SEDIMENTATION of the TORRIDON1AN ROCKS of RAASAY SOALPAY - AND, ADJACENT Rslet4 INVERNESS—SHIRE

SEDIMENTATION OF THE TORRIDON1AN ROCKS OF RAASAY SOALPAY - AND, ADJACENT rsLET4 INVERNESS—SHIRE.

' C. Sell.eyo B. Sc..

Thesis submitted for the Doctor of Philosophy degree of the University of London.

Geology DepartMent' Imperial College of Science d Technology. September 1963 YVC,,i1-:1-I.CS. Looking •from na folaires the hill . of aadday. The '2:orridouitin rocks of ;ieesay can be seen inz fr,minn na to the .ea. oalpay lies to the and *ye to the . . with the Cailins visible far to the 3

ABSTRACT

The islands of Raasay and Scalpay lie between Skye and the mainland region of Wester Ross in Scotland. SOme 25 sq. kms. of these islands are composed of unmetamorphOsed Pre0ambrian sediments of the Torridonian formation. Illtpresentatives of both' the, Lower (Diabaig). 'and. Middle (Applecrois) groups of the Torridonian are present. TheXpper:(Aultbea) group is. absent. In this-thesis five main conclusionsare reached oonoerning the Torridoniin rooks of these islands. The-atratigraphy of the area-proposed,by the Geological SUrvey has been modified and considerably :expanded: It has been found easier to map three lithological facies rather than discrete stratigraphical groups.- These facies are defined According'to the rock types and sedimentary structures-present. They indicate a succession of three different sedimentary environments. A Basal facies of'sUbaerial screei and fanglomerates is succeeded, by, Grey facies rocks of lacustrine or shallow water marine origin. Thee() are in turn overlain by Red facies sediments laid down by braided rivers. Subfaoies end subenvironment6 are also described. . An analysis of palaeocurrente indicates that, ai "previously thought, most -of the sediment came from 4 the Northwest. Local deviations from this direction near the baee of'the succeesion probably indicate control by the irregular topography of the Lewisian gneiss basement. Mineralogical evidence is presented which suggests that nearly all the arenaoeous sediment was derived from the Lewisian gneisii. A considerable amount of rudaceous material is of non—Lewisian origin. Fossilized quicksands, which are common over the whole Torridonian-region are described and their origins discussed in the light of laboratory experiments. 5

TABLE OF CONTENTS

Abstradt , Table of Contehts

Chapter I - INTRODUCTION (i) Introduction (ii) Llajor results of the study (iii) Methods.. of study (iv) Histoiy of previous research (v)GeogriPhy

Chapter- II, - STRUCTURE AND STRATIGRAPHY (i)-Structure (ii) Stratigraphy 'Cliepter _III-- THE BASAL UNCON?ORMITY AND THE

BASAL FACIES (i)-The•basal unconformity (ii)Descriptioh'of 'the Basal facies,. (iii) Discussien.of the Basal" facies' _Chapter IV - THE GREY FACIES (i) Introduction (ii) The Caolas Fladday beds (iii)-The Brothel beds ' Units_ II and III,, Soalpay (v):Otheroccurrehces of Grey facies sediments 6 Chapter V - THE RED FACIES (i)Introduction (ii)Description of lithologies (iii)Primary sedimentary structures Chapter VI - PETROGRAPHY (i)-Introduction (ii)Qualitative mineralogy (iii)Quantitative mineralogy (iv)Discussion of mineralogy (v)Granulometric analyses Chapter VII -•SECONDARY SEDIMENTARY STRUCTURES (1) Introduction (ii)Deformational structures of interbedded sand and clay (iii)Deformational structures of sandstones (iv)Deformational structures of heavy mineral bands Chapter VIII - PALAEOCURRENTS (i)Introduction (ii)Basal facies (iii)Grey facies (iv)Red facies (v)Summary Chapter IV - DISCUSSION OP SEDIMENTARY ENVIRONMENT AND THE RELATIONS OP THE TORRIDONIAN ROCKS OF THE AREA WITH THOSE OF THE REST OF THE FORM- ATION 7

(1) Introduction and historical survey (ii) Phase I (iii) Phase II (iv) Phase III (v) Summary of the history of the area (vi) Synthesis of the origin of the Torridonian formation APPENDIX I. Notes of terminology used APPENDIX II. Summary of lithological and sedimentary environments

Acknowledgements

Bibliography

area studied.

TORRIDONIAN ROCKS STIPPLED.

I KM.

FIG.1. LOCALITY MAPS OF THE AREA STUDIED_

100 KMS• 9

' CHAPTER I. INTRODUCTION

(i),Introduction ? The Torridonian formation crops out in a belt along the western. seaboard of Scotland for a distance 'of some 320 kilo. from Cape Wrath in the North to Islay in the South. It has an estimated maximum' thickness of some 5 kms. (P1iemister,1960„ p.44), and consists, for the most part, of,red'pebbiycoarse sandstones over- and underlain mainly by grey shales and grey sandstones. The.Torridonian rooks rest on the Lewisian gneitilses lvith'a.marked =conformity, and are themselves overlain unconfoimably by Cambrian and later rocks. To the East the formation is terminated abruptly by-a major tectonic distiontinuity, the Moine thrust, and to the West it passes'out'to sea or,has been eroded away to reveal, the gneise basement. The area studied for the p'reparation of this thesis lies_between.Skye and the mainland of bdotland and com- , prises the islands 'of South Rona, Fladday, Raasay and Scalpay (see Pig.:1).- Some 25sq.km. of Torridonian ,rook crop out on these, attaining a maximum estimated thickness of over 2 kms. Rocks of both the Lower (Diabaig) and Middle (Applecross) groups of the formation are presents . The Upper-(Aultbea) group is not represented.

.." 10

On Raasay the unconformity with the Lewisian gneissea is well exposed•. On both Raasay and Scalpay. the Toiridoniwie locally overlain by Triassic rocks, and faulted against other Mesozoic sediments and Tertiary igneous.,roOks.-.It'isaleo, Out, by, a large number of Tertiary dykes and in the southern part of ScUlpay is intruded and. by .a northein extension of

(ii),Maior results- of the..studY The evidente recorded in this thesis contributes to knowledge of the on five main topics. The Stratigraphical succession of the area proposed by theieological_ur4ey has been considerably modified and expanded. More'preoise data is provided to suggest that the Lower Piaiaig,seaiments arc,,of either lacustrine or shallow water marine ()Agin,- and that the succeeding Upper::DiabUig and Applecross deposits were laid down by braided rivers. An analysis Of current, directions has bein'Imide whith provides evidence that, except for local deposits adjacent to the basement, the bulk of the sediments were ' derived from the North and the West., MineralogiCal evidence is presented which suggests that nearly all the arenaceou& -sediments were derived 11

from the'Lewisian gneisses, although, earlier workers advocated a non-Lewisian origin. Pepsi:contemporaneous disturbances of the bedding hive been recorded by previous workers-,from a wide-range of lOcalitiesr but, until recently, they have not received detailed attention. The morphology of these structures, as seenin the area, is described, and their origin and geological significance are discussed in - the light of. laboratory experiments..

(iii)Methods of study Field work was carried out in the summers of 1959- 1963. Geological maps of the area were made on a scale of 1:1,056 using 0.5.6":1 mile topographic maps in con- junction with aerial photographs (see Plate I). Laboratory work in the winters of 1961-62 and 1962-63 consisted mainly of qualitative analysis of the composition and texture of the sediments by study of thin sections. Detailed quantitative work was not carried out. Experiments were devised to test ideas on the origin of the secondary sedimentary structures.

(iv)History of previous research The first account of the Torridonian formation was given by Macculloch (1a9) who toured the Western Isles in'1817. He visited Raasay (Vol.I. p.239-258), Fladday 12

(p.259-260) and Scalpay (p.426-435). On Raasay he recognized the unoonformity between the Lewisian gneiss and the 'primary red sandstone'. He described the sequence, seen at Brothel, of laterally impersistant breccias overlain by 'greywackes and sohists' and followed by red sandstone. He did not notice the unconformity between the Torridonian and Trias at the south end of the island, and so unfortunately thought that the 'primary red sandstone' passed without a break into the 'secondary sandstones'. -During the later part of the nineteenth century many papers were published on the geology of the North- west-Highlands of Scotland. They were mainly concerned with the relative age of the formations. A complete review of this literature was given by Peach,. Horne et alii in 1907 (p.11-32). Only the more important papers relating to the Torridonian will now be mentioned. Murchison and Sedgwiek (1829) correlated the 'red sandstone occurring between Cape Wrath and Durness' with the Old Red Sandstone of Caithness. Salter (in Murohison 1856), on the basis of fossils from the Durness Liie- stone, which he believed to be Silurian, equated the Torridon Sandstone with the Cambrian of Wales. Murchison and A. Geikie (1861a) gave a brief account of a visit to Raasay, but added nothing new to Macculloch's work. The Pre0ambrian age of the Torridonian was demonstrated by 13

Lapworth (1888)' who described the Olenellus fauna from the base of the overlying Cambrian sequence. The fullest account of the Torridonian formation ',available to date is the Memoir on 'The Geological Structure of the North-west Highlands of Scotland', published'by the Geological Survey in,1907. This volume incorporated work by Peach, Horne, Gunn, Clough, Hinxman and • Teail3'giving • a detailed field. account (p.264-362) Of the Torridonian rocks from Cape Wrath to 2kye. The :petrography ie also deeeribed (p.278-290) and particular attention paid t'o the pebbles, of non-Lewisian rocks, which are a characteristic feature of-many of the -coarser ,sandstones (p.279.484). Because of the presence of these exotic pobble6 the authors believed,that the bulk of the Torridonian sediment was deriVed from a non-Lewisian -provenance (p.273-4). These workers divided the formation into three stratigraphic groups (p.274): -

'Aultbea Groups sandstones, flags, dark and black shales and calcareous bands passing down into chocolate and red sandstones, and grey micaceous flags with partings of grey and green shales. Applecross Group: chocolate and red arkoses with pebbles of quartzite, quartz-schist, felsite, jasper, etc., occasional chocolate and red shales. 14

Diabaig Group: hard, fine red sandstones at top, mixed with red mudstones and dark grey sandy shales with calcareous lenticles. At base, conglomerates made out of Lewisian gneiss.'

0f the sedimentary environment of the Torridonian the Vemoir states (p.273): 'it may be inferred that the conditions of deposit were probably those of a rapid accumu- lation in shallow water near a shoreline, subject to violent currents and the influx of flood or stream- borne materials, with occasional intervals of quiescence during which the finer sediments were laid down OOPS The rain-pitted and sun-cracked surfaces of the shales and mudstones seen on Loch Torridon and elsewhere show that the fine sediments of the lowest group are shore deposits, and cannot be regarded as due to deepwater conditions.' The Torridonian strata of Raasay were mapped by Hinxman in the summer of 1896 (A. Geikie 1897, p.15-16) and are described in the I4emoir on pages 340-341. Harker mapped Scalpay in 1900 (A. Geikie 1901, p.15) and his results appear on page 351. Subsequently, very little original material has been published on the Torridonian. Short accounts of Raasay were made by H. Woodward (1913) and C. Davidson (1937). These added little to the earlier account by 15

the Survey. nthin the last six year: interest in the Torridonian formation hss'been renewed. Irving (1957) and Irving and ii.uncorn (1957) conducted investigations of palaoomagnetism which suggested that, for part of the area at least, the Applecross: Diabaig junction was approximately synohronous. :4aumova and Pavlovskii (1961) and Downie (1962) have rodescribed fossils of microorganisms of obscure biological affinities. These Were also noted in the 1907 Memoir (p.287 and 288): Sutton and 7iatson described the sedimentary structures of the Diabdig rooks of Skye, ascribing the 1;pidotic Grits to a fluviatile 'environment (1960) and the Loc4 na Dal, Beinalia :3eamraig and Kinloch 4;eds to be those of tidal mud flat deposits (in press). A. Stewart described the orridonian rocks of Oolonsay and Oronsay: recognizing a greywacke turbidite facies followed by taskoece of suppoeed fluviatile origin (1962a v,Ad b). Mo morphology of the pcnecontomporaneous disturbances of bedding have been desCribed by Sutton and tiatson (1960) and 53elley, Shearman, 7,uttOn and ,Atson (1963), and in the light of ldboratory experiments they have been dttri- buted,to quicksand activity ( Selley and Shearman 1962). The deformation of heavy miner l bands have born described by A. Stewart (1963) and also by .:elley (in press). A

F LA D DAY

• • •

• • FIG.2. MAP OF RAASAY& SCAL PAY. • I •

• P • • • 1 - 1 POST.-TORRIDONIAN • • • •

. . • 1• 1 t / ..."\,...... „.•, rIf J. '.. 'I • I !‘ , \ • I PR E - I t i I \ i i S r i 1 1 I r CONTOURS AT INTERVALS I § I I I I I I 1 OF 500 FEET. I. • 1 I' 1 / &,4 1,6 „-- , t . ,,

I t 0 I 2 5 KM.

1 1 1. 1 0 I 2 e.....),.....% MILES. f, N..-. Ra as Y

...... ♦'pi1'

&46 17

(v) Geggraphy The topography and distribution of Torridonian rocks on the islands studied is shown in Fig. 2. The Torridonian reaches a maximum height on Raasay of some 400 m. at Beinn na'Chapuill, and on Scalpay of some 430 m. at Mullach nan earn. On Raasay there is developed a monotonous topography related to the alternations of sandstones and shales. It is characterized by north - south ridges with east-facing scarps, and dipslopes, each of which descends gently westwards, generally into a peat bog. .The height of these ridges varies irregularly across the Torridonian outcrop which itself consists of one composite cuesta (see Frontispiece). On Scalpay the relief is less regular and reflects the more uniform nature of lithology. The area enjoys a high rainfall, and this, together with the low evaporation rate and the impermeable nature of the rock, gives a large amount of surface drainage. There are a considerable number of lochs and lochans which mainly owe their origin to glacial erosion. The drainage pattern on Scalpay is essentially radial, the monotonous lithology of the Torridonian rocks having little effect on stream traces. On Raasay the streams for the most part drain northward along strike valleys carved in the softer shales. Every now and again they turn to flow abruptly east-west where a fault or dyke 18

has provided a local line of weakness. Over the eastern part of Raasay the main lines of drainage are eastward down. scarp slopes directly to the sea. The coastlines of the islands are for the most pakt of an irregular line of cliffs where the rook is geneially 'magnificentli exposed and often completely inaccessible without the use of artificial climbing aids. These cliffs are on the average about 15 m. high, but on the east'coast of Scalpay ascend to some 30 m. Raised beaches occur locally: the best examples oceuming along the west coast-of.Scalpay. The area is essentially bare of trees, except for small clumps of stunted rowan, hazel and biich, which' grow in sheltered hollows; though there is a Forestry OoMmission plantation stretching along the east coast of Raasay from Broohel to Screapadal. elsewhere the ground , is mostly covered by heather, bracken and peat moss, through which crags of rook protrude. The percentage of bare rock inoreaies with altitude. Above 300 m., perhaps 70% exposure is attained. The Torridonian parts of the islands are uninhabited -with.the exception of three crofts on Fladday and three . around Torran. The land is given over to sheep and deer. FIG.3. Recional structural setting of the area. AFTER THE GEOLOGICAL SURVEY.)

EASTERN LIMIT OF THE Glim "...... STABLE FORELAND.

MOINE THRUST PLANE.

5 KM.

APPL ECRO SS. / /

. / I

SCALPAY,

SK Y E / / / 20

CHAPTER II. STRUCTURE AND STRATIGRAPHY

(1) Structure One of the reasons for selecting the area as being suitable for a study of the sedimentation of the Torridonian rocks was its lack of structural complexity (see Plate ). All of the islands lie on the sub- Kishorn stable foreland west of- the Caledonian thrust belt (Pig. 3), (see Phemister 1960, p.41 and E. Bailey 1955, p.95). Tectonic-deformation, which is practically absent in the northern part of the area, increases in complexity-southwards towards the Caledonian thrust belt and Ouillins volcanic centre. .-The strata of Pladday and Northern Raasay dip at about 20° to the west, with one or two minor undulations. On Hcalpay the regional dip is more variable but averages .20° to the north-east. The dip of a bedding plane, as measured in the field, may be 4 kunotion of three vectors: (i)a deposition41 dip (ii)a compactional dip (iii)a tectonic. dip. : The first of these probably varies according to the style of depoiitiOn. Shales and horizontally laminated sandstones were probably deposited horizontally or nearly so. In:current bedded sandstones the master bedding 21

planes are probably erosional,baokslopes of once migrating.sand bodies (see Chap.V. -Sec.iii). Pettijohn (1962) estimates that these may dip thiee or more degrees up the palaeoslope. Breccias and grits may have depositional dips. of 20° or more, if0 as sometimes happens in the Torridonian, they are deposited as sorees. 'A'compactional dip is. produced by the hydrostatic pressure of overburden on unconsolidated fine-grained sediment. There are only places where this phenomenon is likely to be significant on Fladday and Northern Raasay.'- In the main Torridonian outcropbf Northern Raasay the westerly regional dip is modified by a gentle south-westerly plunging fold. It is possible that this fold is not tectonic in origin but is due to the presence of a' buried Lewisian gneiss ridge which rune along its axis (see Fig. 7). Compaction of the buried shales to the north and south of this ridge probably led to the differential subsidence of the overlying sandstones. In a similar way minor undulations of dip on Fladday can be seen to be controlled by the local shape of the basement. If this explanation of these minor undulations of strata is accepted then the only effect of tectonic dip has been to tilt Raasay and Fladday to the west at about 20°. On Scalpay the effects of compaction, if any,

I-30 M.

0 1.1CM.

I ...,1 I 1 ....,....--...._ • I... ------.1-, I. I , S. N. 11141:1•44 r -.1111311113•3.- X ,

FIG.4. DIAGRAM OF THE EAST COAST OF SCALPAY.

UNIT 7

H UNIT 17 5 WITH SHALE BANDS.

1 MESOZOIC SEDIMENTS.

.______DIRECTION OF YOUNGING OF CROSS BEDDING , INTERPRETATION

MI NOR IGNEOUS INTRUSIONS OMITTED. SOUTH OF

'X' BEDDING IS LARGELY ABSENT, THE ROCK IS WELL

JOINTED & EXTENSIVELY FAULTED. 23

•cannot be assessed,,as the basement,is nowhere exposed. Minor undulations of tectonic origin in the northern part of the island'increase, abruptly in intensity southwards. .A westerly` plunging overfold is deMonstrable (by inverted graded bedding) in the cliffs of the south-eastern part `ofthe'island (Pig.,4): Due to'poor exposure, and the ecaroity of criteria io establish the direction of younging, not been found possible to trace it inland. • The Totridonian' reeks of Raaeay-and Pladday are cut by two sets of faults trending approximately east-, '*estand north-south respectively. The downthrow and *splacement is often imPossibleto estimate due to the scarcity of reliable marker horizons. Where it can be demonstrated for 'the east-west faultsAA is generally seen that they ars-doWnthrown to`the north by only small amounts.''Tba fault breccias'aie cemented by calcite . and, more rarely, by minor amounts of gartz, haematite, siderite and malabhitet' These minerals may have been precipitated from solutions deriyedfrom overlying 'Cambrian ;or Tesozoic liMestones which have since been eroded away. _The age of the faulting here is difficult to establish. ,The exiatance of at-least two phases are established by Tertiary dykes cutting faults and vice- versa. 'The main Torridonian outcrop of Raasay is terminated by a N.B. -S.W: trending fault with a southerly downthrow of some 600 m. which brings downMesozoic 24

sediments. On Scalpay there are at least three ages of faulting. - One probably associated with the overfold- 'ing of the Torridonian and with the development of strain bands in the shales of the Southern part of the island; another post-Mesozoic but pre-Tertiary, and a third post-Tertiary phase. The trends of these faults are variable. The throws, where calculatablepare often considerable, letting down blocks of Mesozoic sediments and Tertiary lavas into the Torridonian basement. The

South7western part of Scalpay was intruded by a northerly extension of the granite of the Red Cuillins of Skye. This has given rise to a metamorphic aureole, some 500 m. wide, within which the Torridonian sandstones have been bleached and hardened to, 4 white 'quartzite' and the shales baked to, a splintery biotite hornfels. 25

(ii) Stratigraphy In common with other PreCambrian'sedimentary formations the erection of time -stratigraphiaal units in the Torridonian presents many problems. Miorofossils of uncertain biological 'affinities were described by (1907; p.287-8-and Plate 52, Figs. 1-1c), Naumova 'and Pavloyskii (1961) and Downie (1962), but ":so far they seem. to.be, of no stratigraphical value. The palaeomagnetio work of Irving and Runcorn (1957) showed that beds could be correlated for a distance .of over 60 kms.,on the assumption that the magnetic fabric of these'sediments was indUced during or very shortly :cater deposition,' On these grounds - they.concluded thatu over the area that they studied,' the Applecrosa:Diabaig . junction was synchronous. No palaeomagnetio work has been done'on the area studied for the preparation of tnia theiis. Thus the erection of time-stratigraphical units its imposeible, Only lithostratigraphical units May be delimited. lateral correlations are based on lithology alone 'and there is no evidence to suggest whether these are.eynchrOnous or diachronous. Mention has. been made already of the three-fold "lithological subdivision of the_TorridOnian" formation proposed by the Geological Survey (Ohap. I, Seo.iv). In this chapter the stratigraphioal divisions of the 26

(Irea proposed by the - ';urvey are described -and reasons aro givon for it3 necd- for modifination. - A new corios of lithostratiraphicra unit; are eroctrid based on the 'recognition of three rain lithofaciou of fundamental siificance,

J:he I°:1 mile G.,.tiap 81 shows the whole Torridonian succession of Pladday as belont;i1v: to the niabaig Group. iuland of Criana to the west ie narked he iundiffcrentiateu' (i.e. of uucertain utrAi6raphical position),. - In northern Vaasay Diabaiz," sediments in the ,,act arc„considered to be overlain by those of the )!.pplecrose croup to the -cot. `_'he bust: of the'Applecrous group was to en to be markEr,i by the first appearanca'of-4 pebble bed containing; phenoolwAs of non -Lewiaian pebbles. Au already mentioned (1:;hp..-1. ;;ect:iv).thie wat:'one of the criteria laid do,,ra by the nurvey In '1907. 1Towevsr, ffinxnan did not record any ApPlecrois rocks on Fla day, but the first occurrence of non-Iowisian phenoeltmitc is in the pebble bed which can be traced aloilg the v:hole of the - eautern aide. of the ieland t few metros inland from the cliff top. Thus all the fl'ortidonian uedim'ento to- the ww5t of this bed (and the bed itself j- should, on their 'criterion ) belonte the Applecroes and not the, Oisbaigroup. 27

The Torridonian rocks of Scalpay are shown on the G.S. Sheet No. 71 4s 'undifferentiated' but a line is draWi dividing them into two groupb defined as follows:' 'Upper Division: Coarse false-bedded feldspathic' sandstones with pebbly seams and beds throughout Lower Division: Fine feldspathic sandstones, never pebbly, passing -down through flaggy beds into Close-grained quartzose grits with little or no feldspar and of a grey.or white colour.' (base not seen.) 'Harier (Geikie 1901, p.15) held that the Upper-Division 'belonged to the Applecross group according to the Survey's .own criteria,-but for some reason this suggestion was dropped' from subsequent Survey-publications.

CenV-Qtli•Zk In the field it was-toundto map' three major sedimentary facies defined on lithology, colour and sedimentary structures. These are as follows:- Facies Colour Lithology Sedimentary structures 'Red Red SandstonesoAlizin Lenticular bedding coarse &pebbly Large-scale cross- 'shales , stratification

Quicksand structures Grey Grey Shales sand- Regular bedding. stones. rare i Small-scale cross- calcareousbande stratification, grading ripplemarks, & desio4' cation cracks. Basal Red or Boulder beds, kassive or poorly Grey. breccias Lt grits bedded

Each of these facies reflects a different sedimentary environment and may in turn be divided into sub-facies and sub-environments. Based on these facies four lithostratigraphioal units may be mapped out on Fladday and Raasay as follows:

Lithos tratigrayhical. Units Facies types Raasay nadday Glame bids Upper FladdNy Mainly Red, but with minor beds amounts of Grey. Junction with the Lewisian gneiss not seen. Lochan Lower nadday. Red, but with no pebble bands Uachdair beds beds of non-Lewisian fragments. Rare Grey facies bands. Red basal facies present adjacent to the Lewisian gneiss.

N.N.E. S.S.W.

A LLT RUDHA RUDH A REIREAG. CHONNAIDH. REIREAG.

•••• 1 I •••• ... t 'I _ •••• .... _. L. _ 1 •••• ... ••••• ..1 1.... • 1 ... •••• I I 1 ••• ••• 1 ; i r ..- --- t t / 1 1 -...... , .... -. ; ;...... - .- ..-I t i ; 1% r / I -- ' I 1 -16 1 ; t I I \ / 1 + ++++ + + + 4- + + + - - - = =-- _- -7 :: -_---- _Z.- .. -- . . . A 1 5 I I I IN a11111 - . I I 1 1 I I II I I I I I I 1 I I 1 I I I I P-I- PP'-..-- ---''" ''.--- - -' - t- -'.-7:"- : :^ ;:=- :-- 1- .-- •. --S = - •••• • 110 ••• •••• •••• 5- .••• ••'• •no ••••• •••

FIG.5 Coast section from Rudh/a/Chonnaidh to Rudha Reireag,Scalpay.

• • • • • • BEACH e•• • •• • UNIT 5 M I , GRANITE TOPPI DONI AN. ZOOM. MINOR INTRUSIONS OMITTED. In 30,

Brochel beds- Caolas Grey, and with grey Basal Fladday beds -facies adjacent to the Lewisian gneiss,

Torran -beds . • • • • 'Red Basal facies. Lewiaian gneiss

On Scalpay it is -possible-to map out five lithostratigraphical unitst .the type section being on the west coast between Rudb' Reireag and Rudh' a' Chonnaidh (Fig. s)'. The sequence-seen here is as follows:-

Red facies "'”IV '7 Red' facies but with no.pebblebands " III - Interbedded- Red and Grey facies - Grey facies - I ? A medium-fi.ne grained sandstone .always bleached because it only crops •'" 'out within the granite aureole. Base not seen.,

Figure 6 Summarizes. these-new lithostratigraphical units, gives their estimated thicknesses,- the Survey correlation, and that proposed by the author. In the following three chapters the rocks will be described facies by facies, :ether than by lithostratigraphical units,- since this would involve much tedious repetition.

The following three cases of lateral facies variation , can be seen in the area: i) Locally on Pladday and Raasay Basal facies rocks pass laterally into those of both Red and Grey facies. '5.0 The junctionOf the Red and Grey facies is abrupt in .the North but.-becomes transitional southwards. iii)'The Red facies rocks of the Applecross group become increasingly coarser southwards, shales become progressively' rarer and pebbly beds more abundant.

The sigtffica4ce_of- ihese changes is discussed in Chapter IX. RAASAY

I KM. 1 BROCH FIG.7 Map of the basement between Fladday & Brochel. : LEWISIAN : TORRIDONIAN UNCONFORMI TY . PRE SENT LEWISIAN C ON T OU R SP5M.INTERVALS) : SUBSURFACE CONTOURS OF THE BASE OF

THE TORRIDONIAN DEDUCED FROM THE GE OMETR Y

OF THE UNCONF ORM' TY . 34

'CHAPTER III. THE BASAL UNCONFORMITY AND THE BASAL FACIES

The bas al uneonformity The unconformity between the Torridonian and the Lewisian:may be traced' discontinuously from the north end of Caolas Fladday to Broohel in north-east Raaeay, a distance of some 5 kms. Prom the south end of Caolas APladdaY to, the south side of Loch Arnish, with the exception of the Torran outlier, it has been removed by .recent marine, erosion. The attitude .of the unconformity may be seen in three.diMineions in the accompanying structure contour map (Fig. 7). This shows that on the east side of Caelds Fladday, on the south side of Loch Arnish and at Brochel, the unconformity_ is very steep. Inland, along the east side of Loch an Uachdair and Loch -Beag between Loch Arnish and Brochel it is almost horizontal. 'Three small 'valleya, exhibiting various etagea ofxemoval of their Torridonian infilling, are seen at the south-east end'of Looh an Uachdair. The Torran outlier looks on the geological map like a long thin valley filled with'Torridonian. In point of fact, as,Fig. I shows, it.occupies the northern slope of a partially re-excavated amphitheatre-shaped hollow in the- gneiss. Thus it can be seen that, in general, a rsteep slope or cliff extends from sea level to a height of 35

some'60 m: before abruptly curving over'to form an undulating plateau incised with several small gullet's. The cliff was breached, however, in the Torran region, -where a much gentler slope prevails. When allowance is made for the regional dip the cliff becomes less spec- tacular, now appearing.as a elope of some 60-70°. It is'difficult to determine how much of the present Lewisian exhumed surface conforms to that of immediate pre-lorridonian times. Small outliers, smears and fissure fillings of Torridonian have only been found adjacent to the present main outcrop. The Lewisian gneiss of South Rona and Raasay, which stretches for 15 kme. to the N.N.E.,,may well be a re-excavated ridge of the old Pre-Torridonian surface, laid bare by subsequent ,erosion of its Torridonian cover. Two possible geomorphological interpretations of the Pre-Torridonian topography may be advanced. The steep slopes may be those of desert inselbergs, implying the subaerial erosion of a landsurface. This was the - interpretation proposed by A. Geikie (1880), who considered that erosion had been by ice action. The Duke of Argyll (1880), however; thought that the glaciated contours of the gneiss were more.probably due to the present Ice Age. Alternatively the irregular topography could be attributable to marine erosion. The unconformity here provides insufficient evidence to solve the problem LE WISIAN GNEISS.

\\\ BOULDER \N- BED S.- - N)to /z y& \_ 3 V -5 ) / _ - " ..---- BRECCIAS & GRITS.

GREY SHALES FALSE-BEDDED GRITS . FIG.8. The Torridonian:Lewisian unconformity at Caolas Fladday,looking East from Eilean Fladday. AFTER A Fl ELD SKETCH . 37

of its' origin. This topic is discussed again in more general terms in Chapter IX.'

(1024e Basal facies a) Introduction - The Basal facies beds occur mainly at the base of the succession and always abut against the Lewisian gneiss. - They in part underly and are in part the lateral equivalent of the Grey facies rocks. Red facies rooks of the Loch an Uachdair beds overstep both of the former to abut against the gneiss between Loch Amish and,Brochel, where they show very little development of breccia. The rocks of the Basal facies are boulder beds, breccias and grits with occasional thin shaley bands. They are exposed on the east side of Caolas Fladday and at Torran. Smaller patches occur on the south side of Leoh Arnish and between the footpath and the sea shore a few metres east of tie-fisherman's bothy at Brochel. 'b) The Basal facies of Caolas Fladday. The east side of Caolas Fladday is a steeply sloping cliff some 50 m. high and 1.5 kms. long. For 0.4 kms, along its base opposite Eilean Fladday the mar-

ginal breccias are well exposed. They dip west, off the gneiss, at about 30°. The unconformity itself dips west at about 80°. The rocks here are seen to be red and green gritty shales and grits in which are embedded PHOTO 1. Basal breccia. 31.ae of Caolas Fladday.20Orn N. of the cause'T;ayo looking S..

PHOTO 2. Large fragment in basal breccias.E.sid- rf Caolas Fladday. 50m N. of the causeway,,looling S. 39

angular fragments of Lewisian gneiss of all sizes from a diameter of 2 m. downwards.(Phoio.1). Gritty shale can be traced for sole 25 m. up the cliff, infilling cracks, and forming thin.smears on the gneiss. The dip of the:shale is, locally very variable and is_closely related to the shape of the underlying gneiss. The _presence of sliokensides on the surfaces of the gneiss boulders suggest that differential movement due, to compaction took place between them and the enclosing matrix. Evidence of compaction is also provided by the cataclas- tic deformation of many of .the sand grade particles in the-matrix, for, as mentioned in the preceeding Chapter, there is no sign of violent tectoniem in this region. In one instance the shale at the base of a boulder is violently contorted (see Photo.2). This may have taken place when the boulder was first deposited into the mud, or again it may have been due to subsequent movement of the boulder during compaction. Some 200 m. north of the causeway the basal brecoias are overlain by some 10 m. of a pale grey-green gritty sandstone. Individual beds, some 6-10 ems. thick are separated by thin partings of red shale. Similar. shale laminae occur between foresets of the planar cross- stratification which is the oharacteristic sedimentary structure of these beds (Photo.3). These sandstones are composed of clean-washed coarse angular grains of l - - - - .'-',.- . •. .- . -. .- . _.7...... iv! --- 0I L..._:...... 7...._

.. - ' . - - • - - • -:- . • . 4, _ ,

PHOTO 3. Cross-bedded epidotic grit.E.side of Caolas Fladday,400m N. of the causeway,looking N. I K M.

/ 20 I/ / 20 2.0 20

19 •`

• • 2.0

LOCH / 20 //

A RNI SH 10

"tk k

FIG.9. MAP OF TORRAN OUTLIER .

UPPER TORRAN BEDS. :- DEPOSITIONAL DIP • • • • • • CORRECTED FOR A

I • 4 • TECTONIC DIP OF • • • • LOWER N.270.E. AT 20. r r • • PRESUMED LEWIS1AN GNEISS. DIRECTION OF SCREE FLOW.

HATCHING INDICATES PRESENT RELIEF, ALLOWING FOR REGIONAL TILTING. 42

quarti,' feldspar and yellow-green epidote. The junction between - the basal beds of the east - side of the Caolas Fladday and the grey shales of the west aide is not seen due to marine erosion. The tombola joining Fladday.to Raasay at low water shows a few exposures of:greTshale. Figure 8 shows.a general view of Caolas Fladday taken from a field sketch. c) The Torran beds. The beds of the Torran outlier crop out for a distance of kmi. along the east aide of Loch iirnish and extend inland as'a south-easterly tapering Strip ending above. the crofts at Arnish. The dip of these beds is- very variable and reflects'a combination of depositional and tectonic dips. The accompanying Map (Fig. 9) shows the directions and amounts of the depositional dip seen after removing a regional tectonic dip of 20° to the west by means of a stercographic net. This map also shows the distribution of the two divisions into whioh the Torran beds are grouped. Illese are: I. Upper Division - Dark red, non-pebbly grits with shaley partings. II.Lower Division - Interbedded dark red grits and brecoias with boulder beds present adjacent to the basement.

Gneiss PHOTO 4. 1 Y Lc A~ In the .:go y., l° Iovran Bed8.E.side of footpath 150m S. of Torran schoolhouse.

PHOTO 5. lieu grits of tae up,oer 11:orran Beds rooting on LewisiEm Gneiss.E.side of Loch Arnish,100m S. of Fishermans botIly.Looking N.. 44

The rocks of the Lower-Division are mainly massive boulder, beds and poorly bedded breccias. The former mainly occur at the base of the succession and adjacent to the gneiss towards the NOrth. They consist Of angular boulders of gneiss, up to several metres in diaieter, embedded in a rubbly dark red gritty matrix. The breccias consist mainly of angular fragments of quartzo-felspathic gneiss in all sizes from'several.centimetres downward, embedded in a matrix 'of dark red grit (Photo.4). In thin section the latter is seen to be composed .of angular grains of quartz and felppar.with subsidiary amounts of mica, epidote'and ;detritalyon ore. ,The cement is a mixture of quartz and haematite. The rockevof the Upper Division of the Torran beds consist almost. entirely of dark red grits. In the south they rest directly on the gneiss, but to the north they'overly the beds of the Lower. Division with an imper- ceptable unconformity. These rocks are massive or poorly bedded. In marked contrast to the Lower Division rocks they are almost'completely devoid of pebbles, or larger fragments,, of Lewidian gneiss, even where they directly ,overly it (Photo.5). Bedding is occasionally picked out by bands of laminated shale a few millimetres thick. These are red or pink in colour, becOming green in the southwestern part Of the outcrop. 45

-(iii) Discussion of Basal facies The coarse poorly sorted nature of these rocks, together with their abundant fragments of Lewisian gneiss, ,suggest that they are of local origin. Their field relationships. lack of bedding and steep depositiomidip lead to'the conclusion that,they originated as screes and fanglomerates at the foot of the Lewisian hills. The basil beds of Caolas Fladday, the south side of Loch Arnish and of Broohel may well have been deposited under water, for - they are:the lateral equivalents of Grey facies muds. During deposition they may thus have resembled' the Subaqueous part of the famous Wastwater screes of the Lake District. The red Basal beds of Torran, however, may well have been true subaerial deposits. They compare favour- ably with fanglomerate deposits both-ancient and modern, e.g• the Tertiary fans of the Big Horn Mountains, Wyoming (Sharp 1948)p - the ,Gerrish facies of the Wolfville Form-. ation (Trias),,Canada (Klein 1962) and recent fans described by McKee (1957). The depositional dips of the Lower Torran beds indicate that they were derived from two fans building out from the north-west and the north- east. The Upper Torran beds represent a fan which built out from the south-east at a later date and partly buried the earlier Lower Torran deposits. Their finer grain size and better sorting may indicate a more distant source. r 30 M. 1 0 LK M .

W. E1 LEAN CA OLA S RAASA Y, E. F LADDAY, FLADDAY,

FIG.I0a. Section of Eilean Fladday across the causeway. UPPER FLADDAY BEDS .

00 0 LOWER 0 0 BASAL BEDS.

CAOLA S 11 11 LEWISIAN GNEISS. 47

CHAPT8R IV.- THE GREY 'FACI'BS

(i) Introduction The rocks ofithe Grey' facies are grey-or grey- green in colour due to a chlotite.rich cement. Lith- °logically they include sandstones and shales, with subordinate quantities of thin breocias, impure limestones and clays. Stratigraphically the Grey facies includes the eaolas Fladday, Brochel and Unit II beds. Thin bands of Grey facies sometimes occur interbedded with Red faciesAmodiMants. This occurs, notably in Unit III of Scalpay, in one or two places in the 'Loch an Uachdair beds of Raasayand,' very rarely, in Applecross sandstones throughout the area. The more important occurrences of the Grey facie's will now be described in detail.

'(ii) The Caolas Fladday beds a) Description ,The Caolas Fladday beds are exposed for a distance of some 1.5 kms.',in the cliffs along the west side of Caolas Fladday. They are at least 30 m. thick. The lower 10 m. or so are not exposed as they crop out below sea level in the Caolas itself. Although the base of the Caolas Fladday beds is .not visible, projection of the bedding planes shows that the bads in part overlap and PHOTO 6. Junction,marked by head of hammer,between the Caolas Fladday Beds, below,and the Lower Fladday Beds,above. W. side of Caolas Fladday,150m S. of causeway. Looking N.. FIG.11. Stratigraphic sections of the Caolas Fladday beds.

100M. SOUTH 200M SOUTH OF CAUSEWAY OF STAC A RUADH.

- 5M. CAOL A S FL AD DAY BEDS

PH RED SHALE LOWER FL AODAY BEDS . ILLij RED S-ST N . J

GREY SHALE.

GREY S-ST N oo PSEUDONODULES.

CALC.S-ST E0-

INTERBEDDED vv WASHOUTS. GREY SHALES & S-ST NS 50

in part pass laterally into the basal breccias on the eastern side of the Caolas (see Fig. 10a). The junction with the overlying Red facies Lower Pladday beds is ,"sharp, and well exposed. at several localities along the cliff top, (Photo.6). The shales and mudstones of the Grey facies are capped by a medium-grained grey massive

sandstone .1 m. thick. This is overlain by 0.6 m. of micaceous shale, the lower part of which is grey and the upper part red. This shale is followed by medium-grained sandstones of the Red facies. Two stratigraphic charts of the Caolas Fladday beds, taken from different localities along the cliff, are given in Fig. 11. These show that there are three main interbedded rook types: a)Grey micaceous shales b) Fine-grained grey sandstones c)Carious weathering fine grained calcareous sandstones. The grey micaceous shales occur throughout the whole sequence, and in the lower part of the section occur to the exclusion of the other two lithologies. They are well laminated and split readily along the bedding into sheets only a few millimetres thick, the surfaces of `vhich show sparse flakes of detrital muscovite. These shales often exhibit polygonal shrinkage cracks of two kinds. There is a large-scale variety having a width of 1 cm. or more, and being infilled with fine-grained

•

. 1

1 1 _ ...

1 , .:-,., NI L. . b..? ..../"! tir-1---.0A,:a )6immAil, -- PHOTO 7. Large scale dessication cracks in the Caolas Fladday Beds..4side of Caolas Fladday,l5Orn S. of Stac a'Ruadh.

PHOTO 8. striking ripples in the Caolas Fladday Beds.4.side of Caolas Fladday 150m S. of Stac a'Ruadh. grey sandstone. They are of the complete, first order, type of ,Shreck (1948. p.192). See Photo 7. The second ' variety are very,much'smaller; they are seldom more than 4 mms-.wide and are infilled with the same material as that 'of the shaleECin,which they are developed. These shrinkage structures are,common in the Diabaig shales of the Torridonian and have- previously been interpreted as , evidence of-subaerial exposure of the muds. (Hinman 1907, p.273, Phemister. 1960, p.45). This may be correct :For the larger variety infilled- with coarser material, but the smaller ones, infilled with a sediment similar to the. surrounding rock,' May-well be subaquemis.synerisis craOks, (White 1961). Another sedimentary structure present in theee_shaliS is symmetrical ripplt marking. Transitional stageS:, ean be found between this and. the assymetrical ripples of the sandstone units. In the ,'upper 10 m. or so of- the Oaolas Fladday Beds the shales areintetbedded with fine grey sandstones. These gen- orally oceur in units 10-20 oms. thick, - separated from one another by a stmilar amount of shale. The base of 'the sandstones are generally flat, but the upper surfaces often weather out to show symmetrical ripple marks. These have a wavelength of 4'-5 oms. and an amplitude of 0.5-1 ems. (Photo.8): .These ripple marks are typieally- compoaed of"continuous undulati4 laminae*, but some examples show obliguefOreiet'lamination. Internally. the sandstones

FIG.12 Ripple structures in the Grey facies.

INCREASING WAVE ACTION. DECREASING SEDIMENT SUPPLY 811 CURRENT ACTION .

DISCONTINUOUS LAM I NATION. SYMMETRICAL WAVE RIPPLES

MODIFIED CURRENT RIPPLES MUD WITH LENTICLES. (SEE ALLEN P ET AL.1960)

ASYMMETRICAL CURRENT RIPPLES.

5 CMS.

ALL EXAMPLES DRAWN FROM COLLECTED SPECIMENS & FIELD SKETCHES. THEY ARE COMPOSED OF VERY FINE SANDSTONE ENCLOSED IN MUDSTONE OR SHALE. 54

show small scale current bedding; this occurs generally in lenticularbands cms. high. The tops of the sandstone are not always sharp for they occasionally show interbedding with the shales. In these instances it can be seen that the current bedded sandstone is -succeeded by perfectly preserved asymmetrical ripples, thin ripple-laminated lenses and finally by impersistant horizontally laminated bands only a few milimetres high. These changes probably reflect a gradually decreasing current velocity,, and alOo suggest that the cross- stratificationwag produced by migrating ripples (see Chap., V, Section iii). Fig. 12 shows the relation between ripples and stratification. Minor structures - morphologically similar to these'have .been described and figured from the -tidal flat sediments of,the Wadden

See by'Van Straaten (1959, Fig. 3, 1954, p.104,Plate 8, Fig._ 15, p.105iFlate'5,Pigs. 20,21). Some sandstone bands Contain irregular nodules of `red earthy haematite, ocoasionally over 1.5 cms.in diaMeter. .0te i'ragment,of sandstone was found, unfortunately not in place,, with an irregular efflorescence of grien malachite. :ThiErie of particular interest in view of the presence of cupriferous nodules in the Diabaig described , shales'of Sto-erxby Fermor (1951). In the cliffs 100- m. south of Stec a' Ruadh the sandstones are seen. to have undergone penecontemporaneous , "' - . _:1 - - ' '• • - . ---- 101.,

,.. !,,rt, ..•.--10 ' - i -;.-- 3

407 1'1".:Z--.-" •

:0 9. Cross-bedded grit fled ashout in grey shales. Caolas Fladday Beds,stui. of Caolas F1adday,150m 8. of Stec a'Rusdh.Looking

PHOTO 10. Limestone bend in Caolas Pladday Beds.150m 3. of Stec alRuadhs .side of Caolas Fladday. 56

deformation to form pseudonodulee (see Chap. VII, Sec.ii). At the same locality, but several metres lower down, there is a horizon in the shales containing a number of washouts. The washouts are generally up to 1 m. or so wide and about 0.2 m. deep. ,They are infilled with false bedded coarse sandetone with gritty lenses (see Photo.9). A very minor part of the Caolas Pladday beds consist Of a series of arenacoous limestone (634 carbonate by'Canalysis of a thin section) bands. These modal ? are. mostly developed in the middle -part of the succession, and occur in impersistant beds equivalent in thickness to those of the sandstones. When fresh they are grey, but weather .to a brown colour with a characteristic carious appearance (Photo.10). Like the sandstones they are characterized by the presence of small-scale current bedding. In one bed 100 m. south of the causeway, incipient loadcasting had developed, controlled by the position of current ripples. A similar phenomenon has been recorded 'by Dzulinsky and I(otlarczyk (1962). In thin section the carbonate, which on staining proved to be Calcite, is seen to be present in two forms.. Some of it has replaced detrital grains of quartz and feldspar. It'is also present as a curious matrix. This

'occurs in - the form of discreet clear structureless bodies, Utah of which, in section,. has a shape like a threepenny 5'7

piece, They are about 0.02 mm. in diameter, and occur singly ,or in grape-like bunches in the spaces between the detrital grains, (Fig. 10b). These curious bodies were described by Teall (1907, p.287) who commented on their possible organic origin.

.:b) Discussion of the'eaolas FladdaLBeds. Taken as a whole the lithology and, sedimentary structures of the. Caolas Pladday beds suggest that deposition took place in quiet shallow water. The presence of large scale desiccation cracks 'suggest that the water periodically receded or evaporated. The laminated shales were possibly deposited by suspension and the current bedded sandstones by gentle traction currents. More violent currents may have been active for a short space of time...to cut and fill the horizon of washouts. To the east the waters lapped against scree flanked Lewisian cliffs. The small crop of current bedded sandstone at the north-east end of the Caelas may represent a beach deposit, but, apart from this there is no evidence of strong wave or current activity'along the shoreline.

(iii) The Brothel Beds a) Descrlition. The BrOchel beds are exposed along the east coast of Raasay, and for a short distance inland for just under COAST SECTION SOUTH FROM BROCHE L,RAASAY. LENGTH. APPROX.I KM. S. F F - t

0 LI C> TERTIARY VENT

..LOCH AN UACHAIR BEDS. .B ROCHE L BEDS.

IF J1 .LEWISIAN GNEISS.

-20 M 0111111110 FIG.1 3. Stratigraphic sections WWI of the Brochel Beds. 0 -

PS E UDONODULES

LEGE ND.

• • • • . • . . COARSE LOCH AN UACHDAIR •.• : . RED SSTN • BEDS.

COARSE GREY S-STN. BROCHEL INTERBEDDED BEDS. GREY SHALES ms S-STNS. MUDFLOW

GREY SHALES. WA MI 0011.11111 CALC. S-STN. 59

1 km. south of Brochel. To the north they abut against the Lewisian gneiss at Broohel, but the junction is not well exposed. small patches of green basal breccia resting on the gneiss may be traced from the coast, east of the salmon fishers' bothy, north-westwards to the bealach above the footpath. The Brochel beds, at least 140 m. thick, are succeeded .by the Loch an Uachdair sandstones of the Red facies, which,at their base,contain one or, two thin bands of Grey facies sediments. From the stratigraphie section (Fig. 13) it can be seen that the Broohel beds are composed of five interbedded rock, types: Three are of common occurrence, two occur only once each in the'whole succession. Major rook types of the Brochel beds: a) Dark grey micaceous shales-with red sandstone partings.

- b) Fine-grained dark grey-green sandstones. c) Medium--coarse grained pink-grey sandstones. Minor occurrences: 0' Grey calcareous sandstone., e) Grit.lense. - These different rock types will now be described in turn. The micaceous shaled are the commonest rock type of the- lrochel beds. They are usually grey in colour, .but tend towards-,purple in the,upper few metres of the Succession. Lamination-is not so well marked as in the O5 M.

•

Cl. yebshale. Red s-st"ne.

FIG.14a. Dessication cracks. Brochel Beds.

14.B.

14 A.

P'NOD. BED SEE F1G.3

LOCALITY SKETCH.LOOKING NW.

shale shale

F 1 G 14b Scour & fill structure.Brochel Beds. 61

Caolas Fladday shales, 1114 thee are often thin, irregular partings og red gritty sandstone a few millimetres thick. The shales often show symmetrical ripple marking of a scale, and type similar to that of the Caolas Fladday beds. Large_scale polygonal shrinkage cracks, generally infilled with red sandstone, are common. Frequently differential compaction of the enclosing shale gives them a ptygMatic appearance (Fig. 14a) similar to that described from the mudflat deposits of the Wadden Sea by Van Straaten (1954, p.76). One speeiMen of purplish shale showed a freshly exposed bedding surface with a number of pits, each some 2 mm. in,diameter. Similai structures described from elsewhere in the Diabaig shales were interpreted. by Iiimeman as rain prints (1907', p.325). Inferbedded with the shales.at various levels are bands of fine-grained,grep-green sandstone which weather out to a bright'red colour. These beds are-generally between 10-20 ems. thick, successive units; being separated by a few centimetres'-of grey micaceous shale.• When traced laterally south-south-westwards along the'coast these sandstones are 'Been to be lenticular (Photo.11). The bases of the units can often be seen to be erosive,' truncating shrinkage cracks and showing scour, and fill structure (Fig. 14E). Frequently these sandatonee are graded; their bases being marked by a line of angular ROCK TYPE. ESTIMATED GRAIN SI ZE.

2MM.

DESSICATION CRACK

SANDSTONE

SHALE

FIG.15b.Grain size variation in a hypothetical section of part of the Brochel Beds. 63

pebbles, up to 2 oms. in length, of quartz, feldspar and reworked shale. In:thin section Concentrations of heavy minerals are sometimes seen near the base (Pig. 15a). The.middle parts of these sandstone units are: generally massive, poorly sorted, and may be petrographically termed a greywacke (see Chap. VI, p.7). The junctions of the sandstones with the overlying shales are generally sharp. The last few centimetres sometimes show very low angle current ripple bedding. Multiple grading is often seen, an influx of coarse material not uncommonly occuring right at the top of the sandstone. Fig. 15b summarizes the vertical variation in grain size seen for a typical sequence or these graded beds. Penecontemporaneous deformation of these beds has only been observed in two instances. One sandstone, not graded, having sunk down into the underlying shale to form paeudonodules (see Chap. VII, Sec. ii), and another exhibits the streamer type of interstratal deformation (Chap. VII, Sec. iii). The third common rock typeim the Brochel Beds is a medium grained sandstone which occurs in massive or poorly bedded units several metres thick. When fresh it has a minutely mottled pink and grey-green colour due to the presence of pink feldspar grains and a chloritic matrix. Within'these sandstones impersistent bands of PHOTO 11. Grey shales & fine-grained lenticular graded sandstones erosivelky overlain by coarse sandstone. Brochel Beds,E.coast of Raasay,500m S. of Brochel. Looking

, •

PHOTO 12. Coarse sLidstone, n erosional base, overlying grey shales and mudstones.Foreshore 25m S. of Brochel Castle.E.coast of Raasay,looking W.. 65

red shale, 1 or 2 ems. thick *occur.at vertical intervals of about 1 m: or so. The bases of the sandstones are often marked'by irregular erosional surfaces (Photo.12). The overlying few centimetres of sandstone, often being coarse-grained and containing angular pebbles of Lewisian rocks and-reworked Torridonian shale, are up to 3 cms. in length. Towards the top of the succession the beds redden, and thus pass up into the Loch an Uachdair sandstones which they otherwise largely resemble. Frequent undulations of the shaley partings indicate that mild penecontemporaneous deformation occurred from time to time (see Photo.12) On the foreshore, 10, m. south of Brochel Castle, there is a band of grey carious weathering calcareous sandstone (30 carbonate by volume, calculated from thin section). Microscopically this rock is seen to consist '- of fine grained sand particles poikilitically enclosed in irregularly shaped calcite crystals about 0.5 mm. in diameter. The enclosed.detrital grains often exhibit signs of replacement along their margins, but not to the same extent as in the Caolas Fladday limestone. The crystalline form of the interstitial calcite of this rock suggests that it is a cement, unlike the calcareous sandstones of the Caolas Fladday beds where it is present as a matrix. -48

4140 't ••• t - • 2 Er-, • 4-1.' _aft

- - • :.... "

PHOTO Pebble-ceoed lent' i; of the Brochel mudflor embedded in grey shales.Foreshore below the eastleolooking N.W..E.coEst of Raasay.

/4744r.N..tt

• --rV. 4t

PHOTO 14. Pebbly ton of the Brochel mudfi ow. Hemmer al igned N. W. -S. E. 67

Also on the foreshore at 'Brochel 1 m. or so above the calcareous sandstone, and interbedded with grey- shales, is a curious tense of crudely bedded grey-green grit. - The lense attains a maximum height of 20 cms.and a'width of 5 m. in A northeasterly to south-westerly direction IPhoto.13). It may be traced inland'for about.20 m. before it is'overlain by shale. The base of the bed is flat for the moat part, but in several places is in the form of.steepsided erosional-channel 2 or:3 ems. deep , and up to.5eme. high; The lowest 2 or 3 ems. of the grit contains fragments of quartz and'feldspar. up to 5 mm. long. , PartiallyeMbedded in the upper part of the bed, is a layer of angular pebbles-up to 10 ems. in length (Photo.14). The overlying shales are draped over the top of these. Estimated relative frequency 'of the rock types cif the pebbles are: Quartt>Quartzo-feldspathic gneiss Micaceous gneiss>reworked basal facies grit. The azimuths of the long axes of 1po pebbles were measured (Pig. 16); they show a marked northwest-southeast orientation parallel to the isopachs'of this bed (Fig.16). The mode of deposition of this bed is perhaps a cause for.considerable speculation. The most likely. possibility is that it was a subaerial mudflow. It shows many of'thecharacteristics of recent mudflows described by Blackwelder (1928), and others, viZ.ua marled lenticularity, normal, to the direction of movement, an 0 1 2 M.

POSE DIAGRAM OF THE LONG AXES OF 100 PEBBLES.

F 1 G .16 MAP OF THE BROCHEL MUDF L OW.

. • . MUDFLOW. ISOPACHS DRAWN PARALLEL BEACH. TO MEAN PEBBLE TREND. SHALE. '69 erosional base, poorly developed bedding, a convex upper surface, and, perhaps most characteristic of all, the rock fragments lie on the surface of the bed, not within it. It is this last feature which distinguishes it from the subaqueous mudflows where•liquefaction and slumping causes the phenoclasts to be randomly dispersed to form a tilloid (see Crowell 1957, Schemerhorn and Stanton 1963, Winterer, in press). The one unexplained fact is the well marked orientation of the superficial pebbles. 'This is hardly likely to have occurred during the movement of the flow for in recent mudflows material, up to the size of large boulders, bounce up and down as they are ,Carried along (Blackwelder 1928). It seems more likely that the orientation was caused by the flood of water which succeeds a mudflow; this may have only had sufficient velocity to arrange the, pebbles into the line of least resistance and not to actually erode them. Not only do the lithological characters of this bed suggest deposition by a mudflow. Three of the four conditions necessary for a mudflow (Blackwelder 1928) were indubitably present, viz.: unconsolidated material, steep slopes, and labk of vegetation. There is no reason why the fourth condition, abundant surface runoff at certain times, was also not available. Thus both the lithology and geological setting of this curious bed point to its having been a subaerial mudflow. 70

b) Discussion of *the Brochel Beds. Taken as a whole the Brochel Beds have many broad lithological similarities with the Caolas Fladday beds. One-may thus assume that they were deposited under broadly , the same conditions, though the presence of more, and coarser, sands suggest somewhat more violent current activity;

(iv) Units II and III, Scalpay. ' Unit II of the succession established on Scalpay crops out along the west coast between Rudha Reiraig and Rudhla''Chonnaidh (Pig. 5) and in Allt Reiraig, consisting of some 70 m. of Grey facies sediments. The overlying Unit III, some 200 m. thick lconsists of interbedded sediments of Grey and Red facies. It.crops out in a belt southwestwards from Rudh' a' Chonnaidh to Loch Cairidh, and for a short distance along the southern shore between Rudha na h'Uanha Dubh and Rudh an Droma Bhain. Unfortunately the Grey facies rocks of these two units occur mostly within the metamorphic aureole of the granite, and_are unfavourable for detailed sodimentologi- cal study. As far as one can ascertain they are broadly similar to the Grey facies rocks already described from Pladday and Raasay. Lithologically they consist of grey micaceous shales (now often hornfels), with red sandy partings often with quartz and feldspar granules. They "....e" . •

ArN -"* .w. • -s4- 1,111g P'HO'TO 1111, Tnterbedded grey laminated shales 6 current-rippled sandstones. Unit t1. 5m l'of the mouth of Jilt Reiratg...mast of Sealpeyflooking,114, 72

show red sandstone filled shrinkage cracks, ripple marks and. occasional current ripple bedding. Interbedded with these shales'are medium-fine grained grey sandstone units 10 ems. or more thick (Photo.15). Internally they show current bedding in sets up to 3 ems. high. Grading has not been observed in these sandstones.

(v) Other occurrences of Grey facies sediment. Thin sequences of Grey facies sediments are occasionally found interbedded in predominantly Jed facies rocks. ' These are seen notably in three places. On the southeide of Loch Arnish, on Raasay, three bands of grey facies sediment occur interbedded with the red Loch an Uaohdair sandstones. The two lower bands die out against the Lewisian gneiss to the south. The upper one can be mapped southwards along the strike until it is lost in Broohel forest. •In no instance do any of these three bands exceed a thickness of 4 m. They consist of interbedded fine-grained current bedded sandstones and rippled grey shales with sandy partings. Two more sequences of Grey facies beds occur right at the top of the red Applecross sandstones on Raasay. These crop out on the west coast opposite Balachairn and may be traced inland northwards for about 5 km. before they are cut out by a fault. Each sequence is between 3-4 m. thick and they are separated by about 5 m. of 73

Red faoies sandstones. They are composed of interbedded grey rippled shales,with,red sandstone partings and fine- grained grey sandstones. One sandstone band in the upper bed has formed pseudonodules,(Ohap.VII, Soc.ii). On Scalbay several bed6 of grey sandstones and shales, seldom'more than 1 m. thick, occur in the. red pebbly sandstones of .Unit V. These occurrences are shown 'in the geological map'(Plate 1i) at the back,of this thesis. 74

-CHAPTER J. THE RED PACIES-

(i)Introduction ' Rocks of the Red facies occupy a greater thickness and cover a larger area than both of the other two facies put together.-. They are exposed on Fladday and the associated islets of Griana Sgeir, Glas Eileen and _Mean Fraoch, and on Raasay and,Scalpay. Stratigraphically they consist of the Upper and Lower Fladday beds, the Loch an Uachdair and Glam. beds of Raasay, and Units IV and V on Scalpay. In Unit III Red and Grey facies beds are intercalated, as mentioned in the previous chapter.

(ii)Description of Lithologies As their.name implies the sediments, of this facies are red,in colour. They vary from pink in the coarser 'sandstones, where'detrital'graine of jasper and feldspar are the' colouring agents, to a dark red-broWn in the fine sandstones and shales where a haematite clay matrix is' abundant. Five lithological types may be distinguished, of which two are of only minor importance: -Manor lithologies in. the Red facies a)coarse-very coarse sandstones, often pebbly. b) medium-_fine-grriined sandstones, c)interbedded very fine sandstones, mudstones and shales. 75

Minor•lithologies d)intraformational conglomerates. e)heavy mineral bands. The vertical and lateral arrangement of these lithological groip8 appears to be random. There is no evidence to suggest,the presence of rhythmic sequences. a) The coarse sandstones contain varying proportions of pebbles, but these are seldom sufficient, except perhaps on Griana Sgeir; to 'warrant the name of conglomerate. On Fladday theAunctions between the coarse pebbly sandstones end the finer non-pebbly ones are mappable and they are seen tO occur in sequences each about 5.m. thick. The bases, wherever, exposed, are marked by a thin intraformational conglomerate. . At the top they pass gradually'into medium-grained non-pebbly sandstones. On Raasay only the base of the lowest coarse pebbly sandstones can be mapped with any degree of confidence. Above this horizon this litholod, though still occurring in discrete units, cannot be mapped with any degree of accuracy. This is due to' insufficient exposures, faults of unknown displacement, and, possibly, lateral imper- sistance of these beds'. On Soalpay, likewise, only the base of the lowest horizon of coarse pebbly sandstone can be clearly,, mapped for similar reasons. Lithologically these beds include pink-coloured coarse and very coarse sandstones and grits. Planar PHOTO 16. Coarse pebbly sand: -och an Uachdair,G1a-ne Beds,looking N.

5.; 00

*4.44 119. PHOTO 17. Bedding plane of coarse pebbly sandstone. Upper Fladday Beds. An Roinne. 77

cross-stratification and festoon beddingare often present, but stratification of any kind is frequently absent. Pebbles - sometimed occur scattered through the whole rock (Photo.16), or may be concentrated on bedding planes (photO.17). The pebbles are variable in size. Generally their long axes are between 2-3 cms. but sometimes attain 5 ems. Theyare.well rounded but of poor sphericity. A variety of shapes occur (Fig. 17a). Some pebbles show a particularly interesting habit. These are elongated and shaped into three or four facets parallel to their long axes, (Fig. 17b). This shape is similar to that of recent ventifacts and to.supposed ventifacts from the basal Torridonian breccia by the shore of Loch Assynt (Fig. 17b). However, there are two significant differ- -ewes: the surfaces of the pebbles are pitted, not polished, and the edges of the facets are rounded, not sharp. Perhaps these curiously shaped pebbles are reworked • ventifacts, formed - by eolian,abrasion, but modified by aqueous transport to their present locality. As already mentioned the pebbles in Red facies rocks are largely of non-Lewisian origin. The relative proportions of different rock types present in a number of samples are presented as percentages in the following table. -This also compares them with collections made by the Geological Survey from two other localities (1907, Peach, Horne, Gunn, Clough, Rinxman & Teall, p.279,333.) 1 SPHEROID. DISC. QUARTZITE FELS1TE, SC AL PA Y. SCALPAY.

PRISM. VOLCANIC. PAASAY 4-- - - - ROLLER QUARTZITE. CID 0 GRIANA SGEIR,

ALL X I.

FIG.17a. Pebbles from the Applecross Group. Shape terms after Zingc(1935). GNEISS. 1

?VENTIFACTS FROM TORRIDONIAN BASAL BRECCIA ) LOCH ASSYNTSI DE , SUTHERLAND.

I -I-- I VOLCANIC. I I 1 VOLCANIC. RAASAY. I 1 i i 1 SC ALPAY

?REWORKED VENTIFACTS. APPLECROSS GROUP.

FIG.17b Curiously shaped pebbles . 80

• Griana Strath :N.Raasay Scalpay Gairloch Si4eir, Lingard Quartz ' 42 51 - 47 50 40 Volcanics- ' 41_ • 34 32 • 6 27 Quartzites .10 -13 14 20 9 Chert and 4 • 3 6 16 10 Jasper Gneiss 2 0 1 0 14 ' Grit ' 0 .0 8 0

•99 .101%, 100% 100% 100%

Thirteen specimens of pebbles were prepared for micro- . scopic examination but will not be described in detail as.little new can be added to the very good account given by Teall12each, et al. 1907, p.279-284), and no hew types were found., One or two observations are worth taking, however. - No pebbles were encountered of spher- ulitic volcanieff. With three exceptions all the pebbles are of sedimentary and acid volcanic rocks not now exposed in the Northwest Highlands. Gneiss pebbles-could well 'have been derived from the Lewisian complex, as could also the quartz pebbles, which,' with their inclusions and-undulose extinction, resemble those of Lewisian - 'pegmatite veins.: One rounded pebble of reworked Torrid- onian sandstone was found in, the lowest pebble bed on Fladday. I •

/ • ,

11111,11:;- ' 1) ' 14404' •

PHC_ ',L6.Lenticular medL.4.ri and c—e-grained sandstones with shley partings.Glame Beds.W.coast of Raasay,2km S.W. of Mantsh Point.Looking W.

PHOTO 19. Finegral. ..6..-YolagG of ,-cl,..L,.,3,mstones and very fine-grained sandstones.Glame beds.N.coast of Raasay,150km E. of Manish Point.Looking S.. 82

b) Making up tho bulk of the Red facies are medium to fine grained non-pebbly red sandstones. They are generally darker in oolour than thecoarse'sandstones due to the _presence.of a haematite clay. matrix. The medium-fine sandstones occur in lenticular beds 1,m. or so thick, sometimes massive or horizontally laminated, but more generally they show planar-pross-stratification. Adjacent beds' are separated by, thin shale bands 2-e3 ems. thick, (Photo.18), o) 'Very fine grained dark ,red sandstonee, mudstones and shales odour in the Red facies of Pladday and Raasay. - They have not been seen on Soalpay. These three rock- types generally occur interbedded with one another in sequences 1 m. or so think (Photo.19). Individual bands of very fine sandstone, mudstone or shale seldom exceed 20 ems. These sequences are generally lenticular when traced over a distance of several metres. Sometimes - this,lenticularity was caused by deposition on a gently undulating Surface,.in other instances it is due to differential erosion by the overlying coarser sandstones . These fine grained sediments are either massive, well laminated or occasionally current-bedded on a small scale; pink lentioles of current rippled very fine sandstone occurring in massive dark red mudstone or shale: - d) Intraformational.conglomerates make up a minor part

04.1.r„._ • ,,.e" - -.1Z_ ... - 1... --, -1±4ourabol, 4A10°- • '1...- - ...,.' a/ " - ••• ' '

. j,:" .„, -. Arti-,ir"-- ` - 1,-,-,'"-".. :-iiia. :. ..'N:•.•.,-4 - -- I-, ,,.4,0.t i ! -t--', -e.'"-- a -, '• N.. — V '-'.1.."..;

PRUTO Lenticular heavy miner,,.1. band interiiedded with sandstones. Unit V.1km up Alit na Criche on the right hand bankt looking N.4.. Scalpay.

PHOTO 21. Intrafor-national conglomerate at the ton of the Lower Fladdey Beds. 200m of An Roinne. 84

of the Red facies sequence on Fladday and Raasay. They have not been seen on Scalpay, presumably due to the absence of shales and-mudstones. They occur at the base ''of washout channels and above flat planes of penecontemporaneous erosional surfaces. They are seldom more than a few centimetres thick and are laterally impenistant. LithologicallY they consist of a matrix of massive poorly .sorted sandstones, containing granules.and small pebbles 'in which are embedded irregular lenticles and rounded pelletV4 red silty shale,,(Photo.21) These generally lie horizontal or subhorizontal to the nargino of the beds. e) Bands rich in detrital heavy minerals are a common minor rock type in the Red facies. They are composed locally of up to WI iron ore, and minor amounts Of other heavy minerald, associated with quartz and feldspar (see Chap. VI, Sec. ii). These bands are black, purple or grey-green in colour, and consist of laminae alternatively rich and poor in heavy minerals. Separate laminae are seldom more than 2mm.-thick, but the heavy mineral bands themselves may up to 1 M. thick, depending on the number of laminae present. •The bands are impersistent and_ can seldom'be traced laterally for more than a few metres (Photo .20 ) . The accompanying graphs (Fig. 18) show the vertical variation in the percentage of heavy minerals present in HEAVY MINERAL CONTENT °/o

0 100 0 100 _ \

5 5

MM. MM. , /

10 10

SLIDE NO. 62/24 8. / \ 15 / 20646.

FIG.18. Vertical variation in heavy mineral content in parts of two heavy mineral bands. 86

two thin sections otheavy mineral bands cut normal to the bedding. Horizontal traverses were run at vertical intervals of 1 mm. and the proportions of light and heavy minerals present in 50 grains were counted. The result- ing graphs show that.the heavy mineral bands consist of laminae with heavy- mineral contents much above the average for the red:facies- sandstones (3% see 0hap.VI,, sec. iii), alternating with laminae containing up to 80% of heavy minerals. Heavy mineral laminae are belieVed to be due to the Selective winnowing 'of sand by a current of suitable 'velocity. It must not be so vigorous as to erode all the sand, nor yet so slow as to deposit fresh sediment. It must be of such a speed as to erode only the less , dense quartz and feldspar grains, leaving behind the denser heavy minerals as a residual,lag deposit. The alternation of thin laminae of normal and enriched sands thus indicate a history Of fluctuating current conditions, periods of deposition alternating with periods of selective 'erosion.. HeaVy.Mineral segregations occur in both alluvial and beach deposits. Alluvial placers are generally lenticular, beach examples are typically more persistent laterally (Thomson 1937). This reflects the difference between locally and widespread uniformity of current I M f \Ft/ IN\ 1 I tI14

FIG 19 Trough bedding seen in plan. COARSE PEBBLY S-STN. UPPER FLADDAY BEDS 2M. INLAND OF HIGH TIDE LEVEL SOLKHWESTCORNER OF FLADDAY.

I M.

FIG. 20. Laterally heterogenous cross strat if icat ion. COARSE S-STN..GLAME BEDS.WESTCOAST OF RAASAY 1 / 2 KM. NORTH OF EILEAN AN INBHIRE. 88

conditions in these two environments.

(iii) Primary Sedimentary Structures The following primary sedimentary structures occur in the rocks of the Red facies: a)Stratification b) Cross stratification c)Washout channels d)Planar erosional surfaces

a)Horizontal stratification is not a very common feature of the Red facies. It is occasionally present in the Medium-very fine grade sandstones.

b) Cross-stratification occurs in three kinds in the Red facies. Each_variety is typically restricted to a specific grade of sediment: Trough bedding - coarse-very coarse sandstones planar cross-stratification - medium-coarse sandstones. Current ripple bedding - very fine sandstones, shales and mudstones. Cosetsytrough bedding are occalonally present in the coarse-very coarse sandstones. then seen in plan individual troughs vary in width from 0.5 to 2 m.; they can generally be traced,downcurrent for more than several metres before being truncated by other troughs (Fig. 19.').

-.it'. 4,, .• ' - -.0ette - ? f-- ,' - I------

11 -4:11.111%. PHOTC Longitudinal section of trough bedding in very coarse sandstone. Unit V.14.coast of Scalpay,Skro S.E. of Camus na Geadaig.Looking S. W.

. •1"`""- -4, • -

. ....ia . 75: • --".L.--- .."- . -773k . — • -----^--. ------(.-- 11...,..„i .....1:1-----4... ..- -.- -,.„.._...... 1.,--... J , • .-.=...... —.. .--, -;-'1 --..-/- 7--,-- ,-5--- -

- • !

- -

PHOTO 23. Section of trough beudLng looking Now. up the axes.Locality and ltthology the same as above. 90

In longitudinal section they are seen to be infilled with dross laminae which curve asymptotically towards the base (Photo.22). In transverse section these laminae run parallel or sub-parallel to the curved bases'of the -trough which, seldom exceed a'depth of 1 m. (Photo.23).

Planar cross-Stratification occurs in sets about 1 m. high. -the foresets are planes which dip, relative to the,base of the'bed, at up to 16° for the medium-grained sandstones, and•26° for the coarse ones (Photo.24). To the eye the grainsize of the sands composing these units, is usually consistent both vertically and horizontally, thoUgh'there are exceptions to this general rule. In the coarse sandstones some beds are vertically graded: the lower part of each lamina consisting of granules which grade up through very coarse grains to coarse ones (Photo.25). This variation is probably caused by the differential movement of sandgrains down the foresets under gravity, 'coarser heavier grains rolling further than smaller lighter ones. Graded cross-stratification like this has been described from the Mississagi quartzite of the Blind River area of Canada (McDowell 1957). Also seen in the coarser planar cross-stratified sandstones are variations in the grain size of adjacent foreset laminae. Thin lenses and stringers of very coarse sandstone and' granules occur parallel to foresets of PHOTO 2:- Planar cross bedding in very coarse sandstone. Unit V. . west of Scalpay 3.5km 8.E. of Camus na Geadaig.i,00king S...

PHOTO 25. Grad(:- cross bedding En very coarse sandstone and grit. Unit V.IL.coast of Scalpay,2km N.N.E. of Camus na Fisteogh.Looking W..N.B.This photogranh is upside down as this example occurs in overturned beds. 92

medium grained rand (fi. 20).-

Current ripple lamination is often seen in the very fine aandu. It occurs of a eimil%r acale and with sirlaar wiriations to the current rip )le laminations flready described from the Grey facies (ehap.IV, Scc. ii).

?he origin and ;eolcO.cal eiLnificance of croos-stratif- _ ictLtion hao attracted considerable interest from the tine of Eorby (1059) onwardu. The presumed mode of formation and i;eological sienificance of the crose-utratification aeon in the rooks of the L.ed facies rocks of the area will now be diecuseed.

The forration of current ripple lamination b the downcurrent migration of asymmetrical rippler was described by 3orby in 1059. ance then it has been studied from vedimento anoient and modern and by means o1 laboratory experiments by Kindl. (1917), Bucher (191A, chee (1957b) tau! J. Allen (19G2b) to mention only u few. Geoloaioally this' structure is of relatively little sicnificanoe, indicatin6 mainly deposition undQr quiet traction -current conditions. It is common in fine grained fossil Sediments and has been deocribed from a wide range Of recent sedimentary environments, e.g, the _deep se Lt Wenard 1952), tidal rulflats (Van :Ara ten 1959) and 93

rivers .(Shantser 1951).

.The deposition of large scale planar cross-stratification by the down current migration asymmetrical megaripples has been described from recent alluvial deposits by Shantser (1951). This seems a reasonable mode of origin for the planar cross-stratified Red facies sandstones. The mechanics.of this process and the geometry of the' resultant cosets and their parent sand, ripples has been deboribid by J. Allen 1962b and a, 1963 a and b (see also Hemingway and-Clarke1963,and Allen and Hemingway and.Clarke 1963). ,.Since the dip of the foresets is markedly unimodal (Chap. VIII) it does not seem likely that they could have been formed by deposition on the banks Of liiterally migrating channels as envisaged . - ‘v„ by M. Wright (1959). Planar cross-stratificationxfrom ancient sediments ascribed to various environments, 'e.g. the fluviatile deposits of the Old Red sandstone (Allen 1962) and the marine Lower Greensand (Schwarzacher 1953). Transverse megaripples have been found in recent sedimentary environments of varying water depth. They occur in rivers (Van Straaten 1953) and' in the sea (Stride 1963). Thus, apart from indicating deposition by fast floWing traction currents of known direction of movement, little can be deduced from the planar cross-stratification

.of the Redfaoies.

FIG. 21. Origin of trough bedding .

PRESUMED SHAPE OF SAND LAMINA DEPOSITED BY MIGRATING LINGUOID RIPPLES.

< CURRENT.

PRESUMED SHAPE OF SAND LAMINA DEPOSITED BY MIGRATING LUNATE RIPPLES. 95

Festoon or trough cross-stratification was first described by S. Knight (1929). He attributed it to the scowelizofl channels followed by their infilling by sand laminae parallel or subparallel to the channel walls. Isolated examples of troughs infilled in this manner have been produced experimentally by McKee (1957a). However, this mechanism does not appear to be adequate to explain the occurrence of Coseto of trough bedding of uniform size. J. Allen (1962a) has suggested that such cosets could be deposited by migrating asymmetrical megaripples, sand laminae being deposited in erosional . hollows at the foot of the advancing ripples. He envis- ages the latter to be either lunate or linguoid bodies. The former seems More likely since the erosional hollows, and hence Ahe sand laminae being deposited in them, are gently'convex upstream, as is seen in -trough bedding. The erosional hollows associated with linguoid ripples on the other hand tend to be more irregular in shape and are occasionally sharply pointed upstreaM. Sand laminae deposited in. them would not be trough shaped (see Fig. 21). Both linguoid and lunate megaripples have been deicribed from recent sediments of various environments. Linguoid megaripples occur in tidal channels of the Wadden Sea, and 'at the Arcachon basin (Van ptraaten 1953) and from the Niger River (Nedeoo 1959). Lunate dunes occur in the Bahaman banks (Illing 1954) and in the PHOTO 26. !ashout channel cut into fine-grained sandstone o & shalest& infilled with coarse sandstone.Lower 21adday Beds.Dubh Camus,100m N.-W. of Stac a'Ruadh.Looking N... 97

Mississippi where it is perhaps significant that they occur associated' withtrough bedded sands (Frazier and Osanik 1961). Trough bedding has been described from both fossil marine formations, e.g. the Pennsylvanian Caspar Formation - S. Knight 1929, and from supposed fluviatile ones, e.g. the Epidotic Grits of the Diabaig Group on Skye (Sutton and Watson 1960). Thus, like planar cross-stratification, trough 'bedding tells one little more than, that the sande were deposited by traction currents of'a known direction of flow. It may, however, in the Torridonian, indicate-more violent currents, since it occurs in a coarser grade of sand. c)• Clearly incised- washout channels'are rare. Only, seven have been recorded. They_are between 2-3 m. wide and seldom more than 1 m. deep. They are generally cut into the finugrained, and hence more easily erodable, sediment, and are commonly infilled with massive medium to coarse grained sandstone. Immediately above the channel floor there is generally a thin development of intraformational conglomerate. Channel walls may be ,gently curved or with flat floors and steep walls. Both symmetrical'and asymmetrical cross-sections occur (Photo. 26). d) In some places, notably on the foreshore between 98

An lioinno and UaMh =tea Daoinc on Pladday,'flii.t pinnoa of crozion occur coverinz hundredo of anuarc nctmo. They truncate cross:-stratifies,laminae and ,:,uichcand otructures, the eroded :surface bein overlain by 2-3 er=. of intra:'orrational cort;lomer%tioll. :.uch coLtt.lipona:eoue crosion'A surfaCe:; nay ropre:3cnt floora of clnl.cio such widt.r than thdsc just deacribod, tiwir ran,inz havin6 never been aeon in the field; altematively.they. could have becn produced by the laterni migration of ouch :mall channela (see Van !;traaton 1961). 99

CHAPTER, VI. PETROGRAPHY

(i) Introduction A detailed quantitative study of the composition and textures of the rooks in the area was not carried out for a number of reasons. The absence of marker horizons and the presence of faulting render the accurate stratigraphio location of samples impossible. The well cemented nature of the rooks Acme not allow them to be broken down without chattering the constituent grains. Thus sieve granulometrio analyses and quantitative heavy mineral separations may not be made. To overcome these problems and to do the topic justice would involve a planned statistical study outside the scope of this work. Thin seotion 'examination consisted mainly of a qualitative Study of the mineralogy and textures,of the 'rocks. A limited number of modes were measured with a point counter, to establish the proportions of _the various minerals present, and to see if any gross .vertical ohangei in their proportions could be recognized. Several grainsize analyses were made by the measurement of the long axes of 'grains in thin section. FIG. 22. Quartz.

62/143 61/91 62 243 6 1 /9 8 61/106 TYPICAL GRAIN SHAPES. I MM.

0-5M

61/105. 62/150. ACICULAR INCLUSIONS. 0.5 MM. RED•BROWN RUTILE I NC LUSION S.

62/150, Quarl-3.

9.5 bubble.

LIQUID INCLUSIONS,

I MM, 62/251. REWORKED SEDIMENTARY GRAIN WITH OVERGROWTH.

0.5 MM. FIGURES REFER TO SLIDE NOS. 101

(ii) Qualitative mineralogy a) Detrital minerals. Quartz'and feldspar are major detrital minerals of' all the teaks examined; with the exception of the heavy mineral bands. -Micas, epidote, lithic fragments and iron ore occur as accessory minerals, and zircon, .tourmaline, garnet and sPhene as minor accessories. The mode of occurrence of these minerals will now be described in detail. Quartz (Fig. 22) occurs in all grades from pebbles 5 ems. or so long dawn to the finest silt. The original edges of the grains ate often not present due to pressure solution but, in general,- they vary. from subangular to well rounded for—the coarse sands down to angular for the silt. (Terminology according to Waddell, 1935). Almost all the,quartz,grains examined showed strain extinction. All stages may be seen from the gentlest undulose extinction to the development of polycrystalline grains. Unstrained quartz.is very rare and possibly derived from an ignedus source. Inclusions are common and fall into three main categories. 'Many grains contain lineally arranged inclusions which, under the highest power available, can be seen to be negative quartz crystals. These are often.liquid infilled and. with a small gas bubble.

Strained quartz crystals with similar*.inclusions are 102 common in the p9gmatites of the Lewisian gneiss which provide a possible source for these (H. Toms personal communication). Some grains contain acicular inclusions generally taken to be rutile. These hairs are commonly oriented in sets which may be up to three in number in one crystal. Such inclusions are characteristic of the blue quartz of the Ccourian pyroxene gneisses (Teall 1907, p.54). A few quartz grains were seen with larger -inclusions of prismatic foxy-red rutiles. -Some compound quartz grains must have been derived from pre-existing_ sediments for they consist of an original strained detrital grain with an optically continuous overgroWth, Both grain and overgrowth form a single rounded sand particle indicating that the overgrowth was not developed by.post-Torridonian diagenesis. The following table summarizes the kinds of quartz grains present in order of relative abundance, together with their possible origin: Abundance Strained Die inguishing Possible Source ' features V. CoMmon Yes No inclusion , Various Bubbles Pegmatitea Lewisian Fairly Granulite gneiss ft common Rutae needles facies Rare No No inclusions Igneous Strained Pre- y. Rare Yee Reworked sediments Torridonian overgrowths FIG.23. Feldspar.

61 /1 03 ORT HOCLASE , MICROCLINE •

61 / 1 03

PLAGIOCLASE .

MICROCLINE PERTHI TE.

PERTHI TE .

UNTWINNED FELDSPAR

1 MM. 104

• ' - Feldspar (Fig.' 23) is a major constituent of almost all :the rocks examined, its content varyingbetween,40-15% of'theiotal number Of grains in the samples'examined. -,Grains of feldspar vary from silt grade to fragments ems. long. A wide variety -of grain-shape and roundness is to be seen. Most grains are angular to subangular, but many, in the coarse-medium grade, are very Well rounded. Many kinds ,of feldipar are present in varying proportions and stages of deOomposition. Orthoclase with 'carlsbad twinning is very rare. Microcline with characteristic cross hatching is.the- commonest variety present, it is generally fresh*and unaltered though often well rounded. Perthites are common; microcline microperthite is abundant, perthitic.intergrowths in untwinned hosts occur, antiperthites are rare. Differential alternation of all these varieties can be seen. Plagioclase occurs showing multiple albite twinning. Extinction angles vary between, 12-20° denoting albite and oligoclase. Most plagioclases are highly altered. Untwinned feldspar, of uncertain composition, is also present. Several kinds of alteration of feldspars are seen. They may have been present in the source rock, or brought about by weathering or diagenesis. Infiltration of red haematite along cleavage traces is .probably the commonest. This imparts a pink colour to the grains When seen with

FIG. 24. Micas.

62/243 62/139 BIOTI TE. EXPLODED B1 OTI T E.

6 1 /9 8 MUSCOVI TE.

_ _ CHLINOCHLORE AFTER MUSCOVITE. 62/243

PENNINE AFTER AN AMPH I BOLE. 62/243

I MM 1 106

the' naLed eye. Such feldspars are common in the 14r4wician gneiss lad this nay well boa predepositional feature. uricitization is' the characti.xistic gore; of alteration of rlagioolnoc. 1;ericite oftezi forlu the cement of the sandstones and in 60M intances- the plagioclascS be seen with racxed edcee ::.djacent to the 8:..ricite matrix, hen cc: it suet somtimen be diagenctic. Eow,:ver, scrie serioitisation was pousiblybrought vlout by r-eathcriuf; Trior to deposition, and some perhaps by retrogrerisivc IT!otamerphiem of the source rock. hnolinization nnL1 epidote inclusions in the untwinned feldspars nay also bo ascribed to the latter eause. In zone rocks near the baucEent,

cpidote occurs, as an authieenic interstitial cemout and some of the feldspars show considerable vpidotination which elay be dinGenctic. iceplacement of feldspar by chlorite and ouloite irj a characteristic diagenetic 'phcnorienon in the Grey facie:: sandstones. Aiscovite, biotite and chlorite (rig. 24) all occur as detrital minerals. :q)scovite is probably the commonest nice, occurrinE, au a minor constituent throui:h- out the.:holy: suecession. In some very Sinc-cyained sandstones of the Red facies it. is sufficiently abun:Iant to give' thera flagty partint;. fisAlity of the Grey

facies shal:s 14 due -to the ScUT cause. '`u. ;cc occurs

ire its characteriatie form of co1ourles5;, platey, ra6,,cd' edged 6raino well developed basal eleavae. it is 107

fresh in the, Red facies rooks, but often shows partial alteration to chlorite in those of the Grey facies Biotite is rarer than muscovite and is generally restricted to the rocks of the Basal facies. It occurs as pleochroic light red-brown platy, ragged edged .grains with well developed basal cleavage. A dark brown colour is sometimes imparted by infiltration of haematite along cleavage planes. Exploded biotites occur in the Torran beds where epidote has crystallized along the cleavage forcing adjacent plates apart. Two types of chlorite occur as detrital minerals in the Grey facies sandstones.. The first is platey with a good cleaveige, strong pleochroism from pale yellow to brownish green, and a moderate birefringence. This is probably chlinochloreafter muscovite. The second variety occurs in irregularly chaped grains with poor cleavage. It has a dark blue-green colour and exhibits 'berlin blue' polarization colours. This is possibly the variety pennine developed as pseudomorphs after mafic minerals (amphiboles and pyroxenes) either as a weathering product or during diagenesis. Detrital iron ore minerals are common as minor accessories throughout the succession, and as major con- stituents in the heavy mineral bands. They occur as angular to well rounded grains generally of medium-fine FIG 25.1ron ores.

61/105. 62/248 , 61/105, SPECULARIT E ALTERING EART HY SPECULAR! TE. TO EARTHY HAEMATITE. HAEMATITE,

0 62 /248, 20646, LEUCOX E NE. SKELETAL ILMENI TE,

Epidote.

62/24 8. 62/2 55 .

TYPICAL ROUNDED QUARTZ; EPI DOTE G RA I N. I NT ERGROWTH .

I MM. i 62/243, TWINNED VARIETY. 109

-sand grade. Foui varieties may be distinguished (Fig.25) i) Black, metallic-lustre in reflected light ii)Purple, tI

iii)Milky white, dull 11 It 11 iv)Red-brown, • 11 11 N It

The first variety has been demonstrated by Irving (1957), using X-ray powder photographs to be specularite (k-haematite), but, as Van Hbuten (1961) has shown, it may well have been produced by the alteration of magnetite. It often has alteration rims of red earthy haematite rhich frequently grade into the matrix of the rock. Variety (ii) is Ilmenite and (iii) Leucoxeno. Transitional stages of the former altering to the latter can be seen. The fourth variety is :1-haeMAtite, but whether it was originally deposited as such, or has subsequently formed from IN-haematite or magnetite is uncertain. 4idote (Fig. 25) is a common accessory mineral occurring throughout the succession. Grains arc generally well-rounded but are angular near the 'base of the sequence. The mineral may be colourless or pleochroic from colourless to yellow-green; some grains are particoloured. Twinning is sometimes present. Poly crystalline grains of quartz and epidoto occur. Sphene (Fig. 26) is a common minor accessory mineral, especially in the heavy mineral bands and near the base FIG 26. Sphene.

62 /243. 62/243. ANGULAR(EUHEDRAL)& ROUNDED 62/243. GRAINS. TWINNED VARIETY.

1 MM. Tourmaline. i

62/243.• 62/243.

BASAL SECTION. LONGITUDINAL 62/14 5.

SECT ION . QUARTZ :TOURMALINE I NTERGROWTH . Garnet.

0 o

62/243 62/243 243 ANGULAR (EUHEDRAL)8, ROUNDED VARIETY WITH GRAINS. QUARTZ INCLUSIONS.

Zircon.

62/9 0.5 MM. 111

of the succession. In the latter position grains are generally-angular and often approach the characteristic diamond shape of the mineral. Elsewhere the grains are generally well-rounded. The mineral varies in colour from light brown, via neutral, to colourless. Parti- coloured varieties and polysynthetically twinned grains occasionally occur. Polycrystalline grains of sphene and metallic iron are seen rarely. Tourmaline (Fig. 26) cveurs as rare well-rounded grains throughout the succession. It is strongly pleo- ohroic, varying,in colour through olive green, blue-green and light brown. In some grains body colour is so strong that it obscures the polarization colours. One grain was seen of a quartz-tourmaline intergrowth in optical continuity. Garnet (Fig. 26) occurs as a minor accessory throughout the succession in the form of rounded grains. It is, colourless or pale pink, probably the variety almandine. Zircon (Fig. 26) occurs as a minor accessory throughout the succession as small grains, often still with well developed crystal faces. b) Cementation textures and authizenic minerals. All the sediments examined are very strongly cemented, preventing the chemical or physical breakdown

FIG.27. Grey facies diagenetic textures.

CALCITE REPLACING DETRITAL QUARTZ & FELDSPAR.

61/95.

DETRITAL FELDSPARS BREAKING DOWN TO A MI CROCRYSTALLI NE QUARTZ, CHLORITE, SERICITE MATRIX.

62/24 3.

QUARTZ FILLED PORE SPACES LINED BY MICROCRYSTALLI NE CHLORITE.

62/2 3 9.

I MM. I t 1 113

of specimens in the laboratory. They are not so strongly cemented though that when broken the fracture outs cleanly across, both detrital grains and cement, i.e. these rocks are not 'quartzites' in the cementation sense of the word. There are two fundamentally different modes of cementation, and, as mentioned in Chapter II these are one of the criteria used to distinguish Red from Grey facies sediments. These two varieties of cement will now be described in turn.

The Grey facies' sandstones, notably those of the Brochel beds, are often poorly sorted (see Section iii). The cement is a finely divided mixture of chlorite, sericite, Waartz and carbonate. In the c/rey facies shales and mudstones a white coloured clay cement is present. The detrital grains of feldspar, especially plagioclase, in these rocks can be seen in various stages of breakdown and alteration to quartz, sericite, chlorite and carbonate. The margins of. such grains often merge into the matrix to which they may, thus have contributed (Fig.27). This process has been suggested by Cummins (19G2) to account for the presence of greywackes, which he believes may have been originally deposited as cleanwashed chemically unstable sands. A thin section of the grey sandstones at the top of the Caolas,Fladday beds shows a 114

somewhat different texture -(Fig. 27). It is a fairly -well- sorted feldspathic sandstone. *.The pore spaces are lined. with a.thin film 'of green chlorite, the remaining spaces'contaiwirg sedondary quartz. This suggests that ,there may,have been three phases in the diagenesis of

the'Grey- facies sandstones: i) pores lined with chlorite films ii)residual cavities filled with quartz iii)break doWn%of pbre cement and unstable detrital grains, mainly feldspar, to _give a finely 'divided quartz; chlorite, sericite, carbonate matrix,. . This zequence may have been halted before the third phase in the top Caolas Fladday sandstone. A change in diagenetic environment may have been caused by the Red facies sandstones which succeed it in less than 1 m.. P. Simony ,very kindly lent the author thin sections of the Pre- Cambrian Lorraine quartzite of the Cobalt series of' Ctinada,', Examination of these support the above hypothesis. Here'a feldspathic sandstone has broken down into a rock with a large amount ,of Chlorite:Sericite matrix generated • by .decaying feldspar. :_/solnted patches occur where quartz-filled pore spaces are lined with chlorite. These may represent enclaveo of unaltered cement which were not made over into a microcrystailine matrix. FIG.28. Red facies diagenetic textures.

DETRITAL QUARTZ GRAINS WITH SECONDARY ,P 1 OVERGROWTHS.

61/106. ... FINEGRAINED SANDSTONE WITH A HAEMATI TIC CLAY MATRIX.

62 /1 43. I MM.

MICROSTYLOLI TIC BOUNDARY BETWEE N QUARTZ & FELDSPAR.

haemolqic clay film.

seyicil'ic film.

M1CROSTY LOL I TIC BOUNDARY BETWEE N TWO QUARTZ GRAINS_

61/102, 116

Diagenesis of the Red facies sediments is closely related to grain size. The coarser dandstones are cemented by a oombinatiOn of pressure solution and secondary quartz, the, finer ones by sericite and a red -haematitic' clay (Pig..28). The pressure welded boundaries of the coarser sandstones are generally straight or gently curved. Sometimes etylolitie sutures are present. The junctions -between grains may be clean or' separated by a-thin haematite clay and sericite film. Bbth quartz:quartz .and quartsifeldePar stylolitic contacts occur. In some *instances of the- latter the quartz advancee.preferentially along:alternating twin planes (Pig. 28). A characteristic feature of these coarser sandstones is that the feldspar grains appear to retain their original, generally well roUnded, shape, whereas the quartz's are welded round them (Fig. 29). Chert, epidote and iron ore minerals often' behave like feldspar in this respect. The greater solubility of quartz may be the factor controlling this texture, though the orientation of adjacent grains may also be important; it'is well known that the solubility of various minerals varies in different directions, controlled by their crystollographic orientation. :Pore spaces,of the coarser sandstones are infilled by secondary quartz. The great amount of pressure solution has prevented the development of crystal faces. FIG.29. Red facies diagenetic textures.

DETRITAL QUARTZ GRAINS PRESSURE WELDED AROUND A RESISTANT FELDSPAR ,

DETRITAL PLAGIOCLASE WITH OPTICALLY CONTINUOUS TWINNED OVERGROWTH. 61/103,

DETRITAL TOURMALINE WITH UNITERMINAL OVERGROWTH.

I MM. 118

A thin pellicle of haematite is often present between a-detrital quartz and its overgrowth. This may have formed at an early stage of diagenesis, or it may have developed as a coating on the grain prior to deposition. Such a coating is often present in recent eolian sands. Other authigenic minerals encountered are: small amounts of feldspar, with twinned optically continuous overgrowths (Fig. 29); epidote, in small quantities in the Torran beds, and tourmaline, one grain with a uniterminal overgrowth (Fig. 29). The medium to fine grained sandstones are cemented by a microcrystalline mixture of quartz, aericite and red haematitic clay. The very fine sandstones, mudstones and shales are cemented by red haematitic clay alone. Much of this must have been deposited as a matrix for occasionally it is SO abundant that the grains float in it without touching 'one another. iii) Quantitative mineralozy To assess the mineralogical differences between the sandstones of the Red'and Grey facies volumetric modal analyses were carried out by point counting twelve thin sections, six from red sandstones and six from the grey sandstones. Two hundred and fifty points were counted for each slide.: The results are plotted on the triangular,diagram drawn up to classify sandstones by matrix & micas.

20 QUA RTZOSE SANDSTONE

25 10 feldspar. quartz

FELDSPATHIC SANDSTONE

FIG.30. Diagram of modal analyses.

-I- RED FACI ES SANDSTONES.

• GREY 11 120

. Pettijohn in 1944 (Krumbein and'Slose 1959, Fig.5-8) (Fig. 30). 'This shows that,:with one exception, all the grey sandstones analysed fall in the greywacke field. All the red sandstones, again with one exception, fall in the arkose field. The. Grey facies sandstones are distinguished, from those of the Red by a greater 'perpentage'of feldspar and matrix. The significance of

• • these variations will become more apparent- later on in this, chapter. iumerioal- grain traverses of 800 grains- per slide

- were made-for a number of thin sections of rocks from the main litho-stratigraphic divisions. The results ark 'seen- plotted .as histograms in Pig. 31. Relative percontagesof the'following five classes, were measured: luartZi feldspar, micas, heavy minerals and lithic frag- . .menti. Per'each, alai's in each litho-stratigraphic division 'the mean values' obtained 'are. shown. The following Changes are seen When comparing the mineral content of ,.. the grey-facies rocks at the base of the Diabaig with the Applecross group sediments on Raasay: Quartz +9%, Feldspar, -8%, Lithio.+3%, Heavy minerals -2%, klicas -2%. These changes in quartz and feldspar were also revealed in the.volumeiric.analyses (Pig. 30) of red and grey sandetones. The relative preportions.of eadh.kind of feldspar present in one -slide of every dtratigraphic unit were

08T FIG. 31.

VERTICAL VARIATION IN MINERAL CONTENT.

GREY FACIES. RED FACIES. RED FACIES.

•DIABAIG GP. DIABAIG GP APPLECROSS GP.

Olo

40-

20-

r 1 I

— ._.J VERTICAL VARIATION OF FELDSPAR TYPES,

0 10 0/0

GLAME BEDS. 1'

LOCH AN UACHDAIR

BROCHEL ) 1

TORRAN 11 .

PLAGIOCLASE & UNTWINNED FELDSPARS. MICROCLINE, PERTHITE & ORTHOCLASE. 122

also measured. The results are presented as histograms in Fig. 31b. These allow a gradual decrease in plagioclase and untwinned feldspar, and a complimentary increase in Microcline and Perthite, moving up the succession.

(iv)Dismission of- mineralogy It would be rash to attach too much importance to the stratigraphic variation of mineral content shown by these results. However, they provide grounds for some interesting speculation'and suggest possible lines for later, more detailed research. It has long been held that the Torridonian sediments, except at the base of the succession, were derived from a non-Lewisian source (Hinxman 1907, p.273-274, Phemister 1960, p.44-45). This conclusion has been based on two observations. Firstly the abundance of pebbles of igneous and sedimentary rooks; both of which are absent in the Lewisian, as it is now exposed. Secondly the ,r f;,, relative scarcity in the Lewisian (Peach, et al., p.270. These two observations are valid, but the data presented in this chapter throws some doubt upon the non-Lewisian' source for the Torridonian sediments of the area discussed. :Up to about 50,6 of the pebbles in the Applecross sand- .6-times are,00mpoaed of vein quartz and gneiss pebbles which may well have been derived from the Lewisian. With

the exception of.the very rare unstrained quartz grains 123

"and a few lithic fragments, all the detrital minerals found in the area have also been reported from the Lewisian. If the Torridonian sediments were derived from a dominantly sedimentary source one might reasonably expect them to be much more mature. Second cycle arkoses appear to be remarkably rare in the world. If the source was of dominantly igneous rocks the large amount of microcline preseht would indicate that these must have been acidplutonio rocks. Such an explanation would not be consietant with the almost complete absence of unstrained igneous quartz. The only remaining possibility is that most of the Torridonian sediment came from an area of essentially high grade quartzo-feldspathic metamorphic rocks. Surely a not unreasonable source is the Lewisian gneiss, or, if one preferred, of an 'archean basement of Lewisian type'. Only minor amounts of lithic sands and pebbles were derived from volcanic and mildy meta- morphosed sediments which perhaps formed the cover of the Lewisian rocks, as suggested by H. Read (in Sutton and Watson 1951). Teall (1907 p.207) compared the volcanic pebbles with those of the Uriconian of Shropshire. J. Gregory (1915) believed the granulite pebbled to have been derived from the Moines. Both these ideas seem suspect since both these sources are as lie in the opposite direction of the supply of'sediment transport for a 24

the area shown by palaeooUrrents (see Uhap.VIII). If this conclusion is accepted then the minor variations in mineral content recorded in the sequence can be explained' in' terms of 'increasing maturity. The later sediments haiing been.derived from, a progressively' more Aistant'source as one would expect from the northerly Overlap -of the TOrridonian strata. Thus the vertical increase in quarts-and the relative increase in the propOrtion of,chemically more stable potash feldspar may be due to their greater resistance to weathering over. other. minerals. The.maturity of the Torridonian sediments has also been discUssed by W. Kennedy (1951) in relation to,the Moine and Sparagmite formations. The minor inorease4n the proportions of lithic fragments and decrease'in'heavy Minerals and mina probably reflect the fact that the Applecrois Aediments are coarser than those • of the Miabeig Group (see SectiOn v). It is possible therefore that,'except for the volcanic and sedimentary pebbles and lithio fragments; and for the rare unstrained 'quartz grains, all the Torridonian 'sediments-in- -the area were derived from an Arahean basement complex of Lewisian type. In Skye and ,further south, where palaeocurrent studies,indicate a westerly or southerly source of supply, the story may perhaps be different (see Ohap.IX, Section'vi). CORRECTED c6 VALUES OF LONG AXES.

0/0 50

0 MEASURED Ch VALUES OF LONG AXES.

RED FACIES SANDSTONES

...• • ..*** ,.." ...... • 1. 0...... • s —.. • r...• • ... • 2 .— 1 --- • f

FIG.32. cumulati ve gray n-size distribution curves. 126

(v) Granulometric analyses Mechanical analyses were carried out on thin sections of seven specimens. Four were of Red facies, and three were of the sandy parts of graded Grey facies sandstones. The longest axes of 250 grains•per slide were measured on an equally speced grid system, using an optioal micrometer and a.traversing stage. The results are plotted as cumulative curves in Fig. 32. The lower abscissae show the lengths actually measured expressed as si values (Krumbein 1934). The upper abscissae show these results corrected to simulate sieve analyses according to Priedman!s graph (Friedman 1958, p.408, Fig.4). These ourves show that the Grey facies sandstones are poorly sorted and bimodal, consisting essentially of fine sand in a silt and clay matrix. In contrast the Red facies samples are better sorted and have less matrix. These differenbes are probably related to the different modes of deposition of the two contrasting rock types. The massive graded Grey facies sandstones possibly having settled out of muddy suspensions, in quiet water, while the cross-etratified sandstones of the Red facies were deposited by strong traction currents. The average grain sizes of the slides subjected to numerical analysis were estimated by the following methqd. , The diameter- of the, field of view of the micro-

FIG.33.

HISTOGRAMS OF THE RELATIONSHIP BETWEEN MINERAL CONTENT & GRAINS! ZE . %

40 - 20- QUARTZ . o 20-

FELDSPAR . 0

20-

LITHIC GRAINS . 0 20n

HEAVY MINERALS. 0

20-

MICAS. 0 1 0 •I .2 -3 .4 -5 •6 MEAN GRAIN SI ZE.(MM) GRAIN SI ZE:MATRIX GRAPH FOR SIX RED FACIE S SPECIMENS. .4

.3 ....

MEAN GRAIN •2 . < -.. -... — -- SIZE. (MM) --— • — •I - , _ ......

0 r I 0 5 10 15' 20, % MATRIX. 128

scope was measured. Then, for each slide, the number of grains cut by the cross-wires for four different fields of view were counted and averaged out. The approximate grainsize could then be found by dividing this figure into, the diameter of the field of view. The average proportions of different classes of minerals are compared with the grain size of the host samples in the accompanying histograms (Pig. 33). The connection between grainsize and mineral content are seen to be consistant with that generally expected (Pettijohn 1957, p.110; Krumbein and Sloss 1959, POW . As mentioned in the previous section this-factor is probably responsible, to a certain extent, for the observed vertical variations of mineral content between the Diabaig and Applecross sediments. Within the red sandstones there is a tendency for the haematitic clay to increase in volume with a decrease in grainsize. This is seen in Fig. 33 which shows the percentage of clay present in the six Red facies samples .used in the preparation of Fig. 30, plotted against estimated grain-size. This suggests that the haematitic material is a true matrix deposited with the sediment, and not a cement infiltrated at a later date. 129

CHAPTER VIII. SECONDARY ' SEDIMENTARY STRUCTURES

(i) Introduction Penetiontemporaneous deformation of bedding has often been recorded from the Torridonian in the past. Until recently the morphology of these structures and their origins and geological significance have attracted • little attention. Harker (1903,p.190-191, Fig.1) described disturbed bedding in the lOrridonian rocks of Rhum, and attributed it to teetonic causes. In the 1907 Memoir similar disturbances were briefly mentioned from a large number of localities (1907r p.331,332,349,361) and one example was figured (1907, Fig.13). In 1913 H. Woodward noted bedding 'most violently contorted' in the Torridonian rocks of Raasay (1913, p.169). E. Bailey (1944) re- interpreted the Rhum structures as penecontemporaneous slump structures (p.171-172). Within the last three years a series of papers has been published describing these structures in considerable detail 'and discussing their origins (Sutton and Watson 1960; Selley and Shearman 1962; Selley, Shearman, Sutton and Watson 1963, Stewart 1963; Salley - in press). One of the main objects of this research project has been to study the penecontemporaneous disturbance t I 4611101 40,1014i: see . •ZN I s - 4ie I -Nei- • s44 ri • •t

• 1-• 'OPL;0' PHOTO 27 .View the pseudonodule horizon. The hammer is resting on top of one pseudonodule.E. coast of Raasay, 200m S.. of Brochel Castle. Looking

PHOTO 28. 3edimentary dykes of pale pink sandstone injected into red shale.Top of the Lower Fladday Beds N. side of Dubh Camus,looking W.. 131

of bedding seen in,the Torridonian sediments of the area studied, and to account for its origin and 'geological significance. To do this it has been necessary to study a wide range of peneoontemporaneously disturbed sediMents, ancient and modern, both in the field and in the literature. Experiments have been devised and carried out, under controlled conditions in the laboratory, to produce structures, similar to those seen in the, Torridonian rocks of the area. There are four main kinds of deformational structures present in the area. They will now be described in turn.

(ii) Deformatilpn structure of interbedded sand and clay a) Pseudonodules. Where sandstones and shales are interbedded, in both Red and Grey facies rocks, lumps of sandstones are often seen'embedded in shales or mudstones, generally devoid of lamination. These lumps vary in length from 5 ems. up to.0.5 m, and in height from 2-25 ems. (Photo.27). The lower part and sides of each body are curved,. the tops are typically flat and parallel to the adjacent bedding planes. The sandstone is generally fine-grained and well laminated. The peripherel laminae are curved parallel to the base and sides, but are truncated or overturned at the top. The cores are often highly con- ""4-•<-___,--:>.%:---'- """'"

----'4...... ------)

i M.

A )GE NE PAL VIEW BETWEEN PTS.A&B -...,,...._ FIG.35The pseudonodule bed, Brochel, Raasay_,

B) DETAIL OF PSEUDONODULE A PINOD. MARKED 1- * BED. 50M. S

IM. 1 I 133

torted (Pig. 35). The shape' of these sandstone masses almost always appears, to be oymmbtrical when viewed in vertical section. There are no Good exposares of them in plan in'the area, but, in the cove 100 m. south-west of Brochel.Oastle,where a shale bed containo' many sandstone bodies, one example can be seen to be cut by two perpendicular joint faces. On both of theee it is symmetrical in section. There is no evidence to wiggeot that, in the area utudiod, theoe sandstone bodies are. linear. As mentioned the eandstone Flauseo are Gregarious and generally occur peraistRntly in u bed of shale which may be traced laterally for -Iany metrou without dying out. Gaps between adjacent mas000 are of variable amount, but in a single shale there is generally only. one sandstone lump in any vertical section (Fig. 35). There is one exception to this in a shale band in the cove east of Manieh Point in.Borthern kaaaay. here many . masses of very fins 'sandstone lie in a' red shale. They often lie one above the other, showing oonsiderable signs of Compaction. This has frequently led to the development of asymmetry but it is not constant in any direction (Fig. 36). ,Sedimentary structures uimilar to those mentioned aboie have been described from the Devonian rocks of the Ardennes, and also in recent deposits by ?Iacar (1948a - und b, 1951) and :,,aaar and Antun (1950. 30 CM.

RED MUDST ONE

VERY FINE RED S-STN. S.E. N.W. FIG.36.Pseudonodules. Glame beds, s.side of 2. bay E. of Manish Point . Raa say.

SEMI DETACHED PSE UDO NODULES

COMPLETE PS EUDONODULES.

FIG.36 Pseudonodules.Caolas Fladday beds b 2 0 Om. S of Stac a'Ruadh.Fladday, 135

These authors call such sandstone lumps pseudonodules, and attribute them to a combination of loading and slid- ing of- unconsolidatingsand 'into a hydroplastic clay, .very shortly after deposition (Macar and Antun 1949, B.140-142). Professor Macar has very kindly shown these.structures to the author in the field. In the Torridonian,pseudonodules described above, the symmetry and lack of' a preferred linear orientation suggest that 'slumping movement with a horizontal component) has not occurred; vertical movement alone, due to the sinking of unconsolidated sand into a water- saturated clay, is all that-is necessary to explain these phenomena. As the sand lumps sink down into the'clay the sides of the nodules are reflected upwards and are sometimes overturned. Both the complete break-up of a single bed, and the partial development of nodules leaving the parent bed intact (though thinner) haie. been observed (Fig. 36). Kuenen in 1958 described a series of experiments in which he reproduced pseudonodules by vibrating a tank of clay capped by laminated sands. The vibration rendered the clay thixotropic and enabled the sand to break up along its length and sink down into it. 'The sand laminae wore deflected upwards. The resultant structures morpho- logidally resembled - pseudonodules-. 136

A similar series of experiments were run by the present author to check this mechanism. China clay was allowed to settle out of suspension into class beakers and tanke. Fine sand was then gently sedimented through water, by means of a teastrainer, onto the clay. Pseudo- nodules could then be generated by tapping the sides of the container. Unless great care was exercised in deposit- ing the, sand evenly over the surface, loading took place without vibration, and it was often not possible for the operator to prevent, the pseudonodules forming before he intended' that they should. One may conclude, therefore, that if the clay is still sufficiently unconsolidated and fluid, then load structures can be formed without vibration. Thus it seems likely that the Torridonian pseudonodules in the area studied were produced by vertical loading of sand into unconsolidated clay; slumping did not take place. b) Sedimentary dykes. Only one set of sedimentary dykes was observed over the whole area studied. This was developed in the upper of two shale horizons at the top of the Lower Fladday beds at Dubh Camua (Photo.27). This red micaceous shale horizon is about 30 ems. high and is succeeded by the coarse pebbly sandstone unit which marks the base 137

of the Upper Pladday sandstones. The junction is marked by 'a thin intraformational conglomerate. The - shale is cut by a number of dykes of pale pink fine- grained sandstone which'dip at about 50° to the north- east. These dykes are vertically, ,lenticular,,having, in some instancest.been pinched off from the underlying sandstone from_whichthey must once have been intruded. Likewise the, tops of the dykes do not always reach the top of the shale. ,One example which does,flaresout at the top,,thesandstone fill containing red shale fragments, merges imperoeptably with the overlying intraformational conglomerate. These sedimentary dykes were probably produced shortly after deposition by the upward injection of quicksand- from the - underlying bed into fissures in the

, shale. 'The overturning of thedikes and their pinched- off bottoms suggest that the shale underwent a certain amountrof plastic movement. The fission of the mud and its infilling by quicksand may have been brought about by stresses set up-by' the violent currents which deposited the succeeding coarse pebbly sandstone.

(iii) Deformational structures of sandstones. a) Interstratal deformation. . One of the most common kinds of deformation in the Red facies of the Torridonian involves contortions PHOTO 29. Gentle undulose bedding,an early stage in the formation of interstratal cusps.Glalle Beds,mouth of the Glame river,'. coast of Raasay,looking N..

,14W A .‘• • 4

PHOTO 30. Well iveloLl Untt IV. W. coast of Scalpaytlkm 3. of Rudh at ehonnaidh. Looking E.. 139

of whole beds of sandstone, These are ubiquitous in the area studied and have also been seen by the, author ',at Stoer, Assynt, Ullapool, Torridon, Applecross and Skye; These disturbances-aregenerally seen in the coarse-medium-grained ,sandstones, and are' rare in the very coarse and. very fine sandstones. They are most commonly seen when,the Sandstone beds are 1 m. or so- thick, lenticular,' and separated from adjacent beds by shale partings 2-3 ems. think. In vertical sections all gradations of shape are seen: from gentle undulations of bedding to vertical' . - • • . diapirio'sitrUctUres several metres-high (Photos.28,2900). The clay ,bands generally- remain continuous over the crests of the:undulations, but may be ruptured or,overturned above sharp peaks, or may. be'trUncated by an overlying, sand bed,, This last 6taervation shows that the structures are sedimentary, and not tectonic in origin. Suoh updoming of sandstone beds may,00cur -in-isolated examples or, as in Biota'', a bed may, form a series of domes separated by broad synclines.' Balled-up sandmaoses such- as-that seen in'Photo.32 are very rare. ,In vertical section the contortions might be mis- taken for folds, but when exposed in plan view they are biroular or oval-(Photos.33 and 34). Linear folds are very rare illioto.35) and When they do occur are oriented

... • — ttat. 0.3 •

, — -

,ttpti

1"46',..- - _

•

IW"tlgiP4fgiltA" PHOTO 31. Gregarious interstratal cuss structures. Glame Beds,. coast of Raasay,3km i+).N.E. of Eilean en inblitre.Looking E..

PHO1 . BallL interstratal structure. Unit V. E. coast of Scalpay.lkm N.N.J. of Carus na i'isteogh, looking N. d. • • • N.

• ____LALI___ea( • -

,;TO 33. undomed interstr _ Jusp structure. Upper Fladday Beds.S.i.V.point of Mean Fladday, at high tide ievel.Looking

. ....,- ----..--v _ tallit fp gleo.-...y. 1 __.'* \ - • -• PHOTO :34. Rupture° cusp. Upper Fladday Beds. Foreshore 011 1,1* 5 1.-ie or Uorrth ma Dao inc... Silean Fladday.Looking E.. PHuru 65. Elongated interstratal fold.Loch an Uachdair Beds, incliffs 1.5km S.E. of Manish Point.Looking N.E.. FIG. 37.

[OM. N.W. I S.E.

- - ' . ' >- • - - - - • ______.L._ ___.A._____I______------)" - - I- - A L_ r .."'Clk:'31 7.177 . 6:. ---?kc;44e--..1. 112-ric';;3-•:!- 1--.•;t'lf.:.:)57'F:A",.- ...•

INTRASTRATAL CONTORTIONS IN COARSE S-STN. OVERTURNED DOWNCURRENT. S. COAST OF RAASAY, I KM. E.OF RUDHA NA CHLOICHE.

SE 15 M. NW I 1

---

OVERTURNED INTERSTRATAL CONTORTIONS. W.COAST OF El LEAN FLADDAY S.S.E. OFGRIANA SGEIR . COARSE S-STN. A. EARLY STAGE .

- — I , ..." 1 ...... " e ••••• GENTLE UPDOMII4G...... , 1 •••• I ..• •••• am ..s......

shale Cap rock.

B. MIDDLE STAGE.

..... - .... I ••• 0. ••• ..... •••• 1 ---,..-- DEVELOPING CUSP. ala ...." •••• 40. 1 ....• I I

bedding begins to be disrupted.

C. FINAL STAGE .

CUSP RUPTURED & ERODED AT APEX.

Plane of erosion.

F1G.38.DIAGRAMS TO SHOW THE ORIGIN OF INTERSTRATAL CUSPS. sf• 145

perpendicular to the prevailing current direction. The control of current on these domeshaped structures is also demonstrated by their occasional overturning down- stream (Pig. 37). Prom the large number of these structures which have been observed it is possible to reconstruct the ctaL;es in their development (Pig. '58) fl The first stage is the formation of a gentle dome. The sandstone in the dome still retains its original bedaing continuous with that of the sandstone on either side of it. In the second stage the height has increased. The top of the cusp is now a sharp apex and the flanks are concave upwards. In the final stage the overlying clay band, which until now has plastically stretched over the cusp, is ruptured and overturned. The top of the structure is generally planed off. By this time the sand in the core of the cusp is generally completely massive, all primary bedding structures having been lost. b) Inctrnstrrttal deforation Associated l'ith and probably genetically related to cusp structures in the hod facies sandstones are disturbances of sandstone laminae within the beds themselves. These usually occur in coarse to mediun-grained sandstones, althonEh they arc very occasionally seen in the very coarse and fine sandstones. PHOT036. Overturned cross-bc,-din6.Unit of Scalpay on righthandside of an unnamed burn,2m above high tide levei.Looking N..

PHOTO 37. Vertical intrastratal disturbances formed in cross-bedded sandstone. Glame Beds. a.coast of Raasay 300m S.4. of Manish Point. Looking W..

FIG. 39.

N. S. j .i...... v...... _...... , I

....••••• ..._ ,_ ...„... 1-.). 6, • j ,,,,.... ,.,....,___„7„,•.:..•• ..,.: .0 ...,_

VERTICAL SECTION OF INTRASTRATAL DEFORMATION IN MEDIUM GRAINED SANDSTONE.GLAME BEDS. W. COAST OF RAASAY 2.75KM N.OF EILEAN AN INBHIRE.

INTRASTRATAL DEFORMATION SEENIN PLAN VIEW IN MEDIUM GRAINED FLAGGY SANDSTONE, UPPER FLADDAY BEDS W. COAST OF EILEAN FLADDAY. (SEE INSET MAP.)

Or- ...... "---.------, - ___ -=--,-

Lv 0-3 M. i I 148

Two main kinds of intrastratal deformation can be distinguished, one occurring in horizontally bedded and the - other in cross-stratified sandstones. The'foreeets'of cross-stratified sandstones, when traced to the top of the bed,- are often seen to be strongly curved and overturned downourrent (Photo.36). These have been desoiibed and figured frok the Torridonian before by .nimmittil (1907, pi333.'Fig.13). . Similar structures have-been recorded from -the Bima sandstones (Cretaceous) of Nigeria-by G. JOnes,(1962) and, under the direction "Of Dr. -Oulink, the author has seen many comparable examples in the Tertiary fluviatile sands of BelgiuM. The seCond>kind-of,intrastratal deformation involve's horizontally laminated sandstones. In vertical section all'stages are seen in the development of a series of tight folds, up to 1 m. high,,separ-ated by, :gentle synclines up:to 3 m. wide (Fig. 39). The folds die Out both. upwards and downwards. In plan these strue— tures.varY -frem circular to linear (Fig. 39). In some in-stances the apexes of the anticlines are overturned downcurrent, as shown by the dip of the foresets of,the overlying.oross-stratified sandstone (Fig. 37).. Transitional forms intermediate between these vertical folds and overturned false bedding occurs: These take,the fOrill.of.oblique folds or kinks of foreset laainae (Photo.37). 149

c) Discussion of sandstone deformation. Inter- and intra-stratal deformation is a common feature of the Red facies sandstones, not only in the area studied in detail, but also wherever else the author has reconnoitred the Torridonian formation. It is therefore very important that the causes of these phenomena are correctly understood if the sedimentary environment of the Torridonian is to be interpreted aright. There are several possible origins of these contorted sandstones. Tectonism, as advocated by Harker (1903), has already been shown to be, an unlikely cause, due to the peneoontemporaneous truncation of many of the structures. Morphologioally some of the deformational shapes, especially the intrastratal folds, resemble air-heave structures described by H. Stewart (1956) and Rolfe (1960). However, the larger scale, and plastic behaviour of the clay partings of the Torridonian structures suggest that such a mechanism is unlikely. Some periglacial contortions show a striking resemblance to those of the Torridonian (see for example Van lieckwijek and Macar 1950, Rlanche I, Photo.2, and Planche II). However, they are seldom so morphologically constant, and are commonly developed in irregularly 150

banded sediment of highly variable lithology, not in monotonously uniform sandstones. It seems most likely that these disturbances were produced by the movement of unconsolidated sand under water as Bailey suggested in 1944 and as their occurrence in aqueously deposited' sediment indicates. However, the dominance of vertical, as opposed to lateral movement, 'and the rarity of a preferred orientation of long axes, make it unlikely that the structures were produced by slumping. One must conclude that the deformational shapes seen in the, Red facies sandstones were produced by the vertical floW of auicksand. Certain problems now present themselves concerning the mechanics involved. A water-saturated clay will flow when subjected to stress because it consists physic- ally of one phase and can behave platically. Load casts of sand in clay arOhus formed by a direct response to differential overhead pressure. A water-saturated sand, however, consists of two phases, a granular solid in a liquid, and for it to be deformed (i.e. have no shear strength) it is essential to reduce the intergranular friction. Loading, per se, cannot explain sand flowage since, instead of reducing intergranular friction, it increases it. A review of literature pertaining to civil engineer- ;51

ing and soil mechanics, e.g. Terzaghi , and Peck 1960, Tolman 1937, Plummer and Dore 1940; suggests that there are two main kinds of quicksand. YirStly intergranular friction may be reduced by the presence, of colloidal clay particles. A good' example is the Knott End quicksand studied by Freundlich and Juliusberger (1935), who showed that this sand was thixotropic. When subjected to stress it was converted r, II CI' from a solid to a liquid oneoof coarse backagain when the stress was removed. The sediment was a well sorted fine-grained sand with a small clay fraction. rhen the latter was removed the sand behaved normally, but regained its thixotropic properties when the clay was replaced. The quicksands of the Torridonian may have behaved in a similar way, especially in the fine-grained ones, where there was a small percentage of ferruginous clay matrit. However, in the coarser sand, clay is almost totally.absent and an alternative mechanism must be found. The second class of quicksand is clean and well sorted: Intergranular friction is reduced by means of a high pore pressure of water which forces the grains apart from one another. The vertical arrangement of the Torridonian quicksand structures suggest that this process may have taken place by the upward movement of water. A high pore pressure can be generated in a sand 152 •

in two ways. Firstly water may be fed into the sand from . a source at a higher -level. Thus a hydrostatic head of pressure ia'foimedwhiCh foicos the water to escape through - the sand. Sand volcanoes may be,formed in this way, and have been, produced experimentally by Rettgcr (1935) and Kuenehlin Gill and Xuenen 1957). Quicksandu, generated by-a hydrostatic head of water, are also to be found on beaches where sands-everliewater bearing gravels which dip and lenee - out seaward. Such a quicksand occurs after heavy tails of'rain on the 'beach at Camus na Geadaig -on Scalpay. v It worke 'both subaerially, and, when the tide 'iS'up,`subaqueoUsly. However, the lenticular; shapeof

the Toriiideniail sandstone ,unite and- the shale _films* which - ' .'deparate- them, would_ , probably'prSvent much lateral per- colation.'` A source df water may be found in the sands them- When ,a sand is -deposited it may settle in a loose packing, especially if, as iii_many of the Torridonian examples depbsition took place on .the- lee slopes of migrating sandbodies.= 'Under certain Conditions the packing may be tightened. This results in a decrease in 'porosity - ind,in the formation of surplus pore water which hasIto be" expelled.. - It flows 'upwards, taking the line of least resistance, and in so doing 'decreases the inter- Pinc.(T n 1t, c. • gtanUlar%rains as it forces them apart. This, upward PHOTO 38.Expertmentaily produced intrastratai disturbances halted at an early stage in their developement. x1.5.

PHOTO 39.H:xperimentally produced intrastratal disturbances halted at a late stage in their developement. x1.5. 154

flow of water could perhaps disrupt or deform the lamin- ationof sand through which it- was passing. Such quicksands occur in recent unconsolidated sand *here they are a considerable'hazard to civil engineers (Casagrande 1936)'. A series,of experiments were devised to see if this mechanism could produoe the kinds of disturbance seen in the red sandstones of the Torridonian. A brief account of them has been already published (Selley and Shearman 1962).' Glass beakers and tanks were filled with water, and sand-allowed to settle gently, into them with a loose packing. To indicate bedding, thin laminae of grinding powder were added at intervals. The containers were then vibrated-by'placing them on the flat lid of a centrifuge which was then turned on. The sand could be seen to be violently agitated as the packing tightened and the level dropped. Through the glass sides of the tanks a complete disruption of the laminae was often visible taking plade. On other occasions the laminae were slowly deformed into a number of tight-crested anticlines separqted by broad flat synclines. This morphology closely resembled the intrastratal deformation seen in the led sandstones of-the Torridonian (Photos.38,39). To simulate intrastratal cusps 2-3 Mme. of kaolin was allowed to settle out above' 2-3 ems. of loosely packed sand layer. A further centimetre or so of sand PHOTO 40. experimentally produced interstratal cusp structure. x1.5 156

was then added. When vibrated the clay-at the side of ,the tank could often be seen doming up over a little blister of water which quickly became filled with quicksand. The clay seal at the top of the dome then often rupturea, and quieksand,flOwed up to the surface of the 'overlying sand-. ,The latter often moved downwards on the- flanks of the cusps to replade the:ejeoted quicksand. The shapes Which .resulted from this process closely

,resemble the intrastratally deformed red sandstones of ,.the TorridonlanPhoto.40). These experiments have several obvious limitations. FirStly the quicksand structures produced in the lab- . oratory are'considerably smaller than those'seen in the field. However, the latter are characterized by a' wide 'range:of scale varying-from a height of several centimetres-(the,size of the experiments) seen in early stages of intrastratal folding, to several metres in some of the larger cusps.- It seems likely that a bed I c1,2tiHmetre thick will 5xpel prOportionally the same amount of water as one 1 m. thick, all other things being equal. A second'poesible' objection to the experiments might be that the structures are seen developed at the interface of the tank-walls with the sediments. It might be argued-that the confining force of the latter may have been significant. However, during vibration centres of quicksand eruption could be seen several centimetres away 157

from the tank walls and when subsequently dissected out the laminae and clay.beneath them showed the same morphology as the structures seen up the walls. In the light of these experiments one may conclude that, at the motent and for want of any better ideas, the quicksand Structures of the red sandstones were possibly formed by the vertical' expulsion of water from looSely packed sand. The remaining problem is the cause of the tighten- ing of the packing of the sands.

G. 4ones'(1962) suggested that earthquakes may have-caused the-overturned foresets in the Bits :sandstone. Areas of seismic activity today are associated with contemporaneous faulting and vulcanicity. There is evidence of'neithek in TorridOniantimeS. Furthermore, the effects of seismic vibration of recent unconsolidated sediments are well documented and.the sedimentary structures which result in no wise resemble those discussed in this section. Fissures:infilled with quicksand occur in recent sediments when subjected to earthquake shocks (e.g. Hobbs 1907, p.78). In fossil sediments such structures are Called sedimentary dykes; they are known for instance' in the'Ooalteasure.sof-the British Isles (Kendal 1922). They do not appear to be common in:the Torridonian; Only ,one bed of them occurs in the area studied and there is'one other recorded example near Gairloch (Hinxman, 158

,1907 p.333). A seoond-feature commonly produced by earthquakes in water-saturated sediments are sandblows. These are conical shaped pits, sometimes several metres deep. They are generated by the,sudden expulsion of sand and water which often infills'them again. Recent sandblows have been described by Grabau (19134 p.883-886), Hobbs (1907, p.78) and Davison (1936, pp.64 And 152); they occur in the Goal Measures (Kendal 1922, p.70)'and in Silurian sediments of Anglesey (Grabau 1913, p.884, Pig. 241). Sandblowc have not been described from the Torridonian formation. A third feature produced by earthquakes on unconsolidated sedimentb are small scale faults. They can generallY'only be traced upwards for short distances before being overlain bi,more sediment. Suoh small-scale faults 'occur in the- CoalMeasures (Kendal 1922). They have not been described from the TorridoniAn. In the- absence of typical sedimentary structures produced by seismic precesses,'and the lack of evidence of contemporaneous faulting or volcanic activity, one must conclude that seismic vibration did not cause the quicksand disturbances of the Torridonian. Oulianoff*(1958) has pointed Out that vibrations are produced in:nature. by a"wide variety of processes

FIG.40. Deformed heavy mineral bands.

_...... -••'-""-.... 15 CM.

W. E

DROPLET STRUCTURES.GLAME BEDS . 2KM. S.S.W, OF MANISH POINT. W. COAST OF RAASAY. VERTICAL SECTION.

I M.

) ‘...... ;..--22) 1/4*--"."

W E

e PSEUDONODULES)IN VERTICAL SECTION . UNIT Ex. 500M. W.N.W. OF MULLACH NAN CARN.SCALPAY. 160

other•than,earthquakes. Thunderclaps, flashfloods and 'breakingWavea are only a few examples. It is perhaps conceivable --that a turbulent sediment-laden fast flowing current,coul& produce sufficient disturbance of the underlying sand to,allow.a reorganilation of its packing to take- plice. In. Chapter UF' the -probably existance of such currents are postulated. The frequent Overturning of the,quicksand strnotilres down Stream shows- that current activity dertain4-pIayed=some part, in their Origin. Furthermore, whilet - carrying 'out the experiments des-. # cribed, in this chapter, it was found that quicksands 'couldfbe generated by manually-created, turbulence in the' overlying water. "All these facts suggest that the fossilized beds of quicksand described in this chapter were geherated - by the same, violent currents that origih- -ally deposited the loosely. packed sediment.

(iv) Deformational structuresofheavymineral _ bands. The heavy mineral bands described in Chapter V. often show some kind of penecontemporaneous deformation. This has been described from Raasay, Scalpay and Stoor by Selley &pf;0 and from Applecroas by A. Stewart (1963). A wide variety of deformational structures occur but four particular kinds appear to be the most characteristic and will now be desCribed. .1111X,W, ,

4 ; • 111",

•4 ,

$1,, •

Yr 5,a $ • "

PHOTO 41.. Balled-up heavy mineral band in Unit V sandstone. lkm S. S. vi. of Mullach nan Carn. Scalpay. Looking N.4. 162

The first,variety may'best be termed droplet structure. Individual laminae - are seen to be broken along their length at intervals of 5-10 ems. Half-way between these' breaks the laminae are seen to thicken into a series of droplets. The two ends of each of these droplets are deflected and thin out upwards (see Pig.' 40). :The second kind of deformation also exhibits'. rupturing along the length of an individual. lamina,, or of several olosely'spaced laminae. The parts of the lamina ,. or laminae adjacent to the break are upturned and occasionally burled over onto themselves (see Fig. 40). In morphology, though not lithology, these structures thus resemble Paleudonodules. In the third kind of deformation a whole mineral band has balled up to form a lump of heavy minerals some 30-40 'ems. in length and maybe' 10-20 ems. high• (Photo.41). Sometimes the heavy mineralaaminae of the origin band are no disturbed and intermixed with the adjacent sands that they are completely indistinguishable and even the boundary, between the ball-and the bed are transitional. In other examples the original laminae are still present though highly contorted and often of irregular thickness (Fig. The fourth main kind of deformation seen in heat?' FIG.41. Deformed heavy mineral bands.

30 CM.

S. N. VERTICAL SECTION OF BALLED UP STRUCTURE .UNIT LY 1,110M. N. OF CAMUS NA F1STEOGH . E.COAST OF SCALPAY,

I M.

S. N

HEAVY MINERAL LENSE(ARROWED)CONTROLLING THE POSITION OF CUSP DEVELOPENT. COARSE S-STN .UNIT y. N.COAST OF SCALPAY 2.5KM E. OF CAMUS NA GEADAIG. 164

mineral bands is the presence of tight anticlinal folds separated by broad flat lying synclines. .The dimensions and morphological characters of these are in all respects identical with the intrastratal deformation described already from the normal sands.

The deformation of heavy mineral bands is con-

. Sistant with the explanation advanced for the origin of the intra- and inter-.etratal deformation of normal sandstones previously described in this chapter. It is reasonable to suppose that if a sand is rendered quick then overlying heavy mineral bands will sink down into the quickened layer by virtue of their higher density. The morphology of these structures clearly suggests that such downward movement has occurred. In one instance

even)a curved lens of heavy minerals lies in a basin between two cusps. It seems as if its weight controlled the positioning of these when the underlying sand went quick (Pig. 41). 165

,CHAPTER VIII. PALAE0OURRENT DIRECTIONS

(1) Introduction Do palaeocurrent studies have been made in the area previously. A certain-amount of work has been carried out elsewhere in the Torridonian formation and this is summarized, together with the authors responsible, in Chapter IX (Fig. 53). During the preparation of this thesis the directional properties'of a large number of various sedimentary - etrUctUres were measured in the field using a compass and an Abney level. The geographic locality, approximate stratigraphic position and lithology were recorded for each reading. The - results are presented in the figures accompanying this chapter as rose diagrams with 10° intervals. All readings have been corrected for magnetic variation. The azimuths of planar structures, such as current bedding, have been corrected for tectonic dip on an equal angle stereographic net. No such cor- rection has been made for linear structures, such as washout axes, since, for the low regional dips encountered, no significant azimuthal variation occurs. Palaeocurrent readings were not recorded by any systematic method of sampling due to the impossibility of knowing their accurate stratigraphic position. The results of these studies will now be described 166

facies by laoies and, in the concluding section of this chapter their significance will'be discussed.,

(ii)Basal facies Pew sedimentary,structures of directional sig- nifiCanoe'are present in,the boulder beds, brecoias and grits of.the basal. facies. The field relationships, 'aithology and depositional dip of- these rocks is sufficient to%show'that they were derived as screes and fanglomerates from the hills 'of the Lewisian gneiss basement to the east lsee Chap. III, Fig. 9).

(iii)Grey taaiee - There is.a considerable amount of data, from various eedimentary'structures, which suggests that, in the northern part of the area, the Grey fades sediment0 came from the west. To the south, on Scalpay, they came from the north-west. In- the Caolas. Fladday bedd-the current ripple lamination, washout axes.and their false-bedded fill, and'the current bedding of the 'beach' sandstone all indicate,an.easterly_source (Fig. 42). In the .Brothel beds of Northern Itaasay. there is - little to show the.-direction of palaeocurrents. The north-4esterly.source of the mudflow'has already been, mentioned (Chap. IV 'Sea. iii). One washout axis with

FIG 42, Palaeocurrent results.

CAOLAS FLADDAY BEDS. A 3 -

WASHOUT AXES. DIRECTION OF DIP OF CROSS-STRATIFIED LAMINATED FILL OF WASHOUTS.

DIRECTION OF DIP OF DIRECTION OF DIP OF

CROSS-STRATI F I CAT ION CURRENT RI PPLE IN THE c BEACH'SANDSTONE. LAMINAE.

SCALE ON NORTHINGS SHOWS NUMBER OF READINGS.

FIG.4 3. Orientation of symmetrical wave ripples in the Caolas Fladday & Brochel Beds.

RESULTS PRESENTED AS ROSE DIAGRAMS AND PLACED IN THEIR CORRECT ,/, POSITIONS RELATIVE TO ,. //i THE SHORELINE.

O 5 10

N° OF READINGS.

1000 M. 4 DIRECTION OF DIP OF CURRENT RIPPLE LAMINAE. 170

a.bearing of 1700 also occurs. Though not'a direction of sediment transport the orientation of the Wave ripple marks will now be des- dribed. These are common in both the Caolas Fladday and Brochel beds and show variations of strike of considerable, interest (see Fig. 43). The relation of the ripples to the Isewisian gneiss at Brochel are not well exposed but at Fladday it can be seen that they run approximately normal to the unconformity in the north. Further south they can be seen to be deflected by the shoreline. This suggests that the waves, which must have formed these ripples, were generated by a northerly wind. In the Grey faeces rocks of Units II and III on Scalpay the foresets of current ripple bedding in the 'sandstone units indicates that they were deposited from southeasterly flowing currents (Fig. 44).

(iv) :Red facies.. There appears to be no difference in the source of supply of the red sandstones of the top of the Diabaig and those of the Applecross Group in any one region, so they will be considered together. The dip of -foresets of planar cross-stratification and the axes of trough bedding indicate that these beds, on Fladday and Northern Raasay, were deposited by south- easterly flowing currents (Figs. 45, 46). The southern FIG.45.Palaeocurrent results.

RED FACI ES :FLADDAY,

DIRECTION OF DIP OF DIRECTION OF DIP OF

PLANAR CROSS BEDDING PLANAR CROSS BEDDING

IN THE LR. FLADDAY BEDS. IN THE UR.FLADDAY BEDS.

DIRECTION OF TROUGH DIRECTION OF TROUGH

BEDDING UR. FLADDAY BEDDING . GRIANA SGEIR. BEDS. FIG.46. Palaeocurrent results.

RED FACE ES -. RAASAY.

PLANAR CROSS BEDDING. LOCH AN UACHDAIR BEDS . PLANAR CROSS BEDDING GLAME BEDS.

TROUGH BEDDING. G LAME BEDS.

PLANAR CROSS BEDDING. SOUTHERN INLI ER. FIG.47. Palaeocurrent results.

RED FACI ES:SCALPAY.

PLANAR CROSS BEDDI NG. TROUGH BEDDING.

RED FACIES:WHOLE AREA.

WASHOUT AXES . 174

inlier of Raasay indicates an east-south-easterly current flow (Fig. 46), whilst further south still, on Scalpay,- they were moving towards a south-south-easterly direction (Fig. 47). As mentioned; in Chap. V., p.//, clearly incised washouts are rare in the area, so they are presented in a single diagram regardless of their geographical location (Fig. 47). Their sense cannot be determined in the field,,but from the evidence of other sedimentary :structures it seems that they must have flowed in a general south-eastern direction rather than to the north-west.

(v) Summary and discussion The Screes and fanglomerates of the basal facies were derived from the east. This is essentially normal -to the hills from whioh they were derived and it is unlikely that this direction is of any regional Signifi- cance since the irregular nature of the topography could giye,rise to scree6 radiating in any direction. In the northernnPart of. the area the Grey facies .rooks were laid'down'by eacterly flowing currents, The direction of these may be of only local significance

being influenced by the topography of the basement which . 'was still exposed at this time. Further south'the south- 175

south-easterly pattern of current flow of the Red facies beds was already established during the deposition of the Grey facies This may perhaps be attributable to the absence of any basement rocks in this area. Unlike in the' north, they were thus probably buried whilst the Unit II and III 'beds were being laid down and hence could not influence the direction of current flow. 176

CHAPTER 1.X.- DISCUSSION OF SEDIMENTARY ENVIRONMENT AND -THE RELATIONS OF THE TORRIDONIAN ROCKS OF THE AREA . WITH OTHER REGIONS

(i) Introduction and historical survey During the past century the.Torridonian sediments have, by ,one author or another, been ascribed to every' kind of sedimentary- environment imaginable. It would be tedious to review alllthe literature available since little of it is based on any kind of detailed field work. VAAHouten (1961) has provided a complete summary_ and bibliography Of . papers published on this topic. Only those papers, the oonolusions of which are based on detailed field'work, will' be reviewed. •Himxman(1.907,-p.273) considered that the 'conditions of deposit' were probably those of a rapid accu- mulation in shallow water near a shoreline, subject to violent currents and the` influx of flood or stream-borne materials, with occasional intervals of ouiescence during which the finer sediments were laid down,... The rain- pitted and sun-cracked -surfaces of the shales and mudstones seen on Lobh Torridon and elsewhere show that the fine sediments of the lowest group are shore deposits,' and cannot be regarded as due to deeper water conditions.' Sutton and Watson, working on Skye, consider the Epidotio'Grit tobe'fluviatile (1960), and the overlying 17'7

Diabaig Sediments to have been deposited in a 'region of broad intertidal sediments and offshore sandbanks.... noting that 'comparison of the assemblages with those sediments deposited near the North Sea coast- of Holland and Germany reveal some resemblances' (1963, in press). P. Allen, J. Allen, Goldring and Maycock (1960) have suggested that the grey shales of Diabaig may have been shallow water marine deposits. A. Stewart (1962a, p.154) considered the red arkose facies to haVe been 'deposited under oxidizing conditions, probably fluvially.' The history of the deposition of the Torridonian sediments seen in the area will now be described, based on deductions from the facts presented in this thesis. Finally an attempt will be made to describe the history and palaeography of the Torridonian formation as a whole, consistant with these results and personal observations and published material from outside the area studied. On Raasay and Fladday it is possible to distinguish three phases of deposition following one another in any one place. There is no evidence to suggest that these phases succeeded one another synchronously over the whole area. Each phase, indicating a different environment of deposition, will now be described in turn.

(ii) Phase I , The'high metamorphic_ grade of the'lsewisian gneiss FIG 48.

RED BASAL FACIES. E.G. THE TORRA N BEDS,

Hypothetical diagram of conditions during the first phase of sedimentation. 179

which forms the basement suggests that prolonged and severe erosion occurred before deposition of the Torridonian Sediments began. The basement was still an irregular surface of hills and valleys with a relief of several tens of metres when the first permanent deposits were formed. The hillsides, as seen from the structure contour map (Fig. 7), were steep and, as demonstrated in the region of Loch an Uachdair, dissected into gullies near the top. The'feet of .the hill slopes were buried in scree; and alluvial fans, such as the Torran beds, , debouohed from the mouths of the gullies. This origin suggested by the field relationships of these deposits to the gneiss is supported by their poor sorting, the angular nature of individual fragments, and their obviously local source. An essentially subaerial environment is further suggested by the red oxidized condition of these first deposits. Fig. 48 is a hypothetical diagram showing the scenery of the area during this first phase. The process of weathering and erosion of the basement and the transportation of rock waste via alluvial fans out of the area had probably continued for a long time. Suddenly though it stopped,and the region began to be buried under quite different deposits which, as will be seen in'the'next section, suggest that the area was now flooded by water. This flooding marked the end of the first phase and the beginning of the second one. 180

(iii) Phase II The existence of the second phase is indicated by the Grey facies rocks of the Caolas Fladday, Brochel and Unit II beds. Due to recent erosion, at no locality in.the area can these be seen overlying the red basal facies of the'Torran beds. 'However, taking into account their relatiie geographical positions and the regional dip, this state of affairs must have originally existed. The wide variety of lithologies and sedimentary structures suggest that the physical processes operating during the second phase must have been diverse. The well laminated grey shales with wave ripples and desiccation cracks imply conditions of deposition under a quiet body of water which ocessionally dried up or receded. The sandstone units with which these beds are intercalated indicate that, from time to time, current activity prevailed. The presence of 'it'll° grains of non-Lewisian rooks suggest that the material of vihich they are composed had travelled from afar and were not derived lobally. In the following section it will be demonstrated that fluvial sandstones were subsequently deposited in the area. It may be that the same rivers that formed these periodically-flooded the waters of the Grey'facies with muddy sand. In the sandstones of the Caolai'Fladday and Unit II beds the presence of current GREY BASAL FACI ES. E.G.THE i:5ASAL BEDS OF CAOLAS FLADDAY. GREY FACES. Hypothetical diagram of conditions durinc the second phase of sedimentation. ✓ 182,

ripple latnf.netior~ shows . that they were transported and

- depoSited-over the bottom by traction currentL3. The massive,6raded tandstdnes of the 1;roche1 beds, however, rere probably transported by, and'oettled out of suspensions or muddy 'snared. Yultiple gra1iw nay have been produced by t'ro moro rivers diashar6ing their load at different times Shorty after an increase in aLrface rtua.-

okr.' by this differenco_should exiat is not clear, 'Perhaps in some areal the river mouths debouched their loadO down uteuli elopOs at: high velocity so that the . Sediment:, wa parried by ouspensiOn far out into-the waters. Llsevihers'the'riVer mouths ray have 11;.d- antler slOpeS'and loner currant velocities, so that the sand Ana transported out.from 'then' by traction currolitu alone. :/urinis this second plhase sone of the blecr.itint. was

still, exposed and contributed marginal :scree:, now soon ao drab oploured boulder beds and brae:ciao' on the east- side of 'aaolas Pladday and elasvhere. Apart fro these local deposita And thi, sus flO of the •Broehel .Bede there ins little ovidence lo sugcmst that those parts of the basement which. still protruded contributed such to oedimentation-. rig. 49 is a hypothetical aiacran of the or44arnnce of the'st_dimentary environmeLt durint, the ..second phase. 'efore pascin,t op to phase three there is 'one- 183

major problem to be raised which, from the evidence presented in this thesis, is insoluble. This problem is the lateral extent'ef the waters in which the Grey faoies sediments of:phase,II were deposited. Exposure in the'area is not sufficient to demonstrate whether or not the Caolas Iladday, Brochel and Unit II beds were . deposited.under one laterally continuous body of water, or, whether, as, the'irregular topography of the'basement could suggestp_they were laid down in separate land- looked basins. This point will be raised again in the second part, of this Chapter, during the discussion of ,the Torridonian formation as a whole. • In the cliffs on the west side of Caolas gladday the' shales of the Grey facies are sharply overlain by Red faoies sandsto4es,. This reflects a major change in environment perhaps, as Hinxman (1907, p.287) observed, as-significant as that between the Rhaetic and the 'Cower marls. This break marks the close of phase II and the beginning, of phase III. -This change, so clearly *seen on Pladday in the north, becomes progressively harder to demarcate southwards, until on Scalpay a dis- tinet:tiansitional sons (Unit III) is Tresent, composed of interbedded Red and Grey facies sediments.

184

(iv) Phase III The Gtey facies sediments of the Caolas Fladday, Brochel and Unit III beds are overlain by the Red facies Sandstones; fine at the base (the Lower Fladday, Lochan Uachdair and Unit IV Bed0 but passing up into coarse often pebbly sandstones (the Upper Fladday, Glame and Unit V Beds). Stewart (1962a) considered this 'red arkose facies' of the Torridonian to be fluviatile in origin, but he gavelno reasons for reaching this conclusion. It will now be argued that the lithologies and sedimentary-structures of the Red fades sediments suggest a depositional environment of braided rivers. The coarse nature of much of the sediments, together with large scale cross-stratification and certain kinds of quicksand activity suggest that transportation and deposition was effected -by subaqueous traction currents of high velocity. The occasional presence of very fine sandstone and shales, heavy mineral bands and intraformational conglomerates suggest that periods of violent current conditions alternated with quieter conditions. The lenticular arrangement of lithological units of varying grainsize implies that variations in current velocity were lateral as well as vertical. The conditions outlined above seem most likely to be found in a fluviatile environment. Ancient deposits 185 similar in lithology and sedimentary structures have been attributed to flUViatile origins; e.g. the Old Red sandstone (Devonian) (Friend 1961 and J. Allen 1962a and c) the Hants facies of the Wolfville Formation, Trias, Canada.(Klein 1962) and the Tertiary formations of the High Plains, U.S.A. (Johnson 1901). Resemblances can also be found between the .Red facies of the Torridonian and Recent _river deposits such as those described by Fisk (1947), Frazier and bsanik (1961), MdKee (1957b), Shantser (1951) and Doeglas (1962)-. Aiversmay be classified into three kinds according to their trace, vii: straight, meandering and braided. It is known that the trace of a river is related to various factors such as regime, load, gradient and velodity (see Leopold-and Wolman 1957). It is also known that the'amount and kind of alluvium deposited is controlled by these factors. Thus Lamakin (1947) recognized what tv),a • he called).thiApossible 'dynamic phases' for rivers. The erosive'phase is illustrated by fast flowing straight rivers with 'instrative' alluvium, generally present only in small amounts and often only temporary. The depositional- phase is shown by braided rivers which deposit considerable amounts of peystrative alluvium. The phase of equilibrium is illustrated by meandering rivers which give rise to constrtive alluvium which is generally present in only small amounts and subject to considerable 186

reworking. Shantser (1951) has classified alluvium into the following three facies:- i)Floodplain facies - essentially fine-grained sands, silts and clays deposited, often out of suspension, at times of flooding. ii)Channel facies - mainly coarse sands and gravels deposited by fast-flowing traction currents in the rivers themselves. iii)Oxbow facies - silts and clays laid down in temporary lakes formed by the cutting off of old river channels.

Kartashov (1961) has pointed out that the difference between braided and meandering river deposits is in the relative proportion of facies present. Thus in the deposits of a meandering river, such as the Mississippi described by Fisk (1947), Turnbull, Krinitzsky and Johnson (1950), floodplain and oxbow facies deposits occur to the almost total exclusion of true channel deposits. Hence the recent alluvium is mainly fine- grained. In braided rivers on-the other hand the alluvium is essentially of the coarse-grained channel facies. Fine grained material is carried further on and not deposited on flOod plains and oxbow lakes which are typically absent. Such-coarse-grained braided river deposits have been described from the Durance and the Ardeche by Doeglas (1962). Thus it appears that the 187

coarse-grained nature of the Red facies sediments suggests that they were, probably deposited by braided rather than meandering rivers. The Red facies attains a thickness of at least 2000 m. in the area, and sometimes two or- three times that elsewhere. 'This Observation supports the previous ' conclusion since meandering rivers are essentially in -equilibrium and deposit relatively little sediment. Braided rivers on the other hand are typically overloaded and can give rise to considerable thicknesses of alluvium. Further support for the braided origin of the Red facies sandstones.i6 provided by the narrow unimodal pattern of the current bedding readings (Fig. 45-47). A wide scatter of current bedding directions suggests deposition by currents of varying direction such as occur when a river meanders. Current bedding directions which plot out with a well developed maxima imply deposition by' almost unidirectional currents such as occur in the mild curves of braided river channels. This criterion of scatter of current bedding results has been used in a similar way before. Mapel and Pillmore (1962) found a wide azimuthal variation in the floodplain deposits of the Lakota Formation (Cretaceous) of Wyoming and Wurster (in press) found a well developed maxima for the current bedding directions'of the Schilfsandetein. Thus the lack of scatter of the Red facies current bedding readings

CHANNEL RACE FACIES. 11

LJ ) SHOAL C)

ABANDONED CHANNEL FACIES. 0

FINE SANDS & SILTS.

MEDIUM GRAINED SANDS COARSE PEBBLY SANDS.

Hypothetical diagram of conditions durinc the third (Red facies) phase of sedimentation. .189

lends further support to their braided fluviatile origin. 'If it is accepted that, for the reasons argued above, these bids were deposited by braided rivers, one major problem remains. This problem is the significance of the distinctive arrangement of lithologies into coarie, medium and fine assemblages. This may perhaps be best interpreted in the light of Shantser's studies (1951) on recent river channel deposits. The- poorly or trough bedded pebbly sands of the coarse Assemblage invite comparison with the alluvium of what Shantser calls the 'race facies'. This is formed as essentially lag depostts in the deep water major channels where current activity' is at its most violent And turbulent. The planar cross, bedded qteditim-grained sands of the Mediut Assemblage compaie with those of the 'river bed shoal facies' which flank the major channels. Here current'activity. ie'less - violent and relatively finer material can be laid down. The very fine sands and silts-of the fine assemblage were probably deposited either in quiet parts of the major channels during times - of low water or in disused channels. (These deposits are thui comparable to the oxbow facies of meandering- 'river belts; but this term is perhaps best avoided to prevent confuSion).

APPLECROSS GP RED FACIES.

DIABAIG GP GREY FACIES.

BASAL FACIES.

Diagram to summarize the relationship between facies and lithostratigraphy. -)K. MARKS THE LOWEST BAND OF NON-LEWISIAN PEBBLES. 191

(v)Summary of the history of the area The Torridonian history of the area may be summarized in a few words. Long continued subaerial erosion orthe basement was halted before peneplanation while there was still considerable relief. This ancientland surface with its screen and fanglomerates was flooded and in theiow lying parts fine-grained, shales were laid down in bodies of quiet water which, from time to time, receded or evaporated. Sands were occasionally carried into the area, probably by rivers rising in a distant source area, Finally these deposits of shale and sandstone, and the parts of the basement which were still exposed, were buried under the sands and gravels of a braided river obmplex which spread out over the area from the North and West. 'The change from basin to fluviatile environment was abriipt-in the North, but in the South interdigitation_of the two facies shows that the change was more gradUal. .The sediments that resulted from this history are seen in the accompanying generalized diagram (Fig. 51).

(vi)Synthesis of the originof the Torridonian formation The facts on which this section is based are taken from various sources including the material in this thesis, published information and observations made by the author in regions outside that covered by this thesis. F1G.52. RHUM_ SKYE. SCALPAY. RAASAY COIGACH. QUINAIG, PARPH. BLACK & 1907.048-362. SELLEY —THIS THESIS . 1907p.320. 1907f .299. 1907x.292. WELSH.1961.

S-S T N S. AULTBEA SSTNS & SHALES GP

SHALES •••

[ T TI ORONSAY GREYWACKE GP 1,000 M.

• Guirdit • • Arkose. APPLECPO GP.

ORONSAY & COLONSAY. S. A.STEWART. •—,C1C)Ci 1962 • • DIABAIG Staosnaig GP. Phyllita. • Loch na Eala ' Arko sQ Kiloran inloth Beds. Flags . LEWISIAN GNEISS .

*Mule Group.

• Rudha na Roinne Grits. !Gleba t ten Be inn na Group . Saamraig Grits. Strati graphical sections of the Torridonian Formation. Loch na Dal Bagh na h'uinha Beds/ Shale. E pi dot lc Grits Machrins ' Group. DATA TAKEN FROM VARIOUS SOURCES:1907 #la sal REFERS TO THE MEMOIR BY PEACH,HORNE, Dun °attain Grits. Epida tic &it. '/ ET AL.. FOR GEOGRAPHI CAL LOCALITIES SEE FIG.53.

MARKS THE LOWEST OCCURRENCE OF PEBBLES OF NON-LEWISIAN TYPE. 193

The general stratigraphy of the Torridonian is •seen summarised in Fig. 52. The aerial distribution of the rocks together with palaeocurrent data so far available are shown in Pig. 53. Prom these two figures it can be seen that the present extent of the formation is in the shape of a strip running across a basin with an approximately east-west axis in the central part of the area. The full extent of this basin is not visible due to the sea to the west and the Moine Thrust plane to the east. In the southern part of the basin the first deposits (the Oronsay Greywacke Group) indicate that infilling began by northerly-flowing turbidity currents. These are succeeded unconformably by deltaic and fluviatile'sediments derived from a similar direction, whioh Stewart (1962a) correlates with the Diabaig sediments of Rhum and Skye. On the former island Black and Welsh (1961) believe these sediments to have been derived from the south and south-west bUt on Skye they come from the north and West (Sutton and Watson - personal communication). Thus the axis of Diabaig sedimentation lay somewhere between Skye and Rhum and perhaps terminated not far to the west. The great thickness and relatively larger aerial extent of the Diabaig deposits of Skye and Rhum suggest Poach & FIG.53. Horne 7907p.233. PAPRH

CO ICA CH

Peach & Horne 1907332 50KM.

Sutton & Watson 7960.

Black & Palaeocurrent directions Welsh 1981. of the Torriconian Formation.

DATA TAKEN FROM VARIOUS SOURCES. TORRIDONIAN ROCKS SHOWN IN RED.

DIRECTIONS IN ROCKS COLONSAY OF THE DIABAIG GP, Stewa 1962, \\4 DIRECTIONS IN ROCKS OF THE APPLECROSS GP.

DIRECTION IN THE ORONSAY GREYWACKE GP 195

that they may lave been deposited in one large basin,

- probably also continuous with the deposits of Colonsay and Oronsay. This, together with the turbidite facies of Oronsay, suggests that this basin may well have been marine. Indeed, as previously mentioned, Sutton and Watson (1963) have compared the Diabaig sediments of Skye (with the exception of the Epidotic grit)'with the tidal mudflat deposits of the Wadden Sea. North of Skye the Diabaig deposits thin out abruptly-au seen in the region of Raatlay and - Loch Torridon. Prom these localities •to as far north as Stoer they occur, as isolated patches in low lying parts of the hilly Lewisian basement. These northern Diabaig sediments may well have been deposited in isolated lakes. Somewhere though a dissected high- land-coastline must have existed between these and the marine southern basin. The position of this coastline is obscure due to the paucity of exposure and the diffi- culty of distinguishing the two kinds of,sediment. North of- Stoer the basement is flat or only gently undulating (1907, p.275) and hence lacustrine Diabaig sediments-are'absent. The sudden appearance of these lakes. in a region which had for a considerable length of time• been undergoing subaerial denudation, is one of • the most interesting problems of the Torridonian. The following account is offered as a tentative explanation.

FIG 54.

SKYE RAA SAY. PARPH.

PHASE I

S. N.

......

PHASE II.

Diagram to show how subsidence of the southern basin may have tilted its northern margin causing a reversal of drainage. 197

Atone time the drainage of, the Lewieian mountains ti between Loch Torridon and Quinaig must have,been northerly :over"the flat peneplane. However, the development of a major basin between Skye and Oronsay to the south would ''lead to a downwirping of its margins. Hende a reversal of the northerly drainage would'occur in the mountains between Loch Torridon and Quinaig and the heads of.the valleys would be flooded to for lakes in which the Diabaig-sediments, were deposited (Pig. 54). A similar flooding of valley heads has been brought about by reversal of drainage on the flanks of the East Africa rift valleys. This marginal subsidence was probably compensated, by uplift to the north-west. Braided rivers flowing south-easterly from the'Lr gradually- buried the area- with-the coarser Applecroaa sandstones. The influence of theSe rivers must have extended as far south as

- Rhum'bilt on Colonsay an 'equivalent facies, though with ,sporadic deltaic affinities,.built out from the south. Thtis during the deposition of:the Appleorosa sediments the axis of the basin had shifted south to somewhere between Rhum and Colonsay (assumiag that the age of these deposits are to any extent synchronous). The Aultbea group of sediments, which succeed thobe'ofthe Applecidss Group, are in many ways comparable to the latter. However, they contain a larger amount of 198

fine ATained. material (1907, p'.274). These may perhaps -then be flood plain deposits, thus indicating that the subsidence of the ToTridonian basin was drawing to a close and that the rivers were now approaching a stage of equilibrium. The explanation of the Torridonian formation in this section can only be regarded as a tentative hypothesis subject to considerable revision by further detailed studies of the sediments, and by the wider aPplication?of palaeoMagnetic studies- (or some other technique) to establish an:accurate chronology. 199

Appendix I

—Notes on some terms used;

1. Particle size terms refer to the Wentworth grade scale (Wentworth 1922).

2: The word-'grit' is used throughout this thesis to describe a rock composed of particles of Granules (between 2-4 -mmd.).

3. The term 'shale' is used to describe a sediment of silt grade possessing a shaley parting, and the term mudstone for its massive equivalent. Argillaceous sediments are absent in the area.

Appendix II

Summary of lithologies and' sedimentary environments

Iithostratigraphic units Facies Environment Subfacies SubenvironmentS

. Units IT and *1 Coarse Assemblage - Race facies Braided Fladday, Glatt() N River bed Red Middle shoal facies and Loch an rivers t Abandoned Uachdair beds Fine channel facies Shallow 1 Medium sandstones-early river .Brochel, Caolas water, channel deposits lacustrine Fine sandstones Fladday and Grey or marine ? lacustrine

Unit II beds cshoreline Shales

Basal,Beds of Caolas Fladday Scree Green Subaqueous scree and Basal and Brochel.Torran Subaerial scree Beds Fanglomerate Red MEP and fanglomerate 201

ACKNOWLEDGEMENTS

The author gratefully acknowledge's the help of the following: The Department of Scientific and Industrial Research for financing the project. Professor J. Sutton,Mr. D.J. Shearman, Dr. J. Watson and Dr. G. Evans for many useful discussions both during the work and the preparation of the manuscript. Mr. E. Hill and-his'etaff-for preparing thin sections for microscopical examination. Mr. J. Gee for 'photographing the experimental sedimentary structures. Mr. J. Walford for permission to work on Scalpay, ,and Mr. D. McNiven, his estate manager, for facilities provided on that island. Finally, the author expresses his gratitude- to the people of Raasay, the dillies family especially, for their many kindnesses to him whilst on that island.

amlmumalmowmpimpftwmir ...... mwmamommolmr.mummumemw

202

BIBL /OGRAPHY.

LLIM,J6114•1962a, The Lower Old Red Sandstone of the southern British 'ales :a facies 'resembl ing - the Alpine MOlasse. Nature,London. Vol.193. No. 4821..p. '1148. 1962b. Asymmetrical ripple marks and the origin of cross-stratification. Nature, London. Vol.194. NC: 4824. p.167. 1962o. Pettology,brigin & deposition of the highest L.ower Old Red .Sandstone of 'Shropshire, . England.J. Sediment. Petrol. Vol. 32. p. 657. ------.----1963S.Asymmetrical ripple marks & the origin of water-laid casette of ,crow-strata.Lpool. . Manchr. geol. J. Vol. 3. Pt2. ALL J...R.1.4, ,BEMINGWAY, 4. ,& OLARIN.1963. Structure of ripple ' marks. Nature.London.. Vol.198. No. 4883. p. 847. ALLEN, P. ,ALLENO.R. L. GOLDRING,R. 4MA1GOOL, I. D. 1960. Festoon , bedding IMud-With-lenticlesilithology. Geol. Mag. Vol: 97. No. 3. p. 261. ARGYLL, Duke 'of4,1880. A fragment of Primeval &trope - a reply. Nature. London. Vol.P”. p.407. BA/LBY,B.1944. The Tertiary igneous tectonics of Rhum. Qivart.J. geol. Soc. Lond. Vol.100. p. 165. , 203

BAILEY, E. 1955. Moine tectonics and metamorphism in Skye. Trans. Edinb. geol. Soo. Vol.16. p. 93. BLACK, G. P.8 WELSH, 1961. The Torridonian succession of the rale- of Rhum. Geol. Mag. Vol. 98. No.4. p. 265. BLACKWIDER,E: 1918. „ Mudflow as 'a geologic agent in semi- arid mountains. Bull. geol. Soc. Amer. Vol. 39. p. 465. :.., BUCHER, W.1919. Ripples & related sedimentary surface forms & their palaeogeographtc interpretations. Pts. Amer. J. Set: Vol. 47. p.149 & 241; CASAGRARDE.A.1936. Characteristics of aohesionless soils affecting -the stability of slopes & earth fills. J. Boiton Soc. Civ. Eng. No. 1. p.13. CROWELL. J. C.1957. Origin of pebbly mudstones. Bull. geol. Soo. Amer. Vol. 66. pa 993. CUMMINS,\I. 1962. The Greywacke problem. Lpool. Manchr. geol. J. Vol. 3. Pt. I. p. 51. DAVIDSON C. 19:W. The geology of the island of Raasay. Proc. Univ. Durham Phil. Soc. Vol.10. p. 246. DAVISON, C. 1956. Great Earthquakes. Thos. Murby Co.. London. DOEGLAS,D._J. 1962. The Structure of Sedimentary Deposits of braided _rivers. Seekimentol ogy. Vol. 1. No. 3.7).16'7. DOWNIE, 0.1962. So-Balled sPores from the Torridonian. Prot.

geol. Soc. Lond. No.1 p 600. p.127. DMA nirSKYi S. & KOTLARC UK, J.1962. 0 pogreznietyoh pregaoh fal s tych. Roc znik Pol et iego Towarzystwa Geologicznego. Tom. 32. Zeszyt 2. - p.147. -204

EMECRY,R, 0.1945. Entrapment of air in beach sand44T. Sediment. Petrol. No.15. p. 39. FERVOR, Sir L.1951. Ox a diScovary, of Copper ore in the Torridonian rocks of Sutherland. Geol. Mag. Vol. 88. No. 34p. 215. FISE,H4/1.19474Fine-grained alluvial deposits and their effects on neat ssipi river activity. Vicksburg. 70 , _ __ A. FRAZISRA 9SAN1/41961. point bar deposits,old river locksite, Louisiana. Trans. Gulf Cst..,Ass.-geol. Soc. Vol.11. p. 21. FREUNDIACE & JULIU8BERGER.1935. '?uicksand as a thixotropic system. Trans. Faraday Soc. Vol. 31. Pt. 5. FRIEDbiAN, G. 1958. 1958. Determinat ion of sieve-size distribution • from thin section- data for 9er3imentery petrology .studies. J. Geol. Vol. 68. p. 394. FRIEND, .1961. The Devonian Stratigraphy of North & Central VestSpitzbergen. Proc. Yorks. Geol. Soc. Vol. 33, Pt* No. 5.p. 77. ' GEMINI/1.1880.A fragment of Primeval Europe. Nature,London. Vol. 22. p. 407. 18976-Annual Report of the Geological Survey for 1898.E. M.S. 0. London. 1901. Summary of Progress of the Geological Survey for 1900. H. M. 3.0. London. 205

KUIENNN.P.11.1958. Sand volcanoes in the ,Carboniferous rocks of County Clare. Ireland. Quart. J. geol. Soc. Lond. Vol.113. p.441. GRABAU0.913. Principles of Stratigraphy. A. G. Seiler & Co. - New York: GRMG0RY.J.W. 1915. Moine pebbles in the Torridonian conglomerate. Geol. Mag. Vol. 42. p. 447. HARKER tA.1905, The overthrust Torridonian" rocks of the Isle of _ithumot& the -associated gneisses. :quart. J. geol. 010: Lond. 59.,p. 189. TIEMINGWAY,J. & CLARKE, A. M. 1963. Structure Of- Ripple Marks. Nature 'Loudon. Vo1.3.97. No.4857. D. 550. HOBBS,111. 04907. Earthquakes — an. 'introduction to seismic * geology.)). Appleton & Coy.. New York. ILL/NG, L.1954. Baiiaman calcareous sands. Bull. Amer. Ass.. patrol. Geol. Vol. 30. p.1. . rtiVTITCy.957* Pti3.aeomagiviitl,c1 investigations in Great Britain. , III. The. origin of the Palaeomagnet ism of the • = Torridonian sandstonas of 'Northwest Scotland. Phil. Trans. Ser. A. No. 974. Vol. esp. p.100. , ,E. ,S, tRVING4t. RUNCORN4957, PalaeoMagnet Jo investigations in Great. Britain. II. Analysis of the palaeomagnet ism 'Of the Toriidontan sandstone series of Northwest •Scotland. Phil. Trans. Ser. A. No.974. Vol. nso. p. 85. 206

JOHNSON, W.D.1901. The High Plains and their utilization. U. Geol. :Air's,. net Ann. Rept. Pt. IV. p. 601. JONM3, G. P.196% Deformed cross—stratification In the Cretaceous Bima Sandstone,Nigeria.J.sediment. Petrol. Vol. 3!1, No. R. p. C*1. KARTASHOV,L. P.1981. The facies, dynamic phases A formations of Alluvium. Izv. Acad• Bei. U. S. S. it. No.9. p. 6'7. FENDAL.M922.The Physiography of Coal swamns•Rep.Brit. Ass. for Div. Sci. Rept. of the 90th meeting. p. 49. gENNEDyIrr. '41951. Sedimentary differentiation as a factor in the Moine—Torridontan florrelation. Geol. Mag. Vol. 88. po. 4. p. 257. KINDLE,114 M.1917. Recent e fossil riprie marks. Bull. Geol. Surv. Cann. No. 25. p. 1. TLETN, G. De V.1f62. Triab sedimentation, ri time province, Canada. Bull. geol. Soc. i,Mer. Vol. '73. p..11217. XlittITIT,84.,H.19!."9. The Fountain A. Caspar formations of the Laramie basin. Univ. ?yor ing _Publ. Sal. Geology No. 1. mil. KRUMBEIN,tv. C.1934. Site 'frequency distrIbution of sediments* J. sediment. Petrol.No14. p. 65. XRUMBETN & 31010.1959. Stratigrarhy & Sedimentation. Freeman A- 'Coy.13an Francisco. 207

RURNRN,P.H.1955.Realms of Water.Oleaver-Hume Press. London. p. 282. 1958. Experiments in Geology.Trans.geol.soc.Olasg. Vol.23.110.1. LAMAKIN04141947.04 the dynamics of alluvial deposits. Dokl.Aoad.801.11.8.8,R. Tom.-67.No.l.p.65. LAPWORTH,C.1888.0n the Discovery of the Dlekellus Fauna in the Lower Cambrian Rocks of Britain. Geol. Mag. Vol. V. P.484. - LEOPOLD,L.B.& WOLMAN,M.0.1957•River Channel Patternsgiraided, Meandering & Straight.Prof.Pap.U.8.geol.Surv. 282-B p.39. MACAR,P..1948.Lea poeudonodules du Fammenian et leur origine. Ann.Soc.Geol.Belg.Vol.72.p.B.47. MACAR4.1948b.rhenoMenes intraformationals dans les couches de 1t Inkisi.Ann.Soc.Geol.Belg.Vol.72.p.B.151.------1951 Pseudonodules en terrains meubles.Ann.5oc.Geo1. Belg.V01.75.p.111. MAOAR,P.& AfiTUR,P4949.Pseudonodules et glissements sous -aquatique dans l'Emsien inferior de l'Oesling. Ann.Soc.Geol.Belg.Vol.73.p.B.121 MACCULLOCH,J.1819.A description of the Western Islands of Scotland. Edinburgh. 208

ioDOWELL,J*P*1957*he Sedimentary Petrology of the Mississagi alartzite in the Blind River area. Geological Circular fio.6.0ntario Dept. of Mines. McKES,E.D.1957a*F1ume experiments on, the production of stratification & cross-stratification*J*sediment* petrol*Vol*27*p.29* 1957b*Primary struatUres in recent sediments. Bull*Amer.Ass*Petrol*Geolaol*41.p.1704* MetEpE*D*11 WETR,0*/.1953.Terminology for Stratification & Cress-Stratification in sedimentary rocks. Bull*geol*Soo*Amer*Vol*G4*p.381* MApELIW•J*& FiLLUORE,CL*1962*Streem directions in the Lakota Formation in-the Northern Blackhills, Wyoming & South Dakota*Prof*Pap*U.S.geol.eurv* 450.-B 13 p.35. H*MENARD*1952*Deep ripple marks in the sea*J*sediment* petrol. No. 22. p. 3. MURCHISON,R*I*185660n the Relations of the Crystalline rocks of the North Highlands to the Old Red Sandstone of that region and on the Recent

Discoveries of Fossils in the formersby Mr C. Peachowith,a note by Salter*Rep*BriteAss*Adv* esi,for 1855.p*85 209 •

MURCHISON& I. & SZlladI014 1829. The Structure & Relations -of the :Deposita contained between the Primary ,Rocks & the Oolitic Series, in the North of ' Scotland. Trans. geol. Soc. Lond. Ser. 2. Vo13. , MURCHISON& I. & GRIRIR,A• 11361a. On the Altered Rocks of the . Western Islands of Scotland the North- Western & Central Highlands. c)uart. J. geol. Soc. -Lond. Vol. 17. p.171. NAUMOVA.„13.11. & R. V. 1961. Plant remains, spores, in the Torridonian Shales of Scotland. Dokl. Acad.

Sc is U. 8. Vol.141. No. 1. p.181. NMDECO,NETTIEULANDSENGLZERING CONSULTANTS.1959.River Studies,Niger & Benue.Lmsterdam. OULIANORP,I.I. N.1958. C. R. Acad. Se 1. L. 247. Geologie: sedtmentologie et geophysique. p. 313.

PEACE, B. 'HORNS, 0; GUNNew. CLOUGH,C• If I WW1, Li & TE.ALL 1907. • ,The Geological Structure of the North-West Highlands of Scotland. Memo ip of the Geologickl SurVey of Great -Britain. Glasgow.

PETTIJOHN,J.1957. Sedimentary Rocks, ' second edition. Harper & Bros. New York.

00.1•1110•11101111••••••••••••••im 1962. Palaeocurrents & Palaeogeography. Amer., i+1313. Petrol. Geol. Vol. 46. No. 8. p.1468. PREthISTER 1960. Scotland: the Northern Highlands. Brit. reg. GeoLp H.1 Si, 0. irdinblirgh• 210

RETTGEtt*R 1935. Experiments in soft rock deformation. ' Bull. Amer. Ass. Petrol. Geol. Vol. 19. p. 271. ROLPEO: D. L1960.g. fine air-heave structure from the Old Red._ Sandstone of Lanarkshire, Scotland. Geol. Mag. Vol. 97:No. 2,73..133. SCHENER/ICRN•L . e. STANTON,'N. 1963. Tilloids in the West Congo gectsynol ine. 'wart. J. Sea. Soo. Vol.119. p. 201. OCHWARZAMIER,W.1953. Cross-bedding and Grains tze in the Lower Cretaceous 'Greeniand of East /441 ia. Geol. nag. vol. 90. No. 322.. SEDthaViCiKaas & KURCH I SON, R. 1.1835. 'On the Structure & •Relations of the Deposits contained between the PritarY Rocks and the Oolitic series in the North of Scotland. Trans. Geol. Soc. Lond. Ser. 2. Vol. 3. p.125. WILLEY, R. 0.-i SHEARMAN. D. J SUTTON, J. & WATSON,J. 1963. Some underwater disturbances in the , Torridonian rooks of Skye et 'ftalisay. Geol. Mag. Vpl. 100. No. 3. p. 224. SELLEYbiti C.8 SfrEq111.ANtiii.T.1962. The experimental production of -sedimentary atruotures in quicksand. Proc. geol. Soo.Lond. Nos-1599. p.101. SELLEY,R. 0. in .preissi. The Peneeontemporaneous deformation - of heaiy mineral bands in the Torr donian Sandstone of 'North-'.nest SeOtland. Proc. '6th. Internat. Sad. Cong. El sev ter..Amsterdam. 211

8RANTSER,Y.V.1951..khe flatland fluvial alluvium of the temperate zone and its significance for the perception of the structural & formative regularitiei of alluvial rivers.Tray.Inst. geo1.11.S.S.R.VYN135,11o.55. 3HARP.R.,P.1948.Earli tertiary fanglomerate,Big Horn Mountainst Wyoming*JeGeol.Vnl.56.11ba.P.1. SHROOK,R.1.1948iSequence in layered rocks. McGraw Hill book .Copaiew'YorkipToronto & London.. S9REY4MC.1859i.Onthe- Structures produced by the currents present during the deposition of stratified rooks. The Geologiet.Vo12.p.137, STEWART,A.D.1962a*On'the Torridontan sediments of Coloneay and theie relationship to the main-outcrop' in NorthWest-Scotland.Lpool Manchr. geol. J. .

••••••■•• ••••••••••••••11111111 l962bwOreiwacke sedimentation in the Mrridonian of Colonsay & dramsaY.Geol.Mag. ' vIro1.99..No.5..P•399• ------1963.On'certain slump structures in the Torridontan Sandstones of Ipplecross.abol.• Mag.Vol.100#No.3.p..206. STEWART H.B.1955,,Contorted sediments in modern coastal .laggons explained by laboratory eperimentg. Bull.Amer;Ass.Petrol.GeolOirol.40.11,153. 212

STRIDE1A.B6190.CUrrent swept sea floors near the Southern half of Great Britain.f4uart.J.geol.Soc.Lond*Vol. .119.Pt26p175. SUTTONid.& WAT3ON,J,1951.The Pre-TorridOnien metamorphic history of the Loch Torridon & Scourie areas in the North-West Highlands' and its bearing on the chronologicalc1nSsitication of the Lewtsian. '.*?uart.J.ge01.3oc4Lond.Vol.106.p.241. 1980.3edtmentary Structures in the Eptdotic Grits of Skye. Geol. Mag. Vol. 97. No. 2. p.108. ------in press. Structures in shallow water Pre-Cambrian sediments fron North-West Scotland. Pros.6thaInternat.Sed.Cong.Elseyier.Amsterdzmu TERZAGHINK.A rECK,R.1960.Soil Mechanics in engineering & sons.New THOMPSON,W.0.1937.0riginal,structures of beaches bars & dunes.Sull.geol.Soc.Amempol.48.p.723. TOLMAN,C.P.1037.GroundwateracGraw Hill. New York & London. TURNBULLows KRINITZSKY.E.& JOHNSON.S.1950.Sedimentaim • geaOgy of the alluvial valley-of the Mississipi .river & its bearing on foundation problems. In: TRASK,P. Applied Sedimentation.J.Wiley & sons. ' New York., VAN HOUTEN.1961.Ferric.oxides in Red beds as Palaeomegnetic ' data.J.sediment.Petrol.No.2.Vol.:51.p.296. 214.

VAN LECXViLTGE,:i.dc MAGAR,P.1950.Phenomenea psaudo-tectoniques dl orogine periglacierepdans les depots'sablo- graveleux dit 'onx' et terrassesfluvialee de la region Liegeoise.Ann.geol.Soc.nelg.Vol.73.p.m.3

VAN STRAATEN,L.M.J.U.1953.Megartpplesin the Dutch sadden

Sea A in thelinsin of Arcachon,France.Geol. en Mijnb.V01.15.p.1. 1054.Compositton & Structure of recent marine sedi-lents in' the Detherlands.Leid.Geol.liededi

------1959. minor Structures of some recent littforal F noritic sediments.Geol.en Mijnb.Vol.21.p.197. ------1961.3edimentation in Tidr.1 flat areus.Journ. laberta 5oc.Petrol.Geol.Vol.9.p.903. aiMELL,114.1035.VOlumessticipo roundness of quartz particles., ' 4.0eol.Vol.43.p.2.59 WRIMORTR#0.14192t.A scale of 'grade t class terns for elastic aeuimonts.J.Gool.Vol.70.p.377. WINMERtE6 in press. Lete Pre-CeMbrian pebbly mudatones in Norrisndy,Prance.Tillite or t*lloid? Proc. NATO' AdiiiStud•Inst.Newcantle. WHITE90.1961.0011oid phenomena in the sedimentation of " 4rgillaceous rocks.J.sediment.Petrol.Vol.31. No. 4.,p, 560e 214

WOOZDVARD,101. 1913. Notes on the geology of Raasay. Trans. Edtnb. geol. Soc. Vol. 10. p.164. D..959. The formation of cross-bedding by a meandering or braided stream. J. sediment. Petrol. Vol. P9. No. 4.3). 610. WUR8TER, P. in press. Delta sedimentation in the German Keuper basin. Pro° 6th. Internat. Bed. Cong. Fa sev i or. Amsterdam. ZINGO.T.1935.Beitrag zur pahotteranalyse. Schweiz. mineralog. pettog. Mitt. Vol.15. p. 39.

4111.11111.4111111.,S1....14•11111/ .1.••••=11,11m401.N.IIIMININIWEEPIIIMPINEMM