The GEOLOGY  of  the SHIANT ISLES (HEBRIDES)

By

FREDERICK WALKER, M.A., D.Sc.
Lecturer in Geology in the University of St. Andrews.

(Communicated by Dr. ALFRED BARKER, M.A., F.R.S., F. G. S.
Read December 18th, 1929.)

CONTENTS.

I. PHYSICAL FEATURES.

The Shiant Isles form a small archipelago in the North Minch, some 5 miles south-east of the Park district of the Island of Lewis. Although grouped with Ross and Cromarty, they are 20 miles distant from Rudh Re - the nearest point of the main­land. Geologically, the islands are connected with the Trotternish district in the north of Skye, showing as their principal featurethick dolerite-sills intruded into relatively thin Mesozoic sediments. The Shiants are, in fact, the northernmost representatives of the ‘trap isles’ of the Hebrides. The group contains three comparatively large islands - Garbh Eilean (the rough island), Eilean an Tighe (the island of the house), and Eilean Mhuire (Mary 1sle). Of these the first two are the largest, both being about a mile in length, with a maximum breadth of about half that distance. They are connected by a shingle-beach (except during unusually high tides), one side or other of which provides a safe landing-place even in rough weather. Garbh Eilean, the northern island, rises to a height of 528 feet, and is bounded on the north and east by high cliffs.  Those of the northern face are particularly impressive, showingat one point a sheer wall 500 feet high rising out of the sea. On the north-east and east the bottom of the cliffs is buried under great scree-slopes which, being partly overgrown, form the nesting place of countless puffins. Eilean an Tighe is 410 feet in height, and is bounded on the east by similar high cliffs, together with occasional scree-slopes. Both islands consist almost entirely of a single dolerite-sill having a visible thickness of over 500 feet, and dipping south-westwards at an angle of 10° to 15°. Thus the south-western side of Garbh Eilean and the western side of Eilean an Tighe have moderately gentle dip-slopes, which are easy of access in calm weather. When a westerly wind is blowing, however, the summit of Garbh Eilean can be reached only by following a very steep route from the shingle-beach, which forms the one safe landing-place in these circumstances.

Eilean Mhuire lies about half-a-mile east of Garbh Eilean, and is surrounded by high cliffs on every side, although its height of 290 feet is inferior to that of the other two islands. Measuring seven-eighths of a mile in length by a third of a mile in breadth, it can only be scaled by landing below the grassy slopes cast of Bid na Faing, or on the rocky platform at the extreme east  of the island whence its smooth green top is reached by a steep climb up crumbling rock. Eilean Mhuire is made up of several almost horizontal dolerite-sills, separated by thin Jurassic sediments.

About half-a-mile west of Garbh Eilean is the easternmost of a string of nine small islands (some of which are mere rocks), stretching westwards and extending over a mile in length. These are named the Galtachean, and the two largest are known as Galta Mor and Galta Beag. These islands also consist of dolerite, the sill or sills of which they are composed dipping southwards at a variable angle. The southern dip-slopes of the two Galtas are grass-clad and easily scaled, although a landing is difficult to effect, even in the calmest weather. None of the Galtachean exceeds 200 feet in height, while Galta Mor, the largest, is about 300 yards long and 150 yards broad.

Two rocks named Seann Chaisteal and Sgeir Mhianuis are separated by a few yards from the extreme south of Eilean Mhuire and Eilean-an-Tighe respectively. They are both less than an acre in extent.

The larger islands have a good water-supply, and are mainly covered with rich grass ; but the group has been uninhabited for the last thirty years. In 1922 it passed into the hands of Mr. Compton Mackenzie, the present proprietor.

II. PREVIOUS LITERATURE.

Although M. Martin devotes over a page to the Shiant Isles in his ' Description of the Western Islands', the first geological account of the group is that of the indefatigable John MacCulloch in 1819. In dealing with the Shiants, MacCulloch shows his customary acute observation, and describes many of the important geological features. He recognized that the islands are formed of sheets of remarkably ophitic 'augit ' trap-rock, and records the presence of Jurassic shales and 'siliceous schist' between the sheets of Garbh Eilean, and again on Eilean Mhuire. The curious botryoidal or spherical structure of some of these sediments did not escape his attention, nor did the impressions of belemnites which prove their ' secondary' or Mesozoic age. Among the minerals found by him were natrolite, stilbite, and wavellite, this last mineral occurring in the joint-cracks of the 'siliceous schist'. He gives, too, a graphic account of the cliffs of Garbh Eilean, and of the natural arch which penetrates the eastern horn of that island. The resemblance of the geological structure of the Shiants to that of parts of Northern Skye and Raasay was also noticed by him.

After the appearance of MacCulloch's account, the Shiants were neglected by geologists for nearly sixty years, although Lord Teignmouth  gives a good general description, which was published in 1836.    It was not, however, until 1878 that .J. W. Judd’s classic paper on the Secondary Rocks of Scotland appeared in print. In this publication he gives a short description of the Jurassic strata of the Shiants, but does not mention on which islands they are found, although a section illustrating their occurrence on Garbh Eilean is printed.  He was successful in finding Anintonites murchisonae Sowerby and A. corrugatus Sowerby, besides the hollow casts of belemnites mentioned by MacCulloch, and on these grounds assigned the strata to the lowest part of the Inferior Oolite.

Shortly afterwards, M. F. Heddle visited the islands, and in 1884 his account of their geology appeared in print. His description of the group shows few advances oil that of Mac­Culloch, but he confirms the presence of wavellite, and also records analcite and mesolite as occurring in the dolerite. He appears to regard the wavellite as permeating the spherules developed in the Jurassic strata. A fairly extensive account of the glaciation of the group is given, his main conclusion being that the ice-flow was from the west. He was unable to land on Eilean Mhuire owing to bad weather ; but he accomplished this on a subsequent visit, which led to a more detailed account including that island.   In this later account he records nepheline from a ‘dolerite’ occurring in a sea-cave at the eastern end of Eilean Mhuire - a significant and suggestive discovery.

In 1885 Judd's well-known paper on the 'Tertiary & Older Peridotites of Scotland ' appeared.  He gives in it petrographical descriptions of the dolerites (probably from Garbh Eilean) com­menting upon the beautifully developed ophitic structure of the augite, He also records for the first time the presence of ultra-basic rocks on the islands ; but, as most of his specimens were collected from fallen blocks, he was unable to make out the mutual relationship of the two types. All varieties from felspathic dolerite to olivine-rock were found ; but, even in the most basic specimens, felspar and augite were present. Thus the term 'dunite' applied to them by Judd is a little misleading. Slightly less basic varieties were named picrite by him. He comments upon the yellow colour of the olivine in thin section, and considers it to be highly ferriferous. The 'dunite', picrite, and the augite from the dolerite are figured in plates. Like Heddle, he was prevented from landing on Eilean Mhuire by bad weather.

Sir Archibald Geikie was the next geologist to turn his attention to the islands, and lie gives a good description of their geology in his '’Ancient Volcanoes of Great Britain’,  based on three visits to the group. He has several new observations to make on Eilean Mhuire. Specimens from one of the thick sills of this island were submitted to Dr. Alfred Harker, who found that they were coarse gabbros without olivine and containing a peculiar pleochroic augite. Geikie notes the occurrence of schlieren of light and dark rock on Eilean Mhuire, and finds that the thickest gabbro-sill is intruded by veins and sheets of olivine-basalt ranging from an inch to 20 feet in thickness. These basaltic intrusions were found to show marked chilling against the gabbro. Specimens of the basalt are also described by Dr. Harker. Geikie was unable to find the ultrabasic rocks recorded by Judd, but suggests with characteristic perspicacity that they have accumulated at the base of some of the sills, 'like the picrite in the Bathgate diabase'.

Dr. Harker, who makes several references to the Shiant Isles in his Geological Survey Memoir on the 'Tertiary Igneous Rocks of Skye', visited the group five years ago, but was unfortunately unable to examine its geology, owing to illness. He succeeded in securing, however, a specimen from the eastern end of Eilean Mhuire, which proved to be a most interesting alkali-syenite. On learning that I thought of making a geological survey of the islands, Dr. Harker with great kindness placed this and other important information at my disposal, and afforded me every encouragement in the work.

III. SUMMARY OF THE GEOLOGICAL STRUCTURE.

The geological structure of Garbh Eilean and Eilean an Tighe is exceedingly simple, the great bulk of both the islands being made up of a single sill of crinanite, which passes gradually into picrite as the base is reached. This sill has a visible thickness of at least 500 feet, and is therefore one of the largest in Scotland. The picrite base of the sill is seen only at the south-east of Garbh Eilean, but its presence all along the eastern and north-eastern coast of the island may be proved by the great numbers of boulders of that rock in the scree-slopes. The western extremity of Garbh Eilean is let down by a large normal fault running north and south across the island, and clearly visible on the northern cliff-face where the rock adjacent to it shows marked brecciation. This fault throws the picrite base of the sill far below sea-level, and gives rise, by denudation, to a very marked hollow, which can be traced across the island from north to south. A similar fault, running north-east and south-west, with a south-westward down­throw, probably occurs between Garbh Eilean and Eilean an Tighe, for the picrite base of the sill is not exposed at all in the latter island. If no dislocation were present, the picrite would certainly be found at the southern end of the shingle-beach.

The great crinanite-picrite sill rests on some 30 feet of Upper Liassic (Whitbian) shales with siliceous bands, exposed on the north-cast of Garbh Eilean, and these are followed downwards by another crinanite-sill of which only the top is seen. A small patch of similar sediments occurs on the north-western shore of Eilean an Tighe. These strata do not exceed 15 feet in thickness, and have probably been caught up by the sill, for they end abruptly on the north against dolerite without any trace of a faulted junction.

The Galtachean appear to be made up of a sill of crinanite petrographically identical with that of Garbh Eilean and Eilean an Tighe, with which it is, possibly, to be correlated. The rock of Galta Mor contains, however, numerous thick bands of alkaline pegmatite which do not occur to the same extent on the two larger islands. While the dip of the 500-foot sill is generally south-westward, and never exceeds 15°, that of the Galtachean is southward, being often much greater and exceedingly variable, as is shown by the disposition of the columns.

The structure of Eilean Mhuire is slightly more complex. The summit of the island is made up of a crinanite-sill very similar to .that seen below the Jurassic strata of Garbh Eilean. This sill rests on at least 60 feet of almost horizontal Upper Liassic shales and siliceous strata, again very much like those of Garbh Eilean.  The Jurassic beds cover a large part of the top of the island, and are well exposed in the highest parts of the cliffs. They are underlaid in their turn by a dolerite-sill which is crinanitic, or even teschenitic, in places, and has been split by numerous basaltic bands of distinctly later age. This sill contains numerous schlieren of alkaline composition towards the eastern extremity of the island. About 200 yards south-west of the summit two north-and-south faults with downthrow to the cast are clearly seen on the cliff-face, and the Jurassic strata east of them dip northwards at a much higher angle than elsewhere (25°).

IV. JURASSIC STRATA.

Sedimentary strata of Jurassic age occur on the shores of Garbh Eilean and Eilean an Tighe, and along the tops of the cliffs of Eilean Mhuire (see map). Those of Garbh Eilean and Eilean Mhuire represent partings between the crinanite-sills; but on Eilean an Tighe the sediments appear to have been floated up from the bottom of the sill, partly by an ebullition of gas generated by reaction with the magma. Traces of this rush of gas-bubbles are now seen in the numerous spherical vesicles filled by radiating zeolites which occur in the igneous rock near the contact.

The strata consist mainly of siliceous mudstones, usually baked lay the sills into a grey flinty porcellanous rock with conchoidal fracture, designated by MacCulloch ' siliceous schist'. Shales more or less hardened by contact-alteration also occur on Garbh Eilean and Eilean Mhuire, mainly towards the bottom of the sequence (although there is no regular order of deposition), and occasionally show the botryoidal or pseudo-pisolitic structure described by MacCulloch.  This structure appears to be a weathering phenomenon, since the spheres are only conspicuous on weathered surfaces or in wave-worn fragments. They run up to 0.6 cm. in diameter, are sometimes separated by greenish saponitic material, and are structureless under the microscope.

The strata attain their maximum. thickness of over 60 feet on Eilean Mhuire. On Garbh Eilean they are about 30 feet thick and only 15 feet on Eilean an Tighe.

MacCulloch and Heddle both record the occurrence of wavellite in the strata of Garbh Eilean, and a careful research which I made near Sgeirean a' Bhaigh - the locality given by Heddle - yielded a mineral that agreed exactly with both descriptions. Optical and chemical tests, however, showed it to be a zeolite (probably stilbite) with mean R.I. = 1.498 ± .003.

The siliceous strata of Eilean an Tighe yielded no fossils, except obscure impressions of the guards of belemnites, but in arenaceous shale on Garbh Eilean there was found an ammonite determined by Mr. J. W. Tutcher. as Elegantuliceras cf. elegantulum (Young & Bird).

The strata of Eilean Mhuire are more fossiliferous, and the following ammonites from shales in 'scrapes' 200 yards west of the summit were determined by the late Mr. S. S. Buckman : Anguidactylites anguiformis S. Buckman, Orthodactylites directus S. Buckman, and Elegantuliceras (near to elegantulum Young & Bird) sp.

The following fossils from shales north of Sgeir na Ruideag and siliceous pebbles on the shore were later determined by Mr. Tutcher: Inoceramus dubius Sowerby (Upper Lias), Tenui­dactylites cf. tenuicostatus Young & Bird, sp. Harpoceratan, and ,Dactylioceras cf. annulatum G. SowerbY, sp. Harpoceratan.

The fossils enumerated above have suffered more or less severe crushing, but Mr. Buckman and Mr. Tutcher agreed that they all represent a very low position in the Whitbian (Upper Lias), and the strata of Garbh Eilean may, therefore, be correlated with the strikingly similar beds of Eilean Mhuire. Judd assigned the strata of Garbh Eilean (presumably) to the lowest part of the Inferior Oolite, but the late Mr. Buckman pointed out in a private communication to me that

"if Judd only saw the Harpocerate (Elegantuliceras) it would not be surprising if he took that for an Inferior Oolite specimen: it has considerable likeness to Inferior Oolite forms."

Owing to the pronounced contact-alteration of the Shiant strata, it is difficult to compare them satisfactorily with the Upper Lias of neighbouring districts ; but they seem to resemble the Upper Liassic beds in North-Eastern Skye, near the entrance of Portree Bay.   Here the strata consist of dark micaceous shales with harder bands, containing, among other fossils, Elegantuliceras  elegantulum (Young & Bird), a species found on both Garbh Eilean and Eilean an Tighe.

V. IGNEOUS GEOLOGY.

[1] The Crinanite-Picrite Sill of Garbh Eilean and Eilean an Tighe.

(a) Field Characters.

The great sill which forms the major part of Garbh Eilean and Eilean an Tighe is the most prominent feature of the Shiant Isles. The dip of 10° to 15° swings round in direction from south at the western extremity of Garbh Eilean to almost due west in the southern portion of Eilean an Tighe. The scarp-slopes north and cast of these islands present nearly vertical cliffs or exceed­ingly steep rockyslopes towards the sea; but the bottom of both .is frequently covered by vast accumulations of big boulders which havefallen from higher parts of the sill. This is to be seen par­ticularly along the shores of the north-cast of Garbh Eilean where the Jurassic strata occur, with the result that the lower junction of the sill is completely buried. Elsewhere the lower contact is below the sea, while the upper junction of the sill has been removed by denudation. The thin Jurassic strata north-west of Eilean an Tighe are seen in contact with the crinanite; but these beds have probably been floated up by the sill, while the junction between them and the igneous rock is abnormal, as will be explained below.

Exposures on the dip-slopes of the islands are naturally less perfect than on the scarp-slopes, but rocky projections rise up through the ,grass very frequently, and there are distinct walls of overhanging columns on the north-western slopes of Eilean an Tighe and the south-western slopes of Garbh Eilean. Along the western shore-line the exposures are almost uninterruptedly perfect.

Columnar structure is well developed throughout the sill, and attracted the attention of all the previous observers. The columns do not display the perfect symmetry of those of Staffa, but have a certain rugged grandeur of their own, being both greater in diameter and very much longer. Though generally quite straight, they show slight curvature near the summit of Eilean an Tighe and on the northern face of Garbh Eilean west of Glaic na Crotha. Round the latter locality some of the columns reach a height. of 350 feet at least, while their average diameter exceeds 5 feet. They are usually hexagonal in form, but pentagonal and even heptagonal examples are frequent. The great scree-slopes on the north-eastern face of Garbh Eilean are formed of the broken fragments of such columns, which have cleaved along joints parallel to the plane of intrusion. The boulders in most cases thus retain the characteristic polygonal outline of the original columns. A white zeolitic substance sometimes fills the joint-cracks between the columns, and has been identified as mesolite by Heddle.

Besides the ordinary columnar jointing, there is a series of vertical major joints running, for the greater part, north-north-east and south- south-west. They are well seen on the cliff-face west of Glaic na Crotha, where the rock has broken off in huge parallel slabs, instead of column by column. On the south-western slope of Garbh Eilean, which is very steep, there is a similar series of major joints, and also a tendency for the columns to overhang and topple into the sea. This has caused some of the major joints to open out, and they now form straight gaping fissures in the grassy slopes, partly choked by soil and vegetation, and with slab-like walls .

Landslip fissure on the south-west side of Garbh Eilean, caused by the opening of a major joint.

The absence of tachylitic selvages and slicken­siding on the sides of these fissures shows that they do not find their origin in the rapid weathering of a basic dyke or of crush-rock along a line of movement. On the south-western shore of Garbh Eilean are numerous caves and fissures running into the rock in a northerly direction, and these are probably due to wave erosion along lines of weakness formed by such major joints. The north-and-south fault which lets down the western part of Garbh Eilean, and the hypothetical fault between Garbh Eilean and Eilean an Tighe letting down the latter island, have already been mentioned, and require little further comment. The hade of the former fault is approximately 30° but no estimate can be formed of its throw.

Microscopic examination of sections taken from all parts of the sill shows that, apart from local segregations. of pegmatitic phases, the composition of the rock does not vary laterally, but is solely a function of the height of the locality above the base of the sill.  Hence a  description of the various types encountered when passing from bottom to top of the most complete vertical section available will apply in fact to any portion of the intrusion.

The best and most complete section is seen on the southern face of Gharb Eilean, which rises with great steepness from the shingle-beach, has numerous excellent exposures, and is accessible to even moderate climbers at nearly every point.  The lowest visible part of the sill is exposed east of the shingle­-beach, where great columns rise out of the sea at high tide. The rock here is a picrite, which has almost the appearance of a dunite in the hand-specimen, so numerous and fresh are the grains of .yellowish-green olivine; but examination with a hand-lens reveals the presence of white plagioclase and plates of black poecilitic augite surrounding the orthosilicate. Where the rock is exposed to wave-action, there is no weathered crust, while elsewhere it is covered by a very thin dark skin, or by lichen. The differential resistance to erosion of the plagioclase and the augite causes pronounced carious weathering.   The picrite is of medium grade, and is well exposed along the shore to a point 300 yards north of the shingle-beach, where it becomes buried under debris from which it does not emerge. It is surprising that acute observers, such as MacCulloch and Heddle, should have failed to detect the presence of this striking ultrabasic rock. Veins of coarse pegmatite composed mainly of augite and felspar traverse the ultrabasic rock nearly, everywhere, assuming a more or less vertical trend, and vary from 1 to 10 inches in breadth.

Columns of picrite on the south-eastern shore of Garbh Eilean, traversed by four pegmatite-veins.

As we pass upwards from the shingle-beach, the rock is seen to assume a grey colour on the weathered crust, while examination of fractured surfaces shows that olivine becomes less abundant with every upward step, while augite and plagioclase show a complementary increase. There is, in fact, a gradual transition upwards from picrite at the beach, through dolerites rich in olivine (15 to 45 feet above the beach), into normal olivine-dolerite (or, more correctly,crinanite) above 45 feet. As we go still higher, the olivine falls off steadily, but much more gradually, and is still quite recognizable in hand-specimens from the highest visible portions of the sill. Not a trace of schlieren or other discontinuous variation is to be seen in the very complete exposures available.  Continuing upwards from basic olivine-dolerite into crinanite, we see that the augite begins to exhibit marked ophitic structure, which becomes more pronounced the higher we ascend. In the upper parts of the sill huge ophitic plates measuring 3 inches across are of no uncommon occurrence, the crystals of felspar and of olivine being minute in comparison. Cavities filled with radiating fibres of zeolites are rare. All the other exposures of the sill in situ consist of crinanite like that above described, showing no change other than a slight upward decrease in olivine.

The veins of pegmatite cutting the ultrabasic rock, but absent from the olivine-dolerite and crinanite, of the type vertical section, have been already mentioned; there are, however, others occurring in the crinanite on the north-western shore of Eilean an Tighe near the rock named Sgeirean Mol na h-Athadh. These latter veins are similar in appearance to those that cut the picrite, but, follow a more sinuous course, and are confined for the greater part to the immediate neighbourhood of the Jurassic strata. Neither the normal crinanite nor the pegmatite shows chilling against the sediments. At this locality spherical cavities filled with acicular radiating zeolites are developed in strong force.

All the rocks of the sill are exceedingly fresh ; but it isa matter of some difficulty to collect satisfactory hand-specimens, for the rock is seamed with irregular cracks in every direction, and tends to split along them. So numerous are these cracks that the preparation of thin sections is far from easy, the slice tending to break up, especially during covering.

(b) Petrography.

Picrite (Although the rock described below contains over 20 per cent. of felspar, the name 'picrite' has been given to it as the most suitable. The rock is distinctly more basic than 'kylite', which appears to be the only other alternative name.)

 Under the microscope the picrite presents a very striking appearance. Rounded grains of olivine make up the bulk of the rock (50 to 75 per cent.), and, except for occasional anasto­mosing veins of serpentine and magnetite, the mineral is unaltered. The grains have an average diameter of 0.5 min., but vary considerably in size, and sometimes form large composite groups, though these are not common. In sections of normal thickness the mineral is practically colourless, but may, in thicker slices, show faint greenish-brown tints. The mineral is optically negative, and has an axial angle of nearly 90°. The average refractive index was found by oils to be in the neighbourhood of 1.68, while measured by the Berek compensator gave the value 0.032. These results indicate a normal olivine containing 11 to 13 per cent. of ferrous oxide. The norm of the analysed rock confirms this determination.  Numerous fluid inclusions are visible in the olivine, and occasional rounded grains of iron-ore (usually about 0.05 mm in diameter), which is probably magnetite, as no chromium was recorded in the analysis.

The olivine is enclosed poecilitically by large crystals of augite and basic plagioclase, the felspar being the more abundant.  The pyroxene is purplish-brown, with very feeble pleochroism; is 0.024 by the Berek compensator, while the extinction­angle Z /\ c is 44°.  This augite forms large crystals measuring up to 1 cm in length, which show idiomorphic boundaries towards the felspar.

The plagioclase-crystals are equal in size to those of the pyroxene, and, although as a rule perfectly fresh, may be partly replaced by serpentinous matter derived from the incipient decomposition of the olivine. The twin lamellae extinguish symmetrically at an-les well over 40°, and β was found by oils to be 1.575 ± .00,5, which indicates bytownite (Ab_l An₄).

Except for a few rounded grains of magnetite, and small needles of apatite, the only other constituents are doubly refracting zeolites, which form thin veinlets in the rock, and will be considered with the pegmatite-veins.

Although the average section contains about 60 per cent. Of olivine, the proportion may increase locally up to 75 per cent., and this more basic variety is probably that designated dunite by Judd, the resemblance to that rock ill the hand-specimen being quite striking. A complete analysis of the average picrite, together with the mode, is given below (p. 371).

Olivine-dolerite -The picrite gradually gives place upwards to a basic olivine-dolerite containing on an average about 30 per cent. of olivine, the transition from this rock to picrite being complete in about 10 feet. This decrease in the proportion of theorthosilicate is accompanied by a striking change in the habit of the augite and the felspar.

The pyroxene in the dolerite assumes a very marked ophitic habit, and encloses numerouslaths of the plagioclase, but the size of the crystals and their optical properties remain unchanged, apart from a tendency towards zonary banding. The felspar, however, is considerably less calcic, the symmetrical extinction-­angles of the albite-lamellae corresponding with a basic labradorite (Ab₁An₃). The laths are elongated and much smaller than those in the picrite, rarely exceeding 0.5 mm. in length.

In the case of the olivine no change is seen in the optical properties, and it is often enclosed poecilitically by augite; but the crystals vary greatly in size, although they always retain their rounded outline.The largest and smallest grains measure respectively 3 mm. and 0.05 mm. in length, and the mineral still contains both fluid inclusions and rounded grains of magnetite. Some crystals of the iron-ore, which is probably titaniferous, have, however, assumed a skeletal habit.

The only other constituents of the rock are inconspicuous needles of apatite, a few patches of interstitial analcite, and still rarer crystals of reddish-brown barkevikitic amphibole, usually associated with the zeolite or the iron-ore.

While the rock is ordinarily very fresh, there may be slight alteration of the olivine to magnetite, serpentine, and iddingsite. A complete analysis and mode of this transition type between picrite and crinanite is tabulated below (p. 371). The rock seems to correspond fairly well with the published description of the type named ‘kylite' by Dr. G. W. Tyrrell, but is somewhat less alkaline, and contains zeolites instead of nepheline.

Crinanite -The picrite at the south of Garbh Eilean to pass into olivine-dolerite 10 feet above the shingle-beach, and the latter rock persists upwards with a steadily diminishing proportion of olivine for 75 feet, above which level it is best described as 'crinanite'. This crinanite extends, with a very gradual upward decrease in the proportion of olivine, to the highest visible part of the sill, and is the only type encountered on Eilean an Tighe. The lower part of the crinanite contains 12 to 16 per cent. of olivine of very different appearance from that seen in the olivine-dolerite and picrite. In the crinanite the mineral is often decidedly ophitic, enclosing numerous laths of plagioclase, while widely separated fragments have the same optical orientation.

Crinanite, 125 feet higher in the sill than the rock at the base of the cliffs at the extreme south of Garbh Eilean. The mineralogical composition of the rock is similar to that of the rock at 70 feet, but the olivine at the bottom of the field is distinctly ophitic and strongly coloured, probably owing to incipient alteration. Magnification approximately X 13.

Some crystals, too, are greenish-brown. and it was thought at first that they were highly ferriferous, and a variety of hyalosiderite, if not of fayalite. The coloured mineral is opticallynegative, with an axial angle of nearly 90°; but measured by the Berek compensator proved to be 0.032, which is exactly the value obtained for the colourless mineral of the picrite and the olivine-dolerite, and indicates an olivine of quite normal character. The colour, then, is not due to a high content of ferrous oxide, but probably to incipient general serpen­tinization. Some crystals are colourless at one end, while the other is a distinct greenish-brown. In view of the high titania content of the rock (2.86 per cent.) it is just possible that the coloured olivine may be a variety of titanolivine, such as that described by Lady McRobert from the analcite-dolerite of Penal Heugh; but this is unlikely, as numerous Tertiary basalts and dolerites, with even higher proportions of titania, contain perfectly normal olivine. The percentage of olivine falls off to 8 in the upperparts of the sill, and occasionally the serpentinization is more advanced there (see illustration below) ; but the mineral is never completely altered. It is invariably enclosed by augite when the two minerals come into contact.

Crinanite, 400 feet higher in the sill than the rock at the base of the cliffs at the extreme south of Garbh Eilean. The mineralogical composition is unchanged, except for the appearance of skeletal ilmenite-crystals in the lower half of the field; but the olivine is loss altered than in the rock at 250 feet. Magification approximately X 13.

The ophitic relationship between plagioclase and augite is even more pronounced than in the olivine-dolerite. In some cases the ophitic '’clusters’ of pyroxene measure 6 to 7 cm. in length, and frequently occupy the whole of a section (2 X 3 cm.). These pyroxene 'clusters' scarcely ever exhibit crystal boundaries, but often show well-marked zoning, changing from a pale purplish-­brown at the centre to a strong purple-madder or reddish-brown at the margin. Where outlying portions of the clusters come into contact with acid plagioclase or zeolites, the aegirine molecule enters into their composition, and they assume a greenish tinge. The maximum extinction angle of the titanaugite is 43° (Z^c), and when the mineral is viewed under the highest powers, the fluid inclusions figured by Judd become visible.

The plagioclase has the same habit as in the underlying olivine­-dolerite, but the symmetrical extinction-angles of 35° and the mean refractive index by oils (1.565 ± .003) indicate a medium labra­dorite (Ab₂ An₃) . A very few phenocrysts of plagioclase with the above properties may also occur. Zoning is always present, but varies in amount, oligoclase (Ab₄ An₁) being the most acid variety detected. Orthoclase was riot found, a fact which is not sur­prising, when the small proportion of potash in the rock is considered (0.40 per cent.). Analcite and natrolite are, however, quite common, and fill many of the interspaces between the felspars outside the pyroxene clusters. The analcite is generally fresh, but sometimes shows alteration to brown decomposition­ products, and in many cases possesses stronger anomalous double refraction than usual. It presents idiomorpbic boundaries towards the natrolite when the two minerals come into contact. The natrolite occurs in radiating fibres with positive elongation, and double refraction slightly exceeding that of quartz. Analcite is usually the more abundant of the two zeolites.

The iron ore of the crinanite appears to be ilmenite, which is both conspicuous and abundant, forming large skeletal crystals up to 3 mm. across ; and small needles ofapatite are ubiquitous.

The ophitic structure of the augite in the crinanite seems com­parable with the ‘ophimottling' described by Prof. E. B. Bailey in the early basic cone-sheets of Mull. There is the same zoning of the augite, refusal of the mineral to exhibit crystal boundaries, and confinement of the olivine to areas between the augite-felspar clusters. The felspars within the clusters are less zonal than those in the interspaces, and it is probable that here, too, ' the relatively early date of the augite is obscured by its almost complete refusal to show crystal boundaries'. The crystallization of the olivine is also thought to be coeval with that of the felspar, despite its ophitic habit, and the significance of these conclusions willbe discussed when the history of the cooling of the intrusion is. considered below.

Eilean an Tighe.  Crinanite practically identical with that of Garbh Eilean makes up almost the whole of Eilean an Tighe, but the pyroxene occasionally shows minor differences. Sections from the summit of the island contain an almost colourless augite, while others from a lower position in the sill show a strongly coloured zonal mineral, which is pleochroic in tints of reddish­-brown and purple. Greenish soda-pyroxene is, on the whole, commoner than on Garbh Eilean, and in places may be observed a fibrous zeolite towards which the analcite is idiomorphic. It is brownish in colour, has lower double refraction and a higher refractive index than natrolite, and is probably stilbite.

A specimen of the crinanite taken from the southern face of Garbh Eilean, 125 feet above the shingle-beach, has been analysed by Mr. E. G. Radley, and the result is tabulated below (p. 371), together with the mode of the rock.

Pegmatite veins in picrite  The pegmatite-veins which. traverse the picrite of the sill exhibit great variety, both in texture and in mineralogical composition. In a single vein several distinct rock-types may be encountered, which presumably represent infillings at different stages during the crystallization of the sill, for there is nowhere any trace of chilling of one variety against another. The felspathic portion of all the veins has suffered more or less decomposition through hydrothermal processes.

The first infilling, which occupies the margins of some of the large veins, crystallized out as a coarse-grained teschenitic rock rich in titanaugite. In this modification the pyroxene is idio­morphic, but has the same optical properties as that of the crinanite. It builds large prisms ranging up to 5 mm. in diameter and 2 cm. in length. These are accompanied by skeletal crystals of ilmenite, often 0.5 cm. across, which have reacted with the magma, to form flakes of biotite, and by a few elongated crystals of serpentinized olivine.   The felspathic portion of the rock has been completely replaced by zeolites, among which analcite and natrolite are conspicuous, the former being invariably the first to crystallize. Apatite is a conspicuous accessory, forming needles which measure up to 1 mm. in length. This variety of infilling, before it underwent zeolitization, must have been very similar to the veins, described below, near the Jurassic strata of Eilean an Tighe.

The second variety to be described sometimes occupies the centre of the composite veins, along with the third. It is clearly of later date than the first variety, is distinctly finer in grain, and poorer in dark minerals. Augite similar to that of the first type is the most conspicuous constituent. Sometimes, however, it assumes a decidedly pink tinge, and shows zoning with well-developed hour­glass structure.

This pyroxene comprises about 20 per. cent. of the rock ; it builds elongated ideomorphic prisms measuring up to 3 mm. in length and 0.8 mm. in diameter, the crystals sometimes radiating outwards from a common termination. Rare elongated prisms of olivine also occur, but have generally undergone almost complete alteration to serpentine. Skeletal crystals of ilmenite constitute over 5 per cent. of the rock, and are exceedingly conspicuous. The felspar is mainly plagioclase (andesine-labradorite), frequently showing pericline twinning, in tables up to 2 min. in length, but a good deal of soda-orthoclase is also present. Clear analcite is abundant, and frequently replaces felspar, the alteration being, however, much less complete than in the previous variety. Apatite is an abundant, but not very conspicuous, accessory.

The third variety is still more felsic, and probably of later infilling, than the second. Augite similar in habit to that in the preceding modification, but considerably paler in colour, is the chief mafic constituent. Occasional idiomorphic crystals of olivine, generally serpentinized, are to be seen, as well as small crystals of titaniferous magnetite. The felspar is recognisable only as ‘ghosts’ up to 5 mm. in length but was probably acid plagioclase : it has been completely replaced by analcite and natrolite, which also occurs in interstitial areas; the former mineral, as always, being the first to crystallize.  Natrolite has also replaced analcite in some cases, and occurs as well defined icositetrahedra composed of radiating fibres.

There are, however, several other exceedingly alkaline con­stituents in the rock. Characteristic hexagonal or rectangular pseudomorphs after nepheline are of frequent occurrence, and occasionally a scrap of the fresh mineral with its straight extinction, low double refraction, and negative sign, may be observed. The mineral varies in its relations towards the felspar, but is generally idiomorphic. There are also small scraps of soda-bearing ferromagnesian minerals, such as biotite, aegirine, barkevikite, and arfvedsonite, which appear to have been formed by reaction of the zeolitic residuum with the iron-ore, as they are commonly associated with the latter. Large needles of apatite are usually very conspicuous.

The zeolitization which all the infillings have undergone probably occurred during the last stage in the crystallization of the crinanite, when the aqueous residuum was expelled through the picrite, using the pegmatite-veins as channels. This will be discussed later, however, when the differentiation of the sill is dealt with. Traces of this expulsion may be noticed in the thin zeolitic veins so often seen in sections of the picrite, and seldom measuring over 0.3 mm. across, These have been forced through the ultrabasic rock under pressure, since olivine-crystals are often seen to have broken in two, one half remaining on either side of the vein. Occasionally, in the neighbourhood of these veinlets, the picrite is seen to contain pyrites, and the presence of rare ,crystals of reddish-brown soda-hornblende may be explained by reaction between the sodic magma and olivine.

The profound hydrothermal alteration which these pegmatitic modifications have undergone renders the measurement of their percentage mineral composition unprofitable. They are all ex­tremely felsic in character, the dark minerals varying from 30 to 10 per cent.

Pegmatite-veins in the crinanite of Eilean an Tighe.  Numerous pegmatite-veins, usually not more than 2 inches across, pursue sinuous courses through the crinanite of Eilean an Tighe, adjacent to the Jurassic beds in the bay south of Sgeirean Mol na h-Athadh. These veins are coarse in texture, containing large prisms of augite which reach a length of 2 cm. or more.   Under the microscope this mineral is the most conspicuous constituent of the rock. It is subophitic towards the equally large plagioclase-laths, and is purplish-brown in colour like the pyroxene of the crinanite. The pleochroism is usually quite distinct, the tints being: X purple, Y greenish-brown, Z brownish­-purple. Occasionally, a small content of the aegirine-molecule imparts a greenish tinge to this pyroxene, and traces may be seen of the early crystallization of an almost colourless diopsidic pyroxene presenting sharp boundaries towards the surrounding pink variety; both, however, have the same optical orientation.  Elongated crystals of fresh, or partly serpentinized, olivine are also conspicuous, and measure  up to 1 cm. in length, but are surpassed in size by abundant crystals of skeletal ilmenite.

The felspar occurs in long laths which may be as calcic as acid labradorite (Ab₁ An₁) at the centre, but become progressively acid towards the margin, which is usually oligoclase. This felspar is replaced by analcite along many of the cracks. The interspaces between the felspars are filled partly by clear isotropic analcite, and partly by a dark brownish indeterminable mesostasis which probably consists largely of devitrified glass, and is extremely sporadic in its occurrence: it contains microlites of felspar, and appears also to be rich in iron-oxide. Natrolite and analcite are rare constituents occupying spherical cavities in the mesostasis; while apatite is an abundant accessory. Both the pegmatite and the crinanite show well-defined, unchilled boundaries against the adjacent Jurassic strata. In the neighbourhood of the pegmatite­-veins the crinanite is spotted with spherical cavities up to half an inch in diameter filled with radiating fibres of natrolite, while close to the junction with the Jurassic strata the olivine assumes a definitely granular habit.