defined by assemblage 1. These factors suggest that the metamorphism was not accompanied by significant chemical mobility. Average analyses of the five groups are given in (Table 24): it is clear that they are chemically very similar apart from a slight rise in K2O, Rb, Li and H2O in assemblage 5, as might be expected since the retrogression of the dykes is greatest in areas of highest deformation where water fugacity and alkali mobility would be highest.

The compositions of the dykes, whose average is given in (Table 23) (analysis 14), show a marked concentration on the ACF diagram (Table 23), analysis 14)." data-name="images/P936490.jpg">(Figure 17) and plot in the cpx-hbl-pl field close to the hbl-pl tie line. This confirms the suggestion made above that hbl-cpx-pl is a more common stable assemblage than hbl-gt-pl.

Crystallisation sequence

Petrogenesis

Very little evidence exists for the petrogenesis of the Younger Basics or indeed for any members of the North Atlantic swarm; Rubie (1972) has suggested that they could represent an important, even primitive, magma type. Some of the higher-grade metadolerites have large crystals of clinopyroxene and plagioclase, which, though corroded, clearly define an ophitic texture, suggesting that the metadolerites formed by the precipitation of clinopyroxene and plagioclase. This is in general accord with the conclusions of O'Hara (1961a) who worked on the Scourie Dykes of the mainland. Soldin (1978) uses Pearce plots and suggests that they show that the metadolerites formed by the co-precipitation of clinopyroxene with a theoretical composition of En32Fs29Wo39 and plagioclase. The large clinopyroxenes of presumed igneous origin have a composition of En33Fs22Wo45, which is very close to Soldin's predicted composition, and would appear to support his conclusions. It should be noted that there is an overlap between igneous and metamorphic clinopyroxene compositions in this part of the clinopyroxene triangle (Fleet, 1974).

Even fewer data are available for the 'Cleitichean Beag' dykes. The only information is from the related 'Ardvourlie' norites, which Soldin (1978) suggests formed by the co-precipitation of orthopyroxene and plagioclase. He suggests that the original compositions were close to the present observed values: orthopyroxene of En70-72 and plagioclase of An53-66.

As discussed above, some metadolerites contain corroded plates of orthopyroxene (En52) which are probably of igneous origin. This orthopyroxene may result from a variation in the crystallisation conditions, or these dykes may belong to a lineage intermediate between the metadolerites and the 'Cleitichean Beag' dykes; indeed there may very well be a complete gradation of petrographical types between the two groups.

Initial recrystallisation

Working on the mainland Scourie Dykes close to Scourie, O'Hara (1961a) shows a variation in assemblage from the centre to the margins. The original cpx-pl igneous assemblage preserved in the central zone passes outwards through cpx-opx-hbl-pl and cpx-gt-hbl-pl metamorphic assemblages to a gt-hbl-pl mineralogy at the margins. O'Hara (1961a) concludes that the dykes had been autometamorphosed on intrusion into a hot crust, a view supported by Tarney (1973) and Park and Cresswell (1973). Tarney (1973) also suggests that the dykes of the Outer Hebrides were emplaced in a similar way, the differences in grade of their assemblages reflecting different temperatures in the country rocks. (On the mainland, granulite-facies dyke cores similar to those of the Outer Hebrides have been noted by one of us (Watson) north of Loch Laxford. They may indicate a change of metamorphic conditions across the mainland Lewisian.) Dearnley (1962a, 1973) and Dearnley and Dunning (1968), however, argue that the dykes were intruded into cold crust and were subsequently subjected to an early Laxfordian granulite-facies metamorphism along with the country gneisses. Retrogression during the Laxfordian produced amphibolite-facies mineral assemblages except in the cores of the thicker dykes. To resolve these various propositions four questions must be addressed:

  1. Were the dykes intruded into hot or cold crust?
  2. Were the dykes and the country gneisses subjected to regional granulite-facies metamorphism?
  3. If not, how do some dykes come to have granulite-facies assemblages?
  4. Was there a variation in the dyke assemblage, across the dyke, produced during the early metamorphism?

1 In relation to the first question, Dearnley (1973) argues that if the dykes had been intruded into hot crust the resultant contact temperature would have caused partial melting of the gneisses, which is not the case. Tarney (1973) however, argues that the dykes may not have been intruded with superheat, and heat transferred to the dyke margins could have been absorbed by crystallisation. In addition, because the gneisses are coarse grained, relatively anhydrous and not of minimum melting composition they would have begun to melt only at the solidus temperature of the dyke. An additional argument against Dearnley's hypothesis of intrusion into cold crust is the implied sequence of events: crustal uplift after the Scourian, rapid burial to produce granulite-facies metamorphism as the primary Laxfordian event, and retrogression during the main Laxfordian events. In addition the structural evidence (see above), which suggests that some of the dykes were intruded synkinematically into a set of shear zones, carries with it the implication that the country rock would not be cold. The most telling argument, however, against intrusion into a cold crust is the lack of chilled contacts. Many dykes have coarse-grained ophitic textures right up to the margins, a feature more compatible with intrusion into hot crust (cf. Tarney, 1973).

2 The answer to the problem of regional granulite-facies metamorphism is more obscure. Dearnley's views imply that during the Laxfordian all the gneisses were first metamorphosed to the granulite facies and subsequently retrograded to the amphibolite facies. There is no textural evidence for this sequence of events. If this metamorphic history occurred, it is perhaps surprising that the gneisses were so uniformly downgraded that the areas of low Laxfordian reworking (with the possible exception of the east coast of Barra) have been as comprehensively downgraded as the areas of intense deformation. On the contrary, strong evidence against a regional event comes from Barra. There, dioritic and granitic rocks in pyroxene gneisses are cut by granulite-facies metadolerites (Francis, 1973). Neither the diorites nor the granites show any evidence of granulitefacies metamorphism; since it is difficult to conceive that they could have been so selectively and completely downgraded, it must be assumed that they have not been, and the assemblages in the metadolerites reflect some specific set of events.

3 What was this set of events? If the dykes were intruded into gneisses within the amphibolite facies why did they acquire pyroxene-bearing assemblages? The most obvious answer is variation in hydration state. Dearnley (1973) argues that the field of crystallisation of the high-grade dyke assemblages on a pressure-temperature plot does not overlap the field of the present gneiss assemblages. Even if water partial pressure was lower in the dykes than in the country rocks, the overlap of the fields is very slight. Dearnley therefore concludes that the present gneiss assemblages are not in equilibrium with those in the dykes. However, Tarney (1973) argues that if the country-rock conditions were close to the top of the amphibolite facies, it would be thermodynamically much easier for the basic rocks, with their largely anhydrous igneous mineralogy, to recrystallise to a granulite-facies assemblage than for the country rock gneiss. Tarney (1973) also argues that the dykes might have had an autometamorphic effect from the latent heat of crystallisation.

4 If it is accepted that the dykes were intruded into hot country rock and as a result many of them recrystallised to a granulite-facies assemblage, what can be deduced about the distribution of facies? If variations in water partial pressure were a significant factor, then after crystallising with an igneous ophitic texture the relatively dry interior could recrystallise to a pyroxene-bearing granulite-facies assemblage, whereas partial equilibration with the gneiss at the margin could lead to an amphibolite assemblage, thus producing the observed variation across the dyke. Similar variations would be produced if autometamorphism were a significant factor. Thus, although further amphibolitisation may commonly have accompanied Laxfordian deformation events, it is probable that even at the onset of the Laxfordian the dykes presented a variety of assemblages. These relate not only to the size of the dyke but also to possible variations in pressure and temperature within the country rock; where the gneisses were relatively hydrated the dykes might have crystallised wholly or largely in the amphibolite facies, whereas where the gneisses were relatively dry the dykes, even those on a small scale, might have crystallised largely in the granulite facies. As stated above, at least some of the metadolerites appear to have crystallised from the igneous melt by the co-precipitation of clinopyroxene and plagioclase to give a coarse-grained ophitic texture. They then began to recrystallise to an equigranular aggregate of cpx-(opx)-(gt)-pl. Garnet appeared at the clinopyroxene/plagioclase boundary, and the orthopyroxene formed as grains around and within the clinopyroxene. What controlled the appearance of orthopyroxene or garnet is uncertain; since many of the dolerites in which the assemblage gt-cpx appears are Q-normative it was probably the precise values of prevailing pressure and temperature (Green and Ringwood, 1967). Many of the metadolerites contain the metastable assemblage opx-cpx-gt-pl. Texturally, however, partly because of the equigranular fabric and subsequent recrystallisation, it is not possible to determine whether garnet or clinopyroxene was unstable or in which direction the reaction was going. If the ideas advanced above are correct, it is probable that the centre of the larger dykes possessed an opx-pl assemblage which passed outwards in turn to cpx-gt, cpx-gt-hbl, and cpx-hbl assemblages. How much hornblende was 'primary' is uncertain.

One problem in this sequence arises from the composition of the small equigranular crystals of metamorphic clinopyroxene, which have an average composition of En33Fs21Wo46 (Appendix 1). This is virtually identical to that of the presumed igneous clinopyroxene (En33Fs22Wo45). It would appear to be impossible for the simple igneous assemblage of cpx-pl to change to a cpx-pl-gt assemblage with the composition of the clinopyroxene remaining unchanged. On the mainland O'Hara (1961a) shows that the original clinopyroxene was a sub-calcic ferroaugite. This initially recrystallised to ferropigeonite (inverting to ferrohypersthene and ferroaugite). Then, with increasing amphibole content, the clinopyroxene became increasingly more calcic, moving through the augite field to the salite field. There seems to be no evidence in the Younger Basics of the

Outer Hebrides for such a sequence of mineralogical changes. Therefore, either the large plates of clinopyroxene have changed uniformly to salite, an unlikely occurrence, or some additional phase was present in the igneous assemblage. This latter suggestion also seems improbable in view of the uniform ophitic texture of cpx-pl seen in the relic igneous assemblages. Perhaps some of the equigranular 'metamorphic' garnet and pyroxene grains themselves represent the residual melt.

Lambert et al. (1970b) regard clinopyroxene coronas around the olivine in the Cleitichean Beag dyke as a late-magmatic phase, the olivine having reacted with interstitial liquids. In addition Park and Cresswell (1973), discussing the garnet coronas found around pyroxenes in some dykes in the Laxfordian areas on the mainland, also suggest that they may be late-magmatic phases. What bearing the development of these coronas has on the problem of the mineralogical evolution of the dykes must await detailed work. Parallels with the mainland (Tarney, 1973; Park and Cresswell, 1973) suggest that if the dykes were intruded over a period of time into an area of active but variable shearing then they may well have been affected by differing degrees of metamorphism.

Laxfordian deformation and metamorphism

During the various phases of the Laxfordian the dykes were affected by considerable deformation and metamorphism. The deformational effects have already been described (this chapter pp.84–85: Field aspects: Metadolerites) and are further discussed in Chapter 13 (Laxfordian strain). Where fabrics associated with the deformation can be identified in the dykes they are defined by recrystallised amphibole and progressively change from equigranular to schistose. It is generally accepted (cf. Dearnley and Dunning, 1968) that Laxfordian deformation greatly promoted the amphibolitisation of the dykes; the imposition of deformational fabrics facilitated the hydration of granulite assemblages. However, as shown above, at the onset of Laxfordian deformation some of the dykes may well have possessed amphibolitic assemblages, at least at their margins. These dykes may have been more susceptible to subsequent deformation, recrystallisation, and the imposition of Laxfordian fabrics (cf. Francis, 1973). The Laxfordian deformation then would have done no more than recrystallise the amphibole and destroy the equigranular texture. Equally, on a larger scale, dykes intruded into relatively wet gneisses might have wholly crystallised to equigranular amphibole assemblages prior to the Laxfordian deformation. This may well be the reason for the widespread amphibolitic assemblages seen in the dykes of western North Harris and western Lewis, especially since in North Harris the Laxfordian deformation was apparently of only moderate intensity. This would agree with Soldin's (1978) evidence of greater hydration towards the west of this area.

It would be wrong, of course, to imply that extensive hydration of the granulite assemblage did not accompany the main Laxfordian events. This was manifested by the replacement of pyroxene by amphibole and of garnet by plagioclase pseudomorphs (in all but a few calcium-poor dykes) and a general movement towards a stable amphibolitic assemblage.

Summary

In summary, therefore, the crystallisation sequence of the Younger Basics may be given as:

1 Initial crystallisation to a coarse-grained ophitic texture of cpx-pl.

2 Recrystallisation, notably at the margins, to an amphibolite- or sub-granulite-facies assemblage, the mineralogy becoming progressively more anhydrous and equigranular towards the core. In the larger dykes, water partial pressure was so low that only partial recrystallisation took place, and the igneous textures and mineralogy survived in part. At this stage some equigranular minerals may have developed by reaction with residual melts.

3 Amphibolitisation during the Laxfordian; it progressed into the cores of the basic bodies as they approached equilibrium. This was aided by the strain related to the development of penetrative Laxfordian deformational fabrics.

4 The growth of lower amphibolite-facies minerals: blue-green hornblende, biotite, epidote, etc.

It is important to emphasise that not all the dykes passed through the whole sequence; many of the dykes, especially the smaller ones, may have recrystallised to uniform amphibolites almost immediately after intrusion, whereas others never recrystallised completely, and igneous textures and mineralogies remain.

Age of intrusion

The results presented above suggest that the Younger Basics may contain a number of magmatic lineages, a conclusion in accord with that of Evans and Tarney (1964) that the dykes of the mainland may have been intruded between 2200 and 2000 Ma, a period of 200 Ma. More recently, however, Chapman (1979) has presented a Rb-Sr whole-rock age of 2390 ± 20 Ma for the Scourie dykes of Assynt, a result which agrees well with the K-Ar whole-rock age of 2440 ± 60 Ma determined by Lambert et al. (1970b) for the Cleitichean Beag dyke.

Relationships of dyke types and ultrabasics

On the mainland the 'Scourie Dykes' are a complex basic- to ultrabasic suite containing dolerites, picrites and norites. Peach et al. (1907) and Tarney (1973) have shown a chronology where early dolerites are cut by norites which are in turn cut by picrites and late tholeiites, the evidence being cross-cutting relationships and different degrees of metamorphism. Although in the Outer Hebrides no such evidence is seen, close petrological and structural similarities would suggest that the metadolerites and 'Cleitichean Beag' dykes belong to the same magmatic episode. Indeed, since the distinction between the two groups is based both on chemistry and the presence of 'primary' orthopyroxene in the norites (as opposed to clinopyroxene in the dolerites), the presence of some two-pyroxene types suggests that the two groups may well grade petrographically into each other. As was pointed out previously (this chapter pp.81, 84: Field aspects: 'Cleitichean Beag' dykes) the initial recognition of the groups was based on their field appearance, and it may be that a more detailed examination and collection would reveal a greater number of transitional types, especially among the large coarse-grained metadolerites of North Uist and Park. Equally, where extensive recrystallisation has occurred, it may be impossible to tell the two types apart either in the field or in the laboratory, and still less to recognise any transitional types.

Less certain (as discussed earlier at the end of Chapter 9) is the relationship of the metadolerites and 'Cleitichean Beag' dykes to the ultrabasics. In the Outer Isles many of the dunites and peridotites seem to be more or less of 'Scourie Dyke' age, and possible correlatives on the mainland are generally of Scourian age. Both Myers (1970b) and Soldin (1978), however, do relate the ultrabasics (e.g. the Maaruig mass: see Chapter 9) and the 'Cleitichean Beag' dykes on petrological grounds.

Conclusions

1 The Younger Basics are a series of continental tholeiites, picrites, and norites; perhaps associated with them are dunites and peridotites. The Younger Basics are equivalent to the 'Scourie Dykes' of the mainland and are part of the great North Atlantic swarm of early Proterozoic dykes.

2 The Younger Basics contain a number of magmatic lineages, most of which are dominantly Q-normative though a few are 64-normative. They were probably intruded over a period of time at c.2400–2200 Ma in a tectonic setting containing active subvertical shear zones.

3 Like their mainland counterparts, the Younger Basics were probably intruded into hot crust. However, the country-rock gneisses in the Hebrides were at higher temperatures than those in the central part of the mainland, possibly in the upper-amphibolite facies.

4 Because they were intruded into a hot crust the dykes recrystallised rapidly from an ophitic or subophitic texture to an equigranular texture with, at least in the larger dykes, granulite-facies assemblages. Marginal amphibolite-facies assemblages may possibly have also formed at this stage.

5 During the Laxfordian the dykes were extensively but heterogeneously deformed and downgraded to a stable amphibolite-facies assemblage with local further retrogression to biotite-bearing lowermost amphibolite facies.

Chapter 11 Lewisian Complex: Late-Laxfordian granites and pegmatites

Uig Hills-Harris granite complex

The Uig Hills-Harris granite complex covers about 420 square kilometres extending from the Uig Hills of Lewis in the north, southwards through western Harris to the edge of the South Harris Complex. The first modern description of the granite complex was given by Jehu and Craig (1927, 1934), who regarded it as the top of an elongated batholith. More recently Dearnley (1963) has described some of the granites in conjunction with his work in the South Harris Complex. Dearnley (1963) suggested that the granites were Late-Laxfordian in age and characterised chemically by high potash/soda ratios. Myers (1971),in an account of the granite complex of Harris, regards them as the upper portion of a dome-shaped granite-migmatite complex consisting of intrusive granite and granitised country gneiss. He also showed that the granite postdated the main Laxfordian deformation events but that it was affected by a late-stage regional shearing and cataclasis. The Laxfordian age of the granite was confirmed by van Breemen et al. (1971), who reported radiometric ages of 1710 ± 35 Ma from a Rb-Sr whole rock isochron and 1715 +20/-10 Ma from U-Pb measurements on zircons.

Field relationships

The complex is magnificently exposed in the high hills of Uig and western Harris and also in the lower, ice-bared hilly interior of South Harris. In addition the indented coastline along the western seaboard provides some excellent three-dimensional sections, the rocks being kept free of lichens by salt spray.

The complex consists essentially of a series of granitic veins and sheets, ranging in thickness from a few centimetres to several hundreds of metres. The largest sheets crop out in the Uig Hills, where their dip and scarp slopes form several outstanding hills (e.g. Suainaval [NB 077 308], Mealisval [NB 022 270] and Griomaval [NB 012 220]). One of the-most spectacular developments of these features is in the area of Beinn Mheadhonach [NB 091 237], where a series of neighbouring hill-tops is marked by granite sheets. The abundance of granite veining is highly variable across the complex. Myers (1971) draws a distinction in Harris between two zones: a central zone characterised by the development of 'granite-migmatite', an intimate association of granite and 'granitised-gneiss' ; and an outer 'nonmigmatite' zone where the granite veins lie in 'nonmigmatic' gneiss. In both zones the granite sheets and veins are generally discrete and separable from the gneiss. Myers (1971) reports that the boundary between the zones is remarkably sharp and can be easily fixed to within 30 m in North Harris. The vertical relief in North Harris has allowed Myers (1971) to define the upper limit of the 'granite-migmatite' zone and to identify it as dome-shaped, the eastern boundary dipping gently to the east, and steepening as it turns towards South Harris. In the Uig Hills a distinction may be made between a central zone dominated by granite sheets and a marginal zone characterised by more widely separated veins.

The complex is elongated roughly north–south, almost at right angles to the regional strike of the gneissose foliation. Although on a local scale the granite veins and sheets appear to be randomly oriented, on a regional scale the large granite sheets have a preferred orientation. Myers (1971) says that there is a variation in orientation outwards from his central zone, from discordant granite sheets lying broadly parallel to the boundaries of the complex through to concordant or sub-concordant granite sheets lying parallel to the regional foliation trends. He suggests that many of the discordant veins may occupy roof-fractures above the main complex. It would appear, however, from an examination of Map 1 that the main feature controlling the regional orientation of the granite sheets was the attitude of the gneiss foliation and that the majority of the large sheets are concordant or subcordant.

The shape of the granite bodies is highly irregular in detail. Within the central part of the complex, veins of all widths follow sinuous courses, commonly branching and varying in thickness. The veins may pinch out into the gneiss or end in diffuse pegmatitic patches. In general the granites have sharp contacts with the gneiss even where the gneiss shows evidence of Laxfordian remobilisation and recrystallisation. This relationship of the granites to the country gneiss is discussed more fully at the end of this chapter.

Texturally and compositionally the veins and sheets range from medium-grained granites which are locally porphyritic to leucogranites which are both aplitic and pegmatitic. The porphyritic granites are largely confined to the central areas of intense veining (Figure 18), whereas the leucogranites dominate the peripheral areas of the complex, particularly at the northern end of the Uig Hills. Field relations show that the porphyritic granites are generally early in the intrusive sequence, usually being cross-cut by a series of nonporphyritic granites which are in turn cut by leucogranites, for example, at Mealisval [NB 021 271] and [NB 018 278], at Aird Brenish [NA 979 265], South of Brenish at [NA 990 250], and at Rubh an Tarain [NB 001 300]. There is, however, at least one example at the north-west end of Loch Roanasgail [NB 033 276] of a raft of leucogranite contained within porphyritic granite, suggesting local variations in the intrusive sequence. The medium-grained granites both grade into the porphyritic varieties as well as sharply crosscut them. At Aird Brenish [NA 980 262] examples can be seen of the intimate association of porphyritic and non-porphyritic granites, both varieties being cut by pegmatite. Contacts between the two types range from sharp to diffuse, with non-porphyritic or sparsely porphyritic granites veining and enclosing the porphyritic granite. Similar relationships are also well seen at Forsgeo [NA 904 229] with porphyritic granite grading into non-porphyritic granite. Also, south of Brenish [NA 990 251] good examples can be seen of coarse-grained porphyritic granite with large phenocrysts (about 2 cm) grading through medium-grained porphyritic granite with small phenocrysts to non-porphyritic granite. Complex relationships of porphyritic and non-porphyritic granite grading into and cross-cutting one another are also well seen on the coast west of Brenish [NA 9800 2625].

The leucogranites generally occur in smaller veins (less than 1 m thick) than the granite, although south of Gallan Head veins up to 5 m are common. Many of the aplitic veins have pegmatitic cores and vice versa. Leucogranite-pegmatites also occur as discrete cross-cutting veins within both the gneiss and the granite. In addition, a series of slightly later pegmatites invades the gneiss and granite causing extensive recrystallisation.

Many of the granite/gneiss contacts are obscured by these later pegmatites which commonly anastomosed as they penetrated the granite in a random manner, causing extensive recrystallisation and coarsening in the granite (for example, in the area north-east of Sgiathan [NB 032 279]). They have also caused extensive recrystallisation of the gneiss, but geochemical studies have shown that there has been no significant metasomatism.

Myers (1971) describes these associations as 'pegmatite-migmatites' and considers them in association with the late-stage dilation pegmatites (this chapter p.110: Pegmatites). The intimate areal and temporal associations of the 'pegmatites-migmatites' and the granite, however, renders such a distinction artificial and suggests they all developed as part of the same sequence of events, the youngest pegmatites being an expression of the late hydrous stage of granite intrusion. This phase of pegmatisation should not be confused with the widespread Laxfordian pre-granite phase of remobilisation and recrystallisation of the gneiss (Chapter 2: Lithology; Lewis and Harris, Laxfordian migmatisation and pegmatites). An excellent example of the complexity of the late intrusive phases is seen at Rubh an Tarain [NB 001 300], where a porphyritic granite is cut by a pegmatite vein which is cut by a younger pegmatite and a strongly foliated aplogranite vein, and these in turn are cut by a still later pegmatite and a quartz vein. At Mas Buaile Chuido [NB 0011 3170] a large aplogranite sheet (about 15 m wide) is cut by a pegmatite vein and a foliated pegmatite vein.

Within the central zone of the complex, xenoliths of gneiss are relatively common in the granite. Some of these appear to be blocks of country rock isolated by the high density of veining but still retaining their original orientation, e.g. at Mealasta [NA 9904 2285], whereas other blocks are much disoriented. Examples of disoriented xenoliths can be seen on the east side of Brinnaval [NB 028 285], north-west of Loch Deireadh Langa [NB 057 307] and at Forsgeo [NA 994 229]. The edges of the xenoliths are usually quite sharp against the granite and there is no evidence of extensive recrystallisation or assimilation. At some localities blocks of coarse sugary gneiss are found in granite, for example on Flodraskarve Mor [NB 045 294]. Myers (1971) also reports xenoliths from Halladale and Scarp which show various stages of Laxfordian recrystallisation with the localised destruction of the fabric. On the north-east slopes of Mealisval [NB 033 276] angular blocks of amphibolite occur in granite. Disoriented xenoliths in biotite granite have also been reported from north-west Scarp (Jehu and Craig, 1934).

As noted above, the granites are later than the main Laxfordian recrystallisation and remobilisation of the country-rock gneisses. They are not folded and may therefore be presumed to postdate the main phases of Laxfordian folding. Myers (1971) describes a foliation defined by the alignment of biotite flakes, which is common in many of the granite sheets (see also Jehu and Craig, 1934, p.858). This fabric he ascribes to a period of late deformation, the foliation being parallel to the axial planes of folds defined by pegmatite veins in the granite and by the banding in the adjacent 'hostgneisses'. As the granite sheets are themselves not folded, however, it is doubtful if this deformation is of more than local extent. The foliation may in fact be a 'flow fabric' related to intrusion. The granites have been involved in an extensive phase of late-stage cataclasis and mylonitisation. This is an important aspect of the general history of the region and is discussed separately at the end of this chapter.

Petrography

Mineralogically the granites are remarkably uniform, consisting of quartz, plagioclase, potash feldspar and biotite, with accessory sphene, muscovite, apatite, epidote, allanite, zircon and ore. A plot of modal analyses (Figure 19) shows that the rocks lie in a small group within the granite field (sensu Streckeisen, 1976). The porphyritic granites, although among the most basic members of the group, are essentially identical to the more basic non-porphyritic granites (Figure 20), their groundmass plotting to the right-hand side of the granodiorite field (Streckeisen, 1976). The percentage modes of the main suite of granites range from quartz 39, K-feldspar 36, plagioclase 21, biotite 4, to quartz 26, K-feldspar 34, plagioclase 30, biotite 10. Some of the aplite and pegmatite veins are more acid.

The granites commonly show evidence of late-stage shearing and cataclasis (see the end of this chapter). This deformation has affected the igneous mineralogy and textures to various extents. The plagioclase is compositionally albite/ oligoclase. It is generally fresh and unzoned with complex twinning, much of which is probably secondary and related to the cataclastic phase. The potash feldspar occurs in two forms: as microcline, and in a form characterised by simple twins and the patchy development of microcline twinning and perthitic intergrowths. The patchy twinning is commoner towards the deformed margins of some crystals, suggesting that it represents orthoclase whose symmetry has been reduced by secondary strain effects (e.g. (S58770), (S58805) and (S57850)). Significantly, microcline twinning is absent around the areas of perthite. The crystals are generally fresh and unzoned with a few inclusions of plagioclase, biotite and quartz. Myrmekite developments are common, projecting into the potash feldspar. Some of the biotite is partly altered to felts of white mica.

The accessory minerals are generally euhedral to subhedral. The allanite (up to 1.4 mm) is typically rimmed by epidote (e.g. (S58802), (S58762)). Where it is not it generally produces anastomosing cracks in the surrounding minerals (e.g. (S57983)). It may also show signs of zoning (e.g. (S57994)).

The textural evidence for the crystallisation sequence is not completely unequivocal. The accessories are commonly euhedral and included within grains of the major minerals, indicating that they appeared early in the sequence. Sphene is particularly interesting in that it normally forms subhedral/euhedral grains (up to 1.3 mm) which are clearly earlier than both biotite and plagioclase. However, in one sample (S61136), the sphene forms a series of anastomosing trails interstitial to all the other minerals. Several examples were also seen of apatite enclosed in sphene (e.g. (S57994)). Plagioclase and biotite crystals are seen included in potash feldspar. In general they abut against and project into each other, indicating that they crystallised at about the same time, with plagioclase possibly having started to crystallise slightly earlier than the other main minerals. The potash feldspar encloses some quartz in addition to the other minerals but also shows good crystal faces against quartz, although it is completely interstitial with respect to plagioclase and biotite. The crystallisation sequence can therefore be given as accessories: plagioclase: biotite: K-feldspar: quartz.

The significance of the porphyritic varieties is interesting. As noted above, many field examples are seen of porphyritic granite grading into non-porphyritic granite. In thin section a corresponding gradation is seen from hypidiomorphic to porphyritic textures, caused only by a relative increase in the size of the potash feldspar crystals, the other textural features of the potash feldspar being unaffected by the increase in size. Potash feldspar crystals of various sizes (0.02 to 8 mm) can be seen in single thin sections (e.g. (S57846)). Potash feldspar is late in the crystallisation sequence, and it appears that crystals of all sizes grew at the same stage in the sequence, for secondary or metasomatic affects are ruled out by the absence of 'anomalous' cores and the paucity of inclusions within the potash feldspar, as well as by the modal and chemical similarity of the porphyritic and non-porphyritic varieties. It appears most likely that some physical control at the time of crystallisation determined the size of the potash feldspar crystals. Whatever the control was, it did not affect the crystallisation of the quartz, which was presumably happening at the same time, for the quartz crystals do not show the same variation in size.

Geochemistry

The granites of the Uig Hills-Harris complex (hereafter termed Uig granites) show a relatively restricted compositional range, SiO2 ranging from 68 per cent to 76 per cent. The spread between the end members is continuous, though the bulk of the samples fall between 68 per cent and 73 per cent SiO2.

Average analyses (Table 25) indicate that the leucogranites are more silicic than expected, and the granites and porphyritic granites are chemically almost indistinguishable. However, the porphyritic varieties are higher in Li and Rb than the granites are. Both the growth of large potash feldspars and high levels of Rb and Li appear to reflect the presence of an abundant fluid phase in the later stages of crystallisation. However, as Li and Rb are highly mobile elements, their enhancement may imply a secondary re-mobilisation rather than a truly magmatic feature.

Magmatic affinities

The restricted range of compositions makes it difficult to determine the magmatic affinities of the Uig granites since most parameters (e.g. Peacock Index, trend on the AFM diagram) require a wider compositional spectrum. However, the criteria of Irvine and Baragar (1971) classify the rocks as calcalkaline, and this is in accord with the peraluminous (c-normative) nature of all but eight specimens (Figure 22).

The Uig granites have many similarities with Scottish Caledonian granites of comparable SiO2 range (Nockolds, 1954; Marston, 1971; Fettes and MacDonald, 1978; Brown et al., 1979) in terms of both major and trace elements. The Rb values are, however, notably high and the K/Rb ratio correspondingly low. The reason for this is uncertain. Since Rb is a mobile element in such systems (see for example, the scatter on a Rb-SiO2 plot, (Figure 21)), the feature may not be primary.

Chemical variation within the suite

On Harker diagrams ((Figure 20) and (Figure 21)), the major elements of the Uig granites show a sufficiently coherent trend for a comagmatic sequence to be probable. There are minor deviations from the main trend that have arisen in two ways: magmatic variants and secondary remobilisation.

For the first, certain rocks with between 68 and 70% SiO2 have distinctly low values for total iron and TiO2, and such rocks must be considered to be relatively low Fe-Ti varieties. Therefore, all the granites cannot have evolved along a single crystallisation path.

For the second, values of Na, K, Rb and Pb are sufficiently scattered, especially in the most silicic rocks, that secondary remobilisation of these elements must be suspected. This may well be the cause of the scattered relationship between SiO2 and the amount of normative corundum C in these rocks (Figure 22). In many magmatic suites, this relationship approaches a linear, positive correlation (Cawthorn et al., 1976; Strong, 1979; Stephens and Halliday, 1979; McCarthy and Groves, 1979). Calculation of C is highly dependent on CaO, Na2O and K2O abundances, however, and any secondary changes are likely to produce the sort of scatter shown in (Figure 22).

The overall trends of decreasing Al2O3, total iron, MgO, CaO, P2O5, TiO2, Sr, Y, Zr, Cu, Ba and La with increasing SiO2 are typical features of calc-alkaline suites and are most commonly explained in terms of crystal fractionation of one or more parental magmas. Further evidence pointing to the importance of crystal-liquid equilibria in the evolution of the suite comes from plots of normative ternary components (Figure 23). The Uig granites plot close to cotectic equilibria for about 5 kbar pressure; this pressure is somewhat greater than estimates for pressures at the initiation of the Outer Hebrides Thrust (see the end of Chapter 15), which, it is believed, closely postdated the granite intrusion. These figures support the view that the granites were intruded during or close to a period of rapid uplift (see also Dickinson and Watson, 1976).

The chemical variations in the Uig granites are compatible with a crystal fractionation hypothesis. This can be shown by a simple extract calculation, relating the average granite to the average leucogranite of (Table 25). The theoretical chemical composition (at 65 per cent SiO2) of the graphically determined extract can be recalculated to give a mineral assemblage (in weight per cent): quartz 25.3, feldspar (bulk composition Or24.5Ab52.1Ant13.1) 60, biotite (Mg/Fe ratio c.1) 10.5, Fe-Ti oxides 3.9, apatite 0.4. This is in good agreement with a typical mode for the least-silica granites (about 68 per cent SiO2)–quartz 26, plagioclase 30, alkali-feldspar 34, biotite 10, and accessory iron oxides, apatite, etc.

The pattern of Rb enrichment and Ba and Sr depletion in the granites is also characteristic of crystal fractionation. These elements have been specifically selected by McCarthy (1976) and McCarthy and Groves (1979) as the changes in their concentrations match crystallisation trends in granitoid rocks. Remobilisation of Rb makes it only of limited value for this purpose in the Uig rocks. Nevertheless, trace-element models, using Rayleigh fractionation, have been calculated for Ba-Rb-Sr distribution (Figure 24). The simplifying assumptions of crystallisation of phases in constant proportions and with constant partition coefficients have been made. Crystal/liquid partition coefficients are those of McCarthy and Groves (1979, table 4) and the assumed fractionating assemblage is that deduced from the major-element data.

The modelling indicates that Ba-Rb-Sr distribution is generally compatible with a fractionation model involving up to 70% crystallisation for a range of suitable parental melts broadly similar to the least silicic granites. The inferred assemblage would not efficiently fractionate La, Y and Zr: allanite, apatite and zircon are the likely mineral phases concentrating these elements.

Granite genesis

Relationship to country-gneiss

In the discussion of gneiss lithology it was pointed out that there was a general increase of Scourian migmatisation towards the west coast of Lewis and Harris (Chapter 2). In addition, Laxfordian recrystallisation and remobilisation of the gneisses was greatest in the area now occupied by the Uig Hills–Harris granite complex. At their most extreme the Laxfordian effects completely destroyed the gneiss foliation, leading to the patchy production of a coarsely crystalline homogeneous rock. In some localities, particularly the Uig Hills, this has led to the widespread crystallisation of a coarse-grained 'sugary' gneiss with a poorly defined foliation. Such rocks contain Laxfordian structures but are cut by discrete granite veins. The contacts of the granites are generally clean-cut and there is no increase of grain size in the gneisses as the contact is approached.

The granites are themselves, however, affected by a late phase of pegmatisation, the pegmatite being the last stage of the sequence of granite intrusion (see above, under Field relationships). The pegmatites permeated the gneiss and granite, causing local recrystallisation and locally caused the contacts between the granite and the gneiss to become diffuse e.g. on the south side of Glen Tealasdale at [NB 0010 2221].

The general areal association of the granite complex and the areas of remobilised and partially homogenised gneiss impressed previous workers. Jehu and Craig (1927, 1934) and Dearnley (1963) do not separate the granite and the migmatite, apparently regarding the migmatite lits and the cross-cutting granite as part of the same intrusive phase.

Myers (1971) makes distinctions between the extensive recrystallisation of the gneisses, which he considers is associated with a general phase of granitisation (or homogenisation), the intrusive granite, and the slightly later phase of pegmatisation. To support the concept of granitisation he cites areas up to several metres across of uniform granitoid rock apparently grading into gneiss. However, two types of such rock can be distinguished on South Harris, and it is important to the present discussion that this distinction is made. First, there are the typical pockets of coarse-grained gneisses (seldom covering more than a few square metres) described above and found throughout West Harris and the Uig Hills. These are attributed to widespread Laxfordian recrystallisation. Secondly there are large masses (commonly covering much more than a few square metres) of uniformly medium-grained granitic rock which pass without significant change in grain size into a granitoid gneiss. An excellent example of this gradation is seen on the roadside near Seilebost school [NG 0051 9718] in South Harris. Grey granite gneiss passes laterally into a massive grey granite, without a significant change of grain size or apparent composition, the constituent minerals simply becoming dispersed. At the southeast end of the Laxfordian granite complex, just north of Flodabay [NG 0984 9957], granitoid gneisses similarly grade into granites. It is difficult to reconcile such textures with the concept of remobilisation or recrystallisation and it is more plausible to interpret these large masses of granitic rock as Scourian granites which became partially foliated (and gneissic) during the Laxfordian.

There appear to be four possible explanations for the association of Laxfordian granite with areas of remobilised and recrystallised gneiss:

  1. The gneiss is exhibiting the initial stages of partial melting which at a lower structural level led to the production of a granitic magma that was the source of the crosscutting veins.
  2. The gneiss was being 'granitised' with an associated intrusion of granite and pegmatite (the view favoured by Myers, 1971). The term granitisation is here used to imply the production of large areas of homogeneous granitoid rock not necessarily accompanied by metasomatism.
  3. The recrystallisation of the gneiss is only indirectly related to the development and intrusion of the granites (e.g. each was driven by a common heat source).

The recrystallisation of the gneiss was unrelated in any way to the granites and the areal association is purely coincidental.

The first explanation seems unlikely, partly because there is no evidence that during the recrystallisation of the migmatitic gneiss an acidic phase separated that was equivalent to a first partial melt. There seems to be no resemblance between the restricted granite composition and probable melts from the gneisses. Extensive recrystallisation leading to the destruction of the pegmatite banded gneiss and to the appearance of uniform 'homogenised gneiss' with general loss of fabric occurs only in small patches seldom exceeding a few square metres. There is no evidence that this homogenised gneiss was separated from its host rock and intruded into adjacent rocks. Although pegmatitic fractions are to be seen in the sheared limbs of folds, boudin-necks, etc., their formation probably relates to a general diffusion of quartz and feldspar into areas of low strain rather than to the development of anatectic melts. However, the most powerful argument against partial melting being the true explanation arises from the chemical data discussed below. These strongly suggest that the granites could not have been derived from the country gneisses.

The nature of these patches of homogenised gneiss also argues against their association with granitisation, which is the second possible explanation. The pink migmatitic lits present in the gneisses are largely of Scourian age (Chapter 2: Lithology; Lewis and Harris). Also, as stated above, the development of homogenised gneiss was generally localised in small patches, forming subordinate parts of the outcrop even in the most intensely recrystallised areas.

The 'sugary gneiss' seen in the Uig Hills and North Harris is coarsely recrystallised but retains a ghost gneiss foliation (Chapter 2: Lithology; Lewis and Harris). There is, therefore, little field evidence to support granitisation per se as an explanation for the association of granite and gneiss, or to indicate that the various recrystallisation phenomena were directly caused by the intrusive granites. The cross-cutting granite veins and xenoliths also show different degrees of Late-Laxfordian recrystallisation, reflecting its in-homogeneous nature rather than indicating that the granites were intruding during the processes of recrystallisation. Many different gneiss types are cross-cut by the granite. Coarse-grained Scourian migrnatites marked by lit-par-lit pegmatitic banding commonly show no obvious effects of Laxfordian remobilisation. They are cut by granite veins, for example, on the shore south of Brenish [NA 990 250], on Lag Macgodrom [NB 962 097] in North Harris, and at the north-west end of Loch Raonasgail [NB 033 276]. Elsewhere, granite veins cut gneisses which have been extensively remobilised during the Laxfordian, for example, at Creag Chleistir [NB 1268 1281].

There is no conclusive evidence of granite sheets rooting in areas of homogenised gneiss. Myers (1971) reports a large cross-cutting granite sheet passing downwards into a series of concordant lenses on Tirga Mor in North Harris. He also suggests that some of the granite veins grade into the host rock and that many flat-lying sheets are fringed by small concordant apophyses. Where, in the present investigation, diffuse contacts were found between intrusive granite veins and the gneiss, they were generally marked by pegmatitic zones up to 10-15 cm wide. This was so even when the pegmatitic zones bordered the margins of thick (about 1–2 m) granite sheets. It is difficult to reconcile the scale of these relationships with the concept that the granite sheet was rooted in the gneiss. We may regard these areas as the terminal points of the intrusive granites, the hydrous fluids associated with the granites gave rise to the pegmatitic patches with localised recrystallisation of the gneiss.

The field evidence, therefore, does not support either of the first two explanations: in situ partial melting, and the production of large areas of granitoid rock by homogenisation (granitisation). Even if we assume that the presently observed phenomena associated with homogenisation, etc. became more intense with depth, there is no reason to assume that this would have given rise to the observable intrusive granite sheets and veins.

The areal coincidence between the granite complex and areas of intense remobilisation is, however, impressive and it is tempting to ascribe both to a common cause. It may be, however, that the connection is more apparent than real, and that the third and fourth explanations are nearer than truth. Although the granite complex is centred on West Lewis and Harris there are extensive developments of granites beyond the areas of strong recrystallisation: for example, the Dalbeg granite [ND 235 455], the zone of veining in central Lewis around Caultrashal, and minor veining in the Park area. Equally, there are areas of intense remobilisation of the gneiss where granite veins are largely absent: for example, around Drineshader in South Harris, and north-east of Hecla (South Uist) at [NB 837 355], where Coward (1973b, fig. 6A) has described an extensive area of completely homogenised gneiss which is apparently free of intrusive granite veins.

As noted previously, there is a general increase in the degree of Scourian migmatisation towards the west as well as an apparent fundamental increase in K2O etc. (Chapter 2: Geochemistry). It may be, therefore, that there is a lithological or compositional control over the development of strongly recrystallised and homogenised gneisses, and the appearance of granites in that area is coincidental (cf. Myers, 1971). At present, however, the field evidence does not enable one to decide which of the various possibilities is most probable.

Geochemical evidence

Earlier it was shown that the chemical variations within the granites were explicable in terms of fractionation of the observed mineral phases. This model is supported by the field evidence that the more silicic members of the suite tend to be the younger. However, if the least silicic granites themselves were derived by fractionation, rocks representing the parental liquids are not to be seen at the present level of exposure. The derivation of the granites must remain a matter of conjecture for now.

It remains to be tested whether there is a plausible alternative mode of origin for the granites; possibilities are the progressive melting of crustal rocks, either the local gneisses or some more deep-seated source materials. Field relationships cast doubt on the possibility that the granites were formed by anatexis of the Laxfordian gneisses at or near the present level of exposure. Aspects of the geochemistry are also inconsistent with this proposal.

Graphs of K/Ba against Ba, and K/Rb and Ba/Rb against Rb ((Table 3), analyses 7 and 8). C and D average values for mainland Laxfordian gneiss ((Table 3), analyses 4 and 5). " data-name="images/P936498.jpg">(Figure 25), (Figure 27) Plot of Ba/Rb against Rb for gneisses and Late-Laxfordian granites. Symbols as for (Figure 25)." data-name="images/P936499.jpg">(Figure 26), (Figure 27)) indicate that the least silicic granites may have been derived by nearly total melting of gneisses having Ba contents in the range 120-150 ppm and Rb in the range 100 -150 ppm. It must be noted, however, that such gneisses occur only to the east of the granite outcrop; on a K/Rb v. Rb plot (Figure 27), the field of the local gneisses scarcely overlaps that of the granites. Furthermore, the least silicic granites have Zr and Ca contents twice those of the gneisses, implying a maximum of 50 per cent melting–a figure at odds with the very much more complete melting required by the Ba-Rb data.

While partial melting of the local Laxfordian gneisses appears to be precluded as a mechanism for the genesis of the granites, deeper-level anatexis of gneissose rocks of different composition may be a possible alternative. One way in which this hypothesis can be assessed is to model the petrological and geochemical characteristics of the potential source rocks using the trace-element data. Since constraints cannot be put on such important factors as source mineralogy, relevant partition coefficients, the role of accessory minerals, the role of volatile phases, mica-rich interactions or the petrological controls of the melting process, such modelling could not provide meaningful solutions (see Shaw, 1977).

Other granite bodies

Outside the Uig Hills–Harris granite complex, Late-Laxfordian granites are generally restricted to minor veins.

These occur most abundantly in Lewis and Harris, only a few having been recorded in the southern isles. Jehu and Craig (1925, p.626) described 'later muscovite-biotite gneiss' in South Uist which they believed to be clearly intrusive. They cite examples from Arnaval, Carnan and Hornish consisting of qz-olig-or-bi-ms assemblages with accessory apatite, epidote and garnet. The rocks also show varying degrees of foliation. Jehu and Craig suggest that the intrusion of these rocks was closely linked to the Late-Laxfordian pegmatites and are thus regard them as members of the granite suite. Coward (1969) also reports sparse granite veins and sheets in South Uist, the largest being on the north-west slopes of Beinn Mhor.

In the northern islands, granites crop out extensively in central North Harris and Lewis, and in the Carloway and Park areas of Lewis. In Great Bernera Watson (1968) describes intrusive 'granite-gneiss' sheets with a predominant NE–SW strike. Watson (1968) suggests that the intrusion of the granites separated Laxfordian phases of deformation and regards the intrusion of the granite veins as related to the remobilisation and recrystallisation of the country gneiss. In the Carloway region Lisle (1974, p.31) reports several granites, the main ones being Dalbeg, Borrowston and Breasclete. Dalbeg is the largest, and has been variously reported in the literature.

Steavenson (1928) describes the Dalbeg mass as having generally steep contacts, locally faulted in the north-west, with surprisingly few apophyses. From the granite Flett (in Peach and Horne, 1930) notes the assemblage qz-ms-bi-kf-ab with epidote, opaques, apatite and particularly abundant allanite. The potash feldspar is principally microcline with subsidiary perthite and orthoclase. Like other Laxfordian granites, the Dalbeg mass locally shows a strong foliation due to late-stage deformation; the foliation is marked by quartz 'granulation'.

Jehu and Craig (1934) note that the granite is the coarsest in Lewis, with feldspar crystals up to 1 cm. Mackie (1932), in a study of heavy mineral accessories in Scottish granites, reports zircon, apatite, sphene, magnetite, allanite, epidote, tourmaline, anatase, enstatite and ilmenite from the 'Loch Roag granite', which may well be the Dalbeg granite. Beer (1952) describes the granite as rich in allanite and suggest that it derives its radioactivity from thorium, uranium being absent.

Granites have also been described from North Harris by Soldin (1978, fig. 3.3), who compared them with the migmatites of South Harris (Skinner, 1970) and with the Uig Hills granites (data from BGS). He found similarities to the migmatites but differences from the Uig Hills granites. In discussing the genesis of the central Lewis granites Soldin (1978) makes three points:

  1. The K/Rb ratios in the Lewis granites are similar to those in the gneiss, whereas those in the Uig Hills granites are lower than those of their host gneisses, which in turn have lower values than the gneisses in the east.
  2. Ternary plots (Q-ab-or and ab-an-or in Soldin, 1978, figs. 3.23 and 3.24) show the Lewis granites to fall within the 'granite field'.
  3. Plots of major oxides against Larsen index indicate similar trends for the Lewis granites and gneisses but markedly dissimilar trends for the Uig Hills granites and gneisses. These points led Soldin (1978) to conclude that the Lewis granites are partial melts and intrusive. It seems, therefore, that these Lewis 'granites' are products of remobilised Laxfordian gneiss and thus different from the intrusive granite veins.

Pegmatites

Although pegmatite development took place throughout the history of the Lewisian (see the discussion in Chapter 6: Late Scourian intrusions; Pegmatites), many of the large pegmatite masses date from the Late-Laxfordian. They typically occur in swarms with a fairly uniform trend, each vein typically being 1-2 m wide but ranging up to 7–8 m in places. Myers (1971) characterises them as dilation pegmatites. Although closely related to the main granite development with its accompanying pegmatisation, the swarms are regarded as a separate and slightly later phenomenon. They have been extensively described in the literature (e.g. Jehu and Craig, 1923, 1925, 1927, 1934; Dearnley, 1963; von Knorring and Dearnley, 1960; Myers, 1971).

In the southern islands the general trend of the pegmatites is easterly to south-easterly as noted by Jehu and Craig (1923, 1925). They have a characteristic reddish colour, with feldspar as the predominant mineral and subsidiary quartz and biotite. The feldspar is a microcline-perthite or oligoclase with antiperthite. Some pegmatites on Pabbay are reported to have feldspar crystals up to 15 cm long which combine with quartz in graphic intergrowths. Pegmatites on the west coast of Benbecula also contain very large feldspar crystals; in one example [NF 756 534] crystals attain 2 m in length. Pegmatites from the southern end of Fuday are reported to contain 'magnetite up to the size of an egg' (Jehu and Craig, 1923, p.429), and in the pegmatite at Rubha na Muireart on Gighay Jehu and Craig (1923) record a colourless pyroxene and allanite.

Pegmatite swarms are particularly abundant on the west coast of Lewis and Harris, for example running across from Gallan Head to the northern tip of Great Bernera, on Scarp, Taransay and in South Harris. The most prevalent direction is east–west, although there is also a cluster around north–south (see Dearnley, 1963, fig. 8, and Myers, 1971, figs. 10, 11). Myers records zoned pegmatites with aplitic margins and quartz-rich cores, with some of the larger pegmatites containing crystals 2 m or more in length. The pegmatites here are again characteristically red or pink, potash feldspar being the dominant mineral and forming up to 80 per cent of the rock (Myers, 1971). The feldspar occurs as orthoclase or microcline, commonly with perthitic intergrowths but also in graphic intergrowth with the quartz. The plagioclase composition is generally close to albite. The main mafic mineral is biotite. Magnetite characteristically forms clusters of crystals up to 15 cm across. Myers suggests that in places the magnetite may constitute up to 50 per cent of the rock in areas several metres in diameter. A pegmatite from near Tarbert has been reported to contain crystals of allanite up to 5 cm long.

Two of the best-known pegmatites are at Chiapaval and Sletteval in South Harris, both of which have been worked as sources of potash feldspar. A more extensive discussion of these and other pegmatites in South Harris is given in Chapter 7: Pegmatites.

Pegmatites have also been described from the Flannan Isles (Stewart, 1933) and North Rona (Nesbitt, 1961).

Late-stage deformation in the granitic rocks of Lewis and North Harris

Most of the granite sheets and veins in South Lewis and North Harris exhibit a deformational fabric of variable intensity. This fabric is less intense or absent in the surrounding gneiss. It is commonly oriented parallel or subparallel to the margins of the granite bodies even when such bodies are discordant to the gneiss foliation. Rare exceptions to this orientation are seen at Cleit an Fhuarain [NB 111 257] and on the south-west side of Teinasval [NB 035 248], where large discordant granite sheets contain a fabric parallel to the adjacent gneiss foliation. Within the Uig Hills the main granite sheets are generally only slightly foliated but the fabric becomes more intense towards the margins of the granite complex, and is locally mylonitic. In places the granite bodies in this area appear to have acted as loci for late-stage deformation and local thrusting. Similar features have been noted in the Scourie/Laxford Bridge area of North-west Scotland, where Peach et al., (1907) recorded six occurrences of partially mylonitised granite or pegmatite in gneisses apparently unaffected by mylonitisation.

At Ard More Mangersta [NB 0100 3336] a foliated granite sheet about 5 m thick has been thrust discordantly over a banded gneiss sequence. The underlying gneiss is foliated for up to 30 cm from the thrust contact. Thin sections ((S58640), (S58641), (S58642) [NB 0083 3324]) of the granite show it to consist of strained, fractured and corroded feldspars in a matrix of recrystallised quartz, biotite (retrograded to chlorite), epidote and muscovite, the last two minerals defining a spaced foliation. The lower thrust contact is cut by aplogranite veins which themselves show only local slight shear effects.

Thrust granite sheets can also be seen at Aird Griamanish [NA 993 207], and on the coast between Clibhe and Valtos, e.g. at Rubha Brataig [NB 069 380], where pegmatite and aplite veins act as loci for small thrusts and shears. In fact, parts of the coastline around Uig Bay may be controlled by thrusts which generally dip about 40° to the south-east, but which locally also dip the the south-west at very low angles. The complex relationship between the age of intrusion and development of a cataclastic fabric is well seen in several parts of west Lewis. For example, west of Mangersta [NB 003 314] a partially foliated porphyritic granite is cut by an undeformed pegmatite. Similarly, at Glen Eilean [NB 0137 3347], a foliated granite 3 m thick is cut by an unfoliated pegmatite; like relationships are also seen at Mala Chaolartan [NB 055 243]. Conversely, there are also examples of non-foliated porphyritic granites cross-cut by foliated aplogranite or leucogranite veins at Aird Brenish [NB 979 265] and at Mealisval [NB 021 261]. Although foliated porphyritic granites are found at Aird Mheadhonach [NB 015 318] and south of Brenish [NA 990 250], in general the porphyritic granites are non-foliated or only weakly affected by later deformation. The close association of late-Laxfordian granite and pegmatite intrusion with local mylonite formation and cataclastic deformation in both time and space is further emphasised by the observations at Aird Fenish [NA 993 294]. Here a 20 cm aplite vein has largely undeformed margins and a progressively more intense fabric towards the centre. A further example is seen on the shore west of Crowlista [NB 027 341] where a 40 m granite sheet has a sheared central zone. This suggests that deformation occurred while the central parts of these intrusives were still ductile.

Thin-section work shows that all granites have undergone some degree of post-consolidation deformation. In the Uig Hills the degree of deformation is greatest in the later leucogranite veins and also in the peripheral granite sheets of the Uig Hills complex (Figure 28). In some granites only the quartz is affected and subgrains have formed; the other constituent minerals are largely unaffected (e.g. (S57841)). In more deformed parts quartz is recrystallised into fine-grained aggregates, biotite altered to felts of white mica, and feldspar shows strain perthite and secondary twinning (S57147). Where the granite is mylonitic, quartz is intensely ribboned and forms a curving fabric around other minerals, except for minor recrystallisation to more equigranular aggregates. Feldspar shows abundant secondary twinning and perthite growth and marginal granulation to fine-grained epidote/clinozoisite-rich aggregates. Epidote and white mica have recrystallised to form a new fabric, but accessories and ore minerals remain unchanged (S57832).

The petrographic character of the mylonitic granite sheets can be shown to change from west Lewis to the east and south-east. (Figure 29) shows the observed occurrence of mylonite and pseudotachylite in gneiss and granite in South Lewis, illustrating the broad transition zone that trends about 020° in central South Lewis. There is good evidence from thin sections, particularly from the granite sheets, that pseudotachylite is more widely developed in minor thrust zones to the east at the expense of mylonite. However, there are still early mylonites in the east, for example at Cipeagil Mhor [NB 241 061] and around Loch nan Deasport [NB 304 217]. Pseudotachylite has been found in gneisses at a few places west of the transition zone, for example at Miavaig [NB 087 347]:

In the more south-easterly granites, the intensity of deformation differs within individual sheets. A typical example is well exposed 2.5 km south-east of Kinlochresort at [NB 1264 1520]. Here a granite sheet 15 to 20 m thick dips gently to the south-east, and there is a gradation from mylonitic granite to marginal ultramylonite ((S62042) (S62043), (S62044) (S62045)). In its central parts the granite consists of abundant highly strained, fractured and broken potash feldspar (commonly perthitic) and plagioclase (oligoclase-andesine) in a fine-grained foliated matrix of quartz, feldspar, biotite, hornblende, white mica and subsidiary magnetite, epidote and allanite. Feldspar and allanite porphyroclasts are commonly rounded and/or fragmented. Quartz has recrystallised preferentially in strain shadows adjacent to slightly rotated feldspars. Where the granite is more mylonitic the proportion of fine-grained recrystallised matrix is greater and the porphyroclasts are smaller and more rounded. Potash feldspar and perthite appear to have been more resistant than plagioclase to this process. Quartz and feldspar form fine granular aggregates with an average individual grain size of 0.0025 mm. Small new epidote/clinozoisite grains with an average diameter of 0.005 mm are abundant in the matrix and epidote overgrowths on hornblende and allanite are also common. Thus, although the deformation in general caused grain size to be reduced, the growth of epidote was promoted, presumably because the movement and diffusion of fluid was enhanced. In places broken and rotated epidote/ clinozoisite grains were noted; they had presumably grown early in the deformation and subsequently some of the larger grains had been rotated and broken. In thin sections of the marginal ultramylomites, there are moderately tight to isoclinal microfolds of the mylonite banding. In most of these, alignment of mica marks a new axial-plane fabric oblique to the banding. It is probable that there was a single major phase of mylonite formation of fluctuating intensity to cause these features. They imply that the granite sheets have been strongly sheared and their component parts have been translated from their original locations.

South of these granite sheets there are several thinner, concordant granite sheets and pods which are locally mylonitised. In thin section one of these ((S61847) [NB 1304 1391]) shows large fractured and fragmented potash and plagioclase (andesine) porphyroclasts lying in a finely-banded matrix of recrystallised quartz, potash feldspar, albite, biotite and sericite. There are epidote overgrowths on allanite and sphene, and limonite coatings on magnetite. This mylonite mineralogy suggests that the granites were deformed in the upper greenschist facies with a high oxygen fugacity (cf. the chlorite-white mica mineralogy seen at Mangersta, which is further to the north-west). Nearby, about 1 km north-west of Rapaire (North Harris) [NB 1425 1425] a granite sheet 6 to 8 m thick with strongly mylonitic margins is directly underlain by a 'pavement' of slightly retrograded pseudotachylite up to 6 cm thick. In thin section (S61849) irregular, almost isotropic zones of devitrified glass with fine-grained margins are seen.

This increase of metamorphic grade relating to the thrusting and consequent mylonite development eastwards is also to be seen in thin sections of mylonitic granite in the Lundale–Grimersta area (West Lewis). Here ((S60006) [NB 1877 3260]) quartz and potash feldspar form recrystallised fine-grained highly lenticular aggregates. Biotite is green, but apparently stable. Microfolds and dislocations are commonly seen, and porphyroclasts (of plagioclase and minor potash feldspar) show a sense of rotation compatible with microfold vergence. In the Grimersta area ultramylonites are also found; they have a dark grey, compact flinty appearance with prominent but small, rounded feldspar porphyroclasts. Locally they contain thin discordant veins of pseudotachylite. In thin section ((S60007) [NB 1885 3253]; (S60009) [NB 1880 3249]) pseudotachylite films mark microthrusts, and movement along them may well have been considerable, for no correlations can be made across them. Two kilometres south-west of Grimersta [NB 2111 2942] pseudotachylite and ultramylonite are developed in locally sheared acid gneiss. Textures similar to those of the granitic mylonites are present (S60010), except that the porphyroclasts in the ultramylonites are oligoclase/albite. However, in places the ultramylonite is extremely fine-grained with a brown colour due to disseminated opaques. The enclosed porphyroclasts contain irregular black cracks and no rotational or banding structures occur adjacent to them (cf. Chapter 15: Pseudotachylitc). These ultramylonites are locally discordant to the mylonitic banding and are similar to retrogressed pseudotachylite seen elsewhere. The gradational nature of the contacts suggests that there has been local melting, probably as a result of a combination of such factors as rapid strain, low water content and compositional differences. Discordant thin pseudotachylite veins also were found in hand specimen. East of the Grimersta area pseudotachylite is more abundant and small thrusts are common. There is thus a gradation from low-grade mylonitisation and cataclasis in granite sheets of the Uig Hills area (where water may have been available) to lower amphibolite-facies conditions in gneisses (possibly drier) of the Ballalan/Seaforth Head area (see also Chapter 15: Pseudotachylite).

Chapter 12 Lewisian Complex: Late-Laxfordian minor intrusions

At three localities (Benbecula, North Harris and northern Lewis) thin dark grey or green dykes, which postdate the Laxfordian folding and the Late-Laxfordian pegmatites, have been mapped.

Benbecula

At Garry a-siar [NF 758 535] a foliated dark grey microdiorite dyke 3 m thick and trending north-west cuts cleanly across both the gneissic banding and a 1 m late-Laxfordian pegmatite that trends east–west. At their junction the pegmatite is displaced sinistrally by 20–30 cm. Three metres farther east, the pegmatite in turn cuts a Younger Basic dyke 8 cm wide, and 9 m farther to the east again, it truncates another Younger Basic dyke 0.5 m wide. In thin section the post-pegmatite dyke is fresh with an average grain-size of about 1 mm. It has a pronounced pervasive fabric defined mainly by dark brown biotites up to 1.5 mm in length. The biotite grows across green hornblende, which is in part dimensionally orientated parallel to the biotite and contains small quartz blebs and inclusions of sphene. Slightly strained quartz, either in single grains or aggregates up to 1.5 mm across, is common. Plagioclase (oligoclase-andesine) occurs in tablets or anhedral crystals showing slight strain-twinning and partial elongation in the fabric plane. Accessory minerals are sphene and apatite, both very common, and isolated, somewhat cubic, opaque magnetite and/or ilmenite, some with relict cores of pyrite. Allanite with narrow rims of epidote is also present.

A chemical analysis of this microdiorite dyke is presented in (Table 23) (analysis 20). The rock is Q-normative, and although similar to the late-Scourian microdiorites it has slightly lower Al2O3 and Na2O : K2O values, and high values of normative apatite and ilmenite (both at 3.8 per cent). It differs radically from the tholeiitic Younger Basics which preceded Laxfordian deformation.

North Harris

On the north side of the road between Tarbert and Husinish, close to the gate leading to Loch Leosaid [NB 052 075], a thin (about 1 m) dark grey microdiorite dyke, foliated parallel to its margins, is seen cutting the gneiss. In appearance it closely resembles the dykes described below from Lewis and on that basis is included in this section.

Lewis

About 1 km WSW of the Butt of Lewis several thin grey dykes have been noted by Watson (1969) and Davies et al. (1975). They describe them as 'possibly representing metalamprophyres' and 'roughly contemporaneous with the pegmatites', the youngest Laxfordian igneous event in that area. The dykes are schistose parallel to their margins and cut across folded supracrustal gneisses of the Ness assemblage. A thin section of one shows it to have a pervasive small-scale augen fabric with a grain size of 0.5 mm. Hornblende is the dominant mafic mineral, occurring as lozenge-shaped dark green crystals with slight marginal recrystallisation into small subgrains. Many crystals are 'dirty', cracked, and twinned, and some are slightly kinked; they are outlined by sheaths of reddish brown ragged biotite which wrap around the hornblende and, like the hornblende, display strain effects and local kink bands. Oligoclase forms distinct augen 0.5 mm long with bent twin lamellae. Quartz is common, with two modes of occurrence: one as highly strained lenticles of similar size to the plagioclase, and the other as elongate strained grains, 1 mm by 0.1 mm, which wrap around the plagioclase and exhibit partial recrystallisation to strain-free fine-grained granular aggregates. Accessories are apatite and sphene, together with smeared-out ilmenite, partially altered to leucoxene. Thin veinlets of hematite randomly cut the rock.

Petrographically the two described dykes are similar, though texturally very different, and both could be classed as quartz-microdiorites. They principally differ from the Late-Scourian microdiorites to the west of the Outer Hebrides Thrust (Chapter 6) in being more mafic and richer in sphene and apatite, as shown in the modal analysis of (Table 26).

The post-pegmatite age of the Garry a-siar dyke is unequivocal, and for this reason it was selected for radiometric dating. Two biotites gave K-Ar ages of 1422 ± 36 Ma and 1396 ± 35 Ma. When combined these give a minimum age of intrusion of 1409 ± 25 Ma (IGS Isotope Geology Unit, Report No. 81/4).

Possible intrusions of this period are described from the island of Rona, close to the Scottish mainland. Here Lyon et al. (1973) distinguish 'basic minor intrusions' emplaced after pegmatites dated at 1740 Ma. The emplacement of these basic intrusions is considered to be the last Lewisian igneous event in that area.

Chapter 13 Structural history

The structural pattern of the Lewisian Complex reflects the repeated deformation during the long period (c.2800–1700 Ma) when the rocks were subjected to high temperatures and pressures. The only important post-Laxfordian structures in the Outer Hebrides are those related to the Outer Hebrides Thrust Zone, which is dealt with later in this memoir. The complexities of the structural pattern seen both in the field and on the maps arise from two main causes: the superposition of structures of several different ages, and lateral variations in structure that reflect differences in the rock's response to stress during successive phases of deformation.

The effects of repeated deformation are demonstrated by the superposition of fold systems, boudinage and shear-zones formed during successive episodes and by the overprinting and modification of early tectonite fabrics by those related to later phases of deformation. Structural sequences based on evidence of this kind have been described from many areas in the Lewisian Complex and have been linked with the general history of the complex by reference to the structural relationships of suites of minor intrusions and to dated metamorphic events (Table 1). The correlation of these sequences from island to island is not everywhere secure. The scheme used in this chapter and summarised in (Table 27) is based on a reappraisal of the published work of Coward, Davies, Francis, Graham, Lisle, Myers and Watson and on unpublished work by these authors and by the IGS (as BGS was then named). An alternative scheme involving different interpretations of regional correlation based largely on presumed constancy of fold geometry and orientation is favoured by Hopgood and Bowes (1972) and Taft (1978); it is not easily related to that set out in (Table 27).

Chronological tabulations of structural events tend to give equal weight to each tectonic episode. In (Table 27) we emphasise the principle phases in the evolution of the structural pattern at the expense of those whose effects were either weak or localised, in order to convey a more realistic impression of the deformation pattern. It will be seen that three events were of crucial importance–the main Scourian (Badcallian) gneiss-forming phase of injection and metamorphism (ds2,), which established the general characters of the gneiss, and the two Laxfordian events (dL2 and dL3) which were responsible both for the formation of most of the mesoscopic folds and related structures and for defining the regional structural pattern.

Large-scale structures

The structures outlined on Maps 3 and 4 are defined mainly by the gneissic banding that largely developed during ds2, the earliest widely visible tectonic phase. (Full descriptions of each phase of deformation follow later in this chapter.) The banding is normally parallel to the narrow belts, screens and lenses of pre-Scourian supracrustal gneiss which occur in the quartzofeldspathic grey gneiss. Many post-Badcallian igneous bodies were emplaced as discordant intrusions which, where subsequently deformed, exhibit structural patterns distinct from those defined by the gneissic banding. These intrusive bodies were affected only by structures developed during or after the period of their emplacement and therefore serve as markers by which earlier and later structures and fabrics may be separated (see (Table 1)).

Over most of the area the tectonic grain defined by the gneissic banding has a WNW to NNW trend that is at least in part a result of tight folding on axial planes that strike north-westerly (Figure 30). The metadolerites of the Younger Basic suite, which are correlated with the Scourie Dykes of the mainland (Dearnley, 1962a), are involved in this folding, so the folding is probably of Laxfordian age. It is commonly attributed to one or both of the principal Laxfordian deformation phases, dL2 and dL3. In Harris and Lewis the north-westerly grain is generally associated with folds on gently inclined axial planes assigned to dL2. In contrast, in the southern islands the dominant north-westerly structures are developed on steep axial planes overprinted on flat-lying structures probably representative of dL2 (Coward et al., 1970). These dominant later structures, assigned to dL3, have a characteristic geometrical form in which broad flat-topped antiforms are separated by steeply dipping 'partitions' of strongly foliated gneisses that represent shear-zones, tight synformal folds or strongly attenuated limbs of major folds.

A distinctive feature of the large-scale structural pattern is the lateral variation in the intensity of strain related to dL2 and especially to dL3 (see insets on Maps 3 and 4 and the detailed discussion of Laxfordian structures later in this chapter). In areas of low Laxfordian strain, up to about 10 km wide, structures formed during Scourian deformational events which are preserved with relatively minor modifications and the primary shapes of Late-Scourian intrusions and Younger Basic dykes are still largely recognisable. Areas which show moderate Laxfordian strain are considerably more extensive, and as a rule are characterised by well-developed structures of the earlier (dL2) set and by a restricted distribution of the later (dL3) set. Areas of high Laxfordian strain commonly show the imprint of strong deformation during dL3, superimposed on earlier Laxfordian and Scourian structures. The highest Laxfordian strain is thus associated with dL3 while the more moderate Laxfordian strain, which is more widespread, is associated with dL2.

In the southern islands and along the Sound of Harris the variations of Laxfordian strain are systematically arranged with respect to the large dL3 antiforms mentioned above (Coward et al., 1970). The broad antiformal areas are regions of low to moderate strain in which the structural patterns date largely from the Scourian or early Laxfordian (dL2). The narrow partition zones that separate adjacent antiforms are characterised by a strongly developed planar foliation in part coinciding with the axial planes of tight to isoclinal dL3 folds. Pre-Laxfordian structures are greatly distorted, and intrusions of the Younger Basic suite appear mainly as concordant amphibolites that may be almost indistinguishable from older mafic gneisses. These 'partition' zones appear to have suffered very high strain during one or more phases of Laxfordian deformation.

Inhomogeneities of Laxfordian strain effects are conspicuous also on a local scale where the relative ductilities of adjacent rock-types controlled their response to stress. Rocks which were more ductile under the prevailing pressures and temperatures were modified by penetrative deformation while the less ductile rocks retained structures dating from earlier phases with only minor modifications. These contrasted responses allowed early fabrics to be preserved locally in competent rocks, even in zones of high strain.

A distinctive aspect of the structure that arose from the in-homogeneous response to stress was the development of linear zones of high strain characterised in the field by close lithological striping, by strong planar foliation and by the scarcity of remnants of early structures. These zones resemble the 'straight belts' described by many authors dealing with gneiss terrains elsewhere (Bak et al., 1975; Bridgwater et al., 1973b; Coward et al., 1973; Watterson, 1975) although here they are considerably smaller in scale than the referenced examples. They were ductile shear zones across which relative movement of adjacent rock masses took place; they are regarded as the deep-seated equivalents of faults. The biggest such zones are commonly marked by Bouguer gravity or aeromagnetic anomalies. Some of these zones of high strain were initiated in Scourian times and strongly influenced the Late-Laxfordian structural pattern.

Mesoscopic structures and fabric systems

Gneissic banding

In almost every outcrop of Lewisian gneisses the most conspicuous structural elements are those defined by small-scale lithological variations. As already described in Chapters 2–12, the variations define structures ranging from a close striping through a crude irregular banding to almost random patterns of more mafic lumps and patches in a more felsic matrix or conversely of felsic veins in a mafic matrix. Individual lithological units range from a few centimetres to a metre or so in thickness, and the structure which they define is commonly emphasised by the alignment of small mineral aggregates and schlieren. This small-scale alignment can be appropriately termed a gneissic foliation; the structural element defined by larger-scale lithological variations is referred to here as a gneissic banding. The banding and foliation, which are in general subparallel, together define the structural patterns outlined on Maps 3 and 4 and (Figure 30).

Teall, whose account of the corresponding structures in the north-west Highlands of the Scottish mainland can hardly be improved upon (Peach et al., 1907, Chapter 4), attributed the heterogeneity of Lewisian gneisses and granulites mainly to the effects of repeated injection of magma under plutonic conditions. An alternative possibility, that the banding is at least indirectly derived from a primary stratification, is preferred by several authors. Most present-day workers, including us, accept both Teall's concept of multiple intrusion and his additional conclusion (Peach et al., 1907, p.71) that the 'secondary and more striking features' of the banding and foliation 'were undoubtedly determined by plastic deformation' (see Chapter 5).

Two aspects of Teall's last point require elaboration. First, the absence of relict primary structures (other than crude lithological stratification) in rocks which are universally accepted as metasedimentary or metavolcanic (e.g., in the Leverburgh and Langavat areas of South Harris) suggest that high strains have been imposed (cf. south-west Greenland, where distorted primary structures are not uncommon: Bridgwater et al., 1976, p.39). In grey gneisses, where repeated injection of acid material can be demonstrated (e.g., central Lewis and North Harris), primary features such as the ramification of vein networks and the discordant contacts of veins have usually been strongly modified; agmatites and vein networks tend to have been flattened, discordant veins have been rotated towards the foliation plane or tightly folded, concordant pegmatites have been boudinaged, and so on. We therefore see that the gneissic banding records the effects of ductile deformation that was shared by rocks of both supracrustal and plutonic origin (cf. Myers, 1970a). Secondly, examples of the whole range of banding styles are cut by Late-Scourian intrusions and by Younger Basic dykes, so it seems clear that the distinctive character of the banding was established at an early stage.

Tectonite fabrics

The gneissic banding and foliation described above define a crude L-S fabric system (Flinn, 1962, 1965), in which linear and planar elements are outlined by thin layers, pods and mineral aggregates. This fabric system originated early in the tectonic history and was repeatedly modified by successive increments of strain during later phases of ductile deformation. Where high Laxfordian stains are recorded, as in much of South Uist, the fabric system defined by the gneissic banding is regarded as the product of two or more independent strain increments (see below and cf. Coward, 1973a; Graham and Coward, 1973; Davies et al., 1975). Variations in the history of the L-S shape fabric are indicated in the summary map (Figure 33).

Fabric elements which are genetically independent of the L-S shape fabric system include axial-plane foliations oblique to the gneissic banding, axial lineations formed by intersection of banding and axial-plane foliation, fabric systems developed in younger intrusive rocks, and schistosities and lineations defined by the preferred orientations of minerals formed after ds,. Many of these elements can be assigned to individual tectonic phases (cf. Bowes and Hopgood, 1975).

Effects of ductility contrasts

The lithological inhomogeneity of the gneisses largely controlled the geometrical forms of the mesoscopic structures. Contrasts in ductility gave rise to many eye-shaped structures (in which foliation and banding in more ductile rock were moulded on irregular or lenticular masses of less ductile) as well as to mesoscopic folds and boudins. The relative ductility depended both on original lithology and on metamorphic environment. In general, pegmatites had low ductility, that is, they behaved competently relative to all other rocks; more mafic rocks also were less ductile than more acid rocks. The competence of mafic rocks was enhanced where they had anhydrous (pyroxene ± garnet) mineral assemblages but was reduced where they contained biotite (cf. Francis, 1973). The presence of biotite as a principal ferromagnesian mineral is almost invariably associated with high ductility in rocks of all compositions. The abundant folds and boudins noted in outcrop are commonly defined by the more competent pegmatitic and mafic layers; folds and boudins of Laxfordian age often involve metadolerites of the Younger Basic suite. The geometrical forms of many such folds indicate that buckling of individual multilayer units was followed, in areas of high strain, by homogeneous flattening (see Coward, 1973a). The scarcity of thick competent units capable of forming buckle folds with amplitudes of a kilometre or more means that there are few major folds with styles the same as minor folds seen in outcrop. Such large folds as have been recognised appear to have formed mainly by mechanisms other than buckling under which the banding acted as a passive marker.

Scourian structures

The earliest structural developments (dS1)

The metasediments and associated basic rocks which appear to be the oldest rocks of the Lewisian Complex (see Chapter 5) locally display structures that may predate the main Scourian gneiss-forming event (dS2). These metasediments are enclosed by grey gneiss (e.g. Myers, 1970a). They commonly show stronger fabrics than the grey gneiss; it is not clear what this signifies.

The main gneiss-forming event (dS2)

The main phase of gneiss-forming activity at around 2700 Ma was associated with the repeated injection of thin concordant sheets, discordant veins and vein networks of tonalitic to granitic 'fluids' during a period of ductile deformation and high-grade metamorphism (cf. Chapter 5). By the end of this phase supracrustal and mafic gneisses had become interleaved with the predominant intermediate to acid gneisses in the form of pods, patches and bands up to a few hundred metres in thickness; the characteristic gneissic banding had been established.

The effects of interleaving of supracrustal and granitic gneisses are illustrated on Map 1 and (Figure 30) and have been discussed in Chapters 2 and 5. These effects are due largely to an injection of felsic magmas into supracrustal rocks while these rocks were undergoing high-grade metamorphism. Evidence that gneiss-forming deformation overlapped with a period of acid injection is seen at many localities; folded veins have been cut by undeformed veins belonging to later phases (Myers, 1970a, fig. 3), veined and distorted mafic gneiss has been agmatised by later influxes of acid material (see Chapter 2: Lithology; various sections on Scourian migmatisation, and Myers, 1970a, plate 3A), and intrafolial folds involving early veins are confined between felspathic stringers of a later generation.

It is uncertain to what extent deformation during dS2 was associated with large-scale folding or reduplication by regional thrusting, as inferred in Greenland (Bridgwater et al., 1974). The Leverburgh and Langavat supracrustal belts were considered by Dearnley (1963) to be isoclinally folded on both a regional and local scale, but this conclusion has not been confirmed by later workers (see Chapter 3: Lithology: Langavat Belt). Possible fold interference patterns outlined by supracrustal gneisses south of Loch Resort, North Harris, involve a fold set predating dL2 according to Myers (1970a). Elsewhere there is little positive evidence of large-scale early folding.

The principal effect of deformation during and after injection was the production of the small-scale gneissic banding already described (this Chapter: Gneissic banding). Myers (1970a) has shown that transitions from more massive veined gneisses and agmatites to gneisses characterised by pegmatitic or mafic lenses, pods and layers are related to deformation. The mesoscopic L-S fabric system (Flinn, 1962, 1965) defined by the gneissic banding which persisted, more or less modified through all the subsequent phases of deformation, appears to have originated during d„, for it is cut discordantly by Late-Scourian intrusions and Younger Basic dykes. In most regions, the planar element of this system is dominant and defines the structure that was folded during later phases. However, there is a weak lineation defined by the elongation of mafic or ultramafic clots and of smaller aggregates of quartz, hornblende, biotite or feldspar (L1 of Hopgood and Bowes, 1972). This crude lineation is generally coaxial with Laxfordian fold axes and other linear structures suggesting that it may have been rotated by Laxfordian deformation. In contrast, over much of central Lewis, where Laxfordian strains were relatively low, there is a strong Scourian ribbon-lineation trending west-south-west that is consistently oblique to the axes of Laxfordian folds (Davies et al., 1975, fig.2).

Late-Scourian structures

Structures that distort the gneissic banding but which are cut by members of the Younger Basic suite are assigned to Late-Scourian phases; these have not, as yet, been firmly correlated from island to island. In East Barra, where minor intrusions dated as being younger than 2600 Ma provide time-markers (Chapter 6 and Francis et al., 1971), at least two sets of small-scale structures have been distinguished (Francis, 1973). One set has distorted the gneiss banding, but the other has only affected early phases of the Late-Scourian intrusives (see Chapter 6: Conclusions). These structures are cross-cut by undeformed Late-Scourian pink pegmatites, which are probably of a similar age to the pegmatite dated at Leanish (c.2610 Ma, see Chapter 6, Pegmatites). In Harris and Lewis, and possibly elsewhere, the development of large asymmetrical folds appears to have been followed by the development of large (Inverian) shear-zones. To interpret these major structures one must consider the effects of subsequent Laxfordian deformation.

The alignment of the gneissose banding prior to the imposition of Laxfordian (dL2 and dL3) folds can be established directly where bulk Laxfordian strains were low, but must be inferred elsewhere by observing fold symmetry and reconstructing the enveloping surfaces of these folds. Through most of Harris (where dL3 is weak), large dL2 folds on NW–SE axial planes have roughly symmetrical M-shaped profiles. The enveloping surfaces of these folds strike north-north-west, as do the boundaries of distinctive gneiss tracts (Myers, 1970a; Soldin, 1978). From Benbecula to central South Uist, where the strike of the foliation is almost north–south, dL2 folds are consistently asymmetrical, defining Z-profiles where the plunge is to the north-west (Coward, 1973b). These relationships are consistent with the strike of the banding having been roughly NNE over large areas prior to Laxfordian folding. The Bouguer gravity contours have a roughly similar trend (Figure 30), and McQuillin and Watson (1973) have inferred that a large-scale lithological zoning running NNE existed in Scourian times, with more granitic low-density rocks concentrated in the west. If there was a gravity gradient from a 'low' over the Uig Hills (Lewis) to 'highs' near the east coast, then the lithological zones dipped steeply. However, Coward (1973b) concludes from the geometry of dL2 folds that dips were low at the onset on Laxfordian deformation.

The old-established north-north-eastward lithological grain is interrupted in anomalous regions where banding and/or gravity contours show north-westerly trends. These anomalous tracts are of two kinds, the first representing the short limbs of asymmetrical folds and the second representing shear-zones.

Evidence for the occurrence of large Late-Scourian asymmetrical folds (collectively designated ds3) comes mainly from areas of low Laxfordian strain. In central Lewis east of Loch Roag, a consistent north-westerly trend with steep dips is emphasised by the outcrops of supracrustal rocks in the Loch Laxavat belt (Lisle, 1974). This trend is crossed obliquely by a group of little-deformed Younger Basic dykes; it was evidently established before the emplacement of the dykes (Davies et al., 1975). The area of north-westerly trend is flanked to the south by a large area (southern Lewis and Harris) in which a north-north-easterly trend predominated in the pre-Laxfordian structure (see above). We suggest that the area of low deformation east of Loch Roag coincides with the short north-west-striking limb of an asymmetrical Z-fold whose long limbs define the regional pre-Laxfordian north-north-east grain (Figure 30). The smaller areas of low Laxfordian strain in north-east Barra, at Ardivachar Point (South Uist) and in west Benbecula coincide with northwest-plunging antiformal fold-crests which have been regarded as Laxfordian structures (Coward et al., 1970; Francis, 1973; Coward, 1973a). Re-examination suggests, however, that Younger Basic dykes which cross these areas do not show systematic variations of orientation in sympathy with the curvature of the gneissic banding such as would be expected if the folding was entirely Laxfordian. We therefore infer that low-deformation areas in east Barra and at Ardivachar and Garry a-siar (Maps 3 and 4) represent the crest and short limb of an open Scourian Z-fold, tightened and modified by Laxfordian deformation.

Shear zones

There are two sets of early shear-zones which appear to have been zones of high strain during both Late-Scourian and Laxfordian times (see also the end of this chapter and Maps 3 and 4). A north-east-trending zone–the only major shear zone with this orientation–skirts west Lewis from the Uig Hills to the Butt of Lewis; it is at least 60 km long. It is characterised by steeply dipping striping that is unusually regular, by closely spaced partings, and by well-developed boudinage of concordant basic layers. The gneissic banding that trends north-west or west in the interior of Lewis curves anticlockwise into the shear-zone as the coast is approached, suggesting a component of sinistral displacement across this zone. This curvature is not shared by discordant Scourie dykes near the margin of the shear zone, which is thus considered to date from a late Scourian (dS4) phase of movement (Davies et al., 1975).

In the Ness area and the Leverburgh belt (South Harris) strongly defined L-S fabric systems are cut by pre-Laxfordian minor intrusions and/or folded in early Laxfordian folds. This suggests that shear-belts in these localities also originated in Late-Scourian times, perhaps during the same event (dS4) as the north-easterly shear-belt of west Lewis. These structures are further discussed at the end of this chapter (Inverian shear zones).

The Younger Basic Suite and first Laxfordian events

The effects of Laxfordian deformation and metamorphism have obscured the original forms and relationships of the Younger Basic suite over much of the Outer Hebrides. Nevertheless, dykes that have been little modified can be seen in areas of low Laxfordian strain, and in areas of moderate strain relatively large intrusions or groups of intrusions can be traced for considerable distances as discontinuous concordant bands, clusters of pods and disrupted fold hinges. The evidence suggests that the original trend of most intrusions was between west and north-west, parallel to that of the Scourie dykes of the mainland. Dyke contacts for which the intrusive relationships have been little modified strike 100–110° in central Lewis and Great Bernera (Watson, 1968; Davies et al., 1975), in west Benbecula, and west at Ardivachar Point (Dearnley and Dunning, 1968; Coward, 1973a). A strike almost due north (Hopgood, 1964; Francis, 1973) in eastern Barra may have been affected by a clockwise rotation of some 30° during the development of the Outer Hebrides Thrust (Hopgood, 1971; Sibson, 1977b).

In parts of south-west and east Greenland and on the Scottish mainland, there are primary irregularities in the forms of dykes that have arisen because the dykes were emplaced deep in the crust and/or during periods of crustal disturbance (e.g. Bridgwater et al., 1973b; Escher et al., 1976a; Jack, 1978). There are indirect indications that some of the Younger Basics in the Outer Hebrides were emplaced in similar conditions; there is an early schistosity in metadolerites of the southern isles, and primary minerals have been partly reconstituted to two-pyroxene or pyroxenegarnet assemblages, notably in Harris and Lewis (see Chapter 10: Mineralogy). The early foliation is parallel to the dyke contacts and to small ductile shear zones (e.g. Coward, 1973a). It was folded and modified by the widespread du structures and was assigned to an early Laxfordian event (dm) by Coward (1930a) and Graham (1970). As the foliation is better developed in the dykes than in the surrounding gneisses, it may date from the period of emplacement. Similarly, the early metamorphic assemblages of Harris and Lewis, which have been assigned by Dearnley (1962a, 1963) to an early Laxfordian event, have been interpreted as products of reactions that immediately followed the period of emplacement (see Chapter 10: Conclusions). We infer that the dykes were emplaced when the crust was hot and subject to regional stresses, as has been suggested for the Scottish mainland (O'Hara, 1961a, 1977, Park and Cresswell, 1973), and not at high levels under stable conditions as envisaged by Sutton and Watson (1951).

Early Laxfordian structures (dL2)

The principal phase of early Laxfordian deformation (dL2) was responsible for much of the mesoscopic and megascopic folding seen over most of the Outer Hebrides as well as the folding, boudinage and disruption of intrusions of the Younger Basic suite. In many regions dL2 was a fabric-forming event and was associated with phases of coarse recrystallisation. The XY plane of the deformation ellipsoid, that is defined by the planar element of the fabric system and by the axial planes of folds, appears to have been sub-horizontal in many areas (Figure 31); the low dips that are prevalent over large parts of central Lewis, North Harris and North Uist, where later deformation was of restricted importance (Figure 33), reflect this initial pattern.

Where Laxfordian bulk strains are low (Maps 3 and 4, insets) du folds are sparsely distributed and their form is usually open (e.g. near Dun Carloway, Lewis). The preexisting banding and L-S fabric in gneisses and metadolerites are folded around these open folds, and a weak axial plane foliation defined by mineral aggregates and crystallographic preferred orientations has developed in places.

In regions of moderate to high Laxfordian strain, dL2 folds are tight to isoclinal. Estimates based on the forms of fold profiles for the minimum flattening strain in two dimensions are around 8:1 (Coward, 1973a) or 10:1 (Davies et al., 1975). Boudinage of early pegmatites and amphibolites (especially those derived from Younger Basic dykes) is conspicuous on the limbs of tight folds; it is well seen in Howmore quarry, South Uist [NF 7662 3649] and along the north-west coast of North Uist and Lewis. Extreme flattening has obscured fold hinges, especially in the less competent acid gneisses, and has led to the development of tracts of gneiss in which the strong foliation parallel to the fold limbs produces a misleading appearance of regular layering. This is well seen in the northern part of the Eye Peninsula (Lewis). The planar element of the Scourian fabric system has been intensified on the fold limbs, for there Laxfordian strain has exaggerated the pre-existing foliation; the linear element of the Scourian system is generally rotated into parallelism with the fold axes. At the hinge-zones of competent layers, the linear element is commonly enhanced, a feature which may result from the imposition of Laxfordian flattening strain on a pre-existing foliation lying at a high angle to the direction of maximum strain (Graham and Coward, 1973; Coward, 1973b).

The regional arrangement of dL2 structures is illustrated in (Figure 31), in which the orientation of fold axial planes and axes and the symmetry of fold profiles are summarised. In areas least modified by later deformation the characteristic features are the low dips of the axial planes and associated fabric, and the predominant NW–SE axial trend. Northwesterly plunges near the west coast of the Uists and Harris and south-easterly plunges elsewhere define an axial culmination along the length of the islands.

Variations in the symmetry of mesoscopic fold profiles relate to the positions of these folds with respect to the hinges of larger folds and (where large folds are not developed) to the angular relationships between the gneissic banding and the dL2 axial planes at the start of the fold phase. Narrow zones of symmetrical M-folds appear to mark the short steep limbs of larger asymmetrical folds in the Ness area of Lewis, (Davies et al., 1975) and on the north-west coast of North Uist (Graham and Coward, 1973). Larger tracts of M-folding are seen north of Carloway and at many localities in central Lewis and North Harris, where fold-stacks with almost vertical enveloping surfaces but subhorizontal axial planes are well displayed in cliffs and steep hill slopes. These structures appear to mark areas which were characterised by moderate to steep dips at the onset of dL2. In central South Harris, fold-profiles are related to a pair of large, tight folds (the Horgabost antiform and Uaval synform (Figure 31)) with axial planes dipping steeply and striking NW–SE (Myers, 1971).

Asymmetrical Z- or S-shaped fold profiles predominate over much of western Benbecula and South Uist where the general trend of the banding is almost north–south. Boudinage of amphibolites and enhancement of the planar element of the L-S fabric-system in these areas record flattening in the plane of the banding and suggest that d„ axial planes made only a small angle with the layering at the onset of deformation (Coward, 1973a, 1973b). The steep dips of the banding and dL2 axial planes in Benbecula and South Uist are attributed by Coward to the effects of subsequent refolding.

Later Laxfordian structures (dL3 and dL4)

The later stages of the Laxfordian event were marked not only by episodes of inhomogeneous deformation but also by the emplacement of granitic vein-complexes and suites of pegmatite dykes, the distribution of which suggests a regional tectonic control (Table 1). The principal phase of ductile deformation, dL3, was roughly coeval with the onset of development of the granite-migmatite complex of Harris (Myers, 1971), and a more local phase, dL4, was associated with migmatisation in South Uist (Coward, 1973b). The main pegmatites postdate these phases, and their emplacement was accompanied and followed by phases of cataclastic deformation, retrogression, localised faulty and open warping (dL5 dL6).

The structural style associated with dL3 has already been outlined at the beginning of this chapter. The broad antiformal regions of moderate to low strain and the intervening north-west-trending partitions characterised by high strains dominate the large-scale structure of the southern islands and the Sound of Harris; there, dL3 was apparently the dominant ductile Laxfordian deformation event. Minor folds in these regions are open in areas of low strain but tight to isoclinal where strains were moderate to high. They form spectacular interference patterns with dL2 structures, notably on the north-western coast of North Uist (Graham, 1970). The planar element of the pre-existing fabric-system is again enhanced on the limbs of isoclinal folds.

In Harris and Lewis, large-scale dL3 structures equivalent to those of the southern isles are poorly developed. There are also comparatively few minor structures; unlike those of the southern islands they appear to have been formed when the rocks were not very ductile except in migmatite areas. The early stages of late-Laxfordian granite injection and migmatisation coincide with dL3, but although the location of the three main sites of injection, namely central South Harris, western North Harris and the Uig Hills, may have been structurally controlled, the structural significance of these complexes is obscure.

Although the orientation of structures assigned to dL3 shows considerable regional variation ((Figure 32) and (Figure 33)), a common feature is the steep dip of fold axial planes and shear-zones, which contrasts with the subhorizontal axial planes and fabrics associated with dL2. This contrast suggests a change in tectonic regime (Graham, 1970, Coward, 1973b) which may have been connected with the rise and unroofing of the Lewisian complex (Dickinson and Watson, 1976). Axial planes mainly strike north-west, as do Late-Laxfordian structures on the Scottish mainland, but in central and northern Lewis, shear-zones and fold axial planes that trend north-north-east are common. In the southern islands, tight dL3 folds are generally coaxial with dL2, whereas most of the more open folds of Lewis are oblique to the d„ axes. These regional variations make correlation uncertain; this is illustrated by the lack of accord between the schemes adopted by Hopgood and Bowes (e.g. 1972) and by Coward et al., (1970) and Davies et al., (1975).

The later stages of the tectonic history are recorded by a number of locally developed sets of minor structures, such as the tight folds associated with an area of migmatitic recrystallisation in South Uist (f4 of Coward, 1973b), by the distribution of the main Laxfordian pegmatite swarm, by widespread but localised cataclastic deformation, and by open warping mainly on north-north-east-trending axial planes.

Late-Laxfordian pegmatites are common over a 200 km tract extending NNE along the entire length of the archipelago. They are most abundant in and around the granite migmatites of Harris and the Uig Hills, where they locally form a significant proportion of the outcrop area (e.g. in Taransay). Beyond the limits of the migmatites, pegmatites are seen every few hundred metres along the west coast from the vicinity of the Butt of Lewis to Barra. They are relatively scarce in central and eastern Lewis and the eastern parts of the southern isles. Outside southern South Harris, almost all of the large pegmatites trend east - west, dip steeply, and are aligned at a large angle to the elongation of the migmatite zone as a whole. If the pegmatites are regarded as dilational bodies (Myers, 1971) they might be interpreted as occupying en-echélon systems of tension-gashes, possibly opened by transcurrent movements parallel to the archipelago. Evidence that deformation continued into or after the late stages of pegmatite emplacement has already been detailed (for example see Chapter 7: Structure: dL4 and later events, and Chapter 11: Pegmatites).

Laxfordian strain

This section is an explanation of the inset diagrams on Maps 3 and 4. During our survey of the Outer Hebrides we paid particular attention to the nature of the gneissic fabric, to minor fold structures and to the field relationships and mineralogy of the Younger Basic dykes. The variations in these features reflect the degree of Laxfordian strain and enabled us to study its pattern and intensity. Such techniques have been used in other gneissic terrains (e.g. by Escher et al., 1976b, in west Greenland; Coward, 1976, 1980, in southern Africa; Sutton and Watson, 1951, on the mainland Lewisian).

The degree of discordance of Younger Basic dykes, whether the gneissic banding is mainly linear or planar, the nature of the Laxfordian folding and the sorts of fabric that have developed are the main features which have been used by Lisle (1976), Coward (1973a), Davies et al. (1975), Myers (1970b) and Watson (1968) to assess Laxfordian strain in detailed studies of parts of the Outer Hebrides. It is difficult to discuss the Laxfordian strain quantitatively for almost nothing is known about the Lewisian gneiss before the Laxfordian deformation. Normally strain in a deformed rock is assessed relative to an undeformed predecessor, but in the Lewisian what the gneiss was like following early- and Late-Scourian ('Inverian') deformation is largely unknown. What the Younger Basic suite was originally like is also unclear (see also this chapter: the Younger Basic suite). It appears that there are concentrations of Younger Basic dykes in particular zones, e.g. in Great Bernera and central Lewis (see Watson, 1968), where they originally formed steeply dipping narrow discordant dykes trending east to ESE. However, Coward (1973a) and Dearnley and Dunning (1968) both show that in the Uists Younger Basic dykes originally trended NW–SE. Myers (1970b) has suggested that although discordant Younger Basic dykes are present in North Harris, many dykes there are concordant and were probably intruded as sheets subparallel to the gneissic banding. There is evidence, both from west Lewis and Barra (see pp.84–85), that dykes were intruded along pre-existing planar zones (?shear-zones). Coward (1973a) states that Younger Basic dykes are thicker and more abundant in low deformation areas, particularly in the Uists. He suggests that these concentrations of Younger Basic dykes were primarily responsible for the establishment of zones of low deformation during the Laxfordian. However, as suggested earlier (this chapter: Late-Scourian Structures), these zones may have started as short limbs of earlier Scourian structures and may subsequently have been resistant to deformation.

The mineralogy of the Younger Basic dykes can be broadly correlated with their structural style and degree of discordance (see Chapter 10: Mineralogy; Feldspar). In some areas, for example North Harris, thin discordant dykes, 2 to 20 cm across, commonly branch and have irregular apophyses (Myers, 1970b). These dykes are simple amphibolites and suggest either widespread Laxfordian amphibolite-grade metamorphism with abundant movement of fluid but little deformation, or possibly retrogression associated with injection of Laxfordian granite.

Laxfordian strain also shows abrupt variations in intensity (Coward, 1973a) commonly on a scale too small to be represented on the inset maps (Maps 3 and 4). In parts the Lewisian gneiss itself lacks features from which Laxfordian strain could be gauged. This is particularly so in South Uist, where Younger Basic dykes are sparse and although there are strongly planar zones, they may well have been already well formed in the early or late Scourian. Previous authors (already cited in this chapter) working in the Outer Hebrides have held that the presence of a dominant planar fabric is generally coincident with high Laxfordian strain areas. Although this relationship is well known it does not exclude the possibility that planar zones or locally planar fabrics were developed prior to the intrusion of the Younger Basic dykes. The superimposition of successive high-strain zones in gneiss terrains has been noted from other parts of the world (e.g. Escher et al., 1976b, in west Greenland). Park and Cresswell (1973) and Tarney (1973) found that in the mainland Lewisian the orientation of Scourie dykes was in part controlled by near-vertical shear zones of Inverian age that were approximately coeval with the intrusion of the dykes.

Such complexities make it difficult to interpret the Laxfordian strain patterns of the Outer Hebrides. Nevertheless, in view of the data already published, in addition to the large quantity of data collected during this survey, we feel that we can make a reasonable broad assessment, with the reservations stated above, that most features could have been formed in more than one way.

Zones of low strain

These are generally lenticular areas or pods typically 1–2 km wide and up to 5 km long in Lewis, Harris and the Uists. The exceptions to this generality are the large meta-igneous bodies of South Harris and South Uist (Corodale Gneiss), and parts of east Barra and the adjacent islands. The granulite-facies meta-igneous bodies are thick, massive and composed of anhydrous minerals, and this has inhibited tectonic and metamorphic reworking. Similar but smaller-scale effects are seen in the central parts of the thicker Younger Basic bodies, such as in the Hosta-Griminish area of North Uist. Dearnley and Dunning (1968) and Coward (1973a) also illustrate minor examples of this feature at Ardivachar Point, South Uist, and Garry a-siar, Benbecula.

An area of pyroxene-bearing acid gneiss occurs north-west of the Outer Hebrides Thrust on Gighay, Hellisay, and the south part of Fuday, immediately north-east of Barra. The south and south-east coast of Fuday provide excellent examples of dyke-gneiss relationships almost unmodified by Laxfordian strain. Coarse-grained, pink augen granitic gneiss with hornblendite pods and extensive, locally discordant, ramifying sheets and veins of Late-Scourian micro-diorite (bi-hbl-pl) are cut by Scourian pegmatite veins up to 3 m wide. This sequence is cross-cut by 'fresh-looking' markedly transgressive Younger Basic dykes (see Chapter 6: Metadiorites and microdiorites; Field aspects of minor intrusions). Similar relationships are also seen in Barra itself above the Outer Hebrides Thrust in the section from Bruernish Point to Rubha Mor. Detailed relationships between deformed microdiorite and Younger Basic intrusives have been recorded on the Leanish Peninsula. Large foliated intermediate meta-igneous dykes and pods are seen in several localities, e.g. the Lamalum Islands [NF 728 033] and the south-east part of Fuday [NF 734 077]. In Fuday microdiorite dykes and associated foliated masses cross-cut coarse-grained pyroxene-bearing acid gneisses. In the Younger Basics of this region mineral assemblages are typically opx-cpx-hbl-pl and cpx-gt-hbl-pl (the hornblende is brown in both assemblages).

Lisle (1976) has enumerated the features of low-deformation zones in north-west Lewis, and we have confirmed them in the present regional study. Younger Basic dykes are commonly discordant, often markedly so, with steep dips and irregular apophyses (see also Dearnley and Dunning, 1968). They are generally not boudinaged, and the irregularities in the dykes are probably largely an intrusive feature. The dykes rarely have a penetrative fabric and where present it is restricted to the marginal zones. The mineralogy of the Younger Basics in such areas of low Laxfordian strain is cpx-opx-pl-brown hbl or cpx-gt-pl-brown hbl. Minor retrogression (more widespread within and adjacent to the Outer Hebrides Thrust Zone) and substantial recrystallisation under uppermost amphibolite-facies conditions is particularly common in the basics. Only very rarely are the original coarse-grained textures preserved (e.g. (S62282) from Oaval, North Uist [NE 8832 7650]).

The gneiss fabric is dominated by linear elements with only few planar zones. Agmatitic structures are abundant, notably where thick Older Basic material is present. Scourian pegmatite-rich gneiss and ramifying migmatitic veins are also seen in many localities (e.g. at Ardivachar Point, South Uist; Lag Macgodrom, North Harris [NB 063 097]; the south-western side of Brevig Bay, Barra). In North Harris, 1.2 km east of Amhuinnsuidhe Post Office [NB 054 084], planar, partly migmatitic, acid gneiss contains anorthosite inclusions and hornblendite pods. Original textures are visible in the anorthosite and the whole sequence is crosscut by Younger Basic dykes.

Major structural features in the gneisses are pre-Laxfordian; in some areas the Scourian folds have axial planes near parallel to the strike of the gneisses. In such areas this structure defines the dominant tectonic grain of the rock. In other areas the dominant linear fabric is accompanied by small-scale, irregular, tight 'dome and basin' fold patterns: these are interference structures from more than one Scourian deformation phase. Such structures are well seen at Aird Fenish [NA 991 293] and west Ardivachar (Coward, 1973a, fig.12).

Zones of moderate strain

These zones constitute a considerable proportion of the exposed rocks of the Outer Hebrides, and include most of North Uist, Benbecula, Harris (except the South Harris Complex) and central Lewis. Within these zones there are localised areas of high or low strain, which are too small to be represented on the inset maps (Maps 3 and 4).

The features of these strain zones are in accord with those described by Lisle (1976) from north-west Lewis. Younger Basic dykes are discordant to the gneissic banding but the angle of discordance is generally small (less than 30°). Marginal irregularities of the dyke are not observed and the dykes commonly contain a new fabric of amphibolite grade. Dykes are commonly boudinaged, and in places there are tight folds, where the initial orientation of the dyke relative to the banding is appropriate. Relict garnet and/or clinopyroxene are commonly seen in thin sections of the basics, and the central parts of thick basic bodies largely retain their original mineralogy; however, the typical assemblages is hblpl-sph (the hornblende is green).

Gneiss develops a more uniform planar fabric than in zones of low strain, although it is still typically well lineated. Scourian structures are generally difficult to recognise, and Laxfordian structures, notably open to tight dL2 folds, are commonly well developed with gently dipping axial surfaces. These structures fold an attenuated banding formed during the Scourian and/or the 'Inverian'. A new Laxfordian amphibolite-facies fabric, related to these dL2 structures, in places becomes a major planar component in the gneisses. Lisle (1976) defines the zones of moderate strain on the basis of the presence of dL2 structures folding a Scourian or Inverian banding and shows that strains of 10:1 to 20:1 are typical in the plane of the fold profile. These zones also show a broad relationship to large-scale dL3 folds in that they typically occur on the more gently north-east-dipping limbs of dL3 antiforms, whereas the steeper western limbs are characterised by zones of high strain.

Zones of high and very high strain

These zones are dominant in South Uist, Barra, south-west and north Lewis and in South Harris and the islands in the Sound of Harris. They are notable for their dominant planar gneissic fabric and concordance or near concordance of Younger Basic dykes. These basics are also commonly strongly podded, amphibolitised and difficult to distinguish from the Older'Basics. Tight to isoclinal, minor to medium-scale folds are observed. Although the lack of discordance has been used to assess the amount of Laxfordian strain, there is some doubt as to whether the Younger Basics were initially as discordant as in the low strain zones. In north and west Lewis the zone of very high strain coincides with a marked change in dip of the gneissic banding from near horizontal to 60° to 90° to the north-west. Assuming that only Laxfordian strains have affected an originally discordant suite of Younger Basics, Lisle (1976) in north-west Lewis calculated that strains of about X: Y:Z = 35:35:1 are necessary to give the observed dyke concordance. This is equivalent to about 90 per cent displacement shortening, assuming no volume loss, over a wide area. Such shortening values are typical of localised shear zones or mylonite zones but seem very high for regional deformation. However the oblate strain ellipsoid (k = 0: Flinn, 1962) is compatible with the abundant boudinage, and this suggests that the idea that large-scale shear (k = 1) is primarily responsible for the major dL2 deformation (Coward et al., 1970) is incorrect.

In the planar gneissic banding individual components can be recognised only with difficulty in many areas because the rocks have been reconstituted and a new penetrative Laxfordian dL2 fabric has developed. Where tight to isoclinal folds of basic dykes are present (e.g. Rubha Quidnish, South Harris [NG 1027 8697]) this new fabric is axial planar. A lineation is variably developed; where present it commonly plunges moderately to steeply to the north-west. Laxfordian agmatisation of Younger Basic dykes and local transformation to hornblende-bearing acid gneisses has been recorded at Ard More Mangersta [NB 004 333] in west Lewis and at Howmore Quarry, South Uist [NF 7662 3649], but this process is not restricted to high strain zones. Davies et al. (1975) note that in north Lewis highly modified pre-Laxfordian relationships are typical of zones of high strain. They observed isoclinal folds of early granite gneiss sheets near the north-eastern tip of the Eye Peninsula in east Lewis.

Graham and Coward (1973) and Coward et al. (1970) suggest that the high strain zones in North and South Uist are coincident with synformal zones of regional dL3 folds. However, we have found in the Bayhead-Tigharry-Hosta and the Pabbay-Berneray areas of North Uist and the Sound of Harris that zones of very high strain form the western limbs of large dL3 west-verging asymmetrical antiforms (see Chapter 14: North Uist). Other north-west-trending zones of high strain in South Uist and South Harris may have been initiated much earlier, and subsequently re-activated during the Laxfordian dL3 episode.

South Harris Complex

The strain variations in South Harris are largely a result of the wide variety of lithologies present. The larger igneous bodies, namely the diorite and anorthosite and parts of the norite and banded gabbros, are little deformed; original banding and/or early granulite-facies fabrics have survived (Horsley, 1978; Dearnley, 1963; Witty, 1975). These bodies were subsequently affected by localised shearing and amphibolite-grade retrogression, particularly in their marginal zones. Graham (1980) and Graham and Coward (1973) have discussed the strain caused by the shearing, and the distribution of shear zones. Using initially randomly orientated feldspathic veins in the sheared zone at the northern margin of the diorite, Graham (1980) obtained X: Y:Z strain values of 10:6:1. However, these values are incompatible with the deflection of the subvertical north-east-striking pyroxene-alignment fabric which formed during the early granulite-facies event. This indicates a total shear value (y) of about 7.5 (implying strains of X: Y:Z equal to 56:7.5:1) adjacent to the diorite contact (Graham, 1980).

The Leverburgh Belt metasediments locally show extensive tight to isoclinal folds and strong penetrative mica fabrics, notably in the zones of marble and graphite schist. Other lithologies, such as the perthite-rich gneisses and the more pelitic units, also show evidence of strong deformation. Although theso rock types formed a locus for Laxfordian strain, early basic pods and ultramafics are discordant to the main fabric and are often little modified internally. This suggests that much of the high strain of the Leverburgh Belt is of Scourian age. The Langavat Belt rocks are also highly strained, but Myers (1968) records that they are cut by slightly to moderately discordant basic dykes. Inclusions of strongly foliated metasediment and amphibolite also occur within the diorite (see Horsley, 1978; Dearnley, 1963; Myers, 1968). Because the rocks of the Langavat Belt are of amphibolite grade, locally biotite-rich, and lie at the junction of migmatitic gneisses and near anhydrous meta-igneous rocks, the belt was affected by considerable Laxfordian and probably Inverian deformation.

Inverian shear zones (dL4)

The term 'Inverian' was used by Evans (1965) and subsequently by Evans and Lambert (1974) to describe a period of post-Scourian and pre- to syn-Scourian dyke age (c.2500 to 2400 Ma) deformation and metamorphism in the Lewisian rocks of north-west Scotland. A record of this event is particularly well seen around Lochinver, where large amphibolite-grade Scourie dykes largely follow regional north-west-trending shear zones. Park and Cresswell (1973) note and discuss other examples from the Loch Torridon, Loch Maree and Loch Laxford areas. They suggest that dyke-gneiss relationships could be explained by varying amounts of structural control during dyke emplacement, a feature first recognised in west Greenland by Watterson (1968). Tarney (1973) notes that members of the Scourie Dyke suite were emplaced over a time-span such that undeformed late picrite dykes cross-cut earlier deformed tholeiitic dykes.

There are several zones in the Outer Hebrides which show evidence of 'shear-zone' deformation approximately synchronous with Younger Basic dyke intrusion. As noted previously (Chapter 10: Petrography: 'Cleitichean Beag' suite), between Aird Fenish [NA 992 292] and Loch Sandavat, a few kilometres to the north-east, there is a steep zone up to 400 m wide of finely banded acid and basic gneisses lying within Scourian agmatitic and lineated gneisses. Amphibolite dykes occur within the zone and along the margins; these Younger Basics show slight discordance to the shear-zone fabric and a weak internal foliation. Their orientation and low deformation are thus compatible with intrusion along an active or pre-existing shear zone. Similar features have been recognised in Greenland where dykes of a similar age have been intruded into Early Nagssugtogidian shear zones (Bridgwater and Myers, 1979; Escher et al., 1976a, b). J. S. Myers (personal communication) states that Younger Basic dykes illustrated by Dearnley and Dunning (1968, figs. 6 and 7) from Garry a-siar, Benbecula, and Pollachar, South Uist, are also related to small shear zones and comments that the overall structure and mineralogy resemble those of the east Greenland dykes. He concludes that the pod-like shape of many concordant Younger Basic dykes may be a product of 'Inverian' deformation at about the time of intrusion.

The problem of the deformation in the Langavat Belt rocks has been discussed in previous sections. The trend of this belt is parallel to Inverian shear zones on the mainland, and it would be surprising if Inverian deformation were not represented in South Harris. If the age of the South Harris Igneous Complex is about 2300 Ma as implied by Nd-Sm isotopic data from the metadiorite (Cliff and Gray, 1983), then its north-eastern margin may have been controlled by such a regional shear zone. At Rubha Charnain in south Barra [NC 680 271] concordant but unfoliated and fresh-looking Younger Basic dykes lie in planar acid gneisses. One to two kilometres farther east, the Younger Basics have a similar appearance but are markedly discordant to the Scourian gneissic structures. It appears that dyke intrusion at the first locality has been controlled by a pre-existing shear zone trending about 160°. Adjacent to this shear zone and its relatively undeformed dykes, however, there is a zone of intense Laxfordian deformation and migmatisation. Here, tightly folded Younger Basic dykes with an axial planar fabric, and partially agmatised Younger Basics, are found in pink, subvertically banded, migmatitic acid and basic gneisses trending about 170°. The relationships can be explained if the pre-existing Inverian shear zone acted in part as a locus for subsequent Laxfordian deformation and migmatisation. It is possible that the trend of these planar zones has been rotated from its original position by the Outer Hebrides Thrust. Sibson (1977b) calculates the rotation to be about 30° clockwise; the criteria for matching features across the main thrust are questionable, but if they are accepted the Rubha Charnain shear zone would have originally trended about 130°. This trend matches similar undeflected zones further north in the Uist and South Harris.

Chapter 14 Regional variations of structure

Regional aspects of the various structures can be considered by reference to three structural domains: the Southern Isles (Mingulay to North Uist); the Northern Isles (dominantly Harris and Lewis); and the Sound of Harris. The last domain lies between the larger ones to the north and south, and is treated separately because it is anomalous in many respects.

The Southern Isles

The threefold regional structural division of the Outer Hebrides proposed by Dearnley (1962a) is based on the inference that North Uist, Benbecula and northern South Uist are largely unaffected by the effects of Laxfordian deformation which are seen in southern South Uist and Barra. Our work indicates that Laxfordian deformation and metamorphism affected almost the entire domain, although variations in the intensity of Laxfordian deformation have resulted in the presence of a number of small, widely separated areas of low strain. However, we regard the domain as an entity.

Mesoscopic structures formed during the two principal Laxfordian phases of deformation, dL2 and dL3 are widespread, and the main features of the mapped structural pattern were established during these phases. The relationships of fabrics and textures suggest that the characteristic coarse-grained gneissose textures date from late-kinematic or post-kinematic recrystallisation associated with dL3. Aspects of the structure are summarised below, working from south to north.

Barra and adjacent isles

The small isles south of Barra and much of western Barra itself are characterised by low dips (10° to 30°) to the east or north-east on the southern limb of a large dL3 fold. Strike variations are due to late open folding on steep axial planes (Hopgood, 1971; Hopgood and Bowes, 1972). The area of low dips is bordered on the north by a narrow NNW-trending zone of steep, strongly foliated gneisses which crosses the Scurrival peninsula; to the south the zone is apparently displaced by the Outer Hebrides Thrust (Francis, 1973). This zone, interpreted by Francis as a tight dL3 synform, is in turn flanked to the north by the broad Scurrival antiform, the axial zone of which is an area of low Laxfordian strain. Late-Scourian microdiorite intrusions and dykes of the Younger Basic suite are preserved in this area and show little subsequent deformation (Francis et al., 1971; Francis and Sibson, 1973; cf. Hopgood, 1964). Pyroxenebearing metamorphic assemblages are common in both the gneisses and the dykes, and positive gravity and aeromagnetic anomalies suggest that granulites occur at depth in and to the east of the antiformal core (Coward et al., 1970;

McQuillin and Watson, 1973). Similar high-grade rocks and relatively unchanged pre-Laxfordian relationships are also well seen above the major thrust, notably at Leanish, Earsary and Rubha Mor.

South Uist

On the north-east limb of the Scurrival antiform, gneisses showing a moderate degree of Laxfordian strain dip northeastward toward the South Uist synformal zone, which extends south-south-eastwards from Rubha Ardvule [NF 710 300] as a tract of steep, strongly foliated rocks. Both dL2 and dL3 folding contribute to define this zone (Coward, 1973a, b). The north-eastern limb of the synform extends to central Benbecula and is crenulated by the effects of late (dL4) folding on steep east–west axial planes (Coward, 1973b). In eastern South Uist, the Corodale gneiss massif above the Outer Hebrides Thrust displays a broad antiformal fold on a NNE-dipping axial plane. This fold, originally assigned to dL3 by Coward (1972), distorts a cataclastic foliation in the granulite-facies metadiorite and is now thought to postdate the formation of the Outer Hebrides Thrust zone (see (Table 27)).

Benbecula

In central Benbecula, mafic and metasedimentary gneisses outline an antiform-synform fold-pair on steep NNW-dipping axial planes which collectively have a Z-profile with north-north-westerly plunge. These relatively open folds may be of dL3 or dL4 age.

North Uist

Strongly developed mesoscopic dL3 folds of the north-west coast (Graham, 1970) around Tigharry and Hougharry collectively define a synform, the main hinge-zone of which can be traced south-east from Hoglan Bay [NF 710 730] towards Kirkibost [NF 750 650].

The corresponding broad dL3 antiform covers much of North Uist. Its north-eastern limb, an area of low to moderate northerly dips, somewhat distorted by later warping, extends to the Sound of Harris.

In western North Uist, forming the attenuated southwestern limb of this antiform, there is a subvertical zone trending about 150° which stretches from near Bayhead [NF 745 683] to the coast north of Tigharry [NF 710 728]. In this zone, banded amphibolites, metasediments and granite-gneiss lenses are all strongly foliated, commonly show boudinage features, and have a pervasive lineation which plunges to the NNW generally between 20° and 40° (ranges from 10° to 70°). This lineation corresponds to the dL3 fold axis orientation. It is also found south-east of the steep zone between Clachan [NF 810 640] and Bayhead, where the and Bayhead, where the regional fold pattern is defined by open to close, medium-scale dL3 folds. The dominant subvertical dL3 fabric in the steep zone has strongly modified an earlier dL2 fabric. However, minor dL3 folds within and adjacent to the steep belt have axial planes trending north–south and dipping at about 60° to the west. This suggests that the steep belt is controlled by a pre-existing lineament. This is also in accord with the abrupt termination of the steep belt at its SSE end.

Graham (1970) has suggested that the two major amphibolite bands in this belt can be matched and that the structure is a large scale subvertical isocline. He envisages it as a 'shoot-through' structure developed in a large-scale boudin neck in an open synformal hinge zone. Such features can locally be seen on a small scale along the north-west coastal section of North Uist. However, no evidence for a fold closure can be seen in the coastal section north of Tigharry, and although Graham (1970) presented evidence for dL3 minor-fold vergence change across the structure, the present study has failed to substantiate this observation. Graham (1970, pp.217–218) recorded strain values from the measurement of deformed pegmatite vein orientations in amphibolite from the coastal section north of Tigharry. X: Y:Z: values of 18:6:1 (k = 0.4) were obtained. If the veins cross-cut the earlier dL2 fabric as Graham (1970) suggested, then taking his value for dL2 strain for X: Y:Z to be 6:3:1 (k = 0.5) gives total finite strain between 108:18:1 (k = 0.29) and 54:36:1 (k = 0.014), dependent on the relative orientations of the two strains. This gives a value for displacement shortening normal to the foliation of 92 per cent, assuming that Z directions are coincident. Unless a considerable proportion of this deformation occurred by shearing (simple shear) there would have been unrealistically large changes in volumes along strike to the southeast, and it is difficult to see how these could have occurred.

In the north-western coastal tract of North Uist, amphibolitised metadolerites generally occur as concordant bands and boudin trains, but they also outline interference patterns between coaxial dL3 and dL2 folds. Graham (1970) concluded that the dL3 axial planes were roughly perpendicular to those of dL2 at the onset of dL3.

The small areas of low Laxfordian strain that occur in western South Uist, Benbecula and North Uist form roughly concordant lenticular massifs enclosed in gneisses showing higher degrees of Laxfordian strain. These rootless massifs are situated at high structural levels with respect to the Laxfordian dL3 fold-systems that plunge to the north-west. The low-strain massif in the core of the Scurrival antiform on Gighay, Hellisay and Fuday, on the other hand, lies at a low structural level and appears from geophysical evidence to be continuous with a tract of granulite-facies rocks which apparently underlie the easternmost part of Barra, South Uist and the adjacent part of the Minch. Coward et al. (1970) attributed the characteristic style of large dL3 folds (broad rounded antiforms contrasting with tight cuspate synforms) to the ductility contrast between such deep-seated granulites and wetter and more ductile grey gneisses at higher structural levels.

The Northern Isles

This domain excludes the southern part of South Harris that is occupied by metasedimentary and metavolcanic gneisses and by the South Harris Igneous Complex. The effects of Laxfordian ductile deformation are present throughout the domain (cf. Dearnley, 1962a), although they are slight in parts of central Lewis (Davies et al., 1975). Mesoscopic structures formed during dL2 have a very wide distribution, whereas those related to dL3 and subsequent phases are more restricted in occurrence. The style of dL3 structures shows regional variations related to the effects of Late-Laxfordian migmatite formation and granite veining. In South Harris, western North Harris and the Uig Hills, dL3 was a phase of ductile deformation, frequently associated with pegmatite development in the axial planes or thinned fold limbs (see Chapter 2: Lithology: Lewis and Harris: Laxfordian migmatisation and pegmatites); the coarse-grained textures of the host-rocks of the gneisses in these regions record post-kinematic recrystallisation with respect to dL3. Elsewhere, features assigned to dL3 include small shear-zones, mono-dines with thinned limbs, and folds disrupted by fractures. The development of these structures was associated with reductions of grain-size and/or cataclasis, and appears to have postdated the formation of coarse gneissose textures. If the correlation with dL3 is correct, it appears that conditions during this phase were less favourable to ductile deformation and recrystallisation in the Long Island (outside the migmatite area) than in the southern domain, perhaps reflecting lower crustal temperatures and/or lower fluid pressures. If, on the other hand, the semi-brittle structures belong to later Laxfordian phases, it would follow that effects of dL3 were of less importance in the north than in the southern islands.

Central South Harris and North Harris

Foliation trends appear to outline a pair of large, almost isoclinal folds with limb lengths of more than 10 km in South Harris (Myers, 1968, 1971) and a number of closed structures possibly representing smaller folds of similar type in North Harris. The axial planes of those folds strike northwestwards and dip consistently south-westwards at angles ranging from moderate over most of the area to low in the hills around Clisham. The axial plunge of associated mesoscopic structures is predominantly south to south-east but shows considerable variation in the axial plane. From the fact that metadolerites and other mafic and ultramafic minor intrusions of the Scourie dyke suite are folded on both a large and a small scale (Myers, 1970b; Soldin, 1978; Thamdrup, personal communication) it appears that the fold-set is Laxfordian. The large Horgabost antiform and Uaval synform predate the development of the granite-migmatite complex, and in North Harris minor folds of the same style are distorted by folds formed during the early stages of migmatisation (Myers, 1971). These relationships suggest that the folding can be correlated with dL2 of the southern domain. The superposed folds of the granite-migmatite complex ('main phase' of Myers, 1971–here equivalent to dL3) are small in scale and are represented outside the migmatites by asymmetrical mesoscopic folds in which the short steep limb is commonly thinned. Younger (dL4 dL5) monoclines on steep north-west-trending axial planes, and open warps on vertical NNE-trending axial planes, are common throughout Harris (Dearnley, 1963; Myers, 1971; Hopgood and Bowes, 1972).

Central Lewis

Over much of the area between Clisham (North Harris), Balallan, Loch Roag and Loch Resort, dips are low and can be seen in many places to be defined by the limbs of close folds on subhorizontal axial planes. These folds, seldom more than a hundred metres in wavelength, appear to correspond to the dL2 folds of Harris (in Great Bernera to the f1 structures of Watson, 1968). Irregularities of strike and dip result from the superposition of late open folds and minor shear zones of one or more generations. At the north-western side of the Uig Hills, the gneiss banding and the axial planes of dL2 curve sharply downward along the boundary of the steep belt of north-western Lewis (see below).

Although the effects of late (post-dL2) Laxfordian deformation are trivial in terms of bulk strain over much of central Lewis, they are of considerable importance in a southsouth-east-trending tract 8-10 km broad extending from Great Bernera to the lower part of Loch Seaforth. Immediately west of Little Loch Roag, the steep limbs of asymmetrical folds (dL3) define north-north-west-trending belts of vertical foliation in which the banding is smeared out and an intersection lineation plunges to the south-east. Chloritisation and cataclasis are widespread within the Loch Roag zone which appears to have acted as a long-standing zone of weakness. It has been etched out by glaciation and is traversed by several post-Laxfordian north-north-west-striking faults that displace Outer Hebrides Thrust features; it is also marked by linear negative aeromagnetic anomalies.

North-east Lewis

To the east of the Loch Roag steep zone the interior of northern Lewis is occupied by coarse-grained gneisses with a north-north-west or north-west strike and various dips. In the areas of low Laxfordian strain, where discordant metadolerites are well preserved (Davies et al., 1975), the dips are generally steep and Laxfordian deformation is recorded by open folds on subhorizontal axial planes. Elsewhere, tight folding on gently inclined axial planes leads to the development of low dips on the fold-limbs. Irregular strike patterns result from later open folding. The massifs characterised by low Laxfordian strain perhaps lie in the short limb of a large asymmetrical late-Scourian fold (Chapter 3: Late-Scourian structures), but are not located at structurally significant positions with respect to dL2. In the Ness region and in the northern part of the Eye peninsula, NW–SE zones of high strain are characterised by a strong foliation and by a relatively fine grain size. These zones of high strain mark shear zones developed in Late-Scourian (perhaps Inverian) or early-Laxfordian times. In the Ness area the fabric planes are folded by one or more (dL3 dL4) systems of folds and monoclines that appear to have been formed during phases of retrogression and grain-size reduction (Watson, 1969; Davies et al., 1975).

The steep belt of north-west Lewis

The region north-west of the Uig Hills and the coastal tract south-west of the Butt of Lewis are occupied by steeply dipping gneisses with a well-developed foliation striking north-east. The steep zone as a whole is oblique to the trend of gneissic banding in the interior of Lewis, and the sense of deflection in horizontal and vertical planes suggests that sinistral displacement across the zone was combined with a north-westward downthrow. The structural relationships of Younger Basic dykes in the area east of Great Bernera indicate that the zone had been formed prior to the emplacement of the swarm (Davies et al., 1975). However, few metadolerites retain primary features within the steep zone itself, and amphibolites that possibly represent metadolerites form concordant layers and boudin trains. The zone therefore also received a large increment of Laxfordian strain. Variations in profile of d„ folds suggest that near Gallan Head the steep belt is occupied by a tight synform of dL3 age.

The Sound of Harris and Southern Harris

The domain which lies between the southern islands and Lewis and Harris is characterised by an assemblage of rocks not seen on a comparable scale elsewhere, as well as by anomalous structural features. The detailed structural sequence in the south-western part of South Harris is described in Chapter 7 (Structure) and we concentrate here on the regional-scale structures. The metamorphosed plutonic rocks of the South Harris Igneous Complex together with the associated Langavat and Leverburgh metasedimentary belts occupy a well-defined NW–SE tract which is inferred from aeromagnetic anomalies to have a total length of some 40 km. This tract is about perpendicular to the 'run' of the gneissic banding through the Lax fordian fold systems of the adjacent domains and to the contours of the 1978 IGS Bouguer anomaly map that define lithological variations at depth.

In broad terms, the rocks of southern South Harris apparently define a synformal structure, for the Langavat and Leverburgh belts dip inwards beneath the main outcrop of the igneous complex. Analysis of aeromagnetic anomalies implies that metagabbroic rocks and associated granulites terminate downwards at depths of 5-7 km (Westbrook, 1974); such units may well be lenticular in low-level gneissic complexes. In the Sound of Harris, grey gneisses on the north-eastern limb of the Sound of Harris antiform appear to lie structurally below the metanorite of Ensay, while in central South Harris migmatised grey gneiss and granite sheets dip beneath the Langavat Belt. If these relationships hold at depth, then the entire assemblage apparently lies at a high structural level with respect to the adjacent grey gneisses (Dearnley and Dunning, 1968, fig.9; Coward et al., 1970, fig.1). Structural analysis of the domain, however, raises doubts concerning its relationships with the adjacent areas. Two aspects of this analysis should be mentioned.

The Roineabhal antiform defined by the Rodel anorthosite is a large early Laxfordian fold regarded by Dearnley (1963) as one of a set developed also in the Leverburgh and Langavat supracrustal belts. This structure folds differentiated stratiform layering notably in the basal units of the anorthosite, and the way-up criteria define it as a syncline facing downward with an eastward-dipping axial plane (Witty, 1975). The fold axis and the early linear elongation of mineral clots both plunge very steeply. Many early mesoscopic fold axes and linear elements also plunge at high angles in the Langavat and Leverburgh belts. Taking these relationships into account, it seems clear that the geometry of the internal fold-structures contradicts the inferred synformal shape of the bounding surface of the assemblage. The sigmoidal curvature of the early mineral fabric about subvertical axes in the diorite and norite, which is caused largely by marginal shearing and related development of very intense deformation (Graham, 1970; Palmer, 1971), may point to a possible answer to this apparent contradiction. Also relevant is the intense foliation in the Langavat and Berneray-Pabbay supracrustal belts that sheath the igneous complex, and the abundance of minor near-vertical shear-zones. These features all suggest that the conflicting geometries may be due largely to the rotation and disruption of competent units during the later stages of deformation; the intervening metasediments accommodated the resulting structures.

Much of the South Harris Igneous Complex displays metamorphic assemblages of granulite facies that commonly define early fabrics but are folded in large-scale structures such as the Roineabhal antiform. Studies of mineral assemblages and compositions have led several authors to conclude that at least one phase of granulite-facies metamorphism took place at pressures of 10-15 kbar (Livingstone, 1976b; Wood, 1975; Dickinson and Watson, 1976; Witty, 1975; Horsley, 1978).

This conclusion (which is supported by evidence of depletion in incompatible elements) implies that the rocks of southern South Harris were held at depths of 30-45 km during the development of at least their early structures. In contrast, the grey gneisses of the domains to north and south, though considered by Dearnley (1963) to have passed through a phase of granulite-facies metamorphism, do not show depletion in lithophile elements (see also Soldin, 1978). Isotopic studies also yield no evidence of very deep-seated metamorphism at the period cited by Dearnley (Moorbath et al., 1975). If this contrast in metamorphic histories is accepted, the granulite-facies massif of southern South Harris was derived from a deeper structural level than the gneisses.

The differences in derivation outlined above suggest that the massif comprising the South Harris Igneous Complex and the supracrustal gneisses associated with it may have been displaced relative to the grey gneisses now lying to north and south at some time between the phase of high-pressure metamorphism and the Late-Laxfordian period of regional (dL3) folding, retrogression and pegmatite injection. The massif could be interpreted either as an allochthonous sheet resting on and folded with the gneisses, or as a lenticular 'pip' squeezed up from a deeper level along a northwesterly-trending zone of weakness. Alternatively the relict granulite facies mineralogy in the Leverburgh metasediments may merely have been preserved by the intrusion of the massive anhydrous igneous bodies at about 2260 Ma (Cliff et al., 1983) with little change in level between the grey gneisses and metasediment/meta-igneous complex. The intrusive bodies themselves show convincing mineralogical and geochemical evidence for the presence of an early granulite facies metamorphic event. This 'hornfels' origin also creates problems, however, because it appears that the igneous rocks were intruded into lower crustal levels and the grey gneisses were affected only by amphibolite-facies metamorphism, though one would expect more intense metamorphism at such depths (see Chapter 7: Synthesis).

Chapter 15 The Outer Hebrides Thrust Zone

A major thrust zone trending north-north-east and dipping 20°-30° to the east crops out along the eastern seaboard of the Outer Hebrides. It has a mapped length of 172 km and continues offshore from Sandray at its southern limit and from Tolsta Bay at its northern limit. The distinctive rocks marking the main thrust zone are chiefly limited to a narrow zone at the eastern margin of the islands, except in Lewis where the zone's outcrop becomes more diffuse, extending over a width of 18 km in the Parc district. Measured values of the orientation of the thrust plane range from strike 024° and dip 25° to the east in Barra to strike 165° and dip 22° to the east in South Uist. Pseudotachylite, a dark glassy material formed when frictional heating on fault planes causes the rock to melt, is particularly well developed along the Outer Hebrides Thrust Zone. In addition to the main thrust zone there are subsidiary thrusts throughout the Hebrides, and pseudotachylite has been formed on these also. Such thrusts are particularly well seen around Greian Head in western Barra and along parts of the west coast of South Uist. In North Uist two minor thrusts marked by pseudotachylite breccia zones up to 15 m thick and with a strike length of 20 km have been mapped. In Lewis subvertical and subhorizontal zones of mylonitic granite and 'grey gneiss' occur locally. Much of northern and western Lewis shows evidence of 'cataclastic' and/or mylonitic deformation. For example, Dougal (1928) records black flinty crush at Cunndal Bay [NB 512 655] near the Butt of Lewis.

The term cataclastic is used strictly to describe minerals which are bent, broken and/or granulated due to severe mechanical stress during dynamic metamorphism. Hence it implies that brittle deformation dominates and mineralogical changes are minimal. Here 'cataclastic' is a field term for rocks, generally gneisses, which show considerable brittle deformation effects but where ductile processes also have undoubtedly been active.

The thrust zone lies mainly within the massive banded grey gneiss typical of the Outer Hebrides, but in South Uist it separates the pyroxene-bearing Corodale dioritic gneiss to the east from the predominantly amphibolite-facies 'grey gneiss' and amphibolite to the west. In Barra the thrust separates granulite- and amphibolitic-facies acid gneisses and meta-igneous gneisses to the south-east from amphibolite-facies gneisses to the north-west. It is possible that the thrusts seen in Barra are merely forethrusts of the main zone, which would then lie offshore to the east.

The thrust zone now forms a strong topographic feature, particularly in the Uists and Barra, where pseudotachylite and brecciated gneiss cap scarps up to 600 m high: e.g. Eaval, North Uist; Beinn Mhor and Hecla, South Uist; Heaval, Barra. Further north in the Sound of Harris and in South Harris the thrust also locally forms topographic features. In much of Lewis the thrust zone and its related structures form few notable topographic features, but in the Parc district (south-east Lewis) around Seaforth Head [NB 295 163] a thick pseudotachylite/breccia development forms a scarp up to 250 m high on Feirihisval [NB 301 147]. Beinn Mholach [NB 356 387] and adjacent hills in the centre of Lewis are also formed of thrust zone products.

The character of the thrust zone and its products changes along its length, and its exposed thickness increases to the north. It can be divided into three differing areas; the southern one between Sandray and South Uist; the central one between South Uist and Scalpay; and a northern one from Scalpay to North Lewis.

In Barra and the smaller isles to the south, a prominent thrust plane overlain by 25–75 m of pseudotachylite, cataclasite and breccia is seen. Pseudotachylite veins also occur extensively below the main exposed thrust in Barra. Further north in Eriskay and the Uists a much greater thickness of pseudotachylite and gneiss breccia lies above the basal thrust plane. Pseudotachylite is thickest in South Uist south of Lochboisdale [NF 840 145], where there is 50 m of black, brittle pseudotachylite and ultracataclasite breccia at the base of the Corodale Gneiss. In northern South Uist, where the Corodale Gneiss is absent, the thrust zone changes to a wider complex zone of retrograded pseudotachylite breccia, mylonite and broken gneiss. These lithologies form the prominent eastern dip slopes of Eaval and the Lees in North Uist, and extend northward to the Lochportain hills and South Harris. From Scalpay northwards the gneiss itself has a dominant 'cataclastic/protomylonitic fabric which is cut by the mylonite and pseudotachylite breccia zones. Prominent pseudotachylite defines thrust features on the hills around the head of Loch Seaforth.

Pseudotachylite is subordinate to mylonite and mylonitic gneiss in Lewis and locally is very difficult to distinguish from ultramylonite once greenschist-grade retrogression has taken place. Mylonites and crushed and retrograde gneisses outcrop on the island of Stuley, east of South Uist, and then to the north at Rubha Bolum [NF 829 283] and on the Usinish peninsula. They are abundant on the island of Ronay and in North Uist. A zone of low-grade hydrous retrogression is associated with the mylonites, and in South Harris such retrogression has affected gneisses lying west of the thrust zone. In the Parc district of Lewis mylonites form thick zones either discordant to the regional structure or sub-parallel to the gneissic banding. They become less abundant to the north, around Loch Erisort and Stornoway.

Previous work

MacCulloch (1819) first noted the existence of 'trap shotten' gneiss in the Outer Hebrides, and Dougal (1928) recorded that a large zone of crushed rocks extended along the eastern coast of Lewis. Between 1923 and 1934 T. J. Jehu and R. M. Craig published detailed field and petrographic descriptions of the rock types along the thrust zone and defined its extent. J. S. Flett, in Peach and Horne (1930), noted the effects of 'crushing' and mylonitisation in detailed petrological observations on gneisses from many parts of Lewis. He tabled degrees of alteration, mylonitisation and retrogression in the acid gneisses.

The thrust zone in South Uist was subsequently mapped in more detail by Coward (1969) and Kursten (1957). Francis (1969) also mapped the thrust zone on Barra and carried out reconnaissance work in the south-west part of South Uist. Bowes and Hopgood (1969a) noted the occurrence of pseudotachylite networks at Mingulay Bay and stated that the gneiss and pseudotachylite had been overprinted by their f2 (dL2) foliation. Our work and detailed observations from Barra by Hopgood (1971), however, show that almost all the pseudotachylites were formed after the Laxfordian deformation, granite intrusion and metamorphism. A possible exception is at Halaman Bay, Barra (Hopgood, 1971), where pseudotachylite is reported to occur parallel to f2(?dL2) fold axes. We examined the exposures cited by Hopgood but could not substantiate Hopgood's relationships. In South Harris early shear zones in granulite-facies rocks contain pseudotachylite (e.g. Witty, 1975); Laxfordian deformation within the meta-igneous bodies is manifested in a different way from in the adjacent metasediments.

Rock types

The distribution of the main rock types found within the thrust zone is shown on the lithological maps (Maps 1 and 2).

Pseudotachylite

Pseudotachylite is a black glassy rock that is formed when frictional heating is caused by seismic faulting along discrete planes in relatively dry rock. Most fault zones throughout the world lack pseudotachylites. This is probably because the normal range of temperature and pore-fluid concentration in fault zones (see Sibson, 1973; 1977a) gives rise to 'wet stick slip' under near-constant volume conditions. In such circumstances frictional heat would be efficiently removed from the fault zone and the temperature rise would be insufficient to cause the rock to melt, unless very large dilation accompanied the movements. In most natural field examples of pseudotachylite the glass is now devitrified, a natural consequence of its age and the pressures and temperatures under which it developed (see Marshall, 1961).

In the field pseudotachylite is typically seen as discordant black veins, which may form a network generally within a limited planar zone either above or below a discrete thrust surface. A good example from the north-west coast of Barra is shown in (Plate 15). For a more detailed description and discussion of the form of pseudotachylite veins in the Outer Hebrides the reader is referred to Sibson (1975). In outcrop, pseudotachylite is finely jointed (Plate 16), and when broken commonly forms small cubes. Coward (1969) has reported that on Ben Kenneth [NF 868 203) in South Uist, such fractures are folded into open warps, kink bands and chevron folds, suggesting that they predated the regional uplift of the area; it is at this stage that joints normally develop.

Where retrograded, pseudotachylite is altered from a black flinty material to a dark or pale grey ultra-fine-grained material resembling ultramylonite; the two rock types can only be distinguished with difficulty even in thin section. Pseudotachylite, however, generally forms networks of intrusive veins with fine-grained margins and may show a distinctive fracture pattern. Vein thickness varies and their orientation changes sharply. Rarely, relict textural features indicative of its glassy origin are to be seen in thin section, e.g. microlites, flow banding and fine-grained margins.

Sibson (1975, 1977b) analysed a pseudotachylite vein and its associated wall rocks (garnet-biotite bearing metasediments) from Grimsay. He found that the wall rocks are slightly the poorer in MgO and Na2O, but otherwise the two rocks are of much the same composition. However, the pseudotachylite contains 9 per cent modal quartz and 11 per cent modal plagioclase (An30) as porphyroclasts, so Sibson recalculated the groundmass composition (melt) to exclude the porphyroclasts, with the results reproduced in (Table 28).

Sibson (1975) relates this melt composition to that of basaltic andesite and considers it to be a product of partial melting. This is superficially plausible; but closer inspection shows that the CaO and K2O values are atypical of basaltic andesite and the melt actually has a rather unusual composition (its CIPW norm includes 7.7 per cent Q., 17.6 per cent hyand 4.7 per cent c), similar to members of the lamprophyre suite (N. M. S. Rock, personal communication).

In fact, the melt merely reflects the parent gneiss composition, although it is not a total melt.

Gneiss with marked 'cataclastic' fabric

Where the Outer Hebrides Thrust Zone crops out over a large area, particularly in south-east Lewis, wide zones of coherent, dominantly acid gneiss show a marked 'cataclastic' fabric of differing intensity. In the field this is seen as a flattening of quartz grains, and more rarely feldspar grains, generally subparallel to the gneissic banding. This feature is particularly marked in some pegmatitic pods. In thin section the biotite is typically recrystallised and the quartz and feldspar show mylonitic features. Most of the feldspar shows signs of cataclastic deformation such as fracturing and comminution, as does some of the hornblende and quartz.

On Eaval [NF 897 606] and at Seaforth Head [NB 301 164] excellent exposures show that locally veins of pseudotachylite cross-cut the earlier 'cataclastic' fabric, which is broken and disorientated. In areas where the 'cataclastic' effects are minimal the general aspects of the gneiss and its overall mineralogy resemble those to the west of the Outer Hebrides Thrust Zone. More rarely the gneiss has developed a dominant new mylonitic fabric and accompanying fissility, a feature seen around Loch nan Deaspoirt [NB 305 217]. These mylonites can be confused in the field with the later mylonites of lower-greenschist grade (see below).

'Mashed' gneiss'

Over much of the thrust zone, and particularly in North Uist, the major dislocations are marked by up to about 160 metres of what Jehu and Craig (1925) called 'mashed' gneiss. The thickest sections have been recorded east of Lochportain [NF 963 724] and on the western slopes of Maol na h'Ordaig [NF 838 148] on southern South Uist. 'Mashed' gneiss consists of a coherent, structurally isotropic microbreccia with abundant pseudotachylite veining, gneiss fragments and much cataclasite. Cataclasite is an aphanitic, structureless, cohesive cataclastic rock in which most of the fragments are smaller than 0.2 mm (see Higgins, 1971); it makes up less than about 30 per cent of the 'mashed' gneiss. 'Mashed' gneiss was formed by multiple fracturing and granulation at moderate crustal depths (i.e. less than 15 km). In South Uist its major development is in the Corodale Gneiss (dominantly metadiorite).

'Mashed' gneiss is generally affected by late-stage low-grade retrogression, particularly north of Benbecula. Rarely, relict pseudotachylite textures such as fine-grained vein margins or pseudomorphs after feldspar microlites (seen in thin section) can be found. Where the thrust zone cuts the South Harris anorthosite there is a zone of intensely brecciated and retrograded anorthosite with amphibolite pods. The large amphibolite bodies appear to have partially controlled where 'mashed' gneiss zones have developed, both on South Harris and further north in south-east Lewis (cf. mylonite). In some areas the breccia contains less cataclasite and has a coarsely 'broken' appearance. The term 'broken' gneiss is probably a better field term for such rocks.

Mylonite

Within the thrust zone mylonites developed later than the products described above, mainly in distinct planar zones 1 m to more than 250 m thick, e.g. at Rubha Bhrollum, south-east Lewis [NB 325 021]. They are fissile to flaggy rocks, fine-grained and generally colour-banded, derived from the local gneiss by intense shear-dominated deformation. They have a greenschist-facies mineralogy and typically range from pale grey-green to dark green, and rarely, in parts, to fawn and cream. Although many of the mylonites are more correctly termed 'phyllonites' (see Higgins, 1971) the more general term is used in this account. The mylonite bands either trend subparallel to the gneissic banding, which mainly dips moderately to the south-east, or dip gently to the east or east-south-east discordant to the pre-existing structures. The thinner mylonite bands are commonly located adjacent to large basic bodies that undoubtedly caused local stress concentrations at the time of mylonite development. Around Crogarry na Hoe [NF 974 724] in North Uist a foliation contiguous with that developed in adjacent mylonites is sporadically developed in the 'mashed' and 'broken' gneiss. Fissile mylonite bands from less than 1 m to several metres thick occur at intervals. The transition from country rock to these fissile mylonites locally occurs over about 1 m. The mylonite fabric cuts across all earlier structures in the brecciated gneiss.

Mylonites are best exposed at several localities along the eastern seaboard of the Uists, such as at Usinish Bay [NF 853 333], Rubha Bolum [NF 828 284], Stuley Island [NF 830 234] and Eigneig Bheag [NF 921 602], and also extensively in the Parc district of south-east Lewis. In some areas they have been folded into open to tight minor 'chevron' folds. This folding was locally so intense that the related axial-plane crenulation cleavage or spaced cleavage became the dominant fabric (e.g. in the Rubha Bhrollum to Gob Rubh 'Uisenis area, south-east Lewis).

Spatially associated with these later mylonites is a zone of low-grade alteration. In it feldspar has been sericitised, hornblende partly altered to chlorite, epidote and more rarely actinolite, biotite altered to chlorite, and ilmenite to sphene. The quartz is strained and in parts has been recrystallised to an elongate mosaic; the microcline is strongly perthitic (Myers, 1971). In south-east Scalpay the gneiss becomes more deformed and retrograded towards the south-east over a cross-strike distance of about 1.4 km. At the south-east coast the gneissic banding is highly attenuated but the original features of the gneiss, though strongly modified, can still be recognised. The fissile mylonite bands are present in parts of the sequence. A detailed map and account of the mylonite features of Scalpay is given by Sibson (1977b), and we have made use of it in our study.

In South Uist the only rocks lying within the zone of alteration are the mylonites, but further north the zone transgresses westward relative to the thrust front and in North Uist the whole exposed thrust zone lies within it. In South Harris pervasive retrograde effects extend north-west of the main thrust for up to 8 km. Epidote developed extensively in joints within the gneisses around Tarbert and in south-eastern South Harris.

Pseudotachylites

Petrography

In 1923 Jehu and Craig published detailed petrographic descriptions and illustrations (pp.435–436) of rocks containing pseudotachylite from Barra and Mingulay. They noted the changes from hornblende acid gneiss to network-fractured and locally cataclastic gneiss, and to brown material of a vitreous nature. This brown material (pseudotachylite) contained comminuted fragments of the original minerals (from 0.01 mm to about 5 mm) and penetrated the surrounding gneiss. There was 'incipient granulitisation' in the included quartz and feldspar and in gneiss adjacent to the pseudotachylite. The quartz porphyroclasts commonly contained numerous fine cracks delineated by magnetite dust. In related rocks Jehu and Craig also described a brown ultramylonite material showing flow textures, and they recognised vitreous material. Several thin sections showed dark brown hornblende (?) and feldspar microlites within the vitreous material, and Jehu and Craig took these features, together with rounded and corroded quartz, to imply partial fusion of the gneiss.

Sibson (1975, 1977b) has also illustrated and described the form and petrography of pseudotachylite veins from various parts of the Outer Hebrides. He reported the occurrence of relict glass within pseudotachylite in the southern part of the Outer Hebrides. In the following section we give only brief descriptions of the features described in detail by Sibson; the emphasis is placed on the problems of interpretation. (Plate 17) illustrates some of the microscopic textures seen in thin sections from the southern islands.

Pseudotachylite is an isotropic glass containing finely divided opaques which in thin section cannot be resolved. (Plate 17)e shows a typical hornblende acid gneiss with patches and stringers of black to brown pseudotachylite. Although radiating feldspar microlites are present, they are abundant only in the slightly coarser now-devitrified zones in the central parts of the veins. Quartz and plagioclase inclusions lie within the pseudotachylite. In some thin sections one sees that the occur rence of pseudotachylite is related to hornblende-rich zones in the acid gneiss.

Folds of flow banding within planar pseudotachylite veins are apparent in thin sections, particularly ones from South Uist, Barra and the smaller southern islands. Rarely, for example (S63008) (North Uist), a single fold structure is confined to the pseudotachylite vein, implying that dilation accompanied movement, as a result either of stress release or of movement across an irregular slip plane or both. In several specimens the margins of pseudotachylite veins are folded whereas the central parts are not (for example (S61310), West Ronay [NF 9042 5647]), again suggesting that movement occurred while the cooling melt-phase remained ductile. In slide (S58137), a hornblende acid gneiss from Sandray, dark to pale brown fine-flow-banded pseudotachylite encloses subrounded fragments of brown pseudotachylite up to 6 mm wide ((Plate 17)c).

(S58036) from Barra shows a similar feature; angular and broken fine-grained pseudotachylite fragments lie within coarser devitrified slightly retrogressed pseudotachylite. The gneiss adjacent to the pseudotachylite consists of highly comminuted green hornblende, quartz and feldspar that is generally finely recrystallised, and highly kinked and/or finely aggregated biotite. In other specimens from North Uist, from sites where small fractures extend from pseudotachylite bodies, some of the hornblende has locally retrogressed from a green to a blue-green form (S63008) or from a brown to a green form (S62215). We presume that this retrogression resulted from the movement of water or another fluid accompanying or subsequent to pseudotachylite formation.

The margins to thicker lenses and veins of pseudotachylite are fine grained, generally isotropic and dark to pale brown. These characteristics imply a pseudo-igneous origin for the material (for example (MC29633), [NF 784 287]; (S61827)). Even where the pseudotachylite is retrograded, for example in south-east South Harris at [NG 0614 8405], this feature is easily recognised. (Plate 17)d shows an example taken from northeast Wiay, in which both dark fine-grained margins and microlite texture can both be seen despite the high degree of retrogression.

It seems that when pseudotachylite veins are formed they are heated and cooled very rapidly (see McKenzie and Brune, 1972), of the order of a few seconds for heating and probably less than a minute for the cooling 'half-life'. The cooling half-life is the time taken for the temperature difference between the centre of the vein and the country rock to fall to half its peak value. Sibson (1975) has discussed the generation and cooling times of pseudotachylite in detail.

Several thin sections (e.g. (Plate 17)a), particularly those from the Uists and Barra, contained feldspar microlites ((S61361), (S63114) and MC 29633). These range from 0.02 mm up to 0.3 mm long, the smaller laths generally occurring in the thinner veins. Some of the microlites form stellate clusters but only a few show a flow alignment. Microlites are typically to be seen in the thicker pseudotachylite veins or pods (more than 1 cm wide), and Sibson (1975) concludes that cooling 'half-lives' of minutes are needed for their development. He reports that their composition commonly lies in the range An30-35 and some of them show nucleation on plagioclase porphyroclasts (e.g. see Sibson, 1975, fig.4(a)). In thin section (S60123) from north-east Wiay a thick body of pseudotachylite lies within highly retrogressed hornblende acid gneiss in which albite, epidote, sericite and chlorite are widely developed. The retrogressed pseudotachylite contains prominent pseudomorphed stellate clusters of feldspar microlites individually up to 0.2 mm long ((Plate 17)b).

Quartz and feldspar porphyroclasts are typically found within pseudotachylite which is derived from acid gneiss, and form up to about 30 per cent of the total volume. The quartz inclusions are usually angular and some show corroded margins (see also Jehu and Craig, 1925, pl. III, fig.3). The inclusions generally show abundant strain shadows and some contain an irregular network of fine cracks in part defined by opaque dust, possibly magnetite. Rarely quartz contains microscopic cataclastic zones, differential movement being taken up by recrystallisation to a fine-grained quartz aggregate (e.g. (S58164), Castlebay, Barra). In retrograded specimens the quartz is totally recrystallised to a fine-grained aggregate of grains 0.01 to 0.02 mm in diameter. The margins of some quartz porphyroclasts (e.g. (S58137) from Sandray) in non-retrograde pseudotachylite are recrystallised. In slide MC 31995 from Ben Kenneth, South Uist, the quartz is totally recrystallised in the thicker veins and lenses of pseudotachylite. More rarely, structures resembling deformation lamellae are present in larger unrecrystallised quartz porphyroclasts ((MC31482), South Uist).

Quartz porphyroclasts are not common in mylonites and ultramylonites in which feldspars form the main relict component. In pseudotachylite, however, quartz and feldspar porphyroclasts are present on average in about equal proportions. In individual specimens one mineral may dominate, for example, (S60123) from north-east Wiay, in which most of the porphyroclasts are quartz. Rarely in pseudotachylite small porphyoclasts are flattened and orientated in a flow foliation (Sibson, 1975, p.780).

Feldspar porphyroclasts in pseudotachylite are rounded to angular, commonly with embayed and rather blurred outlines. They are chiefly plagioclase, but some potash-feldspar inclusions have been noted in pseudotachylite derived from granitoid parent rocks. Irregular twinning, strain shadows and internal fracturing are particularly abundant, and in some specimens feldspar porphyroclasts are marginally recrystallised to an elongate 'wavy-cloudy' aggregate (for example (MC31995), Ben Kenneth, South Uist), suggestive of local melting (see Jehu and Craig, 1923, p.435).

Hornblende and biotite are the first minerals to be absorbed into the pseudotachylite. Pseudotachylite veins commonly cross-cut quartz and feldspar crystals, but biotite, and more rarely hornblende, tend to be ragged adjacent to the vein with embayments of pseudotachylite penetrating the crystal. In some rocks when pseudotachylite is generated within amphibolite relict ovoid hornblendes remain, e.g. at Loch Sgibacleit [NB 3050 1644]. Dougal (1928) figures photomicrographs of stellate clusters of hornblende and feldspar laths from pseudotachylite south of Bayble on the Eye Peninsula in Lewis. The more normal mode of hornblende is described in detail by Jehu and Craig (1923, pp.434–435) from specimens from Barra. Hornblendes are preferentially sheared and drawn out into shreds, and the cleavage is obliterated. Finally the mineral becomes ob scured by black dust (magnetite and some ilmenite–see Sib-son, 1975), and transformed to an opaque isotropic brown mass. Biotite is similarly transformed, but more rapidly than hornblende. The final product is again isotropic brown material with abundant finely disseminated opaques, magnetite being dominant.

Although many rocks were collected as pseudotachylite in the field, where they display the requisite macroscopic features, in thin section they are seen to consist of a microbreccia of pseudotachylite veins and intermixed fine-grained cataclasite ((Plate 17)f). Several generations of pseudotachylite can usually be distinguished. Coward (1969) notes 7 m of apparently pure pseudotachylite on the main basal thrust plane at Ben Kenneth [NF 805 194], but Sibson (personal communication) reports that it is a complex mixture of pseudotachylite, ultracataclasite and ultramylonite. Near Ru Melvick at [NF 823 136], where about 17 m of black 'pseudotachylite' are apparently present, thin section (S63011) observations show that this is in fact a cataclasite-pseudotachylite mixture. Theoretical considerations (McKenzie and Brune, 1972) strongly suggest that pseudotachylite generated on a failure plane during seismic movements has a maximum thickness of about 1 cm. Any thicker bodies must result either from multiple movements localised along a narrow zone, or from melt migration along or away from the movement plane.

Many of the porphyroclasts in pseudotachylites are shattered, and this led Sibson (1975) to suggest that the sudden heating at the time of formation would lead to a large pressure rise in 'trapped' intracrystalline water, such as fluid inclusions, causing explosive hydraulic fracturing. He calculated that the pressure rise ÄP is related to temperature T (°C) by the expression

ΔP ≈ 100T/47 MPa100 megapascals (MPa) = 1 kilobar (kbar)

Observations made during the present survey suggest that the conditions at the time of pseudotachylite formation were not those specified by Sibson (140°C and 130 MPa); however, hydraulic fracturing effects even at great crustal depths would still be significant. Using the tables of Burnham et al. (1969) we have modified the above equation to

ΔP ≈ 100T/79 MPa for initial conditions of 540°C and 360 MPa or 3.6 kbar100 megapascals (MPa) = 1 kilobar (kbar) (see below: Pressure–temperature conditions of pseudotachylite formation).

Mode of occurrence and origin

In the field pseudotachylites occur in two ways:

1 Along near-planar thrust zones, generally dipping 20-30° to the east and up to several hundred metres thick. In such thrust zones there are individual bands of pseudotachylite–cataclasite that are up to 20 m thick. In South Uist thrust movement has been largely localised at the base of the Corodale Gneiss, giving a pseudotachylite-veined cataclasite zone in places up to 40 m thick. In south-east Lewis, however, such pseudotachylite–cataclasite zones are only a few metres thick and form a network over a large area in more uniform acid gneiss.

2 In otherwise unaltered banded gneisses throughout the Outer Hebrides as random networks, locally associated with near-planar 'generation surfaces' along which small-scale dominantly thrust movement has taken place.

Corresponding to the two modes of occurrence, we propose differing modes of pseudotachylite formation. We envisage that the major thrust zones were formed by repeated seismic and ductile movements along the same roughly planar zone over a considerable period of time, resulting in a cataclasite-pseudotachylite breccia on all scales. Once the gneiss begins to be brecciated and its grain size reduced, the zone would act as a locus for further movements. Jackson and Dunn (1974) found that during sliding friction experiments on pre-existing gouge-filled fractures, flakes of welded glassy material formed although displacements were very small (0.5 to 0.7 cm). Wispy pseudotachylite formed at the same time as cataclasite producing an experimental texture not dissimilar from that observed in rocks from the Outer Hebrides Thrust Zone.

The second mode of pseudotachylite formation is as a result of a single seismic-failure event in relatively dry acid gneiss and locally in metabasic rocks. Sibson (1975, 1977b,c) has discussed this process with reference to theoretical considerations and to observations made along parts of the Outer Hebrides Thrust Zone. Once formed, the pseudotachylite is stiffer than the surrounding gneiss, and experimental evidence (Friedman et al., 1974) shows that further movement will then tend to occur adjacent to the original zone. Hence wide zones of pseudotachylite-veined gneiss have developed by repeated minor thrusting, e.g. in the Rueval-Balivanich area, Benbecula, and the Greian Head area, Barra, until the quantity of pseudotachylite renders the whole zone stiff, and seismic movement becomes localised elsewhere. It is possible that in some instances the development of pseudotachylite in this way was a percursor to the development of a major thrust zone in which rock types of different character lay adjacent to one another, such as the Corodale Gneiss and the acid hornblendic gneiss in South Uist.

The two proposed modes of pseudotachylite formation are not mutually exclusive, although it is unlikely that mode 1 would precede mode 2. Sibson (1977c, 1980) has discussed the idea of seismic slip occurring within mylonite zones in the Outer Hebrides.

Pressure–temperature conditions of pseudotachylite formation

Pseudotachylite has been reported from many localities throughout the world (see Higgins (1971) for summary). It is most typically developed in compact, relatively dry, high-grade metamorphic or intrusive igneous rocks; uniform high-grade quartz-rich gneisses are particularly favourable. It forms preferentially in areas of strong uplift, either during or subsequent to thrusting.

In the Seaforth Head area [NB 302 160] of south-east Lewis, fold structures geometrically related to the pseudotachylite veins show no consistent vergence or axial orien tation. Pseudotachylite and ultramylonite veins, developed initially subparallel to the gneiss/'cataclastic' banding, have been folded into open to close minor folds. Also to be seen are pseudotachylite veins across which small normal fault (extensional) movements have occurred. These features are compatible with pseudotachylite generation over a considerable period of time during regional thrusting and uplift. The pre-existing gneiss/'cataclastic' banding has been coarsely brecciated and rotated, apparently randomly, on the west face of Eaval near the summit [NF 897 606]. A similar sequence of 'cataclastic' gneiss–high-grade mylonite–pseudotachylite formation also can be determined in south-east Lewis. These regionally coherent structural patterns imply a complex thrusting history with initial ductile deformation followed by intermittent high-stress seismic failure.

In the 'cataclastic' hornblende acid gneisses from the basal part of the thrust zone at Seaforth Head, small new euhedral and ragged garnets have grown within and across highly comminuted biotites ((S56405), [NB 315 158]) in a strongly mylonitic gneiss. These biotites in part define the dominant mylonitic foliation in these particular rocks. They have changed from large foxy-brown plates to fine-grained aggregates of variably orientated grains. Plagioclase and hornblende grains have also been variously fractured and their size reduced. Quartz is finely recrystallised and granular epidote has developed widely. This phase of cataclastic-mylonitic deformation largely predated pseudotachylite generation; it marks the start of the deformation process which resulted in the formation of the Outer Hebrides Thrust Zone. The presence of garnet is taken to reflect conditions immediately prior to pseudotachylite formation. Garnet formation would have been favoured by the presence of fluid and high rates of ductile strain, and these in turn would have promoted diffusion. Garnet is less likely to form in conditions of falling differential pressure and temperature, such as would follow thrusting.

There is a thin ultramylonite/pseudotachylite zone at Aird an Troim [NB 233 165], south of Aribruach (South Lewis) and some 8 km west of Seaforth Head. In the field the rock shows fine-scale orthogonal jointing and the zone is discordant to the gneissic banding. In thin section (S62000) it consists of roughly circular to elliptic feldspar (oligoclase), hornblende and subsidiary quartz porphyroclasts in an extremely fine-grained matrix dominantly of biotite and quartz. The quartz porphyroclasts are partly or completely recrystallised. Small (0.001 to 0.006 mm across) euhedral garnets, generally zoned, are particularly abundant in the matrix. The zoning is defined by different abundances of inclusions, but electron-probe studies show that there is no significant compositional variation.

The rock was probably initially a pseudotachylite/ultramylonite mixture similar to those sectioned from Grimersta (S60010) and Seaforth Head (S56401). Subsequently, the matrix was altered (perhaps retrograded) to the observed biotite-quartz mixture whose fine grain-size favoured growth of new garnet. We suggest that after the thrusting that formed the pseudotachylite/ultramylonite sufficient water was available to permit recrystallisation of the extremely fine-grained matrix material; garnet thus formed at this time, and so its the prevailing compositions should reflect pressures and temperatures immediately following pseudotachylite formation.

The modified Cambridge Microscan Electron Probe Unit at the Geology Department of University of Edinburgh was used to obtain chemical analyses of garnets, feldspars and biotites from the two specimens (S56405) and (S62000). The analyses were carried out by means of the relatively rapid Energy Dispersive System (EDS) with a cobalt standard (20 kV, beam current 30A, 3 minute count time) to an accuracy of about ± 0.5% . Representative analyses are listed in (Table 29).

The analyses of garnet from the two specimens (17 in total) are all very similar. Spessartine content averages 4.25 per cent in (S56405) and 7.07 per cent in (S62000). Garnets preferentially concentrate MnO relative to the bulk rock composition and the degree of concentration of MnO in garnet is inversely related to the grade of metamorphism, so we may obtain some idea of metamorphic grade by comparing average MnO contents of the garnets with the bulk-rock values. Atherton (1977) has shown that the ratio of MnO in garnet to MnO in the matrix ranges from 81 to 24 with increasing grade in garnet-zone Dalradian meta-sediments (CaO-poor pelites). In both specimens here the MnO content of the bulk rock is not available and the MnO content of the biotites is too low to be recorded. However, we have used the average MnO content of 20 analysed hornblendic acid gneisses from North Harris (Soldin, 1978), rocks very similar to those found around Seaforth Head, to find approximate values for the MnO concentration ratio. Soldin's gneisses have an average MnO value of 0.05 per cent, giving mean concentration ratios of 39 (σ = 22.1) for (S56405) and 63 (σ = 22.4) for (S62000). Even after allowing for increased variance by taking differing values for bulk per cent MnO the results suggest that the garnet grew in the lower amphibolite facies. The high concentration ratio for specimen (S62000) reflects a relatively high MnO content; perhaps this is original and the rock's bulk content of MnO is also relatively high, or perhaps it reflects a lower grade of metamorphism.

In a similar way the concentration ratio for CaO in garnet to CaO in matrix should decrease with increasing grade of metamorphism. Atherton (1977) showed that ratios of 23 to 6 were characteristic of Dalradian metapelites in the garnet zone. An average value of 4% for the matrix (Soldin, 1978) gives concentration ratios for CaO of 2.2 for (S56405) and 1.96 for (S62000). These values lie outside the range for CaOpoor Dalradian pelites and presumably reflect the presence of hornblende and calcic feldspar in the gneisses from North Harris. It is not possible in these rocks to use the CaO content as an indicator of metamorphic grade.

Several authors (Thompson, 1976; Goldman and Albee, 1977; Ferry and Spear, 1978) have used the partition of Fe and Mg between coexisting garnet and biotite to determine the specific temperatures of mineral formation. The partition coefficient (KD)is not sensitive to pressure variations.

Unfortunately, the biotite in specimen (S62000) is too fine grained for accurate probe analysis and only (S56405) could be used for garnet and biotite (2 biotite analyses) determination. This specimen is ideal for garnet-biotite geothermometry as we can be confident that the two minerals are coeval and in equilibrium. Analyses of four adjacent garnets and two biotites from the fine-grained aggregate in which the garnets grew were used to determine the temperature of garnet-biotite growth in (S56405). The following values were obtained using the differing calibrations: Thompson (1976), 535°C; Ferry and Spear (1978), 545°C; Goldman and Albee (1977), 505°C. The probable uncertainty in the values is about ± 50°C. Goldman and Albee's method relies on the determination of an additional constant, generally calculated from 18O/16O ratios in quartz and magnetite. However, this constant can be determined from the value of the partition coefficient from Ferry and Spear's calibration, and the full calculation completed. Further iteration can then be used to refine the value. We take the higher values (545 and 535°C, ± 50°C) to be representative of the temperature of garnet formation, for these are in accord with the observed assemblages.

Electron-probe determination of plagioclase compositions show that there was a progressive decrease in An-content as the larger porphyroclasts broke down to fine-grained aggregates in specimen (S56405). Three sites were probed where porphyroclast and adjacent fine-grained aggregate could be analysed. The results are: An29.6: An29.5 An30.5 An24.8 and An29.7: An26.5 for porphyroclast: aggregate compositions. The decrease in An-content is greatest, however, in the fine-grained ribbon aggregates interbanded with quartz in the matrix of the highly mylonitic[cataclastic' gneiss. A single analysis gives An187. Winkler (1974, p.75), in discussion of the transition from greenschist to amphibolite grade, cites the appearance of plagioclase of An„ coexisting with hornblende as a criterion for the base of the amphibolite facies. The calcium released as a result of these changes is largely taken up by the formation of abundant granular epidote.

Hence all the mineralogical data cited here are compatible with the formation of mylonite and pseudotachylite under lower amphibolite-facies temperatures (around 540°C). The overburden pressure is more difficult to estimate. Shear heating cannot be significant if garnet crystallised as the mylonitic fabric was being formed, and if oligoclase/ andesine has recrystallised to middle/upper oligoclase. If we assume a reasonable normal geothermal gradient of 40° C/km and no extraneous heating such as shear heating, then a temperature of 540° corresponds to a depth of about 13.5 km, equivalent to a lithostatic pressure of about 360 MPa (3.6 kbars).

Stress levels: theoretical considerations

Pseudotachylite forms in a high-stress environment, commonly near tectonic-plate margins associated with regional thrusting, but also associated with major zones of uplift. For example, the Ivrea Zone of north Italy, which is generally regarded as an upturned 'mantle' slab, contains abundant but rather randomly orientated pseudotachylite veins. The major thrusting and uplift probably took place late in the Hercynian orogeny (Koppel, 1974). In the Silvretta Nappe of the Central Alps, Masch (1973) has noted pseudotachylite in conjugate shear joints within paragneiss. One of us (Mendum) has also recorded pseudotachylite in joints in the Adamello tonalite of north Italy. It seems that pseudotachylite generation related to uplift apparently results in irregular networks and abundant subvertical veins whereas that related to specific thrust or fault zones generally yields pseudotachylite that lies in narrow regionally planar zones.

If we take the Mohr-Coulomb criteria as adequate to describe the initial brittle failure of intact rock, then in terms of the principal stresses σ1, σ2, σ3.

σ1– σ3 = σ0 + (k–1)σ3 (after Sibson, 1977b),

where σ0 is the uniaxial compressive strength of the rock and k = (1 + sin Ö)/(1–sin Ö): Öis the angle of internal friction (arctan µ)

If µ = 0.7 (Byerlee, 1968), then σ 1 =  σ 0 + 3.7σ3

In a thrusting environment, σ3 is vertical and taken as lithostatic. This makes σ1 large even at crustal depths as low as 5 km, such as Sibson (1975) proposed. If we take 140 MPa as a typical uniaxial compressive strength for a coherent acid gneiss, then at 5 km depth

σ1 = 140 + (3.7 x 135) = 640 MPa.

Byerlee and Summers (1975) found from experiments on pre-cut surfaces of Westerly Granite that significant stable sliding preceded stick-slip movements at confining pressures of less than 300 MPa. Above 300 MPa such sliding was negligible. The results were independent of strain rate within the range tested (10-4 to 10-6/s). They also found that energy release at confining pressures of less than 150 MPa was slow, which is most unlikely in a zone of pseudotachylite formation. The mineralogical determinations described above from rocks at Seaforth Head imply that a more realistic depth would be about 13.5 km. If so,

σ 1 = 140 + (3.7 x 360) = 1470 Mpa.

This implies an overpressure immediately preceding pseudotachylite formation of 1110 MPa, which is improbably high. If other values for the coefficient of internal friction (µ) are taken, higher values than 0.5 exacerbate the problem, and even the lowest plausible value, µ = 0.5, gives

σ1 = σ0 + 3.6 σ3 = 1440 MPa

Brace and Kohlstedt (1980) have revised Byerlee's law for values of a3 greater than 1100 MPa. The revised equation is

σ1 = 3.1σ3 + 210 MPa.

On this basis a1 here is reduced to 1330 MPa. These values of σ1 are extremely high and imply overpressures of about 1000 MPa.

Using the Mohr envelope we can determine normal and shear stress across a plane dipping at 25° from the horizontal, a likely value for thrust plane formation.

Although overpressures of more than 3σ3 are not impossible, it is difficult to see how they can have occurred repeatedly over a period of thrusting and uplift, for each major seismic movement relieves the stress. R. H. Sibson (personal communication) has also pointed out that if shear stresses (Τf) are as high as 425 MPa, then shear heating must be a significant factor. The resultant temperature rise (Δθ,°C) can be related to time-averaged slip rate (v,cm/ annum), the small time interval during which heating occurs (Δt, s), shear stress (Τf MPa), and total displacement (d, m).

If Δθ = 2.82 x 10-10 Τf v √Δt = 2.82 x 10-10Τf √dv(Sibson, 1977a), then for Τf = 425 MPa and 370 MPa the following temperature rises result:

Time averaged slip rate v Τf = 425MpaTotal displacement d Τf = 370MpaTotal displacement d
1 km 10 km 1 km 10 km
1 cm/annum 169°C 219°C 152°C 190°C
10 cdm/annum 219°C 692°C 190°C 602°C

However, White and others (1980) dismiss shear heating as a significant contribution to the heat budget in their review of mylonites, except possibly in the very earliest stages of shear zone initiation. Moreover, the garnet-biotite palaeotemperatures reported here refer to mylonite which formed early in the history of the thrust zone and to post-thrusting metamorphism. Sibson (personal communication) suggests that this occurrence of garnets in the recrystallised 'pseudotachylite' (S62000) might result from the thermal aftermath of the fault. However, the abundant development of fine-grained biotite and garnet implies the presence of a fluid phase. The garnets in this rock and the mylonite at Seaforth Head are similar in composition, and this suggests fairly consistent pressures and temperatures, prevailed across this limited area, at about the time of the thrusting.

In order to explain the apparent overpressure needed for thrust and pseudotachylite we may look to local stress concentration factors. Inhomogeneities in the gneiss complex, such as basic pods, may cause such stress concentrations, and Sibson (1975) has suggested that these may have localised thrusting. Although this has undoubtedly happened, for pseudotachylite is commonly found adjacent to metabasic pods, there are numerous localities where no basic pods are present but pseudotachylite is abundant. Conversely, in many places metabasic pods are abundant, but have not provided a locus for pseudotachylite formation.

Sibson (personal communication) also suggests that a thrust may have been initiated as a result of stress-tip concentration in a manner analogous to the 'Griffith crack' theory (see Price, 1966, p.29). The validity of this theory is generally accepted on an experimental scale but it is difficult to see how it can apply to a complex regional thrust zone developed by multiple movements in heterogeneous gneiss.

It is known that when rocks are compressed experimentally beyond their yield strength they expand on relief of stress to greater than their original size (Price, 1966). Different modes of stress relaxation in nature are also documented: spalling of massive gneiss from the sides of Norwegian fjords, stress-relief fracturing (which is perhaps how joints develop), explosive spalling and minor faulting in deep mines and during the construction of underground tunnels and power stations, and stress drops following earthquakes (see Hanks, 1977; McGarr et al., 1979).

The Lewisian rocks, which were mainly stiff and relatively dry acid gneisses, remained at moderate to deep crustal levels until the late Laxfordian, when, subsequent to granite intrusion in South Lewis and Harris, they were uplifted regionally and thrust to the west. In this situation the amount of stress relief immediately following faulting would have been high. It is possible that the uplift would have modified the stress pattern at the time of pseudotachylite formation, considerably reducing σ3 and hence the amount of overpressure. It is not possible to quantify this effect but it remains a feasible method to at least partly explain the stress problem.

Sibson (1977b, p.195) puts forward an alternative hypothesis that mylonite was developed at depth in the main thrust zone and pseudotachylite at higher levels. He suggests that pseudotachylite development was initiated at shallower crustal levels and propagated downwards. Sibson (personal communication) states that an earthquake of magnitude 6 has a seismic moment of about 1025 dyne cm (1018N m) and a probable fault plane area of more than 10 km2. However he also notes (1977b, p.166; 1977c) that the thrust features in the Seaforth Head area are typical of the brittle/ductile transition zone, which typically lies in the 10–15 km depth range. This is in accord with the conclusions presented here and with Sibson's own work on the depth of earthquake foci in California (personal communication). This shows that shear resistance peaks sharply around the brittle/ductile transition (seismic-aseismic zone), the depth of which broadly relates to the rate of heat flow. For small earthquakes (magnitude 3 or less), depths of thrust-fault foci range from 10 km in high heat-flow areas to 30 km in low heat-flow areas. Larger earthquakes have even deeper foci.

In conclusion we suggest that thrusting and pseudotachylite development were a result of seismic movement along gentle easterly-dipping thrust faults at midcrustal levels during uplift of the Lewisian craton, following intrusion of the Laxfordian granites and pegmatites. The consequent stress release served to effectively reduce the lithostatic pressure making dry, friction-dominated thrusting characteristic of the Outer Hebrides Thrust Zone.

Mylonites

Mylonites are abundantly developed in parts of the Outer Hebrides Thrust Zone. Their formation was accompanied by lower greenschist grade metamorphism and they are distinct from the earlier-formed 'cataclastic' gneiss, which when strongly developed is mylonitic in appearance (p.138). The latter lithology is only rarely seen; it lacks the fissility of the later main mylonites and has higher-grade mineralogy. It is commonly developed in association with pseudotachylite/ cataclasite-rich zones, for example in the Balallan area of south-east Lewis [NB 305 219]. The low-grade retrogression that accompanies the development of the later main mylonites has in some areas affected non-mylonitic rocks lying a considerable distance west of the thrust front, for example in the north-eastern part of South Harris.

Distribution

In the southern islands mylonites are restricted to parts of the eastern slopes of South Uist and to Stuley Island. They are mainly developed in the acid gneisses that structurally overlie the massive granulite-facies Corodale Gneiss. Locally at its eastern margin, this orthogneiss itself is strongly mylonitic. Kursten (1957) has recorded 35 m of mylonite on Stuley Island [NF 830 234] and it is likely that the offshore linear depression east of Lochboisdale is a submarine continuation of this mylonite zone (Binns et al., 1974). The Usinish mylonites further north are 20 m to 60 m thick (Coward, 1969), and at Rubha Rossel [NF 859 366] and Ornish they dip at approximately 70° to the east.

Mylonite bands are more widely developed northwards of Ronay and North Uist, where individual bands are up to about 30 m thick. They are lenticular and anastomose on several scales. Sibson (1977b) has mapped the exposed parts of the thrust zone in North Uist in considerable detail. He noted that mylonites south of Lochmaddy strike north-east to NNE and dip 15° to 50° to the south-east, whereas north of Lochmaddy zones commonly strike north-east and northwest and dip 15° to 50° to the north-west and south-west respectively. Sibson traces thin mylonite bands across unexposed ground in south-eastern North Uist for 5 km or more, largely by following linear features on aerial photographs. Since so many of the mylonites are lenticular and since there are marked lateral variations in intensity of mylonitic foliation, for example at Beinn na h'Aire [NF 901 589], some of these correlations may not be fully justified. Sporadic development of the mylonitic foliation in largely broken and cataclastic acid hornblendic gneiss is well seen in the Lochportain-Crogary na Hoe area of North Uist.

In South Harris mylonites apparently lie just offshore to the south-east and crop out only in small exposures at Rubha Vallarip [NG 059 831] and Rubha Quidnish [NG 098 864]. Strongly mylonitic gneiss with local thin fissile mylonite bands is seen on south-east Scalpay, and Sibson (1977b) records the occurrence of synkinematic biotite at the extreme south-east margin of the island. Mylonites are most abundant in south-east Lewis, where individual zones are up to several hundred metres thick, for example around Loch Bhrollum. They diminish northwards, but are found at Kebock Head [NB 420 138] and Raerinish Point [NB 430 248]. At Ard Raerinish [NB 428 243] a belt of green platy mylonite more than 150 m wide dips east at 40° to 50°. Sibson (1977b) notes a quartz lineation plunging 44° towards 104° and he also notes Z-profile chevron folds (see the section on folding below).

Possible mylonite bands (green 'schist') were recorded in the Laxdale River section [NB 411 355] by A. D. Haldane of the Geological Survey. The mylonites represented on the lithological map (Map 1) south of Tolsta Head are sheared and retrograded zones within thick amphibolites. The age of this deformation is not known. Fifty metres south of Tolsta jetty [NB 541 462], 50 m of grey fissile calcareous mylonite dipping 25° to the NNE were noted by Sibson (1977b).

Mode of formation

The mode of mylonite formation varies along the length of the thrust zone.

In the Rubha Bhrollum area [NB 321 031] of south-east Lewis several stages of mylonite formation can be distinguished. Generation started in thin mica-rich layers (0.5 to 2 cm thick) in the gneiss, with the new mylonite fabric tracing out a form similar to small-scale cross-bedding (Figure 34). The basal part of the mica-rich band is the more deformed, possibly implying that there is a planar discontinuity at the base. The mylonite zone developed subparallel to the gneissic banding (striking 065° and dipping 35° to the south-east) but the mylonite fabric in the centre of the band strikes around 020° and dips 20° to the east, implying that a down-dip, i.e. normal sense of movement, has occurred across the band. A few hundred metres north-east [NB 327 034], fissile mylonites have developed in acid gneiss adjacent to large basic bodies. Only small lenticular zones retain relict gneissic banding. In detail the mylonites consist of discrete 'flattened' lensoid zones 1 to 15 cm wide within which the strike of the mylonite foliation varies from 010° and the dip 15° to the east to subparallel to the relict and adjacent gneissic banding. These orientations are taken to reflect the differing times of mylonite generation in the various gneiss lithologies, the response of differing lithologies to mylonitisation, and the apparently locally differing sense of shear within such zones (see (Figure 34)). The mylonite fabrics here do not trend subparallel to the contact of gneiss and basic rock, although they do elsewhere.

To the north (for example around [NB 318 118]), at the margin of the Loch Eishken mylonites, the mylonite fabric is initially developed subparallel to the gneissic banding but grades inwards to a weak but penetrative and discordant foliation with a variable orientation. Eventually a more uniform pervasive schistosity completely overprints and destroys the original gneissic banding.

A similar development of mylonite is observed in the Lochportain area of North Uist, where fissile dark green mylonite zones 1.5 to 30 metres thick show narrow transitional boundaries with the adjacent 'cataclastic' and locally 'mashed' gneiss. We interpret the thrust-zone rocks here as being folded into a large late open synform; Sibson (1977b), however, does not recognise a coherent structure in this area but he does record WNW-dipping mylonite bands locally farther south beyond Lochmaddy. Although a mylonite fabric is weakly developed within large areas in the 'mashed' and 'cataclastic' gneisses, it rarely forms a penetrative parting. The mylonite bands represent a localised intensification of this weak regional fabric. The mylonite fabric in the 'mashed' gneisses of this area may have a consistent orientation because there was no pre-existing gneissic banding to influence its development.

In contrast, minor fissile mylonite bands are present on Scalpay only in the central part of the island; the main mylonites in the south-east were produced by progressive attenuation of the existing gneissic banding accompanied by pervasive greenschist-grade retrogressive metamorphism that occurred when water partial pressures were high. The mylonites at Rubha Quidnish in South Harris formed similarly.

In widely separated parts of the thrust zone, for example in the Lochportain area of North Uist and the Eishken area of south-east Lewis, minor thin grey-green ultramylonites developed apart from the main development of late mylonites.

They are generally 0.2 to 2 cm thick and approximately planar, with some bifurcations. They dip gently to the east or south-east and discordantly cut across pseudotachylite/'cataclastic' gneiss zones. Although they lie to the west of the mylonite bands they are probably coeval with their development.

Around Loch Bhalamuis [NB 295 011] a series of thrusts/ reverse faults trends about 070° and dips steeply south, cutting across earlier mylonite fabrics and gently dipping thrusts. The gently dipping thrusts are locally marked by 15-30 m of platy fissile colour-banded mylonite with an internal foliation striking 020° and dipping 30° to the ESE. Adjacent to the late thrusts is an intense mylonitic fabric striking 070° and dipping 60° south. This fabric becomes dominant to the SSE, where there are thick mylonite zones. There were thus two phases of thrust and mylonite formation in this area. There are also late thrusts with mylonites up to 2 m thick 800 m west of Rubha Quidnish in South Harris (foliation 012/23° E). In southern South Uist, on the south-west flank of Roneval [NF 813139], up to 10 m of altered, crushed and sheared gneiss underlies the basal pseudotachylite breccia zone (5–10 m thick) of the Outer Hebrides Thrust Zone. This altered and fragmented gneiss marks a late brittle thrust which appears to have been formed considerably later than the main development of 'mashed' gneiss and pseudotachylite.

Petrography

The mylonites are typically all fine grained and where derived from acid gneisses they are composed of quartz, albite, minor potash feldspar, epidote, chlorite, sericite and phengitic muscovite. The quartz commonly occurs in elongate lenticles composed of near equant grains averaging 0.03 mm in diameter. In the amphibolite-mylonites, tremolite/actinolite laths are widely developed, and the feldspar is retrograded to quartz, albite and epidote. Blue-green hornblende is rare and sphene has commonly replaced ilmenite. Relict brown or green hornblende is common in these basic mylonites whereas mylonitisation has normally destroyed the original components of the acid gneiss. Mylonite mineralogy is fairly consistent along the length of the thrust zone. The foliation is defined by recrystallised chlorite and phengitic muscovite. A mineralogical colour banding is defined in some specimens by variations in concentration of quartz, epidote and the more mafic minerals. This fine-scale banding, which does not necessarily follow the original gneissic banding, has in some specimens been folded into tight asymmetrical microfolds. Such folds have axial plane orientations similar to that of the overall mylonite fabric and probably developed during progressive mylonitisation. More common than these 'mylonite' microfolds are open to close micro-folding and a related later discordant strain-slip cleavage which are readily seen under the microscope (for example, (MC31996)). Rarely (MC31978 from Stuley Island) chlorite defines a foliation that lies at 3° to 6° to the mineralogical banding defined by quartz and feldspar aggregates.

Relict feldspars, both plagioclase (oligoclase to andesine) and potash feldspar, are abundant in some of the mylonites; they have commonly been fragmented, and progressive mylonitisation has made them more circular or lensoid and reduced their size. More rarely large 'newly' crystallised epidotes have also been broken (e.g. (MC 31996)). In mylonite specimens from Rubha Bhrollum (south-east Lewis) the feldspar has been reduced to fine-grained, highly lenticular, recrystallised albite aggregates with much sericite. Rarely new albite porphyroblasts are seen. Poikiloblastic albite porphyroblasts are common in some parts of the thrust zone, e.g. (S57860), east Loch Claidh, Lewis; (S61832), Rubha Quidnish, South Harris; (S61874), Rubha Bhrollum, south-east Lewis; (S59329) Rubha Rossel, South Uist.

Typically ilmenite has been altered to sphene as metamorphism became retrograde, and in many mylonites no ilmenite remains. Exceptionally, sphene has been altered to leucoxene (MC31981) Usinish Bay, South Uist). Although pyrite is present in some rocks (MC31979), Stuley Island), generally hematite or limonite has developed. Original allanite and zircon remain but are granulated in part (for example, (S61876), Rubha Bhrollum, south-east Lewis). Allanite in the mylonites is generally metamict and in parts has epidote overgrowths.

Carbonate is not present in the South Uist mylonite specimens but is found in those from east Benbecula (Greanamul Island) and North Uist. It is particularly abundant in the altered and mylonitic rocks of south-east Lewis and Harris. Epidote veining is extremely common in the gneisses of eastern South Harris.

Mylonitic fabric orientation

The trends of the mylonite fabric are represented on the structural maps as far as possible; however, it has proved imt practicable to separate them from the 'cataclastic' gneiss fabric in many areas. The mylonite fabric generally strikes north-north-east and dips gently to moderately to the east. The pre-existing gneissic banding has in some areas clearly influenced this fabric development, for example in the Parc district of Lewis. (Figure 35) shows two diffuse maxima, one approximately centred around 025/25°SE and the other at about 075/40°S. The latter strike and dip are close to those of the average gneissic banding, particularly in the area around Loch Claidh and Loch Bhrollum. Folding has affected the fabric orientation on both a local and regional scale. Sibson (1977b) recorded the angle between the mylonite foliation and the margin of the mylonite zone at several localities, notably in North Uist. He made crude estimates of the amount of shear strain, obtaining minimum values of 5 for typical mylonite bands. These are estimates only of the minimum amount of translation that has occurred, because brittle thrusting on various scales has been shown to accompany mylonite development.

Folding

Folding in the mylonites may be conveniently divided into three distinct types:

1 Small-scale folding of compositional banding or early-formed mylonite fabric is seen under the microscope. The formation of such folds is thought to be an integral part of the mylonite process, the more strongly banded (and therefore less homogeneous) mylonite having been particularly susceptible. Coward (1972) has described early folds in mylonite banding from Stuley Island and Usinish Bay, South Uist. He notes that a secondary mica fabric lies parallel to the axial plane of these folds and there is commonly a strong lineation parallel to the fold axes. In general, however, no coherent history of deformation can be worked out, and the presence of mylonitic fabrics of differing orientations is not necessarily indicative of widely separated periods of mylonite generation.

2 Open to tight, dominantly chevron-style folds of the mylonite fabric occur at many localities in the Outer Hebrides Thrust Zone. In places there are small kink folds with east–west-trending axes, but medium- to small-scale Z-profile asymmetrical folds are more common. In thick (15–30 m) fissile mylonites on the south-east side of Loch Claidh (south-east Lewis) tight minor folds dominate the outcrop; their wavelength averages about 1 cm but in parts it is as much as 10 m. A related strain-slip or spaced cleavage dips north-west at about 40°, and can be traced for several hundred metres into the overlying gneisses, where it is axial planar to large-scale open to close asymmetrical folds of the gneissic banding. Minor thrusts have been developed where the folding was intense, e.g. around Loch Bhalamuis [NB 295 011] and at Eigneig Bheag [NF 924 601] in the southeast part of North Uist.

At Rubha Bhrollum [NB 241 037] the mylonite is in parts very tightly folded with a spaced axial planar cleavage and an intersection lineation developed parallel to the fold axes. In thin section (S61874) the rock consists of lenticular aggregates of slightly ovoid quartz grains (average diameter 0.04 mm) in a dominantly granular matrix of clinozoisite, muscovite (sericite and phengite), quartz, fine-grained albite, minor potash feldspar and irregular chlorite aggregates and relict hornblendes. The muscovite locally shows a fine-scale strain-slip cleavage with minor recrystallised grains growing in the new cleavage. Small elongate epidotes also define the new fabric. The dominant opaque mineral is hematite, commonly after pyrite.

Both minor and major folds show a dominant sense of vergence to the south-east (Z-shaped profile when viewed from the west or south-west). This reverse or down-dip vergence is particularly well exposed in a fold train at Eigneig Bheag. (Figure 36) shows the orientations of the axes and axial-planes of these crenulation or chevron folds in the Parc district, with additional data from Eigneig Bheag and Rubha Quidnish (Sibson, 1977b). Where folds are marked as–neutral (upright) or vergence (asymmetry) not recorded, most fall into the second category. Although there are minor regional differences in orientation of fold axes and axial planes, in general they plot as a coherent set, possibly in consequence of the fairly restricted range of orientations of mylonite foliation in which these secondary folds developed.

Sibson (1977b) has ascribed the sense of fold asymmetry to the backward (lag) sliding of large masses of crustal material along pre-existing mylonite zones. However, it is difficult to envisage how such a process could occur. There would have to be major extension of the crust, in contrast to the previous thrusting movements and it would entail an increase in volume at depth. However, extension of the crust would favour normal faulting and not backsliding. In the Eigneig Bheag [NF 924 601] outcrops, the folds are progressively tighter from top to base of the mylonite zone. The fold axial planes have been rotated into near parallelism with the dominant basal mylonite fabric. The field relationships thus show that 'extensional' shear strain was highest at the base of the overthrust zone, decreasing upwards. In many late-developed thrust zones it can be shown that thrust blocks developed in a complex manner. Thrusts generally propagate ahead of the advancing thrust front and then cut structurally up-section (Elliot and Johnson, 1980). Field studies of mylonite formation (Figure 34) show that locally thrusts can develop in a normal fault (extensional) regime. We suggest that the fold asymmetry at Eigneig Bheag is due to a series of underthrust slices that formed along the pre-existing mylonite zones. However, this suggestion implies that there are corresponding west-directed thrust zones at the base of such slices. Such zones would have to lie west of the Outer Hebrides.

An alternative explanation is that lag movement may result from low-angle extensional faulting (see White and Glasser, 1987). Extensional faults with accompanying cataclastic products and ductile mylonites have been recognised in the Snake Range in Nevada, USA (Mortimer et al., Stanford University, personal communication). Movement of several tens of kilometres has occurred along a major dislocation accompanied by normal faulting of the material above, and in adjacent rocks. If the Outer Hebrides are analogous, then the initiation of the Minch fault system might have been related to these late-stage extensional movements along the thrust zone and hence be of Late Caledonian age. This hypothesis would accord with seismic reflection data from the MOIST project, which show no obvious displacement by the Minch fault system of the reflections thought to represent the Outer Isles Thrust.

3 Large-scale late-stage open folding on axes plunging gently north to north-north-east has affected the outcrop pattern of the mylonites from South Uist through to the Parc district of Lewis. We interpret this folding as responsible for the westerly dip of the mylonite fabric which is particularly well seen in the Lochportain area (north of Lochmaddy), North Uist. This folding postdates the thrusting, mylonite formation and probably any extensional movements of Caledonian age. It is reflected on the stereoplots by the spread of orientations of foliation poles and axial planes, and to a lesser extent of fold axes (see (Figure 35) and (Figure 36)).

Late-stage alteration

Many of the gneisses in the Hebrides, particularly those in Lewis, show the effects of partial retrogression. These changes probably have occurred at different times in different areas. The gneisses within and adjacent to the main mylonites are the site of low-grade, notably hydrous, retrogression to lower greenschist facies. There is a strong spatial relationship between the retrogression and the mylonites of the Thrust Zone (see Maps 1 and 2), and we take it that they are causally related. In the Grosebay–Stockinish area, south of Tarbert (South Harris), this retrogression has strongly affected acid and basic gneisses, which are cross-cut only by minor thin mylonite bands, and also by zones of broken gneiss near the south-east coast.

The following account is largely taken from Myers (1971), who has described the mineralogical changes caused by late-stage alteration in Harris. A marginal zone to the west of the main thrust ranging from 400 m to 3 km wide stretches south-west from Tarbert to An Coileach [NG 087 926] within which epidote has developed along joints. South-east of this zone there lies a tract of pervasively retrograded Lewisian gneiss where plagioclase has been altered to fine sericite and epidote, and microcline is notably abundant, generally showing perthitic intergrowths and granulated margins. Hornblende has been altered to epidote and chlorite, and biotite to chlorite. New muscovite has locally developed in sericite-rich areas. Sphene is abundant and apatite has been altered to collophane.

Areas of similar, although less intense, low-grade alteration are also found around Kyles Scalpay [NG 213 980] in southern North Harris; on Scalpay itself; within the thrust zone, particularly in the Parc district of south-east Lewis, and in the areas of Rarnish (Benbecula), Wiay, and Loch Skipport (South Uist) (Coward, 1969, p.271). In this last region mineral alteration has occurred within specific zones and in many rocks has only affected particular minerals. For example, in specimen (S60107) [NF 8753 4868], from a basic dyke at Rarnish, the feldspar has been altered to sericite, albite/oligoclase and minor epidote. The clinopyroxene shows only minor alteration and the hornblende is apparently unaffected. In specimen (S60113) [NF 8695 4751], a metasediment from the Keiravagh Islands, Benbecula, biotite has been almost totally altered to chlorite even where it is enclosed in hornblende, and yet other minerals apparently remain unaltered. The scapolite-feldspar analyses from a cross-cutting basic dyke at Rarnish (S60110) show that barium-rich potash feldspar is present and the plagioclase lies in an An29 to An32 range. In thin section the rock contains the assemblage opx-cpx-pl-kf-gt-hbl(brown)-scp-il; the orthopyroxene shows marginal alteration (perhaps to talc, brucite) and the feldspar is slightly altered to sericite. However, other basics (e.g. (S60112) [NF 8756 4813]) at Rarnish show penetrative alteration to epidote, chlorite, sericite, quartz, albite, sphene with only relict hornblende and some ilmenite remaining.

Age of the Outer Hebrides Thrust Zone

It is probable that the thrust zone is older than the Permo-Carboniferous. Four thick quartz-dolerite dykes cut the thrust on Barra and adjacent small islands to the north. Numerous thin camptonite dykes also cut the thrust zone, notably on Eriskay and South Uist. On Stuley Isle, east of South Uist, several camptonite dykes cut across mylonites which became locally silicified up to 60 cm from the dyke margins (Jehu and Craig, 1925). These dyke suites closely resemble known Permo-Carboniferous suites on the mainland of North-west Scotland; they are discussed in detail in Chapter 16. The quartz-dolerites have been dated by Rb-Sr determinations on biotite and by palaeomagnetic methods at about 290 Ma. Steel and Wilson (1975), working on the Permo-Triassic Stornoway Formation, have shown that the conglomerates of the Eye Peninsula contain cobbles of 'mashed' and mylonitic Lewisian Gneiss.

The maximum age of the thrust zone is Late-Laxfordian. It affects Laxfordian granites which in South Harris have been dated by Rb-Sr mineral and whole-rock methods at 1710 Ma, and zircon U-Pb methods at 1715 ± 20 Ma (van Breemen et al., 1971). Biotite and hornblende K-Ar ages on grey gneiss from the Uists fall in the range 1500–1700 Ma (Moorbath et al., 1975). These latter dates apparently record a period of pegmatite development and uplift at the close of the Laxfordian. We postulate that uplift of about 20 km occurred in late- to post-Laxfordian times. Pebbles in the Torridonian Applecross Formation on the mainland of North-west Scotland contain greenschist- to lower amphibolite-grade Laxfordian gneiss assemblages; Laxfordian gneisses exposed at the present time have a considerably higher grade of metamorphism (Williams, 1969). Thus higher levels of the Lewisian Complex were being eroded at the time of deposition of the Torridonian than are exposed now. We suggest that the Laxfordian was being rapidly eroded during the Torridonian and so had high relief at the time, relief caused at least in part by late-Laxfordian thrusting primarily along the Outer Hebrides Thrust Zone.

Sibson (1975, 1977b,c, 1980) has suggested that the development of pseudotachylite and mylonite are related in such a way that mylonite develops at depth in a major thrust zone while pseudotachylite is generated nearer the surface at shallow crustal levels. D. C. Rex (in Sibson 1977b, p.176) has carried out K-Ar whole-rock age determinations on pseudotachylite, cataclasite and fissile mylonite from the thrust zone in the Uists and Barra. He found that pseudotachylite and cataclasite ages (12 in total) range from 442 ± 17 Ma to 2056 ± 80 Ma. He also obtained four mylonite ages ranging from 394 ± 16 Ma to 471 ± 18 Ma with a fifth age at 947 ± 40 Ma. Moorbath (also in Sibson, 1977b) has obtained K-Ar ages of 1120 ± 44 Ma and 1140 ± 44 Ma from pseudotachylite west of the thrust base in South Uist. Sibson accepts the mylonite ages as dating the major thrusting and hence considers the Outer Hebrides Thrust Zone to have been generated during the late Caledonian. Francis and Sibson (1973) draw attention to the similarities of trend and extent of the Moine and Outer Isles Thrust Zones and suggest that they may be approximately coeval.

Park (1961) has described numerous occurrences of pseudotachylite, mainly in subvertical shear zones, from Lewisian amphibolite-facies gneisses of the Gairloch area of North-west Scotland. These zones were noted by Peach et al. (1907) further north where they are of pre-Torridonian age. Moorbath and Park (1972) relate movement on the Gairloch shear zones to dates of 1150 Ma from chloritised biotite in adjacent rocks. Recent work by Consolidated Goldfields (E. Jones, personal communication) in this area has shown the presence of thrusts marked by pseudotachylite, one of which has a minimum displacement of 2 km. These thrusts have an orientation very similar to that of the Outer Hebrides Thrust Zone and are cut by Torridonian sandstone 'dyke' infillings.

A period of thrusting and pseudotachylite development at or before 1150 Ma on the mainland of North-west Scotland would accord well with the evidence for late granite-vein emplacement, thrusting and uplift in the Outer Hebrides following the major Laxfordian deformation. Seismic reflection data from the MOIST line (Pentland Firth to north of Lewis) suggest that the Outer Isles Thrust extends to the Moho with an unchanging dip of about 25–30° (Brewer and Smythe, 1984).

In the light of the data outlined above and the close relationship between granite intrusion and thrusting and local mylonitisation in South Lewis (see the end of Chapter 11), we conclude that the Outer Hebrides Thrust Zone was initiated at about 1700 Ma, when the early mylonites and pseudotachylite began to be formed. Movement possibly continued intermittently for over several hundred million years. We think there was a major episode of thrusting and uplift at around 1150 Ma. The mylonites of lower to middle greenschist-facies and the related alteration we take to be of Caledonian age, formed mainly along the pre-existing thrust zone. If this last conclusion is right, it throws doubt on the role of the Lewisian of mainland Scotland as rigid foreland to the 'Caledonian' orogen.

The recent MOIST, WINCH and DRUM deep reflection seismic profiles which have resulted from marine geophysical work north and west of the Outer Hebrides have shown a major crustal reflector, the Flannan Thrust, dipping 25° to 30° to the ESE (Smythe, 1987). The trace of this thrust lies about 55 km WNW of that of the Outer Hebrides Thrust Zone and coincides at the surface with the Flannan Fault. This latter structure forms the north-west margin of the Flannan Trough and downthrows to the ESE. The Flannan Trough, is a shallow graben some 25 km wide which contains sediments of probable Permo-Triassic age generally dipping 5° to 15° to the north-west. It appears that the Flannan Thrust has been reactivated as a normal fault during the Permo-Triassic in a similar manner to the Outer Hebrides Thrust Zone–Minch Fault complex immediately ESE of the Outer Hebrides. The Minch Fault was undoubtedly active in the Permo-Trias (see Chapter 17) but probably formed in late-Caledonian times (Silurian–Devonian).

Chapter 16 Post-Lewisian minor intrusions

Igneous intrusions, mainly in the form of dykes, cut the basement Lewisian rocks throughout the length of the Outer Hebrides. The vast majority belong to two suites, namely a Permo-Carboniferous camptonite-monchiquite swarm, and several swarms of Tertiary dolerites and basalts. Two additional minor suites are represented: a few hitherto-unknown Caledonian appinites, and a small swarm of Permo-Carboniferous quartz-dolerites. The general distribution of the suites is shown on inset maps on Maps 1 and 2.

Caledonian

Appinite

The term appinite is here used to imply a hornblende-rich coarse-grained rock of gabbroic-dioritic composition. The existence of an appinitic intrusion was first suspected by J. S. Myers while carrying out a reconnaissance study of the Uig Hills area of Lewis in 1965. He discovered an old corn-mill with mill wheels of appinite adjacent to a small burn draining Loch Sandavat [NB 008 303]. Subsequently we discovered seven additional millstone localities within a 3 km radius of Carishader, as well as the source-rock quarry, where there were partly made millstones, near the entrance to Little Loch Roag [NB 124 318]. The source-rock is an east-trending dyke 9 m wide, uniformly massive and coarse grained, with distinct chilled edges. In hand-specimen the appinite is characteristically mottled; dark green to black amphiboles are set in a greenish pink (pyroxene-feldspar) matrix. Isolated poikilitic crystals of biotite are visible in the main part of the dyke and occur as numerous small flakes in the fine-grained, dark grey chilled edge. Westwards the dyke can be traced as far as Geshader but thereafter is lost, presumably displaced by the north-west-trending faults. To the east, it is fully exposed at Aird Chaol [NB 134 320] and again on the east side of Loch Drovinish and 500 m farther to the ENE. A pronounced hollow continues to the east and the chilled edge of the dyke is seen just west of the Bernera road at [NB 175 332]. No exposures of the dyke are seen to the east of this locality across East Loch Roag but an identical dyke, presumably the same one, has been mapped by Lisle on the shore near Breasclete [NB 213 350] and traced intermittently eastwards to [NB 249 360]. Another small exposure, accessible only at low tide, was found by Lisle (personal communication) NNE of Shawbost [NB 268 488].

The only other example of appinite found in the Outer Hebrides is a small isolated exposure of massive, coarse-grained, somewhat altered rock on the shore of Loch Fada, North Uist [NF 881 711].

Thin sections of these appinites consist of bright green, subhedral to euhedral hornblendes up to 3 mm in length which enclose opaque ores, small pyroxenes and apatites (S57837). The very pale green pyroxene of the matrix is much smaller, typically 0.3 X 0.1 mm, and tends to be equigranular. Euhedral to subhedral olivine, completely pseudomorphed, is present in small amounts in the coarse-grained part of the dyke and is the dominant mafic mineral in the chilled marginal zone. Large plates of fresh, pale brown biotite occur in most slices, poikilitically enclosing pyroxene and olivine. The feldspar is typically altered to a turbid reddish brown product. Much of it appears to be orthoclase but some albite is also present. Accessories are opaque ore and apatite, the apatite in needles up to 1.5 mm long. The chilled margin of the dyke is zoned; from the outer edge inwards the zones are successively: a glassy groundmass with olivine pseudomorphs and fresh pyroxene; a very fine-grained zone with abundant olivine pseudomorphs and few fresh pyroxenes; and a fine-grained turbid feldspar zone, with abundant granular or acicular pyroxenes, many olivine pseudomorphs, numerous flakes of olive-green biotite, and much cubic opaque ore (S57906). No hornblende is present in the chilled zones.

Chemical and modal data for a sample from the centre of the dyke are given in (Table 30), analysis 1, and (Table 31) respectively. Two samples from separate localities were taken for K/Ar age-dating. One, using biotite, gave 431 ± 10 Ma; the second, using hornblende, gave 477 ± 11 Ma. The petrography and chemistry of the Lewis appinite are comparable to the type appinites (potassic basalts) of the Appin-Ballachulish area, Argyll (Wright and Bowes, 1979), although the age obtained is significantly older, and probably closer to that of other late-Caledonian hornblendic intrusives in the Northern Highlands that are also termed appinite (Smith, 1979).

Permo-Carboniferous

Quartz-dolerite dykes

The presence of quartz-dolerite dykes which were presumed to be of Tertiary age was noted by Jehu and Craig (1923) in their survey of the Barra Isles. The dykes, which characteristically show a very coarse-grained texture and which weather to a distinctive brown colour, have a restricted occurrence. They crop out only in northern Barra, the islands in the Sound of Barra and in the southern parts of South Uist (Figure 36). A total of ten dykes have been mapped, ranging in thickness from 3 m to 45 m. The four dykes on northern Barra and Fuday have an aggregate thickness of 146 m over a distance of 5.1 km measured at right angles to the dyke walls, showing a crustal extension of 3 per cent across this region.

At one locality in north-east Barra [NF 721 035] a 9 m Tertiary olivine-dolerite cuts a 27 m quartz-dolerite, and on the east coast of Eriskay [NF 804 110] a 30 m quartz-dolerite is cut by a 45 cm thick Permo-Carboniferous camptonite.

Identical field relationships are found on the mainland of north-west Scotland around Arisaig, indicating that the Outer Hebrides quartz-dolerites, like those of the mainland, are Permo-Carboniferous and not Tertiary. A total of twenty whole-rock K-Ar age determinations were carried out on a comprehensive collection of dyke samples. The results were not consistent, ranging from 612 ± 11 Ma at the chilled edges to 1376 ± 28 Ma at the centres. The specimens appear to contain excess radiogenic Ar probably derived from the adjacent gneisses, a feature found elsewhere in basic dykes. However, Rb-Sr ages on biotite separated from two different dykes gave 300 ± 4 and 284 ± 4 Ma (P. J. Patchett, personal communication). Twenty-one palaeomagnetic measurements from three different dykes and from samples from the adjacent baked Lewisian gneiss gave pole positions clustering around 137°E 38°N, which corresponds to the Kiamen direction, i.e. between 295 and 230 Ma (R. Thompson, personal communication). On the basis of the Rb-Sr and palaeomagnetic results we consider these dykes to be Permo-Carboniferous.

The dominant mineral is plagioclase in the form of tablets and laths up to 3 mm in length, averaging 1.5 mm. It has a general compositional range of An50 to An55 and forms a pseudo-ophitic texture with pale pinkish (in thin section) clinopyroxene which occurs as aggregates of crystals up to 3 mm across. Some individual clinopyroxene crystals enclose small plagioclase microlites and apatite. Orthopyroxene (probably enstatite) is much less common but is present in most slices. It forms large crystals partially or completely pseudomorphed by a greenish yellow serpentine mineral. Large irregular crystals or magnetite and/or ilmenite are the main accessory. Small patches of interstitial quartz or quartz-feldspar intergrowths are invariably present associated with brown hornblende, flakes of green to dark brown biotite, pyritous opaque ore and apatite. Typical slices are (S57792), (S58054) and (S59963).

Three analyses from two quartz-dolerite dykes are given in (Table 30) (analyses 2-4) and one mode in (Table 31). The dolerites appear from their chemistry to be typical quartztholeiites and differ from Manson's (1967) worldwide average of 283 quartz-dolerites only in having slightly higher total iron, slightly higher Na/K ratios, and slightly lower Al2O3.

If the Sound of Barra quartz-dolerite swarm is projected across the Sea of the Hebrides, it reaches the Scottish mainland, at the southern, most dense portion of a zone of plugs and thick lensoid dyke-like bodies found between Mallaig and the Sound of Mull ((Figure 37)a). The mainland and Hebridean quartz-dolerites are indistinguishable mineralogically and chemically.

Camptonite-monchiquite dykes

Camptonites and monchiquites (Rock, 1977) are lamprophyres of basanitic composition carrying olivine, titanaugite, kaersutite and/or biotite phenocrysts in a groundmass of clinopyroxene, kaersutite, biotite and plagioclase (in camptonites) or glass/feldspathoids (in monchiquites).

A Permo-Carboniferous camptonite-monchiquite dyke-suite is widespread throughout the Scottish Highlands and Islands (Rock, 1983). Camptonite dykes were first recognised by Jehu and Craig (1923) on Vatersay, where they found a 1 m dyke trending east-south-east that was cut by a Tertiary olivine-dolerite (Jehu and Craig, 1923, fig.4). Subsequently (1925, 1926, 1927) they recognised numerous lamprophyre' dykes on Eriskay and South Uist, together with three examples on North Uist and Benbecula and 'a few' on South Harris. In the present survey we have largely confirmed these findings both as regards distribution and age relations. About 50 dykes on the Outer Hebrides form one of the smaller of the 9 regional swarms into which this suite can be divided. The main part of the swarm extends from Eriskay northwards to Rubha Hellisdale on the east coast of South Uist. A total of 35 dykes have been mapped in this section, ranging in thickness from 0.5 to 9 m with an average of 2 m; the mean trend is 120° (Rock, 1979). The dykes are not evenly distributed; they do not become more common towards a central axis but tend to occur in small groups of about four or fewer. To the south, five individual dykes have been mapped on Barra and Vatersay, but none on the islands further south. To the north, ten widely spaced dykes of variable trend are present between Wiay and Loch Roag in western Lewis. Farther north on Lewis they have not been found, although as rock exposure is poor, we cannot be sure that they do not occur. The distribution of dyke trends is shown on (Figure 37)b.

The dykes can be taken as Permo-Carboniferous on the basis of the now-substantial number of age determinations in the range 330-230 Ma that have been obtained from petrologically similar camptonite-monchiquite swarms in Orkney, Ardgour, the Ross of Mull, and elsewhere (e.g. Brown, 1975; Speight and Mitchell, 1979; Beckinsale and Obradovich, 1973). The dykes belong to a regional alkaline magmatic province which extends from Donegal through Scotland and northern England and over a large part of Scandinavia, where it includes the Oslo plutonic province (see summary in Rock, 1983). The limited field evidence from the Outer Isles is consistent with the Permo-Carboniferous age: two camptonitic dykes on the west coast of Barra [at NF 655 024] and [NF 656 025] are cut by Tertiary dykes, as is the dyke on Vatersay noted above. The camptonite-monchiquite swarm is younger than the PermoCarboniferous quartz-dolerite swarm, as shown by a camptonite cutting a quartz-dolerite on the east side of Eriskay (see the preceding section, on quartz-dolerites); similar field relationships are found in the West Highlands.

Thin sections of 35 of the 50 or so mapped dykes have been examined, and full details are available in a computerised file of data (Rock, 1979). Of these, 28 have proved to be camptonites, composed generally of phenocrysts of forsteritic olivine, clinopyroxene, and rarely kaersutite (in 4 slices), in a matrix of clinopyroxene, kaersutite, plagioclase, biotite (in 14 slices) and subordinate alkali feldspar (in 7 slices). Three specimens are basaltic camptonites, being poor in kaersutite and biotite, one lacking kaersutite altogether. A single monchiquite ((S57913), (S57914)) was identified from Loch Roag, Lewis. It is composed of olivine, clinopyroxene and subsidiary biotite phenocrysts in a feldspar-free base of clinopyroxene, biotite, feldspathoids, zeolites, chlorite, altered glass, etc. This dyke is not dissimilar to the so-called 'nepheline-ouwachitite' of Colonsay (Cunningham-Craig et al., 1911). The remaining three dykes are too altered to be classified.

The groundmass texture is doleritic in most slices, but in one it is 'reverse ophitic', in which plagioclase poikilitically encloses pyroxene and kaersutite (slice NUD2 of Jehu and Craig). Six dykes contain the leucocratic globular structures that are characteristic of camptonite-monchiquites as a whole, and can be used to help distinguish them in Scotland from Tertiary dolerites (e.g. Bailey et al., 1924, p.377). The mineralogy of these globular structures is variable, some containing analcite and carbonates only, some feldspar only, and others brown amphibole in an irresolvable matrix.

The Loch Roag monchiquite carries one of the richest xenolith/megacryst suites of any single camptonite-monchiquite intrusion in Scotland (see Rock, 1983, table 1). Recent sampling by B. G. J. Upton and P. Aspen (Edinburgh University, personal communication) has yielded megacrysts of spectacular hexagonal corundum (up to 25 X 50 mm), biotite, apatite, clinopyroxene and titanomagnetite.

Xenoliths of gneiss, biotite-pyroxenite and anorthosite were also found.

Chemical analyses of five samples of camptonites and monchiquites by A. N. Baxter (Table 30) show that all five have compositions comparable to camptonite-monchiquite world averages (Rock, 1977) and to the compositions of the mainland swarms shown on (Figure 37).

Tertiary

Division into swarms and petrological types

Tertiary dykes are far more numerous than are older dykes, and are present practically throughout the length of the Outer Hebrides. In three areas they are so abundant as truly to deserve the term 'swarm'; namely in the Barra Isles, in South Harris and adjacent islands, and on the east coast of Lewis, where the swarm actually comprises two subparallel concentrations at Loch Odhairn and between Loch Shell and Loch Erisort ((Figure 37)c). A recent summary of Tertiary Hebridean dyke swarms by Speight et al. (1982) includes the Harris and Lewis swarms, but not the dykes on Barra.

Two of these three Outer Hebrides swarms can be tentatively correlated with the better-known swarms on the Inner Hebrides and the mainland ((Figure 37)c). The South Harris swarm lies directly on a north-westward extrapolation of the main Skye swarm through Vaternish (cf. Speight et al., 1982, figs.33.3, 33.5), and the Barra swarm is only slightly offset from a similar extrapolation of the Mull swarm (cf.

Speight and Mitchell, 1979, fig.3). Therefore, assuming there have been no major lateral post-Tertiary displacements on the Minch Fault (nor on other parallel major faults in the Sea of the Hebrides and the Minches), these two swarms are possibly related to the Cuillin and Mull intrusive centres respectively.

The Lewis swarm is more problematical, as the only 'intrusive centre' from which the dyke swarm can be extrapolated is the Shiant Isles ((Figure 37)c). The submerged 'Outer Roag' Centre of Bullerwell (1972) lies on the northwestward extrapolation of the Loch Shell concentration, but the dyke abundance actually decreases markedly across Lewis towards this 'centre', and recent geological studies suggest that the 'centre' may be a lava flow or sill. Extrapolations towards the mainland are equally uncertain. Following Speight et al. (1982, fig.33.5) we tentatively assume that the smaller (Loch Odhairn) concentration on (Figure 37) connects on a curved path with a north–south swarm in the Applecross peninsula (the 'Applecross subswarm' of Speight et al., 1982), given that an analogous curved path can be reasonably conclusively demonstrated for the main Mull swarm through Argyll, Cowal and Bute into Arran (Speight et al., 1982, fig.33.5). Any extrapolation of the large (Loch Shell) concentration, which is the most intense dyke concentration in the Outer Hebrides, is particularly obscure. Speight et al. (1982) show that the swarm bifurcates in the Minch just south-east of the Shiant Isles, with one part curving across Trotternish to join the main Skye swarm, the other twisting sinuously, in part parallel to the 'Applecross subswarm', past Raasay and Scalpay as far as Loch Hourn. However, this is inconsistent with two pieces of data: firstly, the strong aeromagnetic anomaly associated with this concentration on Lewis itself (see sheet 1 of the aeromagnetic map of Great Britain, IGS, 1972) dies out about 10 km north-east of the north end of Trotternish; secondly, very few dykes with the required orientation are known either on Raasay or on Trotternish. Most of the dykes on Trotternish trend north-west like the main Skye swarm, and not due north as required by this extrapolation. On (Figure 37), therefore, the south-eastward extension of the Loch Shell concentration is left open.

Note that although we have correlated the dyke swarms between the Inner and Outer Hebrides as shown in (Figure 37), we have not assumed that dykes necessarily exist beneath the intervening Sea of the Hebrides and the Minches. In fact, offshore work by BGS in these regions has shown them to be underlain largely by Mesozoic sediments intruded by numerous thick Tertiary sills (Binns et al., 1974; Chesher, 1979). It appears that the magma was intruded as dykes into the lavas of Mull and Skye and into the brittle, fractured Lewisian basement but was intruded as sills into the 'soft' Mesozoic sediments. It certainly seems plausible to extrapolate the Loch Shell concentration to the mainland through the submarine sills. A NW–SE basement lineament appears to have existed at Loch Shell at least since Laxfordian times.

Statistics for these three major swarms are given in (Table 32), with comparative data (where available) for the Mull and Skye dyke swarms. Tertiary dykes of the Uists and Benbecula are not included in the table, but are dealt with in the comments below, since it is unclear whether they should be classed with the South Harris swarm to the north or the Barra Isles swarm to the south.

The Tertiary basic intrusives can be grouped according to petrology into two groups: crinanites and olivine dolerites/ basalts, and olivine-free dolerites (tholeiites). The term crinanite is used in the sense of Flett in Cunningham-Craig et al. (1911). The crinanites (sometimes called olivine dolerites of 'Crinan type', or analcite-bearing olivine dolerites) occur as dykes and rarely as sills; the olivine dolerites/basalts and olivine-free dolerites occur exclusively as dykes. The only direct evidence on field relationships between these groups comes from South Harris, where the olivine-free dolerites are earlier than the olivine dolerites; the former were intruded along early Laxfordian shears trending NNW, and are displaced by NNE- and north-trending Tertiary faults along which the olivine dolerites were intruded (Witty, 1975).

Crinanites appear to be the most abundant minor intrusive rock-type in the Uists, Eriskay, North Harris and Lewis. Olivine-dolerites predominate in the Barra Isles and South Harris, while olivine-free dolerites have been recognised only in South Harris. This largely confirms the correlations made in (Table 32) since the abundances of rock-types in the Barra swarm very broadly correlate with those in the Mull swarm while those in South Harris broadly correlate with the Northern Skye swarm. These correlations are based largely on published descriptions, with a minimum of petrographic re-examination.

Crinanite dykes

Jehu and Craig (1923, p.437) compared the crinanite dykes of the Outer Hebrides with those from Jura and Crinan. In the Outer Hebrides these dykes are concentrated mainly on Vatersay, Barra, Gighay, South Uist and Scalpay. In South Uist they average 2-3 m in thickness. Examples 8 m thick occur on Barra, and one of a group of several unusually thick north-south dykes on the most easterly of the Monach Isles is 30 m thick (Jehu and Craig, 1926, p.486).

Petrographically, the crinanites usually consist of fresh to serpentinised olivine, titaniferous augite (usually described as purple- or plum-coloured and sometimes slightly pleochroic), zoned plagioclase, clear or turbid interstitial anal-cite, and accessories such as secondary zeolites, chlorite, apatite and iron ores, with minor biotite and/or brown amphibole in some specimens. An ophitic texture is commonly developed. One dyke from Stuley Island contains two pyroxenes, one purple and presumably titaniferous, the other colourless and presumably diopsidic (Jehu and Craig, 1925, p.635).

No analyses of the crinanites are currently available, but they are clearly of undersaturated alkaline composition and represent an alkali olivine basaltic magma-type.

Crinanite-teschenite sills

Although some of the more irregular crinanite dykes have a sill-like aspect in places, true sills are known only from two localities: near the entrance to Lochmaddy harbour, and the differentiated sills of the Shiant Isles, which are the better known.

Two sills are seen at Lochmaddy: one forming the islets of Madadh Beag, Madadh Mòr and Madadh Gruamach is at least 35 m thick and probably over 45 m; it dips at 20-30° ESE (Mackinnon, 1974). Diving work (Chesher, 1979) shows it to be underlain by a 2.5 m band of baked calcareous mudstones thought to be of Tertiary age and at least 5 m of a vesicular igneous rock, which may be a Tertiary lava or part of another sill. The second sill varies between 4 and 5 m thick and crops out at two localities on the peninsula of Ard nam Madadh, North Uist. It dips roughly parallel to the larger sill (Jehu and Craig, 1926) but is more decomposed and somewhat irregular in course and margins. All the rocks described from these sills have a similar mineralogy to the crinanite dykes. There are slight modal and textural variations within the thick Madadh sill ((Table 31), analyses 16 -20). There is no field evidence of the relations between the crinanite dykes and either sill, though there is a 4 m crinanite dyke nearby.

The Shiant Isles, comprising the main island duplet of Garbh Eilean and Eilean an Tighe, together with Eilean Mhuire to the east and the Galtachean islets to the west, are largely formed of differentiated picrite-teschenite sills. Three sills are probably represented (the precise number depending on how one correlates the sills between the islands), as summarised in (Table 33). Chemical analyses and modes of the various rock-types are included in (Table 30) and (Table 31) (analyses 21 -37) respectively.

Walker (1930) believed that the main (upper) sill is petrographically constant laterally, but varies vertically continuously from picrite at the base through transitional olivine-rich dolerite ('picrodolerite') to crinanite at the top ((Table 31), analyses 37, 21- 24). He recorded a continuous decrease of specific gravity upwards, and attributed the sequence to gravitational settling of olivine within a single magma body. Dreyer (1953) and Dreyer and Johnston (1959, 1965), however, later asserted that the picrite–picrodolerite contact is sharp and undulating, not gradual and planar, and that this structure implies that the two magmas were independent.

The base of the main sill, exposed at one locality on Garbh Eilean, shows a complex banded succession of picrites, teschenites, picroteschenites, crinanites and hybrid rocks occupying some 1.2 m between the main picrite and the metamorphosed Jurassic marls below (Dreyer and Johnston, 1959). The teschenitic rocks appear to have been produced largely through contamination of picritic parental material by marl, involving an endogenous reaction Mg-olivine–titanaugite. The base of this sill has also been located at about 15 m below sea-level on the south-east side of Garbh Eilean (Dreyer and Johnston, 1965). The rocks of all three sills are composed of olivine, clinopyroxene, plagioclase, opaques and analcite in various proportions, with alkali feldspar and perhaps nepheline in the syenitic segregations of the lower (Eilean Mhuire) sill only. Differences in the modal composition of the main sill (Johnston, 1953; Murray, 1954) are summarised in (Table 34).

The olivines exhibit very strong compositional zoning, except in the picrite and lowermost parts of the 'picrodolerite' (Johnston, 1953); the maximum range in a single grain is Fa30 to Fa87 (determined optically) at a height of 120 m above sea-level. The olivine cores are Mg-rich throughout the sill, whereas the margins show a gradual, but irregular, trend towards fayalite, reaching Fa95+. at 115 m above sea-level.

The pyroxenes are titaniferous augites (1.2–1.8 per cent TiO2) which show a crystallisation trend subparallel to the diopside-hedenbergite join in the pyroxene quadrilateral (Murray, 1954), with very little enrichment in the acmite molecule. Walker (1930) reported acmite-rich pyroxenes from the syenitic segregations of Eilean Mhuire, but these have not been analysed.

The picrodolerite-crinanite differentiation trend seen in the main sill is complemented by the crinanite-tescheniteanalcite syenite differentiation trend seen in the lower (Eilean Mhuire) sill. Modal changes in this sill are less well known, but at the east end of the island are approximately as shown in (Table 35). No compositional data for the minerals in this sequence are available.

The Shiant Isles sills together afford an excellent example of a full alkaline–basaltic differentiated sequence from mafic (crinanite/teschenite) to salic (syenite), together with cumulates (perhaps the picrite). As in many other similar bodies (e.g. Cnoc Rhaonastil, Islay; Howford Bridge, Ayrshire), however, there is a notable absence of intermediate (monzonitic) rock-types.

Several sheets of basalt are found intruding the lower sill (Table 33). The only analysis of one of these ((Table 30), analysis 36) is slightly different from any of the analyses of the sills themselves, as it is bynormative.

Olivine-bearing dolerite and basalt dykes

These are concentrated mainly in the islands south of Barra (Vatersay, Mingulay, Sandray, etc.) and in South Harris.

Petrographically they consist of olivine (often replaced by bowlingite and/or serpentine), augite and labradorite which form decussate, ophitic or subophitic textures. Iron ores and apatite are also present, and interstices are filled with quartz, alkali feldspar, biotite and/or chlorite. Phenocrysts of zoned plagioclase and minor augite or olivine occur and some of the plagioclase carries inclusions of brown material that may be glass. A few dykes contain coarse quartz patches.

The dykes thus grouped as 'olivine-dolerites' appear both petrographically and chemically to be somewhat varied, and several distinct magma-types may be present. Petrographically, both olivine-tholeiites, with pale brown augite and interstitial quartz, and alkali olivine-basalts, with titanaugite and interstitial biotite can be identified. Analysed dykes from the Barra swarm show significant normative ne ((Table 30), analyses 10, 11), and from chemistry and petrography together, can be identified confidently as alkali olivinebasalts. By contrast, analyses of 15 dykes from the South Harris swarm (Mattey et al., 1977; see (Table 30), analyses 12–15; a further 11 incomplete analyses are also cited) are generally oland hy normative, and all belong to a near-saturated low-potash magma-type that is transitional to a tholeiitic magma-type. This has a distinctively high CaO content, and is known in Skye as the 'Preshal Mhor' type (Esson et al., 1975). Unfortunately, Mattey et al. (1977) give almost no petrographical data, and it is not therefore possible to determine whether the undocumented olivine-dolerites from the Outer Isles are comparable.

Olivine-free dolerite and basalt dykes ('tholeiites')

These have been conclusively identified only in South Harris, for example south-west of Ha-cleit summit [NG 032 872], at Eilean Quidnish [NG 095 861], and on the west side of Aird Harnasaig, Loch Grosebay [NG 157 925]. They often show chilled margins and slight shearing, and some also show star jointing (Witty, 1975).

They are composed essentially of pale augite and plagioclase, with interstitial brown glass. In some, plagioclase phenocrysts form an ophitic texture with the groundmass. Large amygdales, 3–4 cm wide, are filled with brown glass, chlorophæite (Jehu and Craig, 1927, p.485), or with zeolites such as gismondine, thompsonite and chabazite (Witty, 1975). The zeolite assemblage is consistent with the chabazite zone found by Walker (1960) in Iceland, corresponding to depths of burial of 200–800 m. In conjunction with the abundance of amygdales, it suggests a very high level of emplacement for these dykes.

Chapter 17 Post-Lewisian sedimentary rocks

The Stornoway Formation

Around the town of Stornoway lie outcrops of a distinctive red-brown undeformed conglomerate sequence (see Map 1) which constitutes the Stornoway Formation. These beds lie unconformably on the Lewisian gneiss but are fault-bounded at their western margin. The beds form prominent coastal exposures particularly on the promontories north-east of Stornoway and in the Aignish-Garrabost and Swordale areas of the Eye Peninsula, where cliffs up to 10 m high are commonly found. The beds underlie a total land area of about 50 km2 and consist of massive coarse conglomerate with subsidiary sandstones, cornstones and rare silt and clay horizons. The formation's apparent thickness is nearly 4 km, which makes the Stornoway Formation one of the thickest conglomerates in the British Isles (Steel and Wilson, 1975). The beds are inclined at 15–45° to the north-west. The formation is generally little altered except near some parts of the western bounding faults, and shows no major internal faulting in the exposed sections.

Macculloch (1819) first noted the presence of these rocks and concluded with some misgivings that they were analogous to the Old Red Sandstone, to which age the Torridonian sandstones and conglomerates of the Scottish mainland were originally assigned. Nicol (1844) and Murchison and Geikie (1861) both accepted this correlation with the Torridonian sequence but Murchison and Geikie attributed both sets of rocks to the 'Cambrian'. Morrison (1885) however, concluded that 'the Lewis and Torridon conglomerates have neither a community of origin nor a contemporaneity of formation'. Stevens (1914) compared the conglomerates with rocks of the Triassic sequence of northwest Scotland, an idea first mentioned by Murchison and Geikie (1861).

Steavenson (1928) carried out a comprehensive survey of the conglomerates in the early 1920s and concluded that they were Torridonian in aspect. Jehu and Craig (1934) emphasised the facies variations in the Stornoway Formation and noted the lack of disturbance in the conglomerate adjacent to the bounding faults. Unlike Peach and Home (1930), who doubted the correlation of the Stornoway Formation with the Torridonian, Jehu and Craig (1934) equated the two sets of rocks on the grounds that they had similar environments of deposition. A thick Mesozoic sequence was subsequently discovered in the Minch (see Binns et al., 1974; Smythe et al., 1972), and Steel (1971) related the Stornoway Formation to this sequence, inferring that it is Triassic in age. Steel and Wilson (1975) described the Stornoway Formation (formerly termed Stornoway Beds) in detail and related it to the development of the North Minch basin and particularly to the Minch Fault (Steel, 1977). Storetvedt and Steel (1977) carried out palaeomagnetic studies on alluvial flood plain deposits from the lower parts of the sequence. They obtained reverse and normally magnet ised polarities, and calculated a pole position of 138°E 50°N (uppermost Permian or Trias). The following account is largely taken from Steel and Wilson (1975) and Steavenson 1928).

Contact relationships

The conglomerates dip to the west-north-west except north and east of Coll Sands, where the dip direction is nearer north-north-west. The amount of dip varies from 40° near the base to 10–15° adjacent to the western margin. The base of the sequence is exposed on the Eye Peninsula where it rests unconformably on an irregular, locally steep, northwest-dipping surface, in parts infilling hollows and cracks in the underlying Lewisian gneiss. Steavenson (1928) records that north of Garrabost at about 150 m north-east of the Dun Mor Broch [NB 516 340], and also 100 m east of the mouth of Allt na Muilne [NB 519 342], cakes of red conglomerate 30 m long by 7–9 m thick 'adhere' to cliffs of Lewisian gneiss.

The western margin of the Stornoway Formation is an intersecting series of five normal faults, two of which trend east-north-east and dip 45° to the south-south-east, and two trend approximately north–south and dip 45° to 60° to the east (see Map 1). The east-north-east-trending faults which bound much of the outcrop in the north are exposed in the Coll and Gress rivers, and in two small burns near the coast. The north-trending faults are exposed again in the Coll river section, near Stornoway, and at Arnish. Steavenson (1928) records that the conglomerate is generally shattered for 1 m adjacent to the faults (locally up to 3 m), and reddish gouge and minor fracturing occur in places for some distance away from the fault. The fault surface itself is typically a smooth plane of fracture. The gneiss, in contrast, is more extensively fractured and commonly contains clay gouge-filled zones. The fractured gneiss zones range in width from 2 m to 100 m (Steavenson, 1928). Jehu and Craig (1934) explain this contrast by suggesting that much of the fracturing in the gneiss predates the deposition of the conglomerate. However, the two lithologies may have responded differently to the faulting and its associated stress field.

Lithology and origin

The conglomerates consist predominantly of rounded to subrounded clasts of Lewisian gneiss and vein quartz (less than 10 per cent of the whole), up to 20 cm across, with numerous small pebbles and some larger boulders up to 80 cm across. Steavenson (1928) notes that the large boulders tend to define a crude banding but the overall sorting is very poor. Steel and Wilson (1975) have produced detailed vertical sections of the coastal exposures and conclude that the greater part of the sequence represents alluvial-fan deposits laid down at the foot of escarpments along active normal faults which lay to the west and north. They recognise a Lower, Middle and 'Upper Unit, in all composed of six major fan sequences interdigitated with much thinner fluvial deposits; each fan sequence is composed of conglomerate and minor sandstone (see (Figure 38)). Fine sandstones, silts and clays of fluvial origin occur on the east side of Stornoway Harbour [NB 440315], between Melbost and Aignish [NB 473 331] and near Sandwick (Steavenson, 1928). In these beds, poorly developed, immature cornstones as well as possible root-bearing or burrowed horizons have been recognised. Steel (1974) compared the cornstones to caliche soils in areas of subtropical climate. Steel and Wilson (1975) infer that the alluvial fan deposits are lenticular and during deposition were controlled by an underlying sequence of 'trapdoor' faults. An effect of such faulting would be to restrict the present thickness of the Stornoway Formation at any one point to 1 to 2 km. Steel and Wilson recognise sediments of differing modes of origin within the alluvial fan deposits and these are summarised as follows:

  1. Debris flow or mudflow deposits Sheet-like deposits of poorly sorted conglomerate commonly with a silt-grade matrix and rarely overlain by sandstone. Clasts usually matrix-supported; rare pebble/cobble frameworks. Cross-beds (low planar) in the sandstones.
  2. Streamflood deposits Lenticular, often channel-fill, poorly sorted conglomerates generally overlain by sandstones. Pebble and cobble frameworks. Conglomerates commonly cross-stratified (sets up to 1.5 m thick) and sandstones both cross-bedded and flat-bedded. Marked basal erosion surface.
  3. Braided stream deposits Small-scale lenticular channel-fill, finer-grainea, well-sorted conglomerates passing laterally and vertically into sandstones. Thin siltstone lenses locally present. Intraformational clasts found in conglomerates. Concave-up basal erosion surfaces, and trough cross-bedding in sandstones.

Within the Stornoway Formation the alluvial-fan sequences are either upward-coarsening or upward-fining, with flood-plain sandstones, cornstones and minor silts separating the fan deposits in some areas. Upward-fining sequences (debris-flow–stream-flow–braided stream deposits) are interpreted by Steel and Wilson (1975) to reflect source-ward migration of the locus of sedimentation and/or lowering of relief of the escarpments. Upward-coarsening sequences that show a marked upward increase in maximum clast size suggest increasing rates of uplift in the source area (or subsidence in the basin). Studies of the distribution of clast sizes and sequence thickness were made by Steel and Wilson (1975).

Details of the individual units of the Stornoway Formation and inferred fan radii from the above studies are summarised in (Table 36). The data imply that the locus of alluvial-fan sedimentation migrated westward as the fault systems associated with the Minch fault developed in Permo-Triassic times (Figure 38). The three distinct units (Lower, Middle and Upper) reflect the differing positions of the faults which controlled sedimentation patterns. The relatively thin floodplain deposits, which overlie and in places interdigitate with the fan deposits, mark periods when the alluvial plain of the North Minch Basin, which was drained by rivers flowing to the north-east, encroached on and partly buried the fan aprons along the basin margins.

Jurassic rocks of the Shiant Isles

The Shiant Isles lie in the Minch, 6.5 km south-east of Lewis and 20 km north of the northern tip of Skye. They are largely made up of three thick Tertiary picrite-crinanite-teschenite sills (described in Chapter 16: Crinanite-teschenite sills). Thin beds of baked Jurassic shale are interbedded with the sills on the three main islands (Map 1). On Garbh Eilean they comprise 9 m of hornfelsed calcareous shale, exposed at sea level on the north-east coast beneath the basal picrite of the main (upper) sill (Table 33). On Eilean an Tighe about 5 m of sediment is exposed on the foreshore near the northwest corner of the island. It is hornfelsed to a grey flinty porcellanous rock with a conchoidal fracture, and appears to be part of a raft of sediment caught up within the upper part of the upper sill. On Eilean Mhuire about 19 m of horizontally bedded shale forms a large outcrop on the higher parts of the island. Here only the basal sediment is hornfelsed. This shale overlies the lowest exposed sill on this island (Lower sill of (Table 33)) and is in turn overlain by a crinanite sill which resembles the sill beneath the sediments of Garbh Eilean (middle sill of (Table 33)).

The Jurassic sediments were discovered by Macculloch (1819), and briefly described by Judd (1878), who recorded Ammonites murchisonae and A. corrugatus, probably from Garbh Eilean, as well as hollow casts of belemnites, first noted by Macculloch. Judd assigned the sediments to the lowest part of the Inferior Oolite. Walker's account (1930) of the Shiant Isles contains a detailed description of the sediments and includes a comprehensive list of fossils from Garbh Eilean and Eilean Mhuire. The ammonites were thought by Walker to indicate a Whitbian (Lower Toarcian) age, which suggested a correlation with the Portree Shales of Skye. Further fossils collected by C. B. Wedd for the Geological Survey were recorded by Lee and Pringle (1932), who concurred with the Lower Jurassic age of the sediments, though the collection included an ammonite recorded by Buckman as the Aalenian species Ludwigella cf. attenuata (Buckman).

Recently Penn and Merriman (1978) described the petrology of the hornfelsed sediment and re-examined all the existing fossils from the islands, with a view to establishing the exact age range of the beds. They have shown that the strata are of the same sedimentary facies throughout, having originally been marls with some dolomite. Thermal metamorphism by the sills has altered some of the sediment into a hornfels of pyroxene-hornfels grade, with calcic plagioclase, diopside, and small amounts of quartz and pyrrhotite. Most of the hornfels was subsequently affected by retrograde metasomatism, which produced a diopside-zeolite-potashfeldspar hornfels. The less indurated sediments now consist largely of zeolite (?epistilbite) with subsidiary biotite and chlorite.

The thermal metamorphism has obliterated all traces of calcareous microfauna and organic-walled microplankton. Ammonites, however, are common and their preservation has been enhanced by the development of reaction rims around them. Penn and Merriman (1978) have shown that the ammonites from the lowest strata of Eilean Mhuire are typical of the Harpoceras falciferum zone (lower part of Toarcian). The highest beds from this island are also most likely of H. falciferum age, though fossils collected by C. B. Wedd from the 'top shales' contain a hildoceratacean ammonite resembling Graphoceras, which is indicative of the Graphoceras concavum zone of the Aalenian: The age of these beds may thus range from the H. falciferum zone into the basal Middle Jurassic. The age of the Garbh Eilean strata is still uncertain. Walker (1930) recorded Eleganticeras [Elegantuliceras], which suggests a Toarcian age, but Judd's (1878) record of Ammonites murchisonae [Ludwigia murchisonae] and A. corrugatus would suggest an Aalenian or even Lower Bajocian age.

No systematic collecting with accurate stratigraphical localisation has yet been attempted on these islands, and until that is done it is not possible to ascertain if the zones between the H. falciferum zone and G. concavum zone are present. If these zones are represented there is a marked attenuation of the sequence from the 47 m (mainly 011ach Sandstone, Dun Caan Shales and Raasay Ironstone) in northern Skye (Anderson and Dunham, 1966) to 19 m in the Shiant Islands. No beds resembling the Raasay Ironstone have been recorded on the islands.

Chapter 18 Summary of pre-Caledonian geological history

The Lewisian Gneiss complex of the Outer Hebrides and adjacent isles is a fragment of the ancient North Atlantic craton. As such its geological development bears many similarities to that of Greenland and north-eastern Canada. The main geological events in the Outer Hebrides are given in (Table 1) and summarised as follows.

1 The supracrustal sequence (?2900 Ma)

The earliest rocks represented in the Outer Hebrides appear to be the sequence of metasedimentary and metavolcanic rocks. Remnants of these rocks are found throughout the islands, but they reach their greatest and most varied development in the Leverburgh and Langavat belts flanking the South Harris Igneous Complex. The extent of the rocks throughout the Outer Hebrides is masked by the long history of metamorphism and deformation. The more exotic lithologies can everywhere be relatively easily identified, but there is an extensive series of flaggy quartzitic rocks and finely striped hornblendic rocks whose parentage and association is uncertain. Outside South Harris the most commonly found meta-sedimentary type is garnet-biotite schist, in places with kyanite-, sillimanite- or orthoamphibole-bearing assemblages. In South Harris a more varied assemblage is found, including quartzites, graphitic schists, marbles and finely banded amphibolites. What the supracrustals originally were is uncertain, but they probably represent a series of shales, sandstones, black shales and limestones with associated tuffs and volcanics.

Areally associated with this supracrustal sequence, and probably broadly coeval with it, there is a suite of characteristically banded basics. These rocks are particularly abundant in the southern isles. They show strong compositional banding on a centimetre to metre scale. The composition ranges from ultramafic to felsic. Although the rocks now exhibit only metamorphic textures and minerals (garnethornblende-plagioclase(-clinopyroxene)) it is believed that the banding relates to an original igneous layering. Chemically the suite is Q-normative with a strongly iron-enriched trend similar to modern tholeiites. In the northern isles rocks of this type are uncommon, although the anorthosite at Ness may well be part of the suite.

The age of the supracrustal sequence and associated basics is difficult to assess; similar sequences in Greenland are c.2900 Ma old.

2 The early-Scourian event

There is some evidence that the supracrustal sequence was deformed (dS1) and foliated before the intrusion of the igneous sequence described next. Whether such deformation related to the beginnings of the deep burial of the supracrustals is uncertain.

3 The igneous sequence

The major part of the gneiss now forming the Lewisian complex appears to have been derived from igneous parents, the bulk of which are of granodioritic or tonalitic composition. The igneous rocks are believed to have formed an extensive and complex series of intrusions into the supracrustal pile, the metasediments being largely swamped and surviving as only subsidiary relics. The crustal depth of the rocks that were to become the gneiss at this time is debatable; parallels in Greenland suggest that the crystallisation took place deep in the crust.

A general igneous event appears to have been widespread throughout the North Atlantic craton in the period 2800–2900 Ma. Its cause is a matter for speculation.

4 The main Scourian event

The peak of Scourian activity probably occurred at c.2700 Ma. The metamorphism and deformation (dS2) effectively converted the original rocks to a gneiss and produced the gneissose fabric which, although locally much modified, still dominates the rocks. Although this dominant fabric was formed during the main metamorphism, it seems that no major structures were formed at that time. The rocks became heavily veined and migmatised, with the associated production of basic agmatites, granitic gneiss, etc. The process of gneissification or migmatisation appears to have been polyphasal; there is evidence of 'fits' having been intruded, deformed and intruded by later veins. This process did much to mask the original rock types and their relationships, although it may be presumed that the hornblende-biotite-gneiss was formed from broadly granodioritic or tonalitic parents and the biotite-gneiss from granites. The degree of migmatisation appears to have varied from area to area, with perhaps a general increase towards the north and west where the gneiss is richer in potash, rubidium, etc., then elsewhere.

The Leverburgh Belt metasediments show evidence of granulite-facies metamorphism, possibly of Scourian age, with an associated depletion in lighter elements. Elsewhere the Scourian grade of metamorphism is uncertain; granulitefacies rocks are preserved in the east side of South Uist and Barra, but in other areas no relics have been found. This absence of granulite-facies relics, coupled with the general lack of chemical depletion in the north and west, would appear to indicate that the intensity of metamorphism was never higher than the upper amphibolite facies. The intermediate chemical composition (in particular, the depletion of uranium) of the gneisses in South Uist might indicate an area of granulite or subgranulite facies which has been subsequently downgraded. These differences in metamorphic intensity might have arisen during the Scourian, like the regional variation in gneiss chemistry; if so, the features define a NNE-trending Scourian grain to the Hebrides. Perhaps present-day exposures represent an oblique section through the Scourian crust, those to the south and east being deeper levels.

It is obvious from the high metamorphic grade of the gneisses, as well as from evidence for the pressures and temperatures at the time of Younger Basic intrusion (see below) that the Scourian metamorphism took place in the crust (perhaps at about 30 km depth). However, nothing is known about the overall geometry of ds2 regional structures, and the burial mechanism thus remains highly uncertain. Although modern analogues suggest subduction as the most likely method for rocks to be transported to deep crustal levels, the apparent widespread deep burial which occurred within the North Atlantic craton at this time may suggest some other mechanism.

5 The Late-Scourian intrusives

The end of the main Scourian metamorphism was marked by the emplacement of an extensive suite of minor intrusives. These fall into two groups: first, a series of diorites, monzodiorites and micro-diorites, present throughout the islands but best preserved in Barra; and second, a series of potash-rich granites, monzonites and pegmatites, also best preserved in Barra in areas of low post-Scourian deformation. Elsewhere rocks of the second group are normally rather homogenous biotite gneisses, augen granite, granite-gneiss etc.

These minor intrusions have been dated at c.2600 Ma, and are believed to be part of a major suite found throughout the North Atlantic craton.

6 The Late-Scourian deformation

The Scourian rocks of the Outer Hebrides were probably extensively deformed at this time, although the deformation has been identified in only few localities. There were two major phases of deformation. The first phase (dS3) is poorly documented but may have resulted in the formation of regional-scale asymmetric folds, the long limbs of which imparted a general north-north-east trend to the foliation. The second phase (d„) has been more widely recognised and during it a number of major shear zones developed. Characteristically these have a north-west trend, although a north-east-trending zone has been identified in west Lewis. These shear zones are believed to have been formed at much the same time as the Younger Basics were intruded (see below) and as such are broadly equivalent to the Inverian event of the mainland.

In Barra and South Uist the Late-Scourian microdiorites show evidence of extensive deformation which largely predates the potash-granites; the deformation is therefore of broadly Late-Scourian age and presumably part of the cls, or cl„ events.

7 The Younger Basic intrusives

The Younger Basic intrusives are believed to be broadly equivalent to the Scourie dykes of the mainland, and as such to be part of the great Proterozoic dyke swarm found across the North Atlantic craton and intruded in the period 2200-2400 Ma. The suite consists of picrites, norites and dolerites, the dolerites being by far. the most abundant. Chemically, the suite shows a number of magmatic lineages but is generally Q-normative with a strongly iron-enriched trend, similar to continental tholeiites. Where little affected by later deformation the swarm shows a general west to north-north-west trend. In common with their equivalents elsewhere, the Younger Basic dykes show evidence of having been intruded into gneisses that contained tectonically active shear zones, although the Hebridean dykes show fewer features typical of synkinematic intrusion than those of the mainland and as such may have postdated the shearing to a greater extent. Fabrics associated with early deformation of the dykes have been ascribed to dLl on the strict definitions of Scourian and Laxfordian metamorphic and deformational events. It is possible, however, dLl may be largely equivalent to the later effects of ds4. The dykes show a variety of metamorphic textures and assemblages. These have been interpreted as indicating intrusion into hot crust with an immediate crystallisation of metamorphic minerals in partial equilibrium with the country rocks. Although the cores of the larger dykes have granulite-facies assemblages, it is believed that the prevailing conditions were consistent with crystallisation in the upper amphibolite facies, the dyke magmas being relatively dry. Pressure and temperature at the time of intrusion have been estimated as 700°C and 600–700 MPa (6-7 kb, equivalent to crustal depths of about 25 km).

Associated in time with the Younger Basics is a series of ultrabasic pods. Most are peridotites with subsidiary dunites and pyroxenites. Some of the bodies show evidence of compositional banding.

8 The South Harris Igneous Complex

The South Harris Igneous Complex comprises four major igneous bodies and numerous related minor dykes and pods. These were intruded with regional discordance into the metasediments of the Leverburgh and Langavat Belts at c.2250–2000 Ma. Field relationships, such as the occurrence of angular gabbro xenoliths in hybridised diorite and of an anorthosite xenolith in the norite, indicate that the order of intrusion of the major bodies was gabbro (oldest), anorthosite, norite, and diorite. Although the mineralogy of the complex is almost wholly metamorphic the geochemistry still defines marked igneous trends. Cliff et al. (1983) used Nd-Sm studies to show that the age of anorthosite intrusion was around 2200 Ma.

The igneous suite is thought to have resulted from magmas generated progressively deeper with time from one or more parent magmas of olivine-tholeiite composition. The early gabbros and anorthosite represent gravity-stratified layered basics, and the later diorite and norite plutons are characterised by calcalkaline trends. A small body of tonalite within the diorite is a product of filter pressing of the residual parts of the diorite magma. The last rocks of the complex to be intruded were thin (2–50 cm) shoshonite dykes. These dykes are the only part of the igneous suite to retain an igneous mineralogy. The other bodies have all been metamorphosed to granulite facies, as a result, it is believed, of recrystallisation deep in the crust both during and after intrusion until about 1840 Ma.

The granulite-facies metamorphism of the igneous suite resulted in a local redistribution of CaO from plagioclase to orthopyroxene, clinopyroxene and garnet thus giving the igneous bodies a more mafic appearance in the field. The mineral assemblages within the bodies document a history of uplift. The first metamorphism overlapped intrusion at lower crustal levels of around 35 km (1300 MPa and 825°C) and subsequent recrystallisations indicate progressively lower pressures and temperatures.

9 The main Laxfordian event

The peak of Laxfordian was probably reached at about 1700 Ma. There were two main phases of deformation (dL2 and dL3) both largely imprinting or strengthening the north-west foliation trends that are seen today. The dL2 deformation typically produced folds with flat-lying or gently inclined axial planes and was largely responsible for the rolling foliation orientations of central Lewis. The dL3 deformation formed markedly asymmetric folds on a regional scale with broad hinge zones and attenuated vertical limbs. Such folds predominate in the southern isles. The amount of Laxfordian deformation can be quantified relative to the degree of deformation and recrystallisation exhibited by the Younger Basics. Maps compiled using these criteria show zones of low deformation in the broad antiformal hinges of dL3 folds with high deformation on the limbs. This pattern reflects and was controlled in part by the structures formed during dS3 and dS4.

During dL3 and dL4 the gneiss was recrystallised, mainly in the amphibolite facies. In some areas, particularly those exhibiting intense Scourian migmatisation, the gneiss shows evidence of remobilisation, with localised loss of the gneissose foliation and the development of coarse pegmatitic patches. These features are particularly prevalent in the sheared-out limbs of minor dL3 folds. In areas of extensive remobilisation the Younger Basics have been absorbed and recrystallised in places; coarse patches of hornblende-gneiss have formed in extreme cases.

The relatively anhydrous igneous bodies of the South Harris Igneous Complex have typically been deformed by the development of shear zones, particularly at the margins of the major plutons. However, the main igneous sequence in the Complex has been inverted, and major folds such as the Roineabhal antiform show that the area has undergone major tectonic modification.

There is little evidence for the depths of the gneiss within the crust during the main Laxfordian event. Although there may well have been a gradual rise in level during the Laxfordian the general indications of depth during the dL2–dL3 phases of Laxfordian activity are of still relatively deep burial.

10 The Late-Laxfordian granites

Late-Laxfordian granites are found at several localities; by far their greatest development is in western Harris and Lewis, where they form the Uig Hills Complex. In this area the granite forms a complex network of veins, lenses and sheets ranging in thickness from centimetres to hundreds of metres. The granites contain one variety of mica (biotite) as a major constituent. Three types of granite can be distinguished in the field: leucogranite, granite and porphyritic granite. The leucogranite was generally emplaced late in the intrusive sequence and predominates at the margins of the complex. The porphyritic granite was generally emplaced early in the sequence and is only found in the centre of the complex. The granites and porphyritic granites are geochemically indistinguishable. Although the Late-Laxfordian granites are always found near areas of remobilised gneiss, their chemistry indicates that they were not derived by partial melting of the country rock.

11 Late-Laxfordian deformation

The later phases of granitic injection are closely associated with Late-Laxfordian deformation. There is abundant evidence in the granite veins to show that before they were fully consolidated the region was subjected to strain. Although the resultant deformation was largely localised in the granite bodies, the gneisses also show evidence of widespread brittle and semi-ductile deformation. These movements were most probably associated with rapid uplift of the gneisses and were the forerunners of the Outer Hebrides thrust and associated movements (see below).

12 Thrusting

The Late-Laxfordian deformation in the granite veins caused mylonites to develop in the Uig area, but further east ultramylonite and then pseudotachylite are typically found. This deformation culminated in the major thrust zone that runs down the eastern seaboard of the Outer Hebrides. In Lewis it consists of lenticular zones dipping gently to the east marked by basal pseudotachylite and cataclasite and up to 160 m of 'mashed gneiss'. The intervening Lewisian gneiss has a superimposed 'cataclastic' fabric. Later east and south-east dipping zones of lower greenschist grade mylonites overprint these rocks. These mylonite zones relate to a zone of hydrous alteration; both mylonite and alteration zones extend south onshore as far as South Uist. Southward from Lewis the thrust zone is narrower, and the thrusts are marked by narrow zones of cataclasite and pseudotachylite which form prominent high topographical features in the Uists and Barra.

The Outer Hebrides Thrust Zone is notable for containing pseudotachylite, a black glassy rock resulting from frictional melting of the adjacent gneiss during 'dry' thrust faulting. Pseudotachylite was also developed west of the thrust front, commonly at the site of small thrusts in relatively anhydrous gneisses.

Later mylonites are green platy finely-banded rocks with a pervasive schistosity, and strictly should be termed 'phyllonites'.

Two lines of evidence–the relationships of the Late-Laxfordian granites to the thrusting, and mineralogical evidence from Seaforth Head (which indicates that the thrusting occurred under lower amphibolite-facies conditions)–both suggest that the major movement on the thrust zone took place during the Late Laxfordian. This coincided with a period of major crustal uplift. The thrust zone was reactivated in the later part of the Caledonian orogeny and mylonite was formed. At least part of this later movement was extensional.

Although it is possible to document the main steps in the geological development of the Lewisian, there is little indication of the mechanisms controlling the tectonics. It is clear that either before or during the Scourian the rocks were taken to depths of at least 20 km in the crust. They appear to have remained at these depths for a considerable period of time. Uplift probably commenced only early in the Laxfordian and even then was probably not substantial until later at the time of major granite intrusion. The South Harris Complex shows evidence of crystallisation at very great depths (38 km) and it is possible that the complex was only juxtaposed with the gneisses during the Laxfordian.

It is tempting to interpret the record in the Lewisian gneiss in terms of modern tectonic processes, such as subduction and underplating, but the available evidence is not sufficient to warrant such an approach.

Appendix Mineral analyses

(Appendix Table 1) Amphiboles

1 2
SiO2 43.14 42.15
TiO2 1.12 1.82
Al2O3 11.13 10.82
FeO 20.00 18.55
MnO 0.22 0.12
MgO 7.46 8.96
CaO 11.50 11.51
Na2O 1.41 1.78
K2O 0.76 0.78
Total 96.74 96.49
Si 6.59 6.42
Al 2.01 1.96
Ti 0.13 0.21
Fe3* 0.24 0.35
Fee* 2.32 2.01
Mn 0.03 0.02
Mg 1.70 2.04
Ca 1.88 1.88
Na 0.42 0.53
K 0.15 0.15
Ca + Na + K 2.46 2.52
mg 0.40 0.46

(Appendix Table 2) Garnets

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
SiO2 38.25 38.30 31.43 37.20 37.62 38.45 38.53 38.18 38.78 38.84 37.70 37.91 38.22 38.06 38.31 38.15 36.97 35.86 37.46
TiO2 0.02 0.03 0.01 0.03 0.05 0.00 0.03 0.05 0.03 0.02 0.07 0.04 0.03 0.07 0.06 0.07 0.06 0.04 0.02
Al2O3 20.84 21.03 21.13 20.99 20.74 23.08 22.72 22.61 21.12 21.29 20.22 20.11 20.57 20.41 20.97 21.04 21.35 20.99 21.42
FeO* 26.79 24.82 28.02 28.07 28.21 24.74 25.13 26.07 27.37 27.19 29.39 29.49 26.52 26.70 27.60 27.75 27.29 29.10 27.66
MnO 1.88 1.29 1.72 1.80 1.69 1.07 1.33 1.65 1.71 1.77 1.28 1.26 2.26 2.32 0.93 0.96 1.24 1.21 1.22
MgO 3.66 3.59 2.48 2.35 2.47 4.20 3.95 3.69 4.87 4.93 2.85 2.70 4.46 4.40 4.64 4.61 3.90 3.92 3.97
CaO 8.59 10.73 9.40 9.32 9.36 11.58 11.30 10.93 6.72 6.79 7.37 7.42 7.00 6.95 7.42 7.11 8.51 7.10 7.88
C r2O3 0.02 0.02 0.00 0.00 0.02 0.02 0.02 0.04 0.00 0.06 0.01 0.00 0.00 0.01 0.01 0.00 0.00 0.02 0.00
Total 100.05 99.81 100.19 99.76 100.16 103.14 103.01 103.22 100.60 100.89 98.89 98.93 99.06 98.92 99.94 99.69 99.32 98.24 99.63
Fe2O3* 0.79 0.68 1.04 1.09 1.22 0.72 0.77 1.22 0.48 0.39 0.71 0.60 0.62 0.83 0.74 0.59 1.24 2.01 0.74
FeO 26.08 24.20 27.08 27.09 27.12 24.09 24.43 24.97 26.94 26.84 28.75 28.95 25.96 25.96 26.93 27.22 26.17 27.29 26.99
Si 6.026 6.016 5.938 5.934 5.972 5.831 5.863 5.824 6.047 6.036 6.059 6.091 6.062 6.052
Al 0.000 0.000 0.071 0.075 0.032 0.193 0.156 0.200 0.000 0.000 0.000 0.000 0.000 0.000 6.017

0.000

 6.010

0.000

 5.874

0.143

 5.796

0.232

 5.929

0.080
Al 3.869 3.893 3.880 3.872 3.848 3.932 3.918 3.865 3.881 3.900 3.830 3.808 3.845 3.825 3.882 3.907 3.855 3.766 3.916
Fe3+* 0.094 0.081 0.124 0.131 0.145 0.082 0.089 0.140 0.057 0.045 0.086 0.072 0.074 0.099 0.087 0.070 0.149 0.245 0.089
0.002 0.004 0.001 0.004 0.006 0.000 0.003 0.006 0.004 0.002 0.008 0.005 0.004 0.008 0.007 0.008 0.007 0.005 0.002
Cr 0.002 0.002 0.000 0.000 0.003 0.002 0.002 0.005 0.000 0.007 0.001 0.000 0.000 0.001 0.001 0.000 0.000 0.003 0.000
Fee2+* 3.436 3.179 3.593 3.614 3.600 3.056 3.109 3.185 3.512 3.489 3.864 3.891 3.443 3.452 3.538 3.586 3.478 3.689 3.573
Mn 0.251 0.172 0.231 0.243 0.227 0.137 0.171 0.213 0.226 0.233 0.174 0.171 0.304 0.312 0.124 0.128 0.167 0.166 0.164
Mg 0.860 0.841 0.587 0.559 0.584 0.950 0.896 0.839 1.132 1.142 0.683 0.647 1.055 1.043 1.086 1.083 0.924 0.945 0.937
Ca 1.450 1.806 1.598 1.593 1.592 1.882 1.842 1.786 1.123 1.131 1.269 1.277 1.189 1.184 1.249 1.200 1.449 1.230 1.336
Almandine 57.3 53.0 59.8 60.1 60.0 50.7 51.7 52.9 58.6 58.2 64.5 65.0 57.5 57.6 59.0 59.9 57.8 61.2 59.5
Pyrope 14.3 14.0 9.8 9.3 9.7 15.8 14.9 13.9 1.6 1.3 2.6 2.1 2.0 2.9 18.1 18.1 15.4 15.7 15.6
Grossular 21.7 27.8 23.4 23.1 22.5 29.1 28.2 25.8 17.1 17.4 18.6 19.3 17.8 16.8 18.3 17.8 20.0 14.0 19.9
Andradite 2.5 2.2 3.2 3.4 3.9 2.0 2.4 3.8 18.9 19.1 11.4 10.8 17.6 17.4 2.5 2.2 4.1 6.3 2.3
Spessartine 4.2 2.9 3.8 4.0 3.8 2.3 2.8 3.5 3.8 3.9 2.9 2.9 5.1 5.2 2.1 2.1 2.8 2.7 2.7
Uvarovite 0.1 0.1 0.0 0.0 0.1 0.1 0.1 0.1 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0

(Appendix Table 3) Pyroxenes from the Younger Basics

1 2 3 4 5 6 7 8 9 10 11 12 13 14
SiO2 50.63 50.55 53.05 53.09 52.40 52.35 51.46 50.88 50.54 52.15 51.00 51.47 50.26 51.82
TiO2 0.19 0.26 0.06 0.06 0.13 0.08 0.18 0.19 0.07 0.32 0.08 0.16 0.07 0.20
A12O3 1.72 2.55 1.83 1.79 2.30 1.05 1.71 1.60 1.13 1.18 0.99 1.76 1.25 1.84
FeO 15.08 11.50 22.75 22.81 22.15 11.68 12.50 11.94 30.10 9.58 28.53 12.61 30.85 13.59
MnO 0.22 0.34 0.47 0.48 0.49 0.19 0.27 0.28 0.66 0.10 0.35 0.30 0.77 0.25
MgO 9.40 11.06 22.02 21.48 21.40 11.41 11.11 11.36 15.74 12.74 17.88 11.28 15.59 11.48
CaO 21.15 21.45 0.31 0.37 1.49 21.94 21.46 21.61 0.54 22.68 0.35 21.54 0.56 21.56
Na2O 0.59 0.58 0.00 0.05 0.04 0.56 0.42 0.50 0.00 0.34 0.03 0.41 0.01 0.49
K2O 0.06 0.05 0.03 0.04 0.03 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.07
Cr2O3 0.02 0.01 0.18 0.16 0.20 0.02 0.00 0.00 0.04 0.01 0.02 0.04 0.01 0.07
Total 99.06 98.35 100.70 100.33 100.63 99.28 99.11 98.36 98.83 99.10 99.23 99.57 99.37 101.37
Si 1.97 1.94 1.96 1.97 1.94 1.99 1.97 1.96 1.98 1.97 1.97 1.96 1.97 1.95
Al 0.08 0.12 0.08 0.08 0.10 0.05 0.08 0.07 0.05 0.05 0.05 0.08 0.06 0.08
Ti 0.01 0.01 0.00 0.00 0.00 0.00 0.01 0.01 0.00 0.01 0.00 0.00 0.00 0.01
Fe 0.49 0.37 0.70 0.71 0.69 0.37 0.40 0.38 0.99 0.30 0.92 0.40 1.01 0.43
Mn 0.01 0.01 0.01 0.01 0.02 0.01 0.01 0.01 0.02 0.00 0.01 0.01 0.03 0.01
Mg 0.54 0.63 1.21 1.19 1.18 0.65 0.64 0.65 0.92 0.72 1.03 0.64 0.91 0.64
Ca 0.88 0.88 0.01 0.01 0.06 0.89 0.87 0.89 0.02 0.92 0.01 0.88 0.02 0.87
Na 0.04 0.04 0.00 0.01 0.00 0.04 0.03 0.04 0.00 0.02 0.00 0.03 0.00 0.04
K 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Cr 0.00 0.00 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Formulae based on 6 [0]
En 28.5 33.6 62.9 62.2 61.3 33.8 33.1 33.8 47.7 37.0 52.4 33.3 46.8 33.2
Fs 25.6 19.6 36.5 37.1 35.6 19.4 20.9 19.9 51.1 15.6 46.9 20.9 52.0 22.0
Wo 46 46.8 0.6 0.8 3.1 46.7 46.0 46.2 1.2 47.4 0.7 47.8 1.2 44.8
1 2 3 4 5 6 7 8 9
SiO2 50.79 48.57 51.34 48.33 51.43 48.70 48.60 49.64 45.11
A12O3 26.44 26.90 25.91 26.69 27.42 26.70 27.68 26.35 26.41
FeO 0.09 0.13 0.16 - 0.37 0.49 - -
MgO 0.03 0.05 - - - - - -
CaO 9.79 13.21 12.17 16.91 10.73 15.15 13.41 13.75 17.51
Na2O 7.31 6.28 7.24 3.95 7.27 5.30 6.11 5.63 3.46
K2O 0.66 0.40 0.29 0.27 0.50 0.22 0.40 0.69 -
SO3 - 0.01 - - - - - - 3.42
Cl 1.00 0.67 1.35 0.72 0.34 0.72 0.69 0.71 0.18
Total 96.11 96.88 98.46 96.87 98.06 96.79 97.38 96.77 96.09
Si 7.437

}12

 7.260

}12

 7.524

}12

 7.279

}12

 7.369

}12

 7.290

}12

 7.180

}12

 7.382

}12

 7.142

}12
Al 4.563 4.740 4.476 4.731 4.631 4.710 4.820 4.618 4.858
Fe2+* 0.011

}3.752

 0.016

}4.038

 0.020

}4.042

 -

}3.929

 0.044

}3.748

 -

}4.010

 0.061

}4.009

 -

}3.945

 0.049

}4.058
Mg 0.007 0.011 - - - - - -
Ca 1.536 2.116 1.911 2.725 1.592 2.430 2.123 2.191 2.977
Na 2.075 1.820 2.057 1.152 2.020 1.538 1.750 1.623 0.989
K 0.123 0.075 0.054 0.052 0.091 0.042 0.075 0.131 0.043
S - - - - - - - - 0.413
Cl 0.248 0.170 0.335 0.184 0.083 0.183 0.173 0.179 0.068
Meionite % 41.5 53.1 47.8 69.4 43.7 60.6 54.5 55.5 74.6

References

Most of the references listed below are held in the Library of the British Geological Survey at Keyworth, Nottingham. Copies of the references can be purchased subject to the current copyright legislation.

ANDERSON, F. W., and DUNHAM, K. C. 1966. The geology of northern Skye. Mem. Geol. Sure. GB, Sheets 80 and 90 (Scotland).

ATHERTON, M. P. 1977. The metamorphism of the Dalradian rocks of Scotland. Scott. J. Geol., Vol. 13, Part 4, 331-370.

BAILEY, E. B., CLOUGH, C. T., WRIGHT, W. B., RICHEY, J. E., and WILSON, G. V. 1924. Tertiary and post-tertiary geology of Mull, Loch Aline and Oban. Mem. Geol. Sure. GB.

BAK, J., SORENSEN, K., GROCOTT, J., KORSTGARD, J. A., NASH, D., and WATTERSON, J. 1975. Tectonic implications of Precambrian shear belts in western Greenland. Nature, London, Vol. 254, 566-569.

BAXTER, A. N. 1987. Petrochemistry of late Palaeozoic alkali lamprophyre dykes from N Scotland. Trans. R. Soc. Edinburgh: Earth Sciences, Vol. 77, 267-277.

BECKINSALE, R. D., and OBRADOVICH, J. D. 1973. Potassium-argon ages for minerals from the Ross of Mull, Argyllshire. Scott. J. Geol., Vol. 9, 147-156.

BEER, K. E. 1952. Report on the radioactivity of the gneisses of Northern Lewis, Outer Hebrides. Geol. Sure. GB, Rep. At. Energy Div. No. 125.

BHATTACHARYYA, C. 1971. An evaluation of the chemical differences between igneous and metamorphic orthopyroxenes. Am. Mineral., Vol. 56, 499-506.

BINNS, P. E., McQuiwrq, R., and KENOLTY, N. 1974. The geology of the Sea of the Hebrides. Rep. Inst. Geol. Sci., No. 73/4.

BISHOP, E. 1975. A study of the Lewisian rocks of South Harris. Unpublished BSc thesis, University of Leeds.

BOWES, D. R. 1972. Geochemistry of Precambrian crystalline basement rocks, north-west Highlands of Scotland. Proc. 24th Sess. Int. Geol. Congr., Canada 1972, Sect. 1, 97-103.

BOWES, D. R. 1978. Shield formation in early Precambrian times: the Lewisian complex. 39-80 in Crustal evolution in northwestern Britain and adjacent regions. BOWES, D. R., and LEAKE, B. E. (editors). Geol. J. Spec. Issue, No. 10.

BOWES, D. R., WRIGHT, A. E., and PARK, R. G. 1964. Layered intrusive rocks in the Lewisian of the north-west Highlands of Scotland. Q, J. Geol. Soc. London, Vol. 120, 153-192.

BOWES, D. R., and HOPGOOD, A. M. 1969a. The Lewisian gneiss complex of Mingulay, Outer Hebrides, Scotland. Mem. Geol. Soc. Am., No. 115, 317-360.

BOWES, D. R., and HOPGOOD, A. M. 1969b. Composition of a dolerite dyke, Mingulay, Outer Hebrides. Mineral. Mag., Vol. 37, 427-428.

BOWES, D. R., and HOPGOOD, A. M. 1975. Framework of the Precambrian crystalline complex of the Outer Hebrides. Krystalinikum, Vol. 11, 7-23.

BOWIE, S. H. U. 1962. [Report of] Atomic Energy Division. Summary of progress of the Geological Survey of Great Britain for 1961, 66-68. (London: Her Majesty's Stationery Office.)

BOWIE, S. H. U. 1964. [Report of] Atomic Energy Division. Geological Survey of Great Britain. Summary of progress for 1963, 75-76. (London: Her Majesty's Stationery Office.)

BRACE, W. F., and KOHLSTEDT, D. L. 1980. Limits on lithospheric stress imposed by laboratory experiments. J. Geophys. Res., Vol. 85, 6248-6252.

BREEMEN, O. VAN, AFTALION, M., and PIDGEON, R. T. 1972. The age of the granite injection complex of Harris, Outer Hebrides. Scott. J. Geol., Vol. 7, 139-152.

BREWER, J. A., and SMYTHE, D. K. 1984. MOIST and the continuity of crustal reflector geometry along the Caledonian-Appalachian orogen. J. Geol. Soc. London, Vol. 141, 105-120.

BRIDGWATER, D., WATSON, J., and WINDLEY, B. F. 1973a. The Archaean craton of the North Atlantic region. Philos. Trans. R. Soc., A, Vol. 273, 493-512.

ESCHER, A., and WATTERSON, J. 1973b. Dyke swarms and the persistence of major geological boundaries in Greenland. 137-142 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. G., and TARNEY, J. (editors). (Keele: University of Keele.)

MCGREGOR, V. R., and MYERS, J. S. 1974. A horizontal tectonic regime in the Archaean of Greenland and its implications for early crustal thickening. Precambrian Res., Vol. 1, 170 - 197.

KETO, L., MCGREGOR, V. R., and MYERS, J. S. 1976. Archaean gneiss complex of Greenland. 19-75 in Geology of Greenland, ESCHER, A., and WATTS, W. S., (editors). (Copenhagen: Gronlands Geol. Unders.)

KETO, L. and MYERS, J. S. 1979. Outline of the Nagssugtoqidian of East Greenland. Rapp. Gronlands Geol. tinders., No. 89, 9-17.

BRODIE, K. H. 1975. Some aspects of the geology of the Roneval-Leverburgh area. South Harris. Unpublished BSc thesis, University of Leeds.

BROWN, J. F. 1975. Potassium-argon evidence of a Permian age for the camptonite dykes: Orkney. Scott. J. Geol., Vol. 11, 259-262.

BROWN, G. C., CASSIDY, J., TINDLE, A. G., and HUGHES, D. J. 1979. The Loch Doon granite: an example of granite petrogenesis in the British Caledonides. J. Geol. Soc. London, Vol. 136, 745-753.

BROWN, G. M. 1967. Mineralogy of basaltic rocks. 103-162 in Basalts: the Poldervaart treatise on rocks of basaltic composition. HESS, H. H., and POLDERVAART, A. (editors). Vol. 1. (New York: John Wiley and Sons.)

BULLERWELL, W. 1972. Geophysical studies relating to the Tertiary volcanic structure of the British Isles. Philos. Trans. R. Soc., A, Vol. 271, 209-215.

BURNHAM, C. W., HOLLOWAY, J. R., and DAVIS, N. F. 1969. Thermodynamic properties of water to 1000°C and 10 000 bars. Spec. Pap. Geol. Soc. Am., No. 132.

BURNS, D. J. 1966. Chemical and mineralogical changes associated with the Laxfordian metamorphism of dolerite dykes in the Scourie- Loch Laxford area, Sutherland, Scotland. Geol. Mag., Vol. 103, 19-35.

BUTLER, B. C. M. 1965. A chemical study of some rocks of the Moine Series of Scotland. Q. J. Geol. Soc. London, Vol. 121, 163-208.

BYERLEE, J. D. 1968. Brittle-ductile transition in rocks. J. Geophys. Res., Vol. 73, 4741-4750.

BYERLEE, J. D. and SUMMERS, R. 1975. Stable sliding preceding stick-slip on fault surfaces in granite at high pressure. 63-68 in Earthquake reduction and rock mechanics [special issue]. Wyss, M. (editor), Pure Appl. Geophys., Vol. 113.

CARMICHAEL, I. S. E., TURNER, F. J., and VERHOOGEN, J. 1974. Igneous petrology. (New York: McGraw-Hill.)

CAWTHORN, R. C., STRONG, D. F., and BROWN, P. A. 1976. Origin of corundum-normative intrusive and extrusive magmas. Nature, London, Vol. 259, 102-104,

CHAPMAN, H. J. 1979. 2,390 Myr Rb-Sr whole-rock age for the Scourie dykes of north-west Scotland. Nature, London, Vol. 277, 642-643.

CHESHER, J. A. 1979. Preliminary report on diving project at Loch Maddy, North Uist. Internal Report of the Continental Shelf Northern Unit, Institute of Geological Sciences, No. 79/9.

SMYTHE, D. K., and BISHOP, P. 1983. The geology of the Minches, Inner Sound and Sound of Raasay. Rep. Inst. Geol. Sci., No. 83/6.

CLIFF, R. A., GRAY, C. M., and HUHMA, H. 1983. A Sm-Nd isotope study of the South Harris Igneous Complex, the Outer Hebrides. Contrib. Mineral. Petrol., Vol. 82, 91-98.

COWARD, M. P. 1969. The structural and metamorphic geology of South Uist, Outer Hebrides. Unpublished PhD thesis, University of London.

COWARD, M. P. 1972. The Eastern Gneisses of South Uist. Scott. J. Geol., Vol. 8, 1-12.

COWARD, M. P. 1973a. The structure and origin of areas of anomalously low-intensity finite deformation in the basement gneiss complex of the Outer Hebrides. Tectonophysics, Vol. 16, 117-140.

COWARD, M. P. 1973b. Heterogeneous deformation in the development of the Laxfordian complex of South Uist, Outer Hebrides. J. Geol. Soc. London, Vol. 129, 139-160.

COWARD, M. P. 1976. Archaean deformation patterns in southern Africa. Philos. Trans. R. Soc., A, Vol. 283, 313-331.

COWARD, M. P. 1980. Shear zones in the Precambrian crust of southern Africa. 19-27 in Shear zones in rocks. CARRERAS, J., COBBOLD, P. R., RAMSAY, J. G., and WHITE, S. H. (editors). J. Struct. Geol., Vol. 2.

COWARD, M. P. , FRANCIS, P. W., GRAHAM, R. H., MYERS, J. S., and WATSON, J. V. 1969. Remnants of an early metasedimentary assemblage in the Lewisian Complex of the Outer Hebrides. Proc. Geol. Assoc., Vol. 80, 387-408.

COWARD, M. P. , FRANCIS, P. W., GRAHAM, R. H., and WATSON, J. V. 1970. Large-scale Laxfordian structures of the Outer Hebrides in relation to those of the Scottish Mainland. Tectonophysics, Vol. 10, 425-435.

COWARD, M. P. , GRAHAM, R. H., JAMES, P. R., and WAKEFIELD, J. 1973. A structural interpretation of the northern margin of Limpopo orogenic belt, southern Africa. Philos. Trans. R. Soc., A, Vol. 273, 387-491.

COWARD, M. P. , JAN, M. Q., REX, D., TARNEY, J., THIRLWALL, M., and WINDLEY, B. F. 1982. Structural evolution of a crustal section in the western Himalayas. Nature, London, Vol. 295, 22-24.

CRAIG, R. M. 1928. On the occurrence of flinty crush-rock in the Outer Hebrides. Rep. Br. Assoc. Adv. Sci. for 1928, 543.

CUNNINGHAM, G. J. 1981. Petrology and geochemistry of Lewisian pegmatites and granites, NW Scotland. Unpublished PhD thesis, University of London.

CUNNINGHAM-CRAIG, E. H., WRIGHT, W. B., and BAILEY, E. B. 1911. The geology of Colonsay and Oronsay with part of the Ross of Mull. Mem. Geol. Sure. GB.

DAVIDSON, C. F. 1943. The Archaean rocks of the Rodil District, South Harris, Outer Hebrides. Trans. R. Soc. Edinburgh, Vol. 61, 71-112.

DAVIES, F. B. 1974a. The evolution of two assemblages of metamorphosed igneous and supracrustal rocks in the Lewisian complex. Unpublished PhD thesis, University of London.

DAVIES, F. B. 1974b. A layered basic complex in the Lewisian, South of Loch Laxford, Sutherland. J. Geol. Soc. London, Vol. 130, 279-284.

DAVIES, F. B. 1975. Origin and ancient history of gneisses older than 2,800 Myr in Lewisian Complex. Nature, London, Vol. 258, 589-591.

LISLE, R. J., and WATSON, J. V. 1975. The tectonic evolution of the Lewisian Complex in Northern Lewis, Outer Hebrides. Proc. Geol. Assoc., Vol. 86, 45-61.

LISLE, R. J. and WATSON, J. V. 1977. Early basic bodies in the type Laxfordian complex, north-west Scotland, and their bearing on its origin. J. Geol. Soc. London, Vol. 133, 123-131.

DEARNLEY, R. 1959. Metamorphic petrology and history of the Lewisian rocks of South Harris, Outer Hebrides. Unpublished PhD thesis, University of Leeds.

DEARNLEY, R. 1961. Humite and chondrodite in a Lewisian crystalline limestone from South Harris, Outer Hebrides. Mineral. Mag., Vol. 32, 910-911.

DEARNLEY, R. 1962a. An outline of the Lewisian complex of the Outer Hebrides in relation to that of the Scottish Mainland. Q J. Geol. Soc. London, Vol. 118, 143-176.

DEARNLEY, R. 1962b. Diopside-orthoclase-hornblende rocks from the Lewisian paragneiss of South Harris, Outer Hebrides. Geol. Mag., Vol. 99, 27-29.

DEARNLEY, R. 1963. The Lewisian complex of South Harris, with some observations on the metamorphosed basic intrusions of the Outer Hebrides, Scotland. Q. J. Geol. Soc. London, Vol. 119, 243-312.

DEARNLEY, R. 1973. Scourie dykes of the Outer Hebrides. 131-135 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. G., and TARNEY, J. (editors). (Keele: University of Keele.)

DEARNLEY, R. and DUNNING, F. W. 1968. Metamorphosed and deformed pegmatites and basic dykes in the Lewisian complex of the Outer Hebrides and their geological significance. Q. J. Geol. Soc. London, Vol. 123, 353-378.

DEER, W. A., HOWIE, R. A., and ZUSSMAN, J. 1978. Rock-forming minerals, Vol. 2: Single chain silicates, 2nd edition. (London: Longman.)

DICKINSON, B. B. 1974. Metamorphic evolution of the Leverburgh supracrustal belt, South Harris, with special reference to textural variations. Unpublished PhD thesis, University of London.

DICKINSON, B. B. and WATSON, J. V. 1976. Variations in crustal level and geothermal gradient during the evolution of the Lewisian complex of north-west Scotland. Precambrian Res., Vol. 3, 363 -374.

DOUGAL, J. W. 1928. Observations on the geology of Lewis. Trans. Edinburgh Geol. Soc., Vol. 12, 12-18.

DREVER, H. I. 1953. A note on the field relations of the Shiant Isles Picrite. Geol. Mag., Vol. 90, 159-160.

DREVER, H. I. 1957. A note on the occurrence of rhythmic layering in the Eilean Mhuire Sill, Shiant Isles. Geol. Mag. , Vol. 94, 277-280.

DREVER, H. I. and JOHNSTON, R. 1959. The lower margin of the Shiant Isles Sill. Q. J. Geol. Soc. London, Vol. 24, 343-364.

DREVER, H. I. and JOHNSTON, R. 1965. New petrographical data on the Shiant Isles picrite. Mineral. Mag., Vol. 34, 194-203.

DREVER, H. I. and JOHNSTON, R. 1967. The ultrabasic facies in some sills and sheets. 51-63 in Ultramafic and related rocks. WYLLIE, P. J. (editor). (New York: John Wiley and Sons.)

DRURY, S. A. 1972. The tectonic evolution of a Lewisian complex on Coll, Inner Hebrides. Scott. J. Geol. , Vol. 8, 309 - 333.

DRURY, S. A. 1974. Chemical changes during retrogressive metamorphism of Lewisian granulite facies rocks from Coll and Tiree. Scott. J. Geol. , Vol. 10, 237-256.

ELLIOT, D., and JOHNSON, M. R. W. 1980. Structural evolution in the northern part of the Moine Thrust zone. Trans. R. Soc. Edinburgh, Earth Sci. , Vol. 71, 69-96.

ESCHER, A., JACK, S., and WATTERSON, J. 1976a. Tectonics of the North Atlantic Proterozoic dyke swarm. Philos. Trans. R. Soc., A, Vol. 280, 529-539.

SORENSEN, K., and ZECK, H. P. 1976b. Nagssugtoquidian mobile belt in West Greenland. 76-95 in Geology of Greenland, ESCHER, A., and WATTS, W. S. (editors). (Copenhagen: Gronlands Geol. Unders.)

ESSON, J., DUNHAM, A. C., and THOMPSON, R. N. 1975. Low alkali, high calcium olivine tholeiite lavas from the Isle of Skye, Scotland. J. Petrol., Vol. 16, 488-497.

EVANS, C. R. 1965. Geochronology of the Lewisian basement near Lochinver, Sutherland. Nature, London, Vol. 207, 54-56.

EVANS, C. R. and TARNEY, J. 1964. Isotopic ages of Assynt dykes. Nature, London, Vol. 204, 638-641.

EVANS, C. R. and LAMBERT, R. ST J. 1974. The Lewisian of Lochinver, Sutherland; the type area for the Inverian metamorphism. J. Geol. Soc. London, Vol. 130, 125-150.

FERRY, J. M., and SPEAR, F. S. 1978. Experimental calibration of the partitioning of Fe and Mg between biotite and garnet. Contrib. Mineral. Petrol. , Vol. 66, 113-117.

FETTES, D. J., and MACDONALD, R. 1978. Glen Garry vein complex. Scott. J. Geol., Vol. 14, 335-358.

FLEET, M. E. 1974. Partition of major and minor elements and equilibrium in coexisting pyroxenes. Contrib. Mineral. Petrol., Vol. 44, 259-274.

FLINN, D. 1962. On folding during three-dimensional progressive deformation. Q. J. Geol. Soc. , London, Vol. 118, 385-433.

1965. On the symmetry principle and the deformation ellipsoid. Geol. Mag., Vol. 102, 36-45.

FRANCIS, P. W. 1969. Some aspects of the Lewisian geology of Barra and adjacent islands. Unpublished PhD thesis, University of London.

FRANCIS, P. W. 1973. Scourian-Laxfordian relationships in the Barra Isles. J. Geol. Soc. London, Vol. 129, 161-189.

FRANCIS, P. W. MOORBATH, S., and WELKE, H. J. 1971. Isotopic age data from Scourian intrusive rocks on the Isle of Barra, Outer Hebrides, north-west Scotland. Geol. Mag., Vol. 108, 13-22.

FRANCIS, P. W. MOORBATH, S. and SIBSON, R. H. 1973. The Outer Hebrides Thrust. 95-104 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. C., and TARNEY, J. (editors). (Keele: University of Keele.)

FRIEDMAN, M., LOGAN, J. M., and RIGERT, J. A. 1974. Glass-indurated quartz gouge in sliding-friction experiments on sandstone. Bull. Geol. Soc. Am., Vol. 85, 937-942.

GARSON, M. S., and BRIGGS, D. A. 1971. Field and laboratory investigation of Kyanite-bearing gneiss, Chaipaval Hill, Southwest Harris, Scotland. Report of the Mineralogy Unit, British Geological Survey, No. 89.

GARSON, M. S. and LIVINGSTONE, A. 1973. Is the South Harris complex in north Scotland a Precambrian overthrust slice of oceanic crust and island arc? Nature, London, Vol. 243, 74-76.

GIBB, F. G. F. 1973. The zoned clinopyroxenes of the Shiant Isles Sill, Scotland. J. Petrol. , Vol. 14,.203-230.

GILETTI, B. J., MOORBATH, S., and LAMBERT, R. ST J. 1961. A geochronological study of the metamorphic complexes of the Scottish Highlands. Q. J. Geol. Soc. London, Vol. 117, 233-272.

GLIKSON, A. Y. 1976. Stratigraphy and evolution of primary and secondary greenstones: significance of data from shields of the southern hemisphere. 257-277 in The early history of the Earth. WINDLEY, B. F. (editor). (London: John Wiley and Sons.)

GOLDMAN, D. S., and ALBEE, A. L. 1977. Correlation of Mg/Fe partitioning between garnet and biotite with 180/160 partitioning between quartz and magnetite. Am. J. Sci., Vol. 227, 750-767.

GOLDSMITH, J. R. 1976. Scapolites, granulites, and volatiles in the lower crust. Bull. Geol. Soc. Am. , Vol. 87, 161-168.

GOTTFRIED, D., ANNELL, C. S., and SCHWARTZ, L. J. 1977. Geochemistry of subsurface basalt from the deep corehole (Clubhouse Crossroads corehole 1) near Charleston, South Carolina-magma type and tectonic implications. Prof. Pap. US Geol. Sure., No. 1028, 91-113.

GRAHAM, R. H. 1970. A structural analysis of Lewisian rocks in parts of North Uist and the Sound of Harris, Outer Hebrides. Unpublished PhD thesis, University of London.

GRAHAM, R. H. 1980. The role of shear belts in the structural evolution of the South Harris igneous complex. 29-37 in Shear zones in rocks. CARRERAS, J., COBBOLD, P. R., RAMSAY, J. G., and WHITE, S. H. (editors). J. Struct. Geol., Vol. 2.

GRAHAM, R. H., and COWARD, M. P. 1973. The Laxfordian of the Outer Hebrides. 85-93 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. G., and TARNEY, J. (editors). (Keele: University of Keele.)

GREEN, D. H., and RINGWOOD, A. E. 1967. An experimental investigation of the gabbro to eclogite transformation and its petrological applications. Geochim. Cosmochim. Acta. , Vol. 31, 767-833.

GUPPY, E. 1956. Chemical analyses of igneous rocks, metamorphic rocks and minerals 1931-1954. Mem. Geol. Sure. GB.

HANKS, T. C. 1977. Earthquake stress drops, ambient tectonic stresses and stresses that drive plate motions. Pure Appi. Geophys. , Vol. 115, 441-458.

HARDING, R. R., MERRIMAN, R. J., and NANCARROW, P. H. A. 1984. St Kilda: an illustrated account of the geology. Rep. Br. Geol. Sure., Vol. 16, No. 7.

HEDDLE, M. F. 1878. Chapters on the mineralogy of Scotland. Chapter fourth-augite, hornblende and serpentinous change. Trans. R. Soc. Edinburgh, 28, 453-555.

HEIER, K. S. 1973. Geochemistry of granulite facies rocks and problems of their origin. Philos. Trans. R. Soc., A, Vol. 273, 429-442.

HICKMAN, A. H. 1975. The stratigraphy of late Precambrian metasediments between Glen Roy and Lismore. Scott. J. Geol., Vol. 11, 117-142.

HIGGINS, M. W. 1971. Cataclastic rocks. Prof. Pap. US Geol. Sure., No. 687.

HOPGOOD, A. M. 1964. Structure and tectonic history of Lewisian gneiss-Isle of Barra. Unpublished PhD thesis, St Andrews University.

HOPGOOD, A. M. 1965. Theoretical consideration of the mechanics of tectonic reorietation of dykes. Tectonophysics, Vol. 3, 17-28.

HOPGOOD, A. M. 1971. Structure and tectonic history of Lewisian gneiss, Isle of Barra, Scotland. Krystalinilcum, No. 7, 27-59.

HOPGOOD, A. M. and BOWES, D. R. 1972. Application of structural sequence to the correlation of Precambrian gneisses, Outer Hebrides, Scotland. Bull. Geol. Soc. Am., Vol. 83, 107-127.

HORSLEY, R. J. 1978. Geochemistry of meta-igneous rocks from the South Harris complex, Outer Hebrides. Unpublished PhD thesis, University of London.

INSTITUTE OF GEOLOGICAL SCIENCES. 1978. Bouguer gravity anomaly map. Little Minch. Sheet 57°N-08°W. 1:250 000 (Southampton: Ordnance Survey for Institute of Geological Sciences.)

IRVINE, T. N., and BARAGAR, W. R. A. 1971. A guide to the chemical classification of the common volcanic rocks. Can. J. Earth. Sci., Vol. 8, 523-548.

JACK, S. M. B. 1978. The North Atlantic Proterozoic dyke swarm. Unpublished PhD thesis, University of Liverpool.

JACKSON, R. E., and DUNN, D. E. 1974. Experimental sliding friction and cataclasis of foliated rocks. Int. J. Rock. Mech. Min. Sci., Vol. 11, 235-249.

JAKEŠ, P., and WHITE, A. J. R. 1969. Structure of the Melanesian arcs and correlation with distribution of magma types. Tectonophysics, Vol. 8, 223-236.

JAKEŠ, P. and WHITE, A. F. R. 1972. Hornblendes from talc-alkaline volcanic rocks of island arcs and continental margins. Am. Mineral., Vol. 57, 887-902.

JEHU, T. J., and CRAIG, R. M. 1923. Geology of the Outer Hebrides. Part I-The Barra Isles. Trans. R. Soc. Edinburgh, Vol. 53, 419-441.

JEHU, T. J. 1925. Geology of the Outer Hebrides. Part II-South Uist and Eriskay. Trans. R. Soc. Edinburgh, Vol. 53, 615-641.

JEHU, T. J. 1926. Geology of the Outer Hebrides. Part III-North Uist and Benbecula. Trans. R. Soc. Edinburgh, Vol. 54, 467-489.

JEHU, T. J. 1927. Geology of the Outer Hebrides. Part IV-South Harris. Trans. R. Soc. Edinburgh, Vol. 55, 457-488.

JEHU, T. J. 1934. Geology of the Outer Hebrides. Part V-North Harris and Lewis. Trans. R. Soc. Edinburgh, Vol. 57, 839-874.

JOHNSTON, R. 1953. The olivines of the Garbh Eilean Sill, Shiant Isles. Geol. Mag., Vol. 90, 161-171.

JUDD, J. W. 1878. The secondary rocks of Scotland: third paper. The strata of the western coast and islands. Q. J. Geol. Soc. London, Vol. 34, 660-743.

KNORRING, O. VON, and DEARNLEY, R. 1959. Niobium-zirconium-thorium-uranium and rare-earth minerals from the pegmatites of South Harris, Outer Hebrides. Nature, London, Vol. 183, 255-256.

KNORRING, O. VON 1960. The Lewisian pegmatites of South Harris, Outer Hebrides. Mineral. Mag., Vol. 32, 366-378.

KÖPPEL, V. 1974. Isotopic U-Pb ages of monazites and zircons from the crust-mantle transition and adjacent units of the Ivrea and Ceneri zones (Southern Alps, Italy). Contrib. Mineral. Petrol., Vol. 43, 55-70.

KURSTEN, M. 1957. The metamorphic and tectonic history of parts of the Outer Hebrides. Trans. Edinburgh Geol. Soc., Vol. 17, 1-31.

KWAK, T. A. P. 1977. Scapolite compositional change in a metamorphic gradient and its bearing on the identification of meta-evaporite sequences. Geol. Mag., Vol. 114, 343-354.

LAMBERT, R. ST J., EVANS, C. R., and DEARNLEY, R. 1970a. Isotopic ages of dykes and pegmnatitic gneiss from the southern islands of the Outer Hebrides. Scott. J. Geol., Vol. 6, 208-213.

MYERS, J. S., and WATSON, J. V. 1970b. An apparent age for a member of the Scourie dyke suite in Lewis, Outer Hebrides. Scott. J. Geol., Vol. 6, 214-220.

LEAKE, B. E. 1978. Nomenclature of amphiboles. Mineral. Mag., Vol. 42, 533-563.

LEE, G. W., and PRINGLE, J. 1932. A synopsis of the Mesozoic rocks of Scotland. Trans. Geol. Soc. Glasgow, Vol. 19, 158-224.

LE MAITRE, R. W. 1976. The chemical variability of some common igneous rocks. J. Petrol., Vol. 17, 589-637.

LISLE, R. J. 1974. The application of methods of structural analysis to the study of the basement complex in north-west Lewis. Unpublished PhD thesis, University of London.

LISLE, R. J. 1976. The evaluation of Laxfordian deformation in the Carloway area, Isle of Lewis, Scotland. Tectonophysics, Vol. 42, 183-208.

LIVINGSTONE, A. 1963. The petrochemistry of Lewisian ultrabasic rocks from South Harris, Outer Hebrides. Unpublished PhD thesis, University of Manchester.

LIVINGSTONE, A. 1965. An olivine-bearing sagvandite from Berneray, Outer Hebrides. Geol. Mag., Vol. 102, 227-230.

LIVINGSTONE, A. 1967. A garnet peridotite and a garnet-amphibole pyroxenite from South Harris, Outer Hebrides, and their bearing on the South Harris eclogite facies status. Mineral. Mag., Vol. 36, 380-388.

LIVINGSTONE, A. 1976a. A metamorphosed, layered alpine-type peridotite in the Langavat valley, South Harris, Outer Hebrides. Mineral. Mag., Vol. 40, 493-499.

LIVINGSTONE, A. 1976b. The paragenesis of spinel- and garnet-amphibole lherzolite in the Rodel area, South Harris. Scott. J. Geol., Vol. 12, 293-300.

LYON, T. D. B., PIDGEON, R. T., BOWES, D. R., and HOPGOOD, A. M. 1973. Geochronological investigation of the quartzofeldspathic rocks of the Lewisian of Rona, Inner Hebrides. J. Geol. Soc. London, Vol. 129, 389-404.

MCCARTHY, T. S. 1976. Chemical interrelationships in a low-pressure granulite terrain in Namaqualand, South Africa, and their bearing on granite genesis and the composition of the lower crust. Geochim. Cosmochim. Acta, Vol. 40, 1057-1068.

MCCARTHY, T. S. and GROVES, D. I. 1979. The Blue Tier batholith, northeastern Tasmania. Contrib. Mineral. Petrol., Vol. 71, 193-209.

MACCULLOCH, J. 1819. A description of the Western Islands of Scotland. 3 volumes (Edinburgh: Constable.)

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

MCGARR, A., SPOTTISWOODE, S. M., GAY, N. C., and ORTLEPP, W. D. 1979. Observations relevant to seismic driving stress, stress drop, and efficiency. 2251-2261 in Fault mechanics. HANKS, T. C. (editor). J. Geophys. Res. , Vol. 84.

MCKENZIE, A. J., and PATRICK, D. J. 1971. Regional mineral survey of the Isles of Lewis and Harris, Outer Hebrides. Robertson Research Co. Ltd, Report 666.

BENJAMIN, R. E. K., JONES, E., and DANIEL, E. T. 1973. Special mineral development programme on the islands of Harris and Lewis, Outer Hebrides. Robertson Research Co. Ltd, Report 828, Vols. I and II.

MCKENZIE, D., and BRUNE, J. N. 1972. Melting on fault planes during large earthquakes. Geophys. J. R. Astron. Soc., Vol. 29, 65-78.

MACKIE, W. 1932. The heavier accessory minerals in the granites of Scotland. Trans. Edinburgh Geol. Soc. , Vol. 12, 22-40.

MACKINNON, A. 1974. The Madadh rocks: a Tertiary olivine dolerite sill in the Outer Hebrides. Scott. J. Geol., Vol. 10, 67-70.

McQUILLIN, R., and WATSON, J. 1973. Large-scale basement structures of the Outer Hebrides in the light of geophysical evidence. Nature, London, Vol. 245, 1-3.

MANSON, V. J. 1967. Geochemistry of basalts: major elements. 215-269 in Basalts: the Poldert,aart treatise on rocks of basaltic composition. HESS, H. H., and POLDERVAART, A. (editors). Vol. 1. (New York: John Wiley and Sons.)

MARSHALL, R. R. 1961. Devitrification of natural glass. Bull. Geol. Soc. Am., Vol. 72, 1493-1520.

MARSTON, R. J. 1971. The Foyers granitic complex, Inverness-shire, Scotland. Q. J. Geol. Soc. London, Vol. 126, 331-368.

MASCH, L. 1973. Untersuchung der aufschmelzung and deformation der pseudotachylit der Silvretta (Osterreich, Schweiz) [Investigation of the molten state and deformation of the pseudotachylites of the Silvretta (Austria, Switzerland)]. Neues. Jb. Miner. Monatsh. , Vol. 11, 487-509.

MATTEY, D. P., GIBSON, I. L., MARRINER, G. F., and THOMPSON, R. N. 1977. The diagnostic geochemistry, relative abundance, and spatial distribution of high-calcium, low-alkali olivine tholeiite dykes. Mineral. Mag., Vol. 41, 273-285.

MOORBATH, S., and PARK, R. G. 1972. The Lewisian chronology of the southern region of the Scottish mainland. Scott. J. Geol., Vol. 8, 51-74.

MOORBATH, S. POWELL, J. L., and TAYLOR, P. N. 1975. Isotopic evidence for the age and origin of the "grey gneiss" complex of the southern Outer Hebrides, Scotland. J. Geol. Soc. London, Vol. 131, 213-222.

MOORBATH, S. POWELL, J. L. and PANKHURST, R. J. 1976. Further rubidium-strontium age and isotope evidence for the nature of the late Archaean plutonic event in West Greenland. Nature, London, Vol. 262, 124-126.

MORRISON, W. 1885. Precambrian conglomerate of Lewis. Trans. Edinburgh Geol. Soc. , Vol. 5, 235-242.

MURCHISON, R. I., and GEIKIE, A, 1861. On the altered rocks of the Western Islands of Scotland and the North-western and Central Highlands. Q. J. Geol. Soc. London, Vol. 17, 171-228.

MURCHISON, R. I.1862. Thirty years retrospect of the progress in our knowledge of the Geology of the Older rocks. Am. J. Sci., Ser. 2, Vol. 33, 1-21.

MURRAY, R. J. 1954. Clinopyroxenes of the Garbh Eilean Sill, Shiant Isles. Geol. Mag., Vol. 91, 17-31.

MYERS, J. S. 1968. The tectonic and metamorphic history of the Lewisian migmatite complex of Western Harris, Outer Hebrides, Scotland. Unpublished PhD thesis, University of London.

MYERS, J. S. 1970a. Gneiss types and their significance in the repeatedly deformed and metamorphosed Lewisian Complex of Western Harris, Outer Hebrides. Scott. J. Geol., Vol. 6, 186-199.

MYERS, J. S. (with a note by R. J. LISLE). 1970b. Zones of abundant Scourie dyke fragments and their significance in the Lewisian Complex of western Harris, Outer Hebrides. Proc. Geol. Assoc. , Vol. 82, 365-378.

MYERS, J. S. 1971. The late Laxfordian granite-migmatite complex of Western Harris, Outer Hebrides. Scott. J. Geol. Vol. 7, 254 284.

MYERS, J. S. 1975a. Igneous stratigraphy of Archaean anorthosite at Majorquap qava, near Fiskenmsset, south-west Greenland. Rapp. Gronlands Geol. Unders. , No. 74.

MYERS, J. S. 1975b. Pseudo-fractionation trend of the Fiskenaesset anorthosite complex, southern west Greenland. Rapp. Gronlands, Geol. Unders., No. 75, 77-80.

MYERS, J. S. 1976. Acid and intermediate intrusions, deformation and gneiss formation, north-east of Fiskenmsset. Rapp. Gronlands. Geol. Unders., No. 73, 7-15.

MYERS, J. S. 1978. Formation of banded gneisses by deformation of igneous rocks. Precambrian Res., Vol. 6, 43-64.

MYERS, J. S. and PLATT, R. G. 1977. Mineral chemistry of layered Archaean anorthosite at Majorqap Ova, near Fiskensset, Southwest Greenland. Lithos, Vol. 10, 59-72.

NESBITT, H. C. 1961. The geology of North Rona. Trans. Geol. Soc. Glasgow, Vol. 24, 169-189.

NEWTON, R. C., and GOLDSMITH, J. R. 1975. Stability of the scapolite meionite (3Ca Al2 Si202. Ca CO3) at high pressures and storage of CO2 in the deep crust. Contrib. Mineral. Petrol. , Vol. 49, 49-62.

NICOL, J. 1844. Guide to the geology of Scotland. (Edinburgh: Oliver and Boyd.)

NOCKOLDS, S. R. 1954. Average chemical compositions of some igneous rocks. Bull. Geol. Soc. Am., Vol. 65, 1007-1032.

O'HARA, M. J. 1961a. Petrology of the Scourie dyke, Sutherland. Mineral. Mag., Vol. 32, 848-865.

O'HARA, M. J. 1961b. Zoned ultrabasic and basic gneiss masses in the early Lewisian metamorphic complex at Scourie, Sutherland. J. Petrol., Vol. 2, 248-276.

O'HARA, M. J. 1965. Origin of ultrabasic and basic gneiss masses in the Lewisian. Geol. Mag., Vol. 102, 296-314.

O'HARA, M. J. 1977. Thermal history of excavation of Archaean gneisses from the base of the continental crust. J. Geol. Soc. London, Vol. 134, 185-200.

PALMER, K. F. 1971. A comparative study of two Pre-Cambrian gneiss areas - the Suportoq region, East Greenland, and South Harris, Outer Hebrides - and their bearing on Precambrian crustal evolution. Unpublished PhD thesis, University of Birmingham.

PARK, R. G. 1961. The pseudotachylite of the Gairloch district, Ross-shire, Scotland. Am. J. Sci., Vol. 259, 542-550.

PARK, R. G. 1970. Observations of Lewisian chronology. Scott. J. Geol., Vol. 6, 379-399.

PARK, R. G. and CRESSWRLL, D. 1973. The dykes of the Laxfordian belts. 119-130 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. G., and TARNEY, J. (editors). (Keele: University of Keele.)

PARK R. G., and TARNEY, J. 1973. The early Precambrian of Scotland and related rocks of Greenland. (Keele: University of Keele.)

PARK R. G., and TARNEY, J. 1987. Evolution of the Lewisian and comparable Precambrian high-grade terrains. Spec. Publ. Geol. Soc. London, No. 27.

PEACH, B. N., and HORNE, J. 1930. Chapters on the geology of Scotland. (London: Oxford University Press.)

PEACH, B. N., HORNE, J., GUNN, W., CLOUGH, C. T., HINXMAN, L. W., and TEALL, J. J. H. 1907. The geological structure of the north-west Highlands of Scotland. Mem. Geol. Surv. GB.

PEACOCK, J. D. 1984. Quaternary geology of the Outer Hebrides. Rep. Br. Geol. Surv. Vol. 16, No. 2.

PENN, I. E., and MERRIMAN, R. J. 1978. Jurassic (Toarcian to ?Bajocian) ammonites of the Shiant Islands, Outer Hebrides. Scott. J. Geol., Vol. 14, 45-53.

PIDGEON, R. T., and AFTALION, M. 1972. The geochronological significance of discordant U-Pb ages of oval-shaped zircons from a Lewisian gneiss from Harris, Outer Hebrides. Earth Planet. Sci. Lett., Vol. 17, 269-274.

PRICE, N. J. 1966. Fault and joint development in brittle and semi-brittle rock. (Oxford: Pergamon.)

PRINZ, M. 1967. Geochemistry of basaltic rocks: trace elements. 271-324 in Basalts: the Poldervaart treatise on rocks of basaltic composition. HESS, H. H., and POLDERVAART, A. (editors). Vol. 1. (New York: John Wiley and Sons.)

RAASE, P. 1974. Al and Ti contents of hornblende, indicators of pressure and temperature of regional metamorphism. Contrib. Mineral. Petrol., Vol. 45, 231-236.

RAHEIM, A., and GREEN, D. H. 1973. Experimental determination of the temperature and pressure dependence of the Fe/Mg partition coefficient for coexisting garnet and clinopyroxene. Contrib. Mineral. Petrol., Vol. 48, 179-203.

RAMSAY, J. G., and GRAHAM, R. H. 1970. Strain variation in shear belts. Can. J. Earth Sci., Vol. 7, 786-813.

RITCHIE, W. 1979. Machair development and chronology in the Uists and adjacent islands. Proc. R. Soc. Edinburgh, Vol. 77B, 107-122.

ROCK, N. M. S. 1977. The nature and origin of lamprophyres: some definitions, distinctions and derivations. Earth Sci. Rev., Vol. 13, 123-169.

ROCK, N. M. S. 1979. Camptonite-monchiquite dykes of the Scottish Highlands. Unpublished digitised computer file of field and petrographic data. Held on open file in the IGS [BGS] library at Edinburgh.

ROCK, N. M. S. 1983. The Permo-Carboniferous camptonite-monchiquite dyke-suite of the Scottish Highlands and Islands: distribution, field and petrological aspects. Rep. Inst. Geol. Sci., No. 82/14.

ROCK, N. M. S. and LEAKE, B. E. 1984. The International Mineralogical Association amphibole nomenclature scheme: computerisation and its consequences. Mineral. Mag., Vol. 48, 211-227.

RUBLE, D. C. 1972. The Scourie dykes, a Precambrian iron-rich tholeiitic magma type. 169-170 in Progress in experimental petrology. HENDERSON, C. M. B., and HAMILTON, D. L. (editors). NERC publication series D, No. 2.

SHAW, D. M. 1960. The geochemistry of scapolite. Part 1. Previous work and general mineralogy. J. Petrol., Vol. 1, 218-260.

SHAW, D. M. 1968. A review of K-Rb fractionation trends by covariance analysis. Geochim. Cosmochim. Acta., Vol. 32, 573-601.

SHAW, D. M. 1977. Trace element behaviour during anatexis. 189-213 in Magma genesis: proceedings of the American Geophysical Union Chapman conference on partial melting in the Earth's upper mantle. DICK, H. J. B. (editor). Bull. Oregon Dep. Geol. Miner. Ind. , No. 96.

SHERATON, J. W., SKINNER, A. C., and TARNEY, J. 1973. The geochemistry of the Scourian gneisses of the Assynt district. 13-30 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. G., and TARNEY, J. (editors). (Keele: University of Keele.)

SIBSON, R. H. 1973. Interactions between temperature and pore fluid pressure during earthquake faulting and a mechanism for partial or total stress relief. Nature, London, Vol. 243, 66-68.

SIBSON, R. H. 1975. Generation of pseudotachylyte by ancient seismic faulting. Geophys. J. R. Astron. Soc. , Vol. 43, 775-794.

SIBSON, R. H. 1977a. Kinetic shear resistance, fluid pressures and radiation efficiency during seismic faulting. Pure Appl. Geophys., Vol. 115, 387-400.

SIBSON, R. H. 1977b. The Outer Hebrides Thrust: its structure, mechanism and deformation environment. Unpublished PhD thesis, University of London.

SIBSON, R. H. 1977c. Fault rocks and fault mechanisms. J. Geol. Soc. London, Vol. 133, 191-213.

SIBSON, R. H. 1980. Transient discontinuities in ductile shear zones. 165-171 in Shear zones in rocks. CARRERAS, J., COBBOLD, P. R., RAMSAY, J. G., and WHITE, S. H. (editors). J. Struct. Geol., Vol. 2.

SILLS, J. D., SAVAGE, D., WATSON, J. V. and WINDLEY, B. F. 1981. Layered ultramafic-gabbro bodies in the Lewisian of NW Scotland: geochemistry and petrogenesis. Earth Planet. Sci. Lett., Vol. 58, 345-360.

SILLS, J. D. and WINDLEY, B. F. 1982. Petrogenesis of Lewisian amphibolite dykes from Clashnessie Bay, near Stoer, Sutherland. Scott. J. Geol. , Vol. 18, 167-176.

SKINNER, A. C. 1970. Geochemical studies of the Lewisian of NW Scotland and comparable rocks in E Greenland. Unpublished PhD thesis, University of Birmingham.

SMITH, D. I. 1979. Caledonian minor intrusions of the N Highlands of Scotland. 683-697 in The Caledonides of the British Isles-reviewed. HARRIS, A. L., HOLLAND, C. H., and LEAKE, B. E. (editors). Spec. Publ. Geol. Soc. London, No. 8.

SMITH, J. V. 1974. Chemical and textural properties. In Feldspar minerals. Vol. 2. SMITH, J. V., and SMITH, B. F. (editors). (Berlin: Springer-Verlag.)

SMYTHE, D. K., SOWERBUTTS, W. T. C., BACON, M., and McQuILLIK, R. 1972. Deep sedimentary basin below northern Skye and the Little Minch. Nature, London, Vol. 236, 87-89.

SMYTHE, D. K. 1987. Deep seismic reflection profiling of the Lewisian foreland. 193-203 in Evolution of the Lewisian and comparable Precambrian high grade terrains. PARK, R. G., and TARNEY, J. (editors). Spec. Publ. Geol. Soc. London, No. 27.

SOLOIN, S. R. 1978. The tectonic evolution and geochemistry of the Lewisian complex of North Harris. Unpublished PhD thesis, University of London.

SPEIGHT, J. M., and MITCHELL, J. G. 1979. The PermoCarboniferous dyke-swarm of northern Argyll and its bearing on dextral displacement on the Great Glen Fault. J. Geol. Soc. London, Vol. 136, 3-11.

SKELHORN, R. R., SLOAN, R., and KNAPP, R. J. 1982. The dyke swarms of Scotland. 449-459 in Igneous rocks of the British Isles. SUTHERLAND, D. S. (editor). (Chichester: John Wiley and Sons.)

STEAVENSON, A. G.1928. Some geological notes on three districts of northern Scotland. Trans. Geol. Soc. Glasgow, Vol. 18, 193-233.

STEEL, R. J. 1971. New Red Sandstone movement on the Minch Fault. Nature, London, Vol. 234, 158-159.

STEEL, R. J. 1974. New Red Sandstone floodplain and piedmont sedimentation in the Hebridean province, Scotland. J. Sediment. Petrol., Vol. 44, 336-357.

STEEL, R. J. 1977. Triassic rift basins of north-west Scotland-their configuration, infilling and development. 1-18 in Mesozoic Northern North Sea Symposium (MNNS-77). Spec. Publ. Norwegian Petrol. Soc., Vol. 7.

STEEL, R. J. and WILSON, A. C. 1975. Sedimentation and tectonism (?Permo-Triassic) on the margin of the North Minch Basin, Lewis. J. Geol. Soc. London, Vol. 131, 183-202.

STEPHENS, W. E., and HALLIDAY, A. N. 1979. Compositional variation in the Galloway plutons. 9-17 in Origin of granite batholiths: geochemical evidence. ATHERTON, M. P. (editor). (Orpington, Kent: Shiva Publishing Ltd.)

STEVENS, A. 1914. Notes on the geology of the Stornoway district of Lewis. Trans. Geol. Soc. Glasgow, Vol. 15, 51-63.

STEWART, M. 1932. Notes of the geology of North Rona. Geol. Mag., Vol. 69, 179-185.

STEWART, M. 1933. Notes on the geology of Sula Sgeir and the Flannan Islands. Geol. Mag., Vol. 70, 110-116.

STORETVEDT, K. M., and STEEL, R. J. 1977. Palaeomagnetic evidence for the age of the Stornoway formation. Scott. J. Geol., Vol. 13, 263-269.

STRECKEISEN, A. 1976. To each plutonic rock its proper name. Earth Sci. Rev., Vol. 12, 1-33.

STRONG, D. F. 1979. The Mount Peyton batholith, Central Newfoundland: A bimodal talc-alkaline suite. J. Petrol., Vol. 20, 119-138.

SUTTON, J., and WATSON, J. V. 1951. The pre-Torridonian metamorphic history of the Loch Torridon and Scourie areas in the north west Highlands, and its bearing on the chronological classification of the Lewisian. Q. J. Geol. Soc. London, Vol. 106, 241-307.

TAFT, M. B. 1972. Structure and tectonic history of the Lewisian Gneiss complex north-west of Loch Seaforth, Isle of Lewis, Outer Hebrides, Scotland. Unpublished MSc thesis, University of St Andrews.

TAFT, M. B. 1978. Basic minor intrusions in the Lewisian gneisses of southern Lewis, Outer Hebrides. Scott. J. Geol., Vol. 14, 185 -190.

TARNEY, J. 1973. The Scourie dyke suite and the nature of the Inverian event in Assynt. 105-118 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. G., and TARNEY, J. (editors). (Keele: University of Keele.)

TARNEY, J. 1976. Geochemistry of Archaean high-grade gneisses, with implications as to the origin and evolution of the Precambrian crust. 405-417 in The early history of the Earth. WINDLEY, B. F. (editor). (London: John Wiley and Sons.)

SKINNER, A. C., and SHERATON, J. W. 1972. A geochemical comparison of major Archaean gneiss units from north-west Scotland and east Greenland. 24th Sess. Int. Geol. Congr., Montreal, Sect. 1, 162-174.

TAYLOR, S. R: 1966. The application of trace element data to problems in petrology. Phys. Chem. Earth, Vol. 6, 133-213.

THOMPSON, A. B. 1976. Mineral reaction in pelitic rocks: II, Calculation of some P-T-X (Fe-Mg) phase relations. Am. J. Sci., Vol. 276, 425-454.

TOURET, J. 1971. Le facies granulite en Norvege meridionale. Part I: Les associations mineralogiques. [The granulite facies in southern Norway. Part I: Mineralogical associations]. Lithos, Vol. 4, 239-249.

TOURET, J. 1971. Le facies granulite en Norvege meridionale. Part II: Les inclusions fluides. [The granulite facies in southern Norway. Part II: Fluid inclusions.] Lithos, Vol. 4, 423-436.

TURNER, F. J. 1968. Metamorphic petrology: mineralogy and field aspects. (New York, London: McGraw Hill.)

WAGER, L. R., and BROWN, G. M. 1968. Layered igneous rocks. (Edinburgh: Oliver and Boyd.)

WALKER, F. 1980. The geology of the Shiant Isles (Hebrides). Q. J. Geol. Soc. London, Vol. 86, 355-398.

WALKER, G. P. L. 1960. Zeolite zones and dyke distribution in relation to the structure of the basalts of eastern Iceland. J. Geol., Vol. 68, 515-528.

WATSON, J. V. 1968. Post-Scourian metadolerites in relation to Laxfordian deformation in Great Bernera, Outer Hebrides. Scott. J. Geol., Vol. 4, 53-67.

WATSON, J. V. 1969. The Precambrian gneiss complex of Ness, Lewis, in relation to the effects of Laxfordian regeneration. Scott. J. Geol., Vol. 5, 268-285.

WATSON, J. V. and LISLE, R. J. 1973. The Pre-Laxfordian complex of the Outer Hebrides. 45-50 in The early Precambrian of Scotland and related rocks of Greenland. PARK, R. G., and TARNEY, J. (editors). (Keele: University of Keele.)

WATTERSON, J. 1968. Plutonic development of the Ilordleq area, South Greenland, Part II: Late-kinematic basic dykes. Gronlands Geol. Unders. Bull., No. 70.

WATTERSON, J. 1975. Mechanism for the persistence of tectonic lineaments. Nature, London, Vol. 253, 520-522.

WESTBROOK, G. K. 1972. Structure and metamorphism of the Lewisian of east Tiree, Inner Hebrides. Scott. J. Geol., Vol. 8, 13-30.

WESTBROOK, G. K. 1974. The South Harris magnetic anomaly. Proc. Geol. Assoc., Vol. 85, 1-12.

WESTOLL, N. D. S., and MILLER, J. A. 1969. The age of some kentallenite intrusives in Argyll. Scott. J. Geol., Vol. 5, 11-14.

WHITE, S. H., BURROWS, S. E. CARRERAS, J., SHAW, N. D., and HUMPHREYS, F. J. 1980. On mylonites in ductile shear zones. 175-187 in Shear zones in rocks. CARRERAS, J., COBBOLD, P. R., RAMSAY, J. G., and WHITE, S. H. (editors). J. Struct. Geol., Vol. 2.

WHITE, S. H.and GLASSER, J. 1987. The Outer Hebrides Fault Zone: evidence for normal movements. 175-183 in Evolution of the Lewisian and comparable Precambrian high-grade terrains. PARK, R. G., and TARNEY, J. (editors). Spec. Publ. Geol. Soc. London, No. 27.

WILLIAMS, G. E. 1969. Petrography and origin of pebbles from Torridonian strata (Late Precambrian), northwest Scotland. 609-629 in North Atlantic-geology and continental drift. KAY, M. (editor). Mere. Am, Assoc. Pet. Geol., No. 12.

WINDLEY, B. F., HERD, R. K., and BOWDEN, A. A. 1973. The Fiskenmsset complex, West Greenland. Part I: A preliminary study of the stratigraphy, petrology and whole rock chemistry from Qeqertarssuatsiaq. Gronlands Geol. Unders. Bull., No. 106.

WINKLER, H. G. F. 1974. Petrogenesis of metamorphic rocks. (Berlin: Springer-Verlag.)

WITTY, G. J. 1975. The geochemistry of the Roneval Anorthosite, South Harris, Scotland. Unpublished PhD thesis, University of London.

WOOD, B. J. 1975. The influence of pressure, temperature and bulk composition on the appearance of garnet in orthogneissesan example from South Harris, Scotland. Earth Planet. Sci. Lett. , Vol. 26, 299-311.

WOOD, B. J. and BANNO, S. 1973. Garnet-orthopyroxene and orthopyroxene-clinopyroxene relationships in simple and complex systems. Contrib. Mineral. Petrol. , Vol. 42, 109-124.

WRIGHT, A. E., and Bowes, D. R. 1979. Geochemistry of the appinite suite. 699-704 in The Caledonides of the British Isles–reviewed. HARRIS, A. L., HOLLAND, C. H., and LEAKE, B. E. (editors). Spec. Publ. Geol. Soc. London, No. 8.

WRIGHT, T. L. 1974. Magma mixing as illustrated by the 1959 eruption, Kilauea Volcano, Hawaii. Bull. Geol. Soc. Am. , Vol. 84, 849-859.

WRIGHT, T. L. and OKAMURA, R. T. 1977. Cooling and crystallization of tholeiitic basalt, 1965 Makaopuhi Lava Lake, Hawaii. Prof. Pap. US Geol. Surv., No. 1004.

Figures, plates and tables

Figures

(Figure 1) The Outer Hebrides showing main sources of data for compilation of the 1:100 000 maps.  1 BGS survey. 2 Bowes and Hopgood (1969a). 3 Hopgood (1964, 1971). 4 Francis (1969, 1973). 5 Coward (1969, 1972, 1973). 6 Kursten (1957). 7 Steavenson (1928, unpublished data deposited with BGS). 8 Sibson (1977b). 9 Graham (1970). 10 Dearnley (1963). 11 Palmer (1971). 12 Witty (1975). 13 Horsley (1978). 14 Dickinson (1974). 15 Brodie (1975). 16 Bishop (1975). 17 Thamdrup (personal communication). 18 Myers (1968, 1970a, 1970b, 1971). 19 Taft (1972). 20 Soldin (1977). 21 Steel and Wilson (1975). 22 Lisle (1974). 23 Watson (1968, 1969) and unpublished data; and Davies, (1974a). See text for details and other references.

(Figure 2) Geological sketch map of the Outer Hebrides Metasediments and metavolcanics: A—Ness, B—Laxavat; C—Langavat; D—Leverburgh; E—Sound of Harris; F—North Uist; G—Benbecula; H—South Uist Meta-igneous complexes: a—Ness; b—South Harris; c—Market Stance; d—Corodale Gneiss; e—East Barra

(Figure 3) Classification of quartzofeldspathic gneisses by volume per cent of modal minerals (after Streckeisen, 1976). a—granodiorite field; b—tonalite field.

(Figure 4) Distribution of various elemental ratios in the quartzofeldspathic gneisses, showing low-, medium- and high-range values. The top map shows the number of samples within each quarter.

(Figure 5) Plot of various elemental ratios against SiO2 for the quartzofeldspathic gneisses to show geographical variation across Lewis and Harris: represent samples west of grid easting NB 1100; represent samples east of grid easting NB 1100.

(Table 5). x transitional rocks between basic and metasediment from (Table 5). garnetiferous metasediments from (Table 5), Sheraton et al. (1973), Palmer (1971) and Tarney and others (1972). 0:a average of four Langavat metasediments (Palmer, 1971; (Table 10)); b semipelites from the Langavat group (Table 10). +:D, average of Dalradian slates (Hickman, 1975, table 3); M average of Moine pelites (Butler, 1965, (Table 3)); G average granodiorite (Le Maitre, 1976). average of mainland Scourian gneisses, numbers refer to analyses in (Table 3). average of Outer Hebrides and mainland Laxfordian gneisses, numbers refer to analyses in (Table 3). Dashed lines define general field of metasediments (upper) and quartzofeldspathic gneisses (lower)" data-name="images/P936479.jpg">(Figure 6) AFM plot (weight per cent oxides) of pre-Scourian gneiss units. â–ª layered basics from (Table 5). x transitional rocks between basic and metasediment from (Table 5). garnetiferous metasediments from (Table 5), Sheraton et al. (1973), Palmer (1971) and Tarney and others (1972). 0:a average of four Langavat metasediments (Palmer, 1971; (Table 10)); b semipelites from the Langavat group (Table 10). +:D, average of Dalradian slates (Hickman, 1975, table 3); M average of Moine pelites (Butler, 1965, (Table 3)); G average granodiorite (Le Maitre, 1976). average of mainland Scourian gneisses, numbers refer to analyses in (Table 3). average of Outer Hebrides and mainland Laxfordian gneisses, numbers refer to analyses in (Table 3). Dashed lines define general field of metasediments (upper) and quartzofeldspathic gneisses (lower)

(Table 3). x metasediments from analyses 1 and 2, (Table 5); and from Tarney and others (1972, table 1, analyses 5, 8, 11, 12 and table 2, analysis 9)." data-name="images/P936480.jpg">(Figure 7) Plot of CaO against Y (CaO in weight per cent, Y in ppm) to differentiate between values for the average gneiss and various metasediments. gneiss values from analyses 1, 2, 3 and 5, (Table 3). x metasediments from analyses 1 and 2, (Table 5); and from Tarney and others (1972, table 1, analyses 5, 8, 11, 12 and table 2, analysis 9).

(Figure 8) Rare-earth element data for South Harris metasediments.

(Figure 9) AFM plot (weight per cent of oxides) of the compositional fields of the quartz gabbros and banded ultramafic/mafic gabbros of the South Harris Igneous Complex. The composition of quartzose paragneisses (p), mainly from xenoliths in the gabbros, is also shown (after Witty, 1975).

(Figure 10) AFM plot (weight per cent of oxides) showing various phases of the South Harris Igneous Complex (data from Horsley, 1978; Witty, 1975), and the Corodale Gneiss of South Uist.

(Table 23), analyses 1-5. A ultrabasic bodies, (Table 23), analyses 6-8. layered basics, (Table 5), analyses 7-11. O anorthosite from layered complex, (Table 5), analysis 6. x Late-Scourian 'microdiorites', (Table 5), analyses 12-15. * Late-Scourian 'granites', (Table 5), analyses 16-17. range of members of mainland Scourie Dyke suite from Tarney (1973). Younger Basic metadolerites (Outer Hebrides: this work). 'Cleitichean Beag' basic from Dearnley (1963, table 7, analysis 4). 0 noritic rocks from Soldin (1978, figure 4.11); 'average' of Maaruig ultrabasic from Soldin (1978, figure 4.11)." data-name="images/P936484.jpg">(Figure 11) AFM plot (weight per cent of oxides) of various intrusives. East Gerinish layered complex, (Table 23), analyses 1-5. A ultrabasic bodies, (Table 23), analyses 6-8. layered basics, (Table 5), analyses 7-11. O anorthosite from layered complex, (Table 5), analysis 6. x Late-Scourian `microdiorites', (Table 5), analyses 12-15. * Late-Scourian 'granites', (Table 5), analyses 16-17. range of members of mainland Scourie Dyke suite from Tarney (1973). Younger Basic metadolerites (Outer Hebrides: this work). 'Cleitichean Beag' basic from Dearnley (1963, table 7, analysis 4). 0 noritic rocks from Soldin (1978, figure 4.11); 'average' of Maaruig ultrabasic from Soldin (1978, figure 4.11).

(Figure 12) Distribution of Younger Basic assemblages present in the cores of larger dykes. cpx-(opx)-(gt)- (hb1)-pl (Granulite-facies); cpx-hbl-(gt)-pl (Sub-granulitefacies); o hbl-(gt)-pl (Ainphibolitefacies); A hbl-bi-pl (Lower amphibolite facies)

(Figure 13) Plot of pyroxene compositions from Younger Basics (including data from Soldin (1978)). metamorphic clinopyroxene.; igneous clinopyroxene. Orthopyroxene.; Fields: 1 salite; 2 hypersthene; 3 ferrohypersthene, (after Deer, Howie and Zussmann, 1978, figure 1).

(Figure 14) Plot of meionite content against anorthite content for co-existing scapolites and feldspars respectively, from members of the Younger Basic suite.

(Figure 15) Location and distribution of analysed specimens of Younger Basics.

(Figure 16) Harker plots of various oxides against Mg0 (weight per cent) for Younger Basics. plots of (S58793); and (S59777) (see text for explanation).; x Q-normative dykes. oo/-normative dykes.

(Table 23), analysis 14)." data-name="images/P936490.jpg">(Figure 17) ACF plot (weight per cent of oxides) showing Younger Basic assemblages. Dashed tie lines are for 'granulite-facies' assemblages; solid tie lines are for amphibolite-facies assemblages. Mineral fields are based on data from analyses given in the Appendix. marks average value for Younger Basics ((Table 23), analysis 14).

(Figure 18) Map of the Uig Hills—Harris granite complex, showing areas of porphyritic granite, granite veining and location of analysed samples.

(Figure 19) Classification of Late-Laxfordian granites by volume per cent of modal minerals (after Streckeisen 1976).; granites.; porphyritic granites.; groundmass of porphyritic granite.; The granite, granodiorite and tonalite fields are marked a, b and c respectively.

(Figure 20) Plot of major oxides (weight per cent) against Si02 (weight per cent) for the Late- Laxfordian granites. x porphyritic granite. Granite. o leucogranite

(Figure 21) Plot of trace elements (in ppm) against Si02 (weight per cent) for Late-Laxfordian granites. Symbols as for (Figure 20).

(Figure 22) Plot of normative corundum (c) and diopside (dz) against Si02 (weight per cent) for Late-Laxfordian granites. Symbols as for (Figure 20).

(Figure 23) Compositions of gneiss and Late-Laxfordian granite in normative Q-ab-or and an-ab-or diagrams with the 1 kb and 5 kb cotectics (from Strong, 1979). quartzo-feldspathic gneiss. granite. leucogranite.

(Figure 24) Ba/Sr and Ba/Rb plots for Late-Laxfordian granites. granite. Leucogranite. MFT- modelled Raleigh fractionation trend.

(Table 3), analyses 7 and 8). C and D average values for mainland Laxfordian gneiss ((Table 3), analyses 4 and 5). " data-name="images/P936498.jpg">(Figure 25) Plot of K/Ba against Ba for gneisses and Late-Laxfordian granites. gneiss east of grid easting NB 1100 gneiss west of grid easting NB 1100 granite. leucogranite. trend of granite. -- approximate boundary between the two gneiss groups. A and B— average values for mainland Scourian gneiss ((Table 3), analyses 7 and 8). C and D— average values for mainland Laxfordian gneiss ((Table 3), analyses 4 and 5).

(Figure 27) Plot of Ba/Rb against Rb for gneisses and Late-Laxfordian granites. Symbols as for (Figure 25)." data-name="images/P936499.jpg">(Figure 26) Plot of K/Rb against Rb for gneisses and Late-Laxfordian granites. Symbols as for (Table 3), analyses 7 and 8). C and D average values for mainland Laxfordian gneiss ((Table 3), analyses 4 and 5). " data-name="images/P936498.jpg">(Figure 25). (Figure 27) Plot of Ba/Rb against Rb for gneisses and Late-Laxfordian granites. Symbols as for (Table 3), analyses 7 and 8). C and D average values for mainland Laxfordian gneiss ((Table 3), analyses 4 and 5). " data-name="images/P936498.jpg">(Figure 25).

(Figure 27) Plot of Ba/Rb against Rb for gneisses and Late-Laxfordian granites. Symbols as for (Table 3), analyses 7 and 8). C and D average values for mainland Laxfordian gneiss ((Table 3), analyses 4 and 5). " data-name="images/P936498.jpg">(Figure 25).

(Figure 28) Histogram to illustrate degree of cataclasis in different Late-Laxfordian granite types. The vertical axis shows the percentage of granite specimens in which more than 50% of the feldspar crystals show evidence of granulation in thin section. P, porphyritic granite; G, granite; L, leucogranite.

(Figure 29) Map of South Lewis and North Harris showing pseudotachylite and ultramylonite/mylonite localities.

(Figure 30) Sketch maps to show principal geological and geophysical features of the Outer Hebrides.Left, main lithological units. Metasedimentary and metavolcanic rocks; A, Ness; B, Laxavat; C, Langavat; D, Leverburgh; E, Sound of Harris; F, North Uist; G, Benbecula; I-I, South Uist. Meta-igneous complexes; a, Ness; b, South Harris; c, Market Stance; d, Corodale Gneiss; e, East Barra. Centre, main geophysical features (after McQuillin and Watson, 1973). Right, main structural features.

(Figure 31) dL2 structures. Left, fold axes. Right, axial planes and fold profiles.

(Figure 32) dL3 structures. Left, general features. Right, fold axes.

(Figure 33) Left, dL3 axial planes and shear zones. Right, significance of linear elements.

(Figure 34) Diagrammatic sketch of small-scale mylonite zone formation taken from field localities in the Loch Bhrollum area, south-east Lewis. Above: mylonite zone parallel to gneissic banding [NB 3253 0331]. Below: complex geometry resulting from mylonite formation at different times and with differing shear-senses [NB 3262 0337].

(Figure 35) Lower hemisphere stereoplot showing poles to the mylonite fabric in the Parc district of southeast Lewis.

(Figure 36) Lower hemisphere stereoplot showing orientation data for the crenulation ' folds of the mylonite fabric in south-east Lewis, South Harris, and North Uist. Data from Sibson (1977b) is included.

(Figure 37) Azimuthal distributions and regional correlations of post-Lewisian dyke-swarms. a Permo-Carboniferous quartz-dolerite suite. Map showing relationship of the Sound of Barra swarm to the equivalent swarm on the Scottish mainland. Dashed lines mark the limit of Outer Hebrides swarm and the limit of abundant plugs and dykes on the mainland. b Permo-Carboniferous camptonite-rnonchiquite suite. Azimuth distribution rose diagrams for the Outer Hebrides swarm compared with the other major camptonite swarms in the West Highlands (adapted from Rock, 1983, fig.1). Lengths of arms are proportional to the percentage of the total number of dykes within a given range of azimuths, and the areas of circles are proportional to the total number of dykes in each swarm. Number of dykes in individual swarms ranges from 22 for Coll and Tiree to 1127 for Eil- Arkaig. Centres of circles and rose diagrams are positioned as near the geographical centre of each swarm as space permits. c Tertiary basalt/dolerite swarms. Rose diagrams, constructed as in b and superimposed on the map showing major axes of dilation for the Inner Hebridean swarms. Data are adapted from Mattey and others, (1977, figure 1), Speight and Mitchell (1979, figure 3), and Speight et al. (1982, figure 33.5). The thickness of the axis is proportional to crustal dilation at each point along its length. Note: The axis of the Lewis rose diagram is offset to the south-west for reasons of space. Its actual position coincides with the dilation axis through the Shiant Isles.

(Figure 38) Palaeogeography of the Stornoway Formation showing the position of fault scarps and progressive alluvial fan development in Permo-Triassic times (after Steel and Wilson, 1975).

Plates

Cover photograph Eaval from Baymore, Grimsay (Benbecula). The upper part of the scarp and the dip slope to the east are formed of resistant pseudotachylite-gneiss breccia. This marks the position of the Outer Hebrides Thrust in southern North Uist. (D03197)

Frontispiece Castlebay and Heaval (383 m), Barra. The feature running upwards from the building on the right marks the position of a thrust within the Outer Hebrides Thrust zone (D 2936)

(Plate 1) Looking south from Achmore across Central Lewis (D02679)

(Plate 2) Grimsay and Eaval (347 m), North Uist, showing the dip slope of the Outer Hebrides Thrust (D03196)

(Plate 3) Folded and boudinaged Younger Basic dykes. Howmore Quarry, South Uist [NE 7659 3645] (D02918)

(Plate 4) Photomicrographs of textures in Lewisian gneiss a Highly strained quartzofeldspathic gneiss. Glen Langadale, Harris, (S71999). X 9.5. (PMS 318) b Garnetiferous metasediment. Claddach Illeray, North Uist. (S61278). X 9.5. (PMS 319) c Typical hornblende-biotite gneiss. Eilean na Cille, Benbecula. (S60106). X 9.5. (PMS 314) d Foliated biotite gneiss, Loch Laxavat Ard, Lewis. (S63150). x9.5. (PMS 313) e Hornblende-pyroxene gneiss. Fuiay, Barra. (S58051). X 9.5. (PMS 320) f Foliated hornblende-biotite gneiss. Tiumpan Head, Lewis. (S63154). X9.5. (PMS 316)

(Plate 5) Banded Basic body with alternating mafic and feldspathic layers. Rubh' Aird-mhicheil, South Uist [NF 7347 3390] (D02922)

(Plate 6) Mafic concentrates defining a folded fabric in Late-Scourian dyke. Leanish Point, Barra [NL 7032 9876] (D02941)

(Plate 7) Undeformed Younger Basic dyke (D) cutting a Late-Scourian pegmatite (P) which itself cuts a late-Scourian dyke (I). Leanish, Barra [NL 7026 9863] (D02945)

(Plate 8) Undeformed Younger Basic sheet (D) cutting a strongly folded Late-Scourian dyke (I). Rubha Carraig-chrom, Fuday. [NF 7435 0840] (D02951)

(Plate 9) Undeformed Younger Basic dyke (D) cutting a late pegmatite (P) which itself cuts a deformed Late-Scourian dyke (I). South-east Fuday [NF 7429 0843] (D02908)

(Plate 10) Younger Basic dyke cutting Scourian migmatitic gneiss. Loch Leosaid, North Harris [NB 055 083] (D02704)

(Plate 11) Folded Younger Basic dyke cutting Scourian foliation. South of Creagan Rudha, North Harris [NB 019 096] (D02702)

(Plate 12) Boudinaged and migmatised Younger Basic dyke. Howmore Quarry, South Uist [NF 7659 3645] (D02915)

(Plate 13) Edge of shear zone in mixed gneiss, with shear zone marked by planar gneiss on right of photograph. The Younger Basic dyke is slightly transgressive to the shear zone fabric. Aird Fenish, Lewis [NA 992 294] (D02693)

(Plate 14) Photomicrographs of textures in Younger Basics a 'Spongy' amphiboles sieved with quartz. Carishader, Lewis. (S57915) X 9.5. (PMS 305) b Garnets rimmed by plagioclase. Kearnaval, Lewis. (S61962) X 9.5. (PMS 302) c Foliated amphibolite. Fornaval, Lewis. (S57157) X 9.5. (PMS 306) d Equigranular two-pyroxene basic with relict (?) igneous plagioclase. Orosay, Barra. (S58905) X 9.5. (PMS 311) e Equigranular basic showing relic ophitic texture. Oaval, North Uist, (S62282) x 9.5. (PMS 308) f Plagioclase rimmed by garnets. The rock shows a relict igneous texture retained. Ard More Mangersta, Lewis. (S58589) x 9.5. (PMS 309)

(Plate 15) Folded quartzofeldspathic gneiss net-veined by pseudotachylite. Greian Head, northwest coast of Barra [NF 6582 0402] (D02962)

(Plate 16) Typical close-jointed vein of pseudotachylite cutting quartzofeldspathic gneiss. Northwest coast of Barra [NF 6571 0474] (D02959)

(Plate 17) Photomicrographs of textures in rocks from the Outer Hebrides Thrust Zone a Pseudotachylite vein with quartz fragments and feldspar microlites. Beinn Bheag Deas. MC 29633 X 42, (PMS 331) b Retrogressed pseudotachylite with relic microlite texture. North-east Wiay. (S60123) X 24. (PMS 330) c Flow-banding and enclosed pseudotachylite fragments within a pseudotachylite vein. Sandray. (S58137) X 23. (PMS 327) d Retrogressed pseudotachylite vein with darker margin and microlite texture. North east Wiay. (S60123) x 9.5. (PMS 322) e Pseudotachylite veining in hornblende-gneiss. Sandray. (S58137) x 7. (PMS 324) f Mixed pseudotachylite and cataclasite (different ages). South Uist. MC 31481 x 7. (PMS 325)

Tables

(Table 1) List of main Lewisian events in the Outer Hebrides

(Table 2) Modal analyses of quartzofeldspathic gneisses

(Table 3) Average chemical analyses of quartzofeldspathic gneisses

(Table 4) Comparison of various elemental ratios for 'non-migmatic' gneisses both outwith and within the general 'zone of migmatisation'. Arbitrary values of K2O are chosen for comparative purposes

(Table 5) Chemical analyses of various Scourian and pre-Scourian rock types

(Table 6) Mineral assemblages from the Langavat Belt

(Table 7) Mineral assemblages from the Leverburgh Belt (after Dickinson, 1974)

(Table 8) Chemical analysis of garnet (average of six) from kyanite-garnet-gneiss at Traigh na 'h Uidhe, Northton (after Dickinson, 1974).

(Table 9) Chemical analyses of metasediments from the Leverburgh Belt

(Table 10) Chemical analyses of metasediments from the Langavat Belt

(Table 11) Modal analyses of Late-Scourian intrusions from South Uist Barra and South Uist

(Table 12) Mineral compositions from the gabbro rim of the anorthosite complex, South Harris Igneous Complex (after Witty, 1975).

(Table 13) Average chemical analyses of gabbros from the South Harris Igneous Complex (mainly from Witty, 1975).

(Table 14) Mineral compositions from the anorthosite complex of the South Harris Igneous Complex (after Witty, 1975)

(Table 15) Chemical analyses of members of the anorthosite, South Harris Igneous Complex (after Witty, 1975).

(Table 16) Mineral compositions from the norite of the South Harris Igneous Complex (after Horsley, 1978).

(Table 17) Chemical analyses of the norite, South Harris Igneous Complex.

(Table 18) Mineral compositions from the diorite, South Harris Igneous Complex (after Horsley, 1978)

(Table 19) Chemical analyses of the diorite and tonalite of the South Harris Igneous Complex.

(Table 20) Average chemical analyses of ultramafic rocks intruding the norite and diorite of the South Harris Igneous Complex.

(Table 21) Clinopyroxene composition from the late potash-rich basic dykes (average of 6 points) after Horsley (1978)

(Table 22) Average chemical analysis of the late potash-rich basic intrusions (3 specimens) from the South Harris Igneous Complex; after Horsley (1978, p.395)

(Table 23) Chemical analyses of various Older and Younger Basic and ultrabasic intrusions (including a Late-Laxfordian microdiorite).

(Table 24) Average chemical analyses of Younger Basics exhibiting different degrees of amphibolitisation.

(Table 25) Average chemical analyses of Late-Laxfordian granites.

(Table 26) Modal analysis of quartz-microdiorite dyke, Garry a-siar, Benbecula ((S59796)A)

(Table 27) Structural history

(Table 28) Chemical analyses of (1) garnet-biotite meta-sediment (mean of 2 analyses from wall rocks adjacent to pseudotachylite vein) (2) pseudotachylite melt plus quartz and feldspar porphyroclasts (3) pseudotachylite with 9% quartz and 11 % plagioclase (An30) subtracted (from Sibson, 1977)

(Table 29) Representative mineral analyses from mylonite and retrograde pseudotachylite

(Table 30) Chemical analyses of Outer Hebrides post-Lewisian minor intrusions

(Table 31) Modal analyses of Outer Hebrides post-Lewisian minor intrusions

(Table 32) Statistics for the major Tertiary dyke-swarms of the Outer Hebrides, with comparative data for the Mull and Skye swarms

(Table 33) Geological succession on the Shiant Isles

(Table 34) (image) Modal variation in the main sill on Garb Eilean, Shiant Isles * Heights are given above sea-level instead of above the base of the sill because the base is not exposed on Garbh Eilean

(Table 35)  (image) Modal variation in the lower sill, Shiant Isles

(Table 36) Stornoway Formation - detailed stratigraphy and palaeogeography (after Steel and Wilson, 1975)

Maps

(Map 1) Lewis and Harris (north and south), 1: 100 000 solid edition.

(Map 2) Dist and Barra (north and south), 1: 100 000 solid edition.

(Map 3) Lewis and Harris (north and south), 1: 100 000 structure.

(Map 4) Dist and Barra (north and south), 1: 100 000 structure.

5 South-west part of South Harris, 1:25 000 (solid) .

Tables

(Table 2) Modal analyses of quartzofeldspathic gneisses

Thin Section No. Grid reference Quartz Plagioclase K-feldspar Biotite Hornblende Epidote Others
(S61122) [NG 0551 9718] 35.4 56.2 3.4 4.1 – 0.7 0.2
(S61135) [NB 0633 0934] 32.7 48.2 2.2 11.8 – 4.9 0.2
(S58832) [NB 2415 3175] 30.0 49.3 8.8 11.0 – 0.7 0.2
(S58768) [NB 0741 3642] 29.6 59.0 1.3 8.6 – 1.4 0.1
(S61144) [NB 0761 0651] 27.0 49.0 0.7 13.1 6.0 3.8 0.4
(S61139) [NB 0168 0983] 25.2 55.8 2.6 10.7 4.9 0.3 0.5
(S58835) [NB 2045 2890] 22.9 44.9 7.5 8.9 12.0 3.6 0.2
(S58825) [NB 0327 2756] 22.4 57.7 2.2 8.0 0.6 8.5 0.6
Mean 28.1 52.6 3.6 9.6 3.0 3.0 0.3
RL* 27.2

57.8

12.7

2.3

(Table 3) Average chemical analyses of quartzofeldspathic gneisses

1 2 3 4 5 6 7 8
SiO2 68.72 68.00 67.1 69.4 66.72 61.0 61.22 64.56
TiO2 0.32 0.36 0.34 0.4 0.34 0.6 0.54 0.47
Al2O3 15.07 15.29 15.48 14.7 16.04 15.7 15.64 15.74
Fe2O3 0.73 1.70 1.26 0.9 1.94 2.9 3.07 2.53
FeO 1.87 1.65 2.38 2.1 1.47 3.0 2.57 2.00
MnO 0.04 0.04 0.05 0.05 0.04 0.1 0.08 0.06
MgO 1.19 1.20 1.44 1.6 1.44 3.60 3.36 2.23
CaO 2.99 3.51 4.81 3.1 3.18 5.8 5.57 4.50
Na2O 4.27 4.20 4.62 4.4 4.90 4.3 4.42 4.60
K2O 2.70 2.27 1.5 2.0 2.09 1.0 1.03 1.15
P2O5 0.11 0.13 0.1 0.14 0.2 0.18 0.16
Ba 797 787 809 795 713 720 757 779
Ce 65 71 50 48 42
Co 28 19 66 35 45
Cr 39 25 26 <50 32 100 88 48
Cu 31 15 25 35 50
Ga 12 10 15 15 17.5
La 53 50 43 55 32 20 20
Li 29 14 22 10
Nb 4 6 5 5 5
Ni 17 21 23 25 20 65 58 37
Pb 19 46 6 22 13 12
Rb 122 83 41 85 74 10 11 13
Sr 564 466 370 530 580 545 569 565
Y 8 7 11 7 10 9 8
Zn 48 71 35 45 80
Zr 171 176 185 135 193 190 202 197
K/Rb 184 227 304 195 234 763 737
K/Sr 39.7 40.4 33.6 31.3 30.5 15.5 17.0
K/Ba 28.1 23.94 15.39 20.9 24.3 11.3 12.3
Rb/Sr 0.22 0.18 0.11 0.16 0.13 .019 .023
Ba/Rb 6.5 9.5 19.7 9.4 9.6 68 60
Ca/Sr 38 54 93 39 70 57
Ba/Sr 1.41 1.69 2.19 1.3 1.3 1.4
Ca/Y 2669 3580 3122 3200 4400 4000

(Table 4) Comparison of various elemental ratios for 'non-migmatic' gneisses both outwith and within the general 'zone of migmatisation'. Arbitrary values of K2O are chosen for comparative purposes

K2O 1.5% 1.5% 2.0% 2.0% 2.5% 2.5% 5.7%
1 2 1 2 1 2
K/Rb 140 190 178 212 200 360 640
Ba/Sr 0.4 0.5 0.5 1.1 1.2 1.2 0.4
K/Sr 26 15 26 30 33 28 152
Ca/Sr 35 45 37 72 32 45 32
Ba/Rb 4 7 4 7 7 15 2

(Table 5) Chemical analyses of various Scourian and pre-Scourian rock types

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
SiO2 68.1 – 77.1 65.4 60.5 47.5 49.6 57.3 49.9 52.7 47.30 47.0 54.8 55.6 50.2 55.98 60.10 – 54.0
TiO2 0.53 – 0.22 0.75 0.98 0.07 1.76 1.4 0.77 1.53 1.94 1.20 0.82 1.9 1.4 1.03 0.66 – 0.84
Al2O3 14.4 – 11.1 12.4 12.7 28.6 12.5 11.3 14.9 12.2 14.26 18.29 18.80 17.6 18.4 18.26 18.68 – 16.9
Fe2O3 1.3 – 0.7 1.9 2.6 1.3 3.9 5.6 3.6 2.7 2.87 4.14 3.83 3.0 3.7 3.23 2.43 – 4.3
FeO 4.0 – 3.7 6.8 6.7 0.7 13.1 10.3 8.5 12.6 10.78 6.40 3.44 5.6 6.7 3.70 2.06 – 4.2
MnO 0.09 – 0.06 0.15 0.19 0.06 0.24 0.20 0.22 0.23 0.18 0.14 0.12 0.09 0.13 0.11 0.04 – 0.14
MgO 2.7 2.24 0.4 2.1 2.9 0.4 5.5 2.6 7.7 3.6 7.17 5.77 2.76 2.8 4.0 2.45 1.11 2.98 4.4
CaO 3.1 3.50 3.1 3.2 7.4 16.0 9.6 8.3 11.5 9.0 10.59 8.46 4.24 5.2 7.7 4.50 2.63 4.65 6.6
Na2O  3.4 – 1.3 2.3 2.0 3.0 1.1 1.6 1.9 1.9 2.63 3.83 5.00 4.7 4.8 4.90 4.22 – 5.0
K2O 1.8 2.08 1.3 1.3 0.4 0.2 0.6 0.20 0.10 0.40 0.49 1.73 3.39 2.5 1.2 2.88 6.34 3.74 2.2
P2O5 0.10 – – – – – – 0.15 0.04 – 0.18 0.49 0.82 0.88 0.67 0.75 0.24 – 0.41
CO2 – – – – – – – – – – 0.05 0.45 0.25 – – 0.53 0.15 – –
Ba 511 632 100 340 185 35 10 42 12 90 85 1100 2500 1272 729 2500 3500 1832 929
Co 32 18 60 70 70 40 85 26 81 90 80 40 130 35 45 20 23 22 8
Cr 201 102 <10 <10 15 <10 105 19 346 <10 205 25 <10 1 0 10 <10 19 30
Cu 24 42 45 85 90 <10 140 37 88 20 310 65 24 58 61 10 27 45 10
Ga 13 18 – – – – – 18 5 – 21 24 16 18 22 16 18 17 21
La 49 46 – – – – – 0 0 – – – – 137 130 – – 39 45
Li 22 21 – – – – – 21 18 – 9 10 12 1 30 19 8 10 16
Mo 1 0 – – – – – 0 5 – – – – 0 0 – – 0 0
Ni 25 46 <10 <10 <10 30 65 5 215 30 120 41 <10 15 4 <10 <10 14 9
Pb 0 13 – – – – – 0 0 – – – – 0 0 – – 8 44
Rb 84 55 – – – – – 5 3 – 17 – – 102 41 – – 74 39
Sr 408 584 55 170 150 550 60 161 102 260 205 1100 1500 1282 928 1000 950 1037 1016
V 95 81 50 65 155 60 515 47 472 365 310 180 79 173 218 100 82 128 61
Y 14 14 – – – – – 56 18 – 24 – – 18 19 – – 19 32
Zn 50 81 55 45 100 <10 125 80 80 100 120 – – 140 160 – – 93 90
Zr 308 85 – – – – – 354 69 – 150 100 200 847 394 50 300 355 89
CIPW norm 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Q 28.54 – 54.62 32.54 25.76 – 8.15 24.08 2.19 11.47 – – 0.96 3.79 – 4.11 4.77 – –
or 10.64 – 7.68 7.68 2.36 1.18 3.55 1.18 1.18 2.36 2.90 10.22 20.03 14.77 7.09 17.02 37.46 – 13.0
ab 28.76 – 11.00 19.45 16.92 20.25 9.30 13.53 16.07 16.07 22.24 25.36 42.29 39.75 39.35 41.44 35.69 – 42.29
an 14.72 – 15.38 15.87 24.50 63.98 27.40 23.06 31.54 23.58 25.66 27.61 15.68 19.55 25.12 17.43 10.53 – 17.18
c 1.47 – 0.92 1.39 – 2.78 – – – – – – 1.16 – – 0.70 1.02 – –
ne – – – – – – – – – – – 3.81 – – 0.68 – – – –
wo – –– – – –
di – -- – – 10.17 2.54 16.89 14.48 20.56 17.86 21.11 9.0 – 0.41 7.12 – – – 10.28
hy 12.28 – 6.96 15.19 11.04 – 23.61 11.49 20.92 18.70 7.9 – 8.89 11.60 – 8.73 3.52 – 3.38
ol – – – – – – – – – – 10.43 12.02 – – 9.97 – – – 4.09
mt 1.89 – 1.02 2.76 3.77 1.89 5.66 8.12 5.22 3.92 4.16 6.00 5.55 4.35 5.37 4.68 3.52 – 6.24
il 1.01 – 0.42 1.42 1.86 0.13 3.34 2.66 1.46 2.91 3.68 2.28 1.56 3.61 2.66 1.96 1.25 – 1.60
ap 0.24 – – – – – – 0.36 0.09 – 0.43 1.16 1.94 2.09 1.59 1.78 0.57 – 0.97

(Table 6) Mineral assemblages from the Langavat Belt

Calcareous rocks
cc-do-serp-di (green hbl)-scp- phl-act-sph-mt-hum Bàgh Steinigie (S63080)
fo(serp)-cc-do-(phl)-(gt) Sta Bay and Bagh Steinigie (Jehu and Craig, 1927; Dearnley, 1963)
cc-do-fo-(br)-(trem)-(hum)-(gt) Bàgh Steinigie (Dearnley, 1963)
cc-di-clz Sta Bay (Dearnley, 1963)
cc-and-sph-phl Allt Sta (Dearnley, 1963)
Pelitic and psammitic rocks
qz-or-olig-bi-mt-py Borve Lodge area (S63072)
qz-or-olig-bi-hbl-ep-sph-ap Borve Lodge area (Jehu and Craig, 1927)
qz-olig-bi-gt-ep Borve Lodge area (Dearnley, 1963)
qz-olig-gt-bi-(ms)-(to)-(mt) Bàgh Steinigie inlier near Bleaval (Myers, 1968)
qz-olig-gt-bi-st-co-sil Bàgh Steinigie inlier near Bleaval (Myers, 1968)
qz-or-olig-bi-ep/clz-mt-sph-ap Stetteval (Dearnley, 1963)
qz-olig-gt-trem-(ep) Borsham (Dearnley, 1963)

Minor mineral components are shown in brackets.

(Table 7) Mineral assemblages from the Leverburgh Belt (after Dickinson, 1974)

Marbles
cc-do-fo-di-sph-(phl)-(pa)
cc-do-fo-di-(phl)
cc-do-fo-phl-(pa)-(scp)
Calc-silicates di-trem
di-scp
phl-pa-(scp)
Quartzose gneisses gt-bi-pl-qz
gt-hbl-scp-di-pl-qz-(bi) hy-qz-pl-bi
gt-hbl-qz-pl-scp-(bi) gt-bi-(scp)-(sil)-qz-pl-(hb)
qz-pl-mic-bi-(ms)-(chl)
Pelitic gneisses Rodel - Leverburgh area
ky-bi-co-pl-kfp-qz-grn Rodel - Leverburgh area
ky-bi-qz-gt-(sil)-(p1)-(kf) Rodel - Leverburgh area
qz-bi-pl-(gt) Rodel - Leverburgh area
qz-grn-ph1-(ky)-(p1) Rodel - Leverburgh area
ky-gt-bi-sil-qz-p1-(mic)-(ms) Chaipaval - Northton area
gt-bi-qz-pl-(ky)-(mic) Chaipaval - Northton area
bi-qz-pl-(gt)-(sil) Chaipaval - Northton area
bi-qz-pl-ms-(sil) Chaipaval - Northton area
Perthite gneisses
kf (mic)-ky-gt-bi-sil-qz-pl
kf (mic)-gt-bi-qz-pl
kf (mic)-gt-bi-sil-qz-pl-(ky)
Minor mineral components are shown in brackets.

(Table 8) Chemical analysis of garnet (average of six) from kyanite-garnet-gneiss at Traigh na 'h Uidhe, Northton (after Dickinson, 1974).

SiO2 36.79 Si 5.480

{6.000

 Al 68.0
Al2O3 20.69 Al 0.160 Py 27.2
FeO 33.35 Al 3.712

{6.512

 Gr 4.7
MnO 0.05 Fe 4.427 Sp 0.1
MgO 7.50 Mn 0.007
CaO 1.79 Mg 1.774
Ca 0.304
Total 100.17

(Table 9) Chemical analyses of metasediments from the Leverburgh Belt

1 2 3 4 5
SiO2 67.36 66.35 70.80 71.80 –
TiO2 0.64 0.65 0.40 0.39 0.40
Al2O3 14.61 14.47 14.38 14.20 –
Fe2O3 2.45 2.37

4.14†

 1.37

2.73†
FeO 3.31 4.98 1.05
MnO 0.06 0.08 0.06 0.02 0.025
MgO 2.86 3.15 1.20 1.17 1.80
CaO 3.20 3.06 4.48 1;86 3.55
Na2O 2.52 2.56 3.53 3.08 2.22
K2O 2.75 2.07 0.63 3.98 2.26
P2O5 0.13 – nd – –
Total 99.89 99.74 99.62 98.92 –
1 2 3 4 5
Ba 1100 809 967 823
Cl 180 200 100 –
Cr 120 166 90 23
Ni 36 38 27 9
Pb 20 – – 5
Rb 59 57 97 57*
Sr 405 323 308 521
Th 7 – – 6.7
Y 19 – – 15
Zr 288 396 – 139
To Wt. oxides 1 2 3
SiO2 67.30 – –
TiO2 0.34 1.00 0.15
Al2O3 15.40 – –
Fe2O3 1.96 7.31* 0.77*
FeO 0.95 – –
MnO 0.03 0.02 –
MgO 1.04 2.37 0.56
CaO 2.90 1.83 0.30
Na2O 5.40 2.99 0.97
K2O 1.83 2.80 2.72
Total 97.15 – –
Trace elements (ppm) 1 2 3
Ba 808 498 1600
Cl 83 – –
Cr 12 170 60
Ni 12 39 8
Pb – 37 5
Rb 91 86 39
Sr 570 197 136
Th – 6.2 1.6
Y – 12 1
Zr – 211 212

(Table 11) Modal analyses of Late-Scourian intrusions from South Uist Barra and South Uist

Quartz Potash-Feldspar Plagioclase Hornblende Biotite Apatite Opaques
Microdiorite dyke, Leanish, Barra (S57801) 0.3 11.4 56.6 4.8 20.7 1.9 4.3
Microdiorite dyke, Leanish, Barra (S57802) 0.1 0.0 45.8 37.8 10.2 2.1 4.0
Microdiorite, Ludac, South Uist (S60375) 8.0 1.3 60.2 7.7 19.6 2.5 0.7
Intermediate dykes, east coast of Barra. Average of 5 modes (Francis, 1973) 0.0 0.0 53.4 28.1 14.8 0.8 2.9
Metadiorite, Rudha Charnain, Barra (S58101) 1.9 0.0 68.6 6.6 15.4 1.7 5.8
Metadiorite, Earsary, Barra, (S59979) 2.7 5.0 67.9 11.9 6.2 2.2 4.1
Quartz-monzodiorite, Earsary, Barra (S57805) 5.7 13.6 55.7 5.1 17.0 0.6 2.3
Quartz-monzonite, east coast of Barra (S57804) 5.4 40.8 39.2 2.7 8.6 0.2 3.1
Monzonite, east coast of Barra (S59977) 3.0 52.3 36.1 2.7 2.8 0.5 2.6

(Table 12) Mineral compositions from the gabbro rim of the anorthosite complex, South Harris Igneous Complex (after Witty, 1975).

1 2 3
SiO2 40.88 49.90 39.49
TiO2 0.85 0.38 0.13
A12O3 14.45 7.17 20.76
Fe2O3 – – 3.14
FeO 12.00 8.68 16.56
MnO 0.09 0.19 0.41
MgO 13.12 13.31 9.37
CaO 12.30 20.17 10.66
Na2O 1.98 0.90 –
K2O 1.13 – –
Cr2O3 0.21 0.12 0.04
Total 97.01 100.82 100.56
1 2 3
Si 6.093

8.000

 1.838

2.000

 Si 5.958

6.000
Al 1.907 - 0.162 Al 0.042
Al 0.631

5.173

 0.149

}2.027

 Al 3.650

4.027
Fe2+ 1.496 0.267 Cr 0.005
Mn 0.011 0.006 Ti 0.015
Mg 2.915 0.731 Fe3+ 0.357
Cr 0.025 0.003 Mg 2.107

5.973
Ti 0.095 0.011 Fee* 2.090
Ca 1.954- 0.796 Mn 0.053
Na 0.572

0.787

 0.064 Ca 1.723
K 0.215 0 24.000
0 23.000 6.000
Fe/Mg 0.517 0.374 1.017
Fe/Fe + Mg 0.341 0.272 0.504
Niggli mg 0.659
En 44.73 Al 34.99
Fs 14.90 Py 35.28
Wo 44.37 Gr 19.78
An 8.71
Sp 0.88
Others 0.36

(Table 13) Average chemical analyses of gabbros from the South Harris Igneous Complex (mainly from Witty, 1975).

1 2 3 4
SiO2 43.08 50.46 44.2 47.56
TiO2 0.86 1.07 1.1 1.34
Al2O3 16.69 13.71 13.4 13.97
Fe2O3 4.47 4.41 2.6 6.30
FeO 9.32 8.31 14.4 9.40
MnO 0.21 0.21 0.2 0.25
MgO 9.36 7.00 9.6 7.32
CaO 14.67 10.87 12.7 12.38
Na2O 1.00 2.82 1.2 0.84
K2O 0.15 0.47 0.07 0.20
P2O5 - 0.05
Total 99.81 99.33 99.52 99.56
1 2 3 4
Ni 40 125
Cr 354 103
Sr 88 125
Rb – 1
Ba – 96

CIPW norm

1 2 3 4
Q – 0.47 – 5.95
or 0.88 2.78 0.39 1.11
ab 6.13 23.86 10.17 7.34
an 40.61 23.36 30.94 33.64
ne 1.26 – 26.09 22.48
di 26.09 24.86 – –
by (En) – 15.57(63) 3.77(51) 17.17(65)
ol (Fo) 16.72 (63) – 21.93(49) –
mt 6.48 6.39 3.78 9.05
il 1.63 2.03 2.10 2.58

(Table 14) Mineral compositions from the anorthosite complex of the South Harris Igneous Complex (after Witty, 1975)

1 2 3 4 5 6
SiO2 51.00 49.13 39.18 38.72 47.02 48.79
TiO2 0.46 0.88 0.14 n.d. 1.55 0.67
Al2O3 6.87 5.96 21.17 21.17 9.10 8.23
Fe2O3* – – 1.82 2.19 – –
FeO 5.57 12.29 20.26 20.33 11.94 12.25
MnO 0.12 0.02 0.49 0.36 0.06 0.07
MgO 13.24 11.80 9.86 5.91 13.46 13.99
CaO 22.12 18.78 6.79 11.76 11.97 12.07
Na2O 0.79 0.47 – – 2.11 1.26
K2O – – – – 0.52 0.18
Cr2O3 0.11 0.08 n.d. – 0.11
Total 100.28 99.33 99.74 100.44 97.73 97.62
1 2 3 4 5 6
Si 1.867

2.000

 1.858

2.000

 5.973

6.000

 5.955

6.000

 6.864

8.000

 7.071

8.000
AlIV 0.133 0.142 0.027 0.045 1.136 0.929
AlVI 0.163

1.998

 0.124

2.009

 3.777

4.011

 3.792

4.045

 0.430

4.993

 0.477

5.079
Ti 0.013 0.033 0.016 – 0.170 0.073
Cr 0.003 – 0.010 – – 0.013
Fe3+ – 0.208 0.253 – –
Fe2+ 0.170 0.390 2.583

5.989

 2.615

5.995

 1.458 1.485
Mn 0.004 0.001 0.063 0.047 0.007 0.009
Mg 0.722 0.665 2.241 1.385 2.928 3.022
Ca 0.867 0.761 1.103 1.908 1.872 1.874
Na 0.056 0.035 – – 0.597

0.716

 0.354

0.387
K – – – – 0.119 0.033
O 6.000 6.000 24.000 24.000 23.000 23.000
Fe/Mg 0.241 0.506 1.181 1.965 0.498 0.494
Fe/(Fe+Mg) 0.194 0.370 0.541 0.663 0.332 0.331
1 2 3 4 5 6
En 41.03 En 36.64 Al 43.12 Al 43.91 Niggli mg 0.666 Niggli mg 0.669
Fs 9.69 Fs 21.43 Py 37.41 Py 22.75
Wo 49.28 Wo 41.93 Gr 12.94 Gr 26.19
An 4.96 An 6.36
Sp1.06 Sp 0.79

(Table 15) Chemical analyses of members of the anorthosite, South Harris Igneous Complex (after Witty, 1975).

1 2 3 4 5 6
SiO2 48.42 50.76 52.49 50.76 52.24 50.00
TiO2 0.28 0.11 0.13 0.20 0.17 0.23
Al2O3 21.24 25.15 28.90 30.11 27.80 25.33
Fe2O3 1.48 1.14 0.44 0.62 0.24 1.98
FeO 5.25 4.29 1.01 0.53 0.09 2,92
MnO 0.11 0.08 0.02 0.02 – 0.07
MgO 7.34 4.06 0.05 0.28 0.07 3.71
CaO 14.08 9.97 12.86 14.07 16.07 12.67
Na2O 2.12 3.44 3.43 2.64 1.58 2.74
K2O 0.28 0.83 0.35 0.19 1.10 0.39
P2O5 – – 0.02 – – –
Total 100.60 99.83 100.48 99.44 99.36 100.04
1 2 3 4 5 6
Sr 189 206 361 266
Rb 23 7 18
Zr 6 10 3 17
Cr 620 190 50 110
CIPW Norm 1 2 3 4 5 6
Q – – 1.99 3.79 7.50 –
c – 0.47 – – – –
or 1.65 4.91 2.07 1.12 6.50 2.30
ab 17.75 29.11 29.02 22.34 13.37 22.08
an 47.67 49.46 62.68 69.76 65.52 56.76
ne 0.10 – – – – –
wo – – – – 5.37 –
di 17.83 0.52 0.04 0.38 5.78
hy (En) – 5.96 (59.4) 3.04 (60.7) 1.40 (49.7) – 9.45 (72.66)
ol (Fo) 12.85 (67.7) 8.07 (57.1) – – – 0.35 (70.70)
mt 2.15 1.65 0.62 0.90 0.26 2.87
il 0.53 0.21 0.25 0.38 0.19 0.44
sp – – – – 0.17 –
cc – – – – 0.14 –
FeO/FeO + MgO 0.473 0.528 0.678 0.832 0.788 0.559
Total normative feldspar (An) 67.02 (70.94) 82.45 (59.26) 93.77 (64.94) 93.21 (74.33) – 81.14 (69.95)

(Table 16) Mineral compositions from the norite of the South Harris Igneous Complex (after Horsley, 1978).

1 2 3
SiO2 50.15 50.89 61.20
TiO2 0.50 – –
Al2O3 5.21 3.11 25.04
FeO 10.15 24.82 –
MnO 0.55 –
MgO 11.62 21.00 –
CaO 20.77 0.47 6.38
Na2O 1.04 0.02 8.02
K2O – – 0.24
Total 99.44 100.86 100.88
1 2 3
Si 1.864

}2.000

 1.924

}2.000

 2.699
AlIV 0.136 0.076 Al 1.301
AlVI 0.092

}1.966

 0.063

}2.070
Ti 0.014 –
Fe 0.316 0.785
Mn – –
Mg 0.643 1.184
Ca 0.827 0.019 0.301
Na 0.074 0.001 0.685
K – – 0.013
O 6.000 6.000 –
Fe/Mg 0.491 0.663 –
Fe/Fe + Mg 0.330 0.399 –
En 36.0 59.5 An 30.1
Fs 17.7 39.4 Ab 68.5
Wo 46.3 1.0 Or 1.4

(Table 17) Chemical analyses of the norite, South Harris Igneous Complex.

1 2 3 4
SiO2 54.94 53.24 57.34 53.20
TiO2 0.68 0.63 0.68 0.86
Al2O3 14.88 14.57 14.89 14.10
Fe2O3 1.61 1.76 2.05

8 50*
FeO 6.04 6.57 5.21
MnO 0.14 0.16 0.14 0.11
MgO 7.08 7.35 6.56 7.95
CaO 7.71 8.78 7.31 7.69
Na2O 3.99 3.78 3.73 3.68
K2O 1.55 1.06 2.04 1.69
P2O5 0.32 0.46 0.26 0.29
H2O 1.09 1.56 1.26 –
Total 100.03 99.92 101.47 98.07
1 2 3 4
Ba 781 822 864 1090
Cr 467 258 326 516
Cu 79 124 159 –
Ni 65 60 44 555†
Rb 19 6 41 19
Sr 1010 1229 749 1050
Y 17 16 20 13
Zr 154 146 185 86
CIPW Norm 1 2 3 4
Q – – 3.54 –
or 9.16 6.26 12.05 9.95
ab 33.75 31.99 31.56 31.13
an 18.12 19.66 17.86 16.96
di 14.60 16.95 13.47 15.50
hy (En) 17.63 (66) 15.36 (71) 16.85 (75) 16.21 (80)
ol (Fo) 1.32 (64) 3.33 (70) – 0.84 (79)
mt 2.33 2.55 2.97 5.48
il 1.29 1.20 1.29 1.64
ap 0.76 1.09 0.62 0.67
FeO/FeO +MgO 0.52 0.53 0.52 0.52
Total normative feldspar 61.03 57.91 61.47 58.04
% Normative anorthite 35 37 35 35

(Table 18) Mineral compositions from the diorite, South Harris Igneous Complex (after Horsley, 1978)

1 2 3 4 5
SiO2 49.47 49.70 41.11 39.05 57.10
TiO2 0.54 0.16 2.12 0.35 –
Al2O3 5.58 3.99 13.17 20.41 27.22
FeO 10.36 25.00 15.34 26.46 –
MnO 0.19 0.64 0.16 1.44 –
MgO 11.75 20.01 10.09 6.48 –
CaO 20.94 0.63 11.98 6.49 9.31
Na2O 0.90 0.14 1.11 – 6.39
K2O – – 1.46 – 0.29
Total 99.73 100.27 96.54 100.68 100.31
1 2 3 4 5
Si 1.847

}2.000

 1.891

}2.000

 6.199

}8.000

 6.162

}6.162

 2.536
AlIV 0.112 0.089 1.801 – –
AlIV 1.1477

}2.050

 0.089

}2.081

 0.541

}5.005

 3.796

}9.593

 Al 1.464
Ti 0.015 0.005 0.241 0.042 –
Fe 0.324 0.796 1.935 3.492 –
Mn 0.005 0.021 0.021 0.192 –
Mg 0.655 1.135 2.267 1.522 –
Ca 0.839 0.025 1.936 0.549 0.464
Na 0.065 0.010 0.324

}0 605

 0.520
K – – 0.281 – 0.016
O 6.000 6.000 23.000– 24.000 8.000
Fe/Mg 0.495 0.701 0.854 2.294 –
Fe/(Fe+Mg) 0.331 0.412 0.460 0.696 –
En 36.0 En 58.0 Niggli mg 0.537 Al 55.4
Fs 17.8 Fs 40.7 Py 24.1 An 46.4
Wo 46.1 Wo 1.33 Gr 17.4 Ab 52.0
Sp 3.0 Or 1.6

(Table 19) Chemical analyses of the diorite and tonalite of the South Harris Igneous Complex.

1 2 3 4 5
SiO2 55.88 58.13 57.80 54.20 66.46
TiO2 0.84 0.76 0.82 0.85 0.29
Al2O3 18.48 17.69 17.47 20.10 17.46
Fe2O3 1.68 1.53 1.94 3.23 0.72
FeO 5.23 4.90 4.76 3.90 2.12
MnO 0.11 0.11 0.10 0.11 0.05
MgO 4.20 3.70 3.92 4.40 1.46
CaO 7.78 6.89 7.05 7.55 4.70
Na2O  3.67 3.66 3.69 3.90 5.34
K2O 1.12 1.36 1.43 1.38 0.92
P2O5 0.34 0.26 0.30 0.33 0.13
H2O 0.83 0.95 0.95 – 0.57
Total 100.16 99.94 100.23 99.95 100.22
1 2 3 4 5
Sr 681 577 638 720* 1356
Ba 720 758 795 690 900
Rb 10 23 26 25 7
Zr 110 134 136 164 206
Y 9 11 12 18 3
Th 2 2 2 5 –
Ni 25 20 26 26 10
Cu 19 29 18 – 28
Cl 171 203 301 250 –
F 334 329 419 – –
CIPW Norm 1 2 3 4 5
Q 6.33 10.07 9.39 3.32 19.02
Or 6.62 8.03 8.45 8.12 5.44
ab 31.04 30.95 31.21 32.96 45.16
an 30.65 27.83 26.89 33.27 20.96
di 4.67 3.81 4.98 1.77 1.23
hy (En) 15.20(60) 13.03(62) 13.30(62) 13.38(77) 5.94(56)
mt 2.44 2.22 2.81 4.71 1.04
il 1.59 1.44 1.56 1.61 0.55
ap 0.81 0.62 0.71 0.77 0.31
Fe/FeO + MgO Total 0.620 0.633 0.628 0.609 0.659
Normative Feldspar 68.31 66.82 66.55 74.35 71.56
Normative Anorthite % 48 47 45 50 32

(Table 20) Average chemical analyses of ultramafic rocks intruding the norite and diorite of the South Harris Igneous Complex.

1 2 3
SiO2 46.29 44.97 53.54
TiO2 0.87 1.47 0.29
Al2O3 8.86 10.21 3.53
Fe2O3 2.79 3.86

9.42*
FeO 9.19 9.79
MnO 0.20 0.16 0.19
MgO 15.88 11.80 16.96
CaO 11.38 13.83 13.86
Na2O 1.62 1.64 0.73
K2O 0.65 0.41 0.11
P2O5 0.23 0.06 0.02
H2O 1.63 1.71 -
Total 99.59 99.91 98.65
1 2 3
Ba 172 140 66
Cr 605 267 833
Cu 96 20 –
Ni 183 55 236
Rb 4.5 3 1.6
Sr 137 99 77
Y 19 20 11
Zr 51 55 28
CIPW Norm 1 2 3
Q – – 2.60
Or 3.84 2.41 0.67
ab 13.70 9.45 6.18
ne – 2.38 –
an 14.99 19.27 6.01
di 32.04 39.40 49.78
hy (En) 0.25 (74.5) – 29.07 (79.7)
ol (Fo) 26.92 (72.7) 16.55 (67.7) –
mt 4.05 5.61 3.78
il 1.65 2.80 0.55
ap 0.55 0.14 0.05
FeO/FeO +MgO 0.428 0.530 –

(Table 21) Clinopyroxene composition from the late potash-rich basic dykes (average of 6 points) after Horsley (1978)

SiO2 52.58 Si 1.974

}2.000

 Wo 47.4
TiO2 0.09 AlIV 0.026 En 38.6
Fs 14.0
Al2O3 1.04 AlVI 0.020

}2.021
FeO 8.73 Ti 0.003
MnO 0.30 Fe 0.273
MgO 13.51 Mg 0.753
CaO 23.09 Mn 0.009
Na2O 0.54 Ca 0.924
Na 0.039
Total 100.12

(Table 22) Average chemical analysis of the late potash-rich basic intrusions (3 specimens) from the South Harris Igneous Complex; after Horsley (1978, p.395)

CIPW norm
SiO2 53.51 Ba 2445 Q 1.67
TiO2 1.31 Cr 154 or 38.11
Al2O3 11.88 Ni 55 ab 8.63
Fe2O3 2.26 Rb 188 an 8.79
FeO 5.36 Sr 1576 di 20.95
MnO 0.11 Y 53 hy (En) 11.10 (71.6)
MgO 6.09 Zr 499 ol -
CaO 9.18 mg 3.28
Na2O 1.02 il 2.49
K2O 6.45 ap 3.89
P2O5 1.64
H2O 1.27
Total 100.08
FeO/FeO + MgO = 0.559

(Table 23) Chemical analyses of various Older and Younger Basic and ultrabasic intrusions (including a Late-Laxfordian microdiorite).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
SiO2 50.3 52 52.5 51.2 53.4 54.4 53.9 41.8 45.5 47.6 50.8 50.01 50.6 50.2 50.2 50.3 51 51.4 49.6 51.7
TiO2 0.45 0.58 0.84 0.86 0.92 0.29 0.31 0.28 0.22 0.81 0.37 0.55 1.47 1.46 1.32 1.76 1.6 1.5 1.98 2
Al2O3 5.6 13.7 14.4 17.2 15 7.3 6.61 3.87 4.21 7.53 11.7 6.84 13.2 13.6 13.7 14 13.6 13 14.8 14.4
Fe2O3 1.9 1.8 3 3 1.9 0.63 0.8 2.01 3.45 3.27 2.16 1.05 6.68 3.6 4.04 2.76 2.9 3.7 3.38 5.4
FeO 8 6.8 6.1 6.2 6.8 8.13 8.07 8.43 8.93 8.74 7.85 9.35 8.59 10.2 10.6 12 10.5 11.1 8.03 6.7
MnO 0.19 0.15 0.15 0.14 0.14 0.18 0.18 0.16 0.17 0.19 0.16 0.17 0.21 0.25 0.21 0.22 0.21 0.25 0.18 0.17
MgO 16.7 8.6 7.3 4.9 6.6 19.9 21.1 34.8 31.8 20.4 15.1 23.88 5.58 5.86 6.43 5.7 5.4 5.3 7.3 4.2
CaO 11.9 8.6 8.4 8.2 8 6.49 5.12 2.84 4.62 7.79 9.17 5.63 9.44 9.63 9.86 9.97 9.4 9.8 10.4 6.5
Na2O 1.1 3 2.8 3.8 2.8 1.25 1.09 0.63 0.64 2.31 1.68 1.56 2.6 2.65 2.8 2.12 2.8 1.8 2.37 3.2
K2O 0.51 1.6 1.2 1.2 2 0.26 0.21 0.22 0.21 0.35 0.29 0.46 0.78 0.84 0.83 0.33 0.8 0.4 0.43 2.7
P2O5 0.1 0.15 0.29 0.29 0.29 0.04 0.04 0.04 0.03 0.02 0.04 0.03 0.26 0.13 0.1 0.14 0.19 0.12 0.24 1.6
CO2 – – – – – 0.06 0.11 0.26 – – – – 0.17 – – 0.21 – – 0.03 –
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Ba 119 547 400 432 886 105 80 40 83 60 347 110 330 199 181 121 217 48 – 1719
Co 96 62 42 30 48 90 95 160 - – – 55 44 67 30 70 72 – 35
Cr 1551 585 341 5 327 2650 3430 6750 6325 2313 1843 >1000 54 128 127 260 76 138 – 61
Cu 117 13 11 19 29 35 25 45 – – – – 150 173 – – 166 119 – 69
Ga 5 10 14 17 12 9 7 5 – – – <1 19 7 – 5 18 14 – 22
La 0 0 14 9 7 – – – – – – – – 7 – – 0 0 – 142
Li 16 26 25 14 41 16 22 6 – – – <5 8 15 – 8 12 – – 49
Mo – – – – – – – – – – – – – – – 60 – – – 0
Ni 224 112 62 0 77 440 495 2050 1305 779 296 >1000 40 78 103 – 65 87 – 39
Pb 0 0 0 0 0 – – – – – – – – 120 – – 5 0 – 0
Rb 18 69 34 25 49 14 8 13 13 10 11 – – 27 22 177 21 8 – 92
Sr 117 517 541 717 583 95 85 40 61 87 155 95 270 222 188 – 220 158 – 1341
V 448 282 241 189 217 155 130 90 – – – 95 250 352 – 333 319 – 214
Y 14 16 26 18 16 5 5 2 – – – – – 22 – – 22 32 – 35
Zn 60 70 70 60 60 60 65 90 – – – – – 121 – – 119 101 – 140
Zr 176 99 35 49 73 40 20 <20 31 18 76 45 150 125 97 – 165 191 – 310
CIPW norm 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Q – – 3.7 – 2.23 0.2 0.88 – – – – – 5.9 1.58 – – 2.37 8.55 2.39 5.93
or 3.01 9.45 7.09 7.09 11.8 1.54 1.24 1.3 1.24 2.07 1.71 2.78 4.61 4.96 4.9 1.95 4.73 2.36 2.55 16
ab 9.3 25.4 23.7 32.1 23.7 10.6 9.22 5.33 5.41 19.5 14.2 13.62 22 22.4 23.7 17.9 23.7 15.2 20 27.1
an 8.84 19.2 23.2 26.3 22.5 13.5 12.5 7.08 8 9.15 23.6 10.01 22.1 22.7 22.8 27.8 22.2 26.2 28.5 17
di 39.6 18.2 13.4 10.2 12.5 14.7 10.1 5.33 11.7 23.4 17.3 14.14 18.8 20 21.4 17.3 19.3 18 17.2 3.91
hy 24 14.9 19.4 13.6 20 56.8 61.6 10.5 20.6 4.98 30.8 26.54 13 18.4 17.9 22.9 18.5 19.6 18.7 13.4
ol 8.21 5.82 – 0.99 – – – 52.1 37.4 33.6 8.86 29.55 – – 0.85 – – – – –
mt 2.76 2.61 4.35 4.35 2.76 0.91 1.16 2.91 5 4.74 3.13 1.62 9.69 5.22 5.86 4 4.21 5.37 4.9 7.83
il 0.86 1.1 1.6 1.63 1.75 0.55 0.59 0.53 0.42 1.54 0.7 1.06 2.79 2.77 3.34 3.34 3.04 2.85 3.76 3.8
ap 0.24 0.36 0.69 0.69 0.69 0.09 0.09 0.09 0.07 0.05 0.09 – 0.62 0.31 0.24 0.33 0.45 0.28 0.56 3.79

(Table 24) Average chemical analyses of Younger Basics exhibiting different degrees of amphibolitisation.

1 2 3 4 5
SiO2 50.9 50.8 50.4 49.9 48.9
TiO2 1.6 1.4 1.4 1.4 1.8
Al2O3 14.1 13.7 13.2 13.5 13.2
Fe2O3 3.2 2.6 4.0 3.5 4.9
FeO 10.6 11.2 10.1 10.1 10.3
MgO 5.8 5.7 5.7 6.4 5.3
MnO 0.21 0.21 0.23 0.21 0.23
CaO 9.8 9.9 9.3 10.1 9.3
Na2O 2.6 2.2 3.0 2.5 2.7
K2O 0.6 0.6 1.0 0.8 1.2
P2O5 0.2 0.1 0.2 0.1 0.2
H2O+ 0.6 1.0 1.2 1.11 –
H2O- 0.1 0.2 0.1 0.2 –
CO2 0.2 0.05 0.06 0.04 –
Cr 122 93 133 141 83
Co 72 83 68 65 69
Ni 82 76 67 88 70
Cu 202 184 175 138 178
Ga 14 12 15 6 8
Rb 12 24 14 19 54
Sr 241 186 204 178 254
Y 22 26 28 21 37
Zr 158 107 123 75 139
Li 10 11 12 18 19
Zn 113 118 140 106 144
Ba 152 184 193 163 250
La 7 0 11 3 11
Pb 100 128 177 122 104
V 328 407 279 363 460

(Table 25) Average chemical analyses of Late-Laxfordian granites.

Leucogranites Granites Porphyritic granites Mean of three types
n = 18 n = 59 n = 8 n = 85
SiO2 75.0 70.9 71.3 71.8
TiO2 0.10 0.28 0.33 0.25
Al2O3 13.6 14.0 14.3 14.0
Fe2O3 0.6 1.0 1.0 0.9
FeO 0.6 1.5 1.5 1.3
MnO 0.03 0.03 0.03 0.03
MgO 0.2 0.52 0.54 0.45
CaO 1.2 1.42 1.56 1.4
Na2O 3.8 3.7 3.6 3.7
K2O 5.0 5.2 5.1 5A
P2O5 0.02 0.09 0.09 0.08
Total 100.15 98.64 99.35 99.01
Leucogranites Granites Porphyritic granites Mean of three types
Cr 52 34 39 38
Co 20 22 12 21
Ni 7 3 7 4
Ca 6 11 13 10
Ga 18 12 12 13
Rb 462 288 401 243
Sr 125 334 357 292
Y 8 10 11 9
Zr 120 357 341 305
Li 16 20 35 19
Zn 28 45 59 43
Ba 409 1191 1359 1041
La 28 132 94 112
Pb 23 17 21 19
Ca/Sr 68 30 34 38
K/Rb 90 150 106 133
Ba/Rb 0.89 4.14 3.4 3.4
Rb/Sr 3.70 0.86 1.12 1.49
K/Sr 332 125 119 168
K/Ba 101 36 31 49

(Table 26) Modal analysis of quartz-microdiorite dyke, Garry a-siar, Benbecula ((S59796)A)

Quartz Potash feldspar Plagioclase Hornblende Biotite Apatite Sphene Opaques
7.0 0 40.3 16.5 25.7 3.0 5.7 1.8

(Table 27) Structural history

Movements on Outer Hebrides Thrust Zone (c.450 Ma)
Mylonite formation
dL5–dL6 Formation of open warps, restricted cataclasis and retrogression. Initiation of thrusting on Outer Hebrides Thrust Zone. Main pseudotachylite formation.
dL4 Penetrative deformation of restricted extent Formation of the Uig Hills–Harris granite complex and related migmatites. Late intrusion of pegmatites

(c.1700 Ma)
dL3 Extensive penetrative deformation on steep (mainly NW-trending) axial planes.
dL2 Extensive penetrative deformation, gently inclined axial planes. Anorthositegabbro (Harris) (c.2250 Ma)
dL1 Opening of dyke fractures, early fabric in dykes Younger Basic dyke swarm (c.2400 Ma)
dS4 Major shear zones Diorite-granite suite (Barra) (c.2600 Ma)
dS3 Large asymmetrical folds
dS2 Regional ductile deformation development of planar–linear fabric system Regional gneiss-forming episode with accompanying intrusions (c.2750 Ma)
dS1 Penetrative deformation in supracrustal gneisses
<

(Table 28) Chemical analyses of (1) garnet-biotite meta-sediment (mean of 2 analyses from wall rocks adjacent to pseudotachylite vein) (2) pseudotachylite melt plus quartz and feldspar porphyroclasts (3) pseudotachylite with 9% quartz and 11 % plagioclase (An30) subtracted (from Sibson, 1977)

1 2 3
SiO2 61.55 60.57 55.40
Al2O3 17.06 17.45 19.11
TiO2 0.59 0.64 0.78
Fe2O3 7.11 7.28 8.91
MnO 0.12 0.11 0.13
MgO 4.11 4.93 6.04
CaO 3.45 3.64 3.17
Na2O 3.00 3.45 3.26
K2O 2.32 2.51 3.07
P2O 0.10 0.11 0.13
Total 99.41 100.68 100.00

(Table 29) Representative mineral analyses from mylonite and retrograde pseudotachylite

(S56405) (S56405) (S56405) (S56405) (S56405) (S62000) (S62000) (S56405) (S56405) (S56405)
Biotite Biotite Garnet Garnet Garnet Garnet Garnet Plagioclase 1 Plagioclase 2 Plagioclase 3
SiO2 36.78 36.54 38.31 38.36 38.00 37.80 38.02 60.33 61.17 63.75
TiO2 4.04 4.63 – – – – – – – –
Al2O3 15.36 15.46 21.46 21.63 21.38 20.79 21.14 24.46 24.22 21.90
FeO* 17.82 17.52 27.74 25.92 27.69 28.58 27.35 – – 0.11
MnO – – 1.90 0.59 2.77 2.90 4.31 – – –
MgO 11.42 11.32 2.71 3.05 2.91 2.71 2.67 – – –
CaO – – 8.57 10.75 8.14 7.42 7.04 6.17 5.42 3.83
Na2O – – – – – – – 8.04 8.24 9.06
K2O 9.64 9.44 – – – – – 0.09 0.10 0.18
Cl 0.24 0.20 – – – – – – – –
95.30 95.01 Total 100.69 100.30 100.89 100.20 100.53 99.09 99.15 98.84
O≡Cl‡ 0.05 0.05
Total 95.25 94.51
(S56405) (S56405) (S56405) (S56405) (S56405) (S62000) (S62000) (S56405) (S56405) (S56405)
Si 5.586

8 00

 5.546

8 00

 Si 6.026 6.009 5.986 6.019 6.025 Si 10.825 10.935 11.377
Al 2.414 2.454 Al – – 0.014 – – Al 5.173 5.105 4.607
Al 0.336

5.65

 0.314

5.63

 Al 3.980 3.996 3.956 3.904 3.950 Fe3+ – – 0.019
Ti 0.462 0.528 Fe3+ 0.005 – 0.054 0.112 0.046 Ca 1.186 1.038 0.732
Fe2+ † 2.263 2.224 Mg 0.635 0.712 0.685 0.644 0.631 Na 2.794 2.854 3.136
Mg 2.584 2.561 Fe2+ 3.645 3.397 3.599 3.695 3.583 K 0.022 0.022 0.041
K 1.868 1.828 Mn 0.253 0.078 0.370 0.392 0.579
Cl 0.062 0.051 Ca 1.444 1.804 1.374 1.266 1.195
Al 61.0 56.7 59.7 61.3 59.7 Ab 69.8 72.9 80.2
Py 10.7 11.9 11.4 10.7 10.5 An 29.6 26.5 18.7
Gr 24.0 30.1 21.7 18.7 18.9 Or 0.6 0.6 1.1
An 0.1 – 1.3 2.8 1.2
Sp 4.2 1.3 6.1 6.5 9.7

(Table 30) Chemical analyses of Outer Hebrides post-Lewisian minor intrusions

Caledonian

Permo-Carboniferous

Tertiary
Appinite

Quartz-dolerite dykes

Camptonite and monchiquite dykes

Olivine-dolerite dykes

Shiant Isles sills

Barra Swarm

South Harris Swarm

Main (Garbh Eilean) Sill

Eilean Mhuire Sill
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

37

 21 22 24

34

 36
SiO2 47.46 50.10 49.80 49.90 44.90 43.20 45.20 45.90 44.30 44.28 43.34 44.93 45.53 46.40 46.97

41.65

 40.62 45.07 47.83

58.36

 46.48
TiO2 0.75 2.42 2.40 2.40 2.73 2.55 1.93 2.79 1.96 2.15 2.21 0.37 0.54 1.02 1.24

0.62

 0.82 0.83 2.86

0.48

 2.00
A12O3 10.00 13.60 13.30 13.40 14.10 13.00 14.50 12.40 11.20 16.33 14.13 20.72 19.66 14.78 13.87

12.95

 8.93 14.43 15.31

15.82

 15.59
Fe2O3 4.12 3.15 3.10 2.90 3.46 3.98 3.19 – 4.81 2.60 4.09 1.50 1.50 1.50 1.50

1.62

 0.57 0.80 1.15

4.87

 4.54
FeO 5.86 10.20 10.00 10.00 4.88 5.61 5.93 11.10* 6.77 13.49 9.40 5.33 6.37 9.02 10.87

10.70

 12.61 10.69 9.22

2.53

 8.62
MnO 0.18 0.19 0.16 0.15 0.19 0.22 0.17 0.16 0.18 0.21 0.19 0.10 0.13 0.18 0.21

0.18

 0.39 0.33 0.36

0.27

 0.28
MgO 13.93 6.12 6.20 6.40 8.35 11.85 10.45 8.35 12.75 6.80 9.98 8.52 8.45 8.25 7.42

21.98

 26.31 14.61 6.60

0.59

 9.19
CaO 10.14 9.30 10.20 10.20 10.82 10.08 10.69 10.08 9.43 8.20 9.08 13.08 13.31 12.25 12.37

6.51

 5.64 9.74 12.38

1.99

 8.98
Na2O 2.20 2.55 2.60 2.50 1.65 0.95 3.15 3.10 2.45 3.62 3.20 1.96 1.78 2.00 2.38

1.32

 1.32 1.75 2.53

7.47

 2.79
K2O 2.00 0.62 0.63 0.43 3.88 3.40 1.35 1.20 1.20 0.54 0.25 0.21 0.14 0.13 0.13

0.16

 0.13 0.34 0.40

4.31

 0.71
H2O 2.13 1.08 1.20 1.20 4.31 4.50 2.28 2.46 3.20 1.59 2.21 3.10 2.43 3.55 1.84

1.60

 2.19 1.05 1.28

2.62

 0.85
P2O5 0.39 0.26 0.28 0.26 1.16 0.92 0.33 0.43 0.26 0.32 0.26 0.07 0.08 0.11 0.14

0.05

 0.15 0.10 0.16

0.35

 0.11
CO2 0.01 0.01 0.02 0.02 0.12 0.09 0.87 0.76 1.66 0.25 0.05 – – – –

–

 0.03 – 0.05

–

 –
Total 99.17 99.60 99.89 99.76 100.27 99.95 100.05 98.73 99.67 100.38 98.89 99.89 99.92 99.19 98.94

99.34

 99.71 99.74 100.13

99.66

 100.14
CIPW NORMS
Q – 3.40 2.15 2.88 – – – – – – – – – – –

–

 – – –

–

 –
ab 15.70 21.88 22.27 21.44 7.38 5.30 17.23 27.22 20.73 23.87 23.34 13.87 15.43 17.67 20.72

10.25

 5.11 14.57 21.63

51.11

 23.75
or 12.16 3.71 3.77 2.58 22.93 20.09 7.98 7.36 7.09 3.23 1.53 1.28 0.85 0.80 0.79

0.97

 0.79 2.03 2.39

26.22

 4.22
an 11.87 24.20 23.07 24.43 19.61 21.17 21.44 17.02 16.02 27.05 24.27 48.68 46.41 32.38 27.58

29.61

 18.52 30.92 29.58

–

 28.13
ne 1.88 – – – 3.57 1.48 5.10 –– - 3.85 2.51 1.76 – – –

–

 – 0.22 –

4.35

 –
di 29.53 17.11 21.50 20.53 20.25 17.54 19.02 21.15 14.71 8.63 16.29 14.34 16.75 24.81 28.58

2.46

 7.17 14.04 25.49

6.51

 12.87
by – 19.72 17.38 18.55 – – – 4.411 6.08 – – – – 8.27 2.86

–

 – – 5.01

–

 4.88
ol 20.25 – – – 9.36 17.33 15.96 6.11 17.27 24.06 20.80 16.88 17.04 11.46 14.43

52.34

 62.10 35.17 8.18

0.08

 15.39
nit 6.16 4.64 4.56 4.27 5.02 5.77 4.63 8.36 6.97 3.82 6.14 2.25 2.23 2.27 2.24

2.40

 0.85 1.18 1.69

4.66

 6.63
it 1.47 4.67 4.62 4.63 5.19 4.85 3.67 5.51 3.73 4.13 4.34 0.73 1.05 2.03 2.43

1.21

 1.60 1.60 5.50

0.94

 3.83
0 0.93 0.61 0.66 0.61 2.69 2.13 0.76 1.03 0.60 0.75 0.62 0.17 0.19 0.27 0.33

0.12

 0.36 0.23 0.37

0.83

 0.26
cc 0.02 0.02 0.05 0.05 0.27 0.20 1.98 1.79 3.77 0.58 0.12 – – – –

–

 0.07 – 0.11

–

 –
ac – – – – – – – – – – – – – – –

–

 – – –

5.23

 –
TRACE ELEMENTS (ppm)
B nd <5 – - – – – – – – <5
Li 16 4 – – – – – – – 8 4
F 880 350 – – – – – – – – 280
V 165 240 – – 201 223 187 – 193 192 290
Sc – – 32 32 24 31 28 – 20 – –
Cr 2200 68 – – 170 142 294 208 388 38 630
Co 53 40 48 45 – – – – – 52 58
Ni 260 61 73 – 93 93 111 148 401 59 300
Cu 92 190 – – – – – – – 86 190
Ga <10 15 – – – – – – – 28 16
Rb – – 16 8 80 75 70 50 55 63 -
Sr 480 260 305 280 1830 1354 750 659 549 295 340
Y – – 30 32 32 26 24 - 20 11 -
Zr 100 160 142 139 400 317 149 - 125 206 180
Nb – – 16 16 – – – – – – –
Ba 1100 130 180 140 3175 2209 581 640 407 225 66
La – – 12 12 115 86 29 - 21 <35 -
Hf – – 3.6 3.5 – – – – – – –
Th – – 1.1 1.2 – – – – – – –

(Table 31) Modal analyses of Outer Hebrides post-Lewisian minor intrusions

Qz Kf Plag An F'oids Ol Cpx Opx Hbl Bio Op Ap Ze G'mass
CALEDONIAN
1 – 17.3 – – 3.2 25.5 – 36.1 13.0 3.3 1.6 –
PERMO–CARBONIFEROUS
2 5.5 – 44.9 – – 30.4 0.4 6.5 3.1 9.3 – –
TERTIARY
Olivine–dolerite dykes
Barra Swarm
10 – – 55.5 58 – 20.2 18.3 – – – 6.0 – –
11 – – 51.4 – 15.8 21.8 – – 1.3 9.7 – –
South Harris Swarm
12 – 39.6 (p) 5.5 (p) 54.9
13 – 34.3 (p) 4.9 (p) 60.8
14 – 1.2 (p) 7.9 (p) 90.7
Crinanite sills
Madadh rocks
16 – – 63 55–64 1 (Ac) 1 31 – – – 4 – –
17 – – 63 55–64 2 (Ac) 5 28 – – – 3 – –
18 – – 67 – 2 27 – – 3 – – –
19 – – 53 58 1 (Ac) 5 35 – – – 6 –
20 – – 54 55–60 2 (Ac) 10 28 – – – 6 – –
Shiant Isles
Main Sill
21 – – 26 – 59 10 – – – 2 – 3
22 – – 50 – 31 17 – – – 2 – –
23 – – 54 – 20 21 – – – 3 – 2
24 – – 60 – 12 24 – – – 3 – 1
25 – – 54 – 11 29 – – – 4 – 2
26 – – 61 – 8 24 – – 4.5 – 2.5
27 – – 38 – 6 38 – – – 11 – 7
28 – – 23 80 – 66 8.5 – – – – –
29 – – 29 80 – 62 4 – – – – – –
30 – – 23 – 62 9 – – – – – –
31 – – 34 55.80 – 44 19 – – – – – –
32 – 34 59.77 – 40 12 – – – – – –
33 – – 58 – 4 23 – – – 7 – 8
Lower Sill
34 – 50 – 3 – 21 – 1 – 5 – 20
35 – 62 – 4 – 11 – 1 – 4 – 18
36 – – 61 – 13 15 – – – 6 – 5
37 – – 40 4 2 50 – – – 6 – –

(Table 31) (notes) Qz quartz Kf potash feldspar Pl plagioclase An anorthite content of plagioclase F'oids feldspathoids Ol olivine Cpx clinopyroxene Op opaques Ap apatite Ze zeolites G'mass groundmass (p) phenocrysts (Ac) analcime

(Table 32) Statistics for the major Tertiary dyke-swarms of the Outer Hebrides, with comparative data for the Mull and Skye swarms

SWARM BARRA ISLES (Berneray-Gighay) SOUTH HARRIS (including Taran say, Scalpay, etc.) LEWIS

(Loch Shell/ Loch Erisort

 MULL NORTHERN SKYE
Equivalent mainland swarm and intrusive centre Mull? Skye main swarm (Cuillin Centre) ?Applecross sub- swarm (Cuillin Centre?) – (Cuillins)
Approximate distance (km) from related intrusive centre 120 75 ? – 50
Total dykes recorded 234 199 275 375 87 [1]
Average number of dykes per kilometre [2] 4 (maximum 14) 5 20 (maximum 60) 19 14 [3]
Mean thickness and standard deviation (m) [4, 5] 3.3± 3.3 (152) 4±3 (72) No data 2 No data
Thickness range (m) [4] 0.15-30 0.3-16.5 No data 0.3-60 0.3-60
Thickness class (m) [4, 6] 0 2-3 No data No data No data
Crustal dilation (%) 5 (maximum) 1.2 (average) 15 (maximum) 4 3
Modal azimuth (Figure 36) 141- 150° 161- 170° 155 - 164° 135° 140°
Abundance of petrological types [7]
crinanites abundant fairly common, especially on Scalpay very common extremely common 12
olivine dolerites most numerous most numerous some common 24 [3]
olivine-free not recorded a few not recorded common 24
dolerites
sources of data BGS; Jehu and Craig (1923) BGS; Jehu and Craig (1927); Mattey et al. (1977) Speight et al. (1982); Jehu and Craig (1934) Bailey et al. (1924) Anderson and Dunham (1966); Mattey et al. (1977)

(Table 33) Geological succession on the Shiant Isles

Succession Dip Notes Location
Lithology Thickness (m)
Main (Upper) sill >150

15-20° to the S and W (20-90° to the S on Galtachean)

 Picrodolerite gradational with crinanite, but shows sharp

contact with picrite

 Garbh Eilean, Eilean an Tighe
Crinanite & picrodolerite >120
Picrite 30 ?Galtachean
Basal hybrids 1 Basic pegmatite veins locally Essexite/syenite pegmatites on Galtachean only Top not seen
Upper Liassic shales and marls 9 15° to the SW Complex interaction with magmas at base of main sill Garbh Eilean
Middle sill Crinanite >30 Uniform crinanite lithology Top seen at NE tip of Garbh
Eilean. Denuded remnants of base on summit of Eilean Mhuire
Upper Liassic shales 19 25° to the N Eilean Mhuire
Lower sill >60
Horizontal Entirely crinanite at W end of island Eilean Mhuire
Upper dolerite > 15 Base not seen
Analcite syenite 20 Schlieren present in all rock
Teschenite

>6

 types
Crinanite or analcite dolerite Sheets of basalt (up to 6 m) have invaded syenite and teschenite.
No junctions are exposed: they are probably gradational

(Table 36) Stornoway Formation - detailed stratigraphy and palaeogeography (after Steel and Wilson, 1975)

Lower Unit (SE) Middle Unit Upper Unit (NW)
Possible contemporaneous NE- trending fault (unexposed) downthrowing to SE in Branahuie Banks area. (Active during Lower Unit deposition providing clastmaterial). Swordale (1370 m thick) - Garrabost (960 m thick) area

Two coarsening-upward alluvial fan sequences separated by a flood plain (sandstone, locally abundant cornstone) sequence. Fans 5-10 km across. Palaeocurrents imply NW derivation for conglomerate but palaeoslope of flood- plain shown by braided stream cross-bed foresets which dip to the NE and ENE.

 Possible con- temporaneous but now unexposed ENE trend- ing fault scarp in Broad Bay and extend- ing up Lax- dale Valley (postulated as source of debris for Middle Unit). Branahuie Banks- Stornoway, and Melbost area. Total 1520 m thick.

Two fining-upward alluvial fan sequences. Fans 5-10 km across. Palaeocurrents show northerly derivation. Alluvial plain sloped to the NE.

 Currently exposed marginal faults appear to have been a major source of fan material for the Upper Unit Arnish, Stornoway, Vatisker-Coll and Shellavaig areas. 770 m maximum in Stornoway area. (500 in–VatiskerColl). Two coarsening-upward fan sequences. Fans 5 km or less across. Palaeocurrents suggest derivation from present bounding faults. Alluvial plain slopes to the NE and NNE.
Upper fan sequence - 315m to 210 m thick (thins towards NE over 2 km). Average maximum clast diameter also decreases from 38 to 20 cm over same range. Dominantly debris-flow and stream- flood coarse conglomerates in Swordale area but abundant floodplain sandstones and braided stream deposits at Garrabost- Aignish. Upper fan sequence - 770m in Holm Point area. Thins to NNE. Mainly stream flood conglomerates with minor graded units. Subsidiary braided stream and debris flow conglomerates. Maximum clast size is about 30 cm. Capped by thin floodplain deposits, but these are also abundant in the lower part in Melbost area (fan margin). Also cornstones and root structures. Upper fan sequences. + 290 m. Various se-quences of coarse debris flow and streamflood conglomerates. Minor braided stream deposits. Locally very coarse angular conglomerate, clasts up to 1.8 m across. Prominent local unconformities and conglomerates with angular clasts in braided stream deposits. Local tectonism controlled lithology.
180 m of floodplain sandstones thickening in Garrabost area to 250 m where cornstone horizons are developed.
Lower fan sequence 935 m at Swordale thinning north-eastwards to 410m at Garrabost. Corresponding maximum clast sizes 45 to 18 cm. Massive coarse debris-flow conglomerate particularly thick in Swordale section. Minor basal streamflood conglomerate. Upper parts in Garrabost- Aignish area composed mainly of stream flood and floodplain deposits. Lower fan sequence. 250 m exposed in Branahuie Banks- Holm Point area, 210 m around Melbost. Lower 500 m largely unexposed except for minor debris flow conglomerate outcrops SE of Melbost. Debris flow and stream flood conglomerates dominant - minor floodplain and cornstone horizons at Melbost. Maximum particle size decreases laterally from 38 to 20 cm towards the NE over 3 km. Lower fan sequences 480 m thick around Stornoway. Debris flow and subsidiary stream flood conglomerates. Maximum clast size increases upwards from 30 to 60 cm. Minor braided stream deposits at Vatisker-Coll.