CHAPTER XLIII

THE BOSSES AND SHEETS OF GABBRO

Petrography of the Rocks--Relations of the Gabbros to the other members of the Volcanic series--Description of the Gabbro districts--Skye

In singular contrast to the nearly flat basalts of the plateaux, another series of rocks rises high and abruptly above these tablelands into groups of dome-shaped, conical, spiry, and rugged hills. It is these heights which, more than any other feature, relieve the monotony of the wide areas of almost horizontal stratification so characteristic of the volcanic region of the north-west. Their geological structure and history are much less obvious than those of the bedded basalts. Their mountainous forms at once suggest a wholly different origin. Some portions of them have even been compared with the oldest or Archæan rocks.[325] That they are really portions of the Tertiary volcanic series, and that they reveal a wholly distinct phase in the history of volcanic action, is now frankly admitted. Whether we regard them from the petrographical or structural point of view, they naturally arrange themselves into two well-defined groups. Of these one consists of highly basic compounds, of which olivine-gabbro is the most prominent. The other comprises numerous varieties--granite, granophyre, felsite, quartz-porphyry, pitchstone and others--all of them being more or less decidedly acid, and some of them markedly so. For reasons which will appear in the sequel, the former group must be considered as the older of the two, and it will therefore be described first.

[Footnote 325: This was my own first impression, when I began, as a boy, to ramble among them. The remarkable resemblance of some parts of them to ancient gneisses will be afterwards dwelt upon. Macculloch had correctly grouped them with the other overlying rocks, and this conclusion was afterwards confirmed by Prof. Zirkel.]

i. PETROGRAPHY OF THE GABBRO AREAS

Since the publications of Macculloch, the occurrence of beautiful varieties of highly basic rocks among the igneous masses of the Western Isles has been familiar to geologists. They were named by him "hypersthene rock" and "augite rock,"[326] names which continued in use until 1871, when my friend Professor Zirkel published the results of his tour through the West of Scotland, and showed that the rocks in question were mostly true gabbros.[327] Since his observations were published some of these rocks have formed the subject of important papers by Professor Judd.[328]

[Footnote 326: _Western Islands_, vol. i. pp. 385, 484.]

[Footnote 327: _Zeitschrift. Deutsch. Geol. Gesellsch._ xxiii. (1871), p. 1.]

[Footnote 328: _Quart. Jour. Geol. Soc._ xli. (1885), p. 354; xlii. (1886), p. 49.]

The general petrographical characters of the gabbro areas of Western Scotland may be summarized as follows:--A very considerable variety of petrological structure and chemical composition is observable among the rocks. At the one end of the series are compounds of plagioclase and augite, which, though wanting in olivine, have the general structure and habit of dolerites. At the other end are mixtures wherein felspar is scarce or absent, and where olivine becomes the chief constituent. Between these two extremes are many intermediate grades, of which the most important are those containing the variety of augite known as diallage and also olivine. These are the olivine-gabbros, which form so marked a feature in the central parts of the great basic bosses. That some of these varieties of rock pass into each other cannot be doubted. Their distinctive composition and structure appear to have been largely determined by their position in the eruptive mass. The outer and thinner sheets are in great measure dolerites, with little or no olivine. Coarse gabbros are abundant in the inner portions. Rocks rich in olivine, however, occur at the outer and especially the lower part of the gabbro masses of Rum and in some parts of Skye. The following leading varieties may be enumerated:--

Dolerite.--This rock varies from an exceedingly close grain (when it approaches and graduates into basalt) up to a coarse granular crystalline texture, in which the component minerals are distinctly visible to the naked eye. An average sample is found to consist of plagioclase, usually lath-shaped, and crystals or grains of augite with or without olivine. Under the microscope, the different varieties are distinguished by the presence of more or less distinct ophitic structure, the felspar being enveloped in the augite. For the most part they are holocrystalline, but occasionally show traces of a glassy base. Ilmenite is not infrequent, with its characteristic turbid decomposition product (leucoxene). In other cases, the iron-ore is probably magnetite. Between the dolerites and gabbros no line of demarcation can be drawn in the field, nor can a much more satisfactory limitation be made even with the aid of the microscope. As a rule, the thickest and largest intrusive masses or bosses are gabbro, those of less size are dolerite, while the smallest (and sometimes the edges of the others) assume externally the aspect of basalts.

Gabbro.--Under this term I arrange, as proposed by Professor Judd, all the coarse-grained granitoid basic rocks of the region without reference to the variety of augite present in them. Under the microscope, they are found to be holocrystalline, but with a granitic or granulitic rather than an ophitic structure, though traces of the latter are by no means rare. To the naked eye their component minerals are usually recognizable. Professor Zirkel, from his examination of the Mull gabbros, believed them to consist of three parts of plagioclase, two parts of olivine, and one part of diallage.[329] Olivine, however, is not invariably present.[330] The pyroxene also does not always show the peculiar fibrous structure of diallage. Professor Judd, indeed, maintains that the diallagic form is due to a deep-seated process of alteration (schillerization), and that the same crystal may consist partly of ordinary augite and partly of diallage.[331] Ilmenite (with leucoxene), magnetite, apatite, biotite, and epidote are not infrequent constituents.

[Footnote 329: _Zeitschr. Deutsch. Geol. Gesellsch._ xxiii. (1871), p. 59.]

[Footnote 330: Professor Judd (_Quart. Jour. Geol. Soc._ xlii. p. 62) believes that originally all the gabbros contained olivine, and that where it is now absent, it has been altered into magnetite or serpentine. But in some coarse massive gabbros this mineral does not appear to have been an essential constituent. See _op. cit._ vol. l. p. 654.]

[Footnote 331: _Op. cit._ xli. In a later paper he insists on the gradation of the coarse granitoid varieties (gabbros) into holocrystalline compounds, where the felspar appears in lath-shapes with crystals or rounded grains of augite and olivine (dolerites), and thence into true basalts, magma-basalts, and tachylytes (_op. cit._ xlii. p. 62).]

In a recent study of the gabbros of the Cuillin Hills of Skye by Mr. J. J. H. Teall and myself, four characteristic types have been recognized.[332]

[Footnote 332: _Quart. Journ. Geol. Soc._ vol. 1. (1894), pp. 645-659, and Plates xiii. xxvi.-xxviii. See also Prof. Judd's paper, _op. cit._ (1886), p. 49.]

(1) _Granulitic Gabbros._--These are dark, fine-grained rocks which externally resemble some of the altered basalts of the plateau-series. They occur in bands or sheets which, so far as can be made out, are the oldest portions of the whole gabbro mass. Under the microscope they are found to possess a finely granulitic structure, and to consist of grains of pyroxene (augite, but more usually with the inclusions characteristic of diallage and pseudo-hypersthene), and of felspar allied to labradorite, with green pseudomorphs agreeing in form and size with the pyroxene-grains, but made of minute prisms and fibres of green hornblende and a little chlorite.

(2) _Banded Gabbros._--These are characterized by a remarkable arrangement in parallel bands of different mineral composition like the banding of ancient gneisses. This structure will be more

## particularly described in later pages. They are coarse-grained rocks

composed of pyroxene, plagioclase, olivine and magnetite. But these minerals are not distributed equally through the mass. The pale bands contain much felspar; the dark bands are largely composed of the ferro-magnesian minerals and magnetite. The pyroxene, occurring as ordinary augite, not uncommonly shows a tendency to ophitic structure. The felspar, a variety closely allied to labradorite, occurs as grains, as irregular ophitic patches, and also in forms that give broad rectangular sections. Olivine in an unaltered condition has been detected by Mr. Teall in only one specimen, and he thinks that this mineral probably never played an important part in the original constitution of these rocks. Its rounded grains may be observed to have the other minerals moulded round them, whence it may be inferred to be of older consolidation. Magnetite is generally present, either in rounded grains or in large irregular masses. Though it occurs also in strings traversing the other minerals as a secondary product, it must undoubtedly have entered largely into the original composition of these rocks. It is found enclosing the augite grains and behaving like a groundmass between the felspars. Among the dark bands there occur narrow lenticular black layers ('schlieren') composed entirely of augite and iron-ore.

The extraordinary differences between the composition of the pale felspathic and the dark ultra-basic bands are well brought out in the following analyses by Mr. J. Hort Player, No. 1 being from a light-coloured band consisting mainly of labradorite with some augite, uralitic hornblende and magnetite; No. 2 from a dark band composed of augite, magnetite and labradorite; and No. 3 from a thin ultra-basic layer mainly formed of augite and magnetite. All these specimens were taken from the ridge of Druim an Eidhne, on the eastern side of the Cuillin Hills, Skye.[333]

I. II. III.

Silica 52·8 40·2 29·5 Titanic acid ·5 4·7 9·2 Alumina 17·8 9·5 3·8 Ferric oxide 1·2 9·7 17·8 Ferrous oxide 4·8 12·2 18·2 Ferric sulphide ··· ·4 ·4 Oxide of manganese ··· ·4 ·3 Lime 12·9 13·1 10·0 Magnesia 4·8 8·0 8·7 Soda 3·0 ·8 ·2 Potash ·5 ·2 ·1 Loss by ignition 1·2 ·5 1·0 ---- ---- ---- 99·5 99·7 99·2 ---- ---- ---- Spec. grav. 2·91 3·36 3·87 ==== ==== ====

[Footnote 333: _Quart. Journ. Geol. Soc._ vol. 1. (1894), p. 653. Banded structures have been recognized in many gabbros of different ages. See the references in this paper; also Mr. W. S. Bayley, _Journ. Geol._ Chicago, ii. (1895), p. 814, and vol. iii. p. 1.]

(3) _Coarse-grained massive Gabbros._--These rocks, so abundant among the great basic bosses of the Inner Hebrides, are characterized by their coarse granitic structure, their component crystals being sometimes more than an inch long. They occur as sheets, veins and irregular masses traversing the varieties of gabbro already mentioned. They consist of the same minerals as the banded forms, and indeed are themselves sometimes banded. They are more uniform in composition than the typical banded gabbros, though showing also some variation in the relative proportions of their constituents. The specific gravity of three specimens was found to be 2·82, 2·97, and 3·06.

[Illustration: Fig. 330.--Granulitic and coarsely foliated gabbro traversed by later veins of felspathic gabbro, Druim an Eidhne, Cuillin Hills, Skye.]

(4) _Pale Gabbros of the Veins._--These occur abundantly as irregular branching veins, from less than an inch to several yards in width, and cross all the other varieties (Fig. 330[334]). Their whiteness on weathered surfaces makes them conspicuous by contrast with the dark brown or black hue of the rocks which they traverse, and shows at once that they must be poorer in bases than these. They are found on microscopic examination to consist of the same minerals as the more coarsely crystalline gabbros, but with a much greater abundance of the felspar. They contain also apatite, and hornblende appears to predominate in them over augite. They are to be distinguished from the pale veins that form apophyses from the intrusive granophyres.

[Footnote 334: Figs. 330, 336 and 337 are from photographs taken for the Geological Survey by Mr. R. Lunn.]

Troctolite (Forellenstein).--This beautiful variety of plagioclase-olivine rock occurs as a conspicuous feature on the east side of the gabbro-area of the island of Rum. It forms a sill on the side of the mountain Allival, in which the component minerals are drawn out parallel with the upper and under surfaces of the bed (Fig. 341). So marked is this flow-structure that hand-specimens might readily be taken at the first glance for ancient schistose limestone. "The felspathic ingredient (probably labradorite or anorthite) is white, and its lath-shaped crystals have ranged themselves with their long axes parallel to the line of flow. The olivine occurs in perfectly fresh grains, which in hand-specimens have a delicate green tint. Under the microscope they appear colourless, and are penetrated by the felspar prisms in ophitic intergrowth. There is a small quantity of a pale brownish augite, which not only occurs in wedge-shaped portions between the felspars, but also as a narrow zone round the olivines."[335] Considerable differences are visible in the development of the flow-structure, and with these there appear to be accompanying variations in the microscopic structure. Dr. Hatch, to whom I submitted my specimens, informed me that in one of them, where the flow-structure is so marked as to give a finely schistose aspect to the rock, "there is a larger proportion of augite, some of which exhibits a distinct diallagic striping; the olivine grains show no ophitic structure, but are sometimes completely embedded in the augite." To this remarkable flow-structure I shall again refer in connection with the light it throws on the bedded character of much of the gabbro bosses.

[Footnote 335: MS. of Dr. Hatch.]

Between the different basic intrusive igneous rocks of the Inner Hebrides, as Professor Judd has shown, there are many gradations according to the varying proportions of the chief component minerals. Thus from the olivine-gabbros, by the diminution or disappearance of the augite we get such rocks as troctolite; where the plagioclase diminishes or vanishes, we have different forms of picrite; where the olivine is left out, we come to compounds, like eucrite; while by the lessening or disappearance of the felspar and augite, we are led to ultra basic compounds, consisting in greatest part of olivine, like lherzolite, dunite and serpentine. To some of the features and probable origin of these chemical and mineralogical diversities in the same great eruptive mass further reference will be made in later pages.

ii. RELATIONS OF THE GABBROS TO THE OTHER MEMBERS OF THE VOLCANIC SERIES

Various opinions have been expressed regarding the connection between the amorphous eruptive rocks of the hill-groups and the level basalt-sheets of the plateaux. Jameson, though he landed at Rudh' an Dunain, in Skye, where this connection can readily be found, does not seem to have made any attempt to ascertain it. He noticed that the lower grounds were formed of basalt, and that the mountains "appeared to be wholly composed of syenite and hornblende rock, traversed by basalt veins."[336] Macculloch, in many passages of his _Western Islands_, alludes to the subject as one which he knew would interest geologists, but about which he felt that he could give no satisfactory information, and with characteristic verbiage he refers to the impossibility of determining boundaries, to the transition from one rock into another, to the inaccessible nature of the ground, to the almost insuperable obstacles that impede examination, to the distance from human habitation, and to the stormy climate,--a formidable list of barriers, in presence of which he leaves the relative position and age of the rocks unsettled.[337]

[Footnote 336: _Mineralogical Travels_ (1813), vol. ii. p. 72.]

[Footnote 337: See his _Western Islands_, vol. i. pp. 368, 374, 385, 386. With much admiration for the insight and zeal, amounting almost to genius, which Macculloch displayed in his work among the Western Islands, at a time when, with poor maps and inadequate means of locomotion, geological surveying was a more difficult task than it is now, I have found it impossible to follow in his footsteps with his descriptions in hand, and not to wish that for his own fame he had been content to claim credit only for what he had seen. His actual achievements were enough to make the reputation of half a dozen good geologists. It was unfortunate that he did not realize how inexhaustible nature is, how impossible it is for one man to see and understand every fact even in the little corner of nature which he may claim to have explored. He seems to have had a morbid fear lest any one should afterwards discover something he had missed; he writes as if with the object of dissuading men from travelling over his ground, and he indeed tacitly lays claim to anything they may ascertain by averring that those who may follow him "will find a great deal that is not here described, although little that has not been examined" (p. 373). Principal Forbes long ago exposed this weak side of Macculloch and his work (_Edin. New Phil. Journ._ xl. 1846, p. 82).]

Von Oyenhausen and Von Dechen, who wrote so excellent an account of their visit to Skye, and who traced much of the boundary-line between the gabbros and the other mountainous eruptive masses ("syenite"), seem to have made no attempt to work out the connection between the former and the rest of the volcanic rocks.[338]

[Footnote 338: Karsten's _Archiv_, i. p. 99. They frankly admit that "the relation of the hypersthene rock to the other trap rocks was not ascertained."]

J. D. Forbes, in his able sketch of the _Topography and Geology of the Cuchullin Hills_, was the first to recognize the superposition of the "hypersthene rock" upon the "common trap rocks"--that is, the plateau-basalts. He was disposed to consider the "hypersthene mass as a vast bed, thinning out both ways, and inclined at a moderate angle towards the S.E."[339]

[Footnote 339: _Edin. New Phil. Journ._ xl. (1846), pp. 85, 86.]

Professor Judd regarded the bosses of basic and acid rocks that rise out of the bedded basalts as the basal cores of enormously denuded volcanic cones. He believed the granitoid rocks to have been first erupted, and that after a long interval the basic masses were forced through them, partly consolidating underneath and partly appearing at the surface as the plateau-basalts.[340] That the order of appearance of the several rocks has been exactly the reverse of this supposed sequence was fully established by me in the year 1888, and has since been amply confirmed.[341] Professor Zirkel recognized that the gabbros are a dependence of the basalts, that they overlie them, and that on the naked flanks of the mountains they are regularly bedded with them.[342]

[Footnote 340: _Quart. Journ. Geol. Soc._ xxx. (1874), p. 249.]

[Footnote 341: _Trans. Roy. Soc. Edin._ xxxv. (1888), pp. 122 _et seq._; _Quart. Journ. Geol. Soc._ vol. 1. (1894), pp. 216, 645; vol. lii. (1896), p. 384, and Mr. Harker, _ibid._ p. 320.]

[Footnote 342: _Zeitschrift. Deutsch. Geol. Gesellsch._ xxiii. (1871), pp. 58, 92.]

Up to the time of the publication of my memoir in 1888 no one had traced out in more detail the actual boundaries of the several rocks on the ground, so as to obtain evidence of their true relations to each other as regards structure and age. Some of the numerous impediments recorded by Macculloch no doubt retarded the investigation. But, as Forbes so well pointed out, there is really no serious difficulty in determining the true structural connection of the amorphous rocks with each other and with the bedded basalts of the plateaux. I have ascertained them in each of the districts,[343] and have found that there cannot be the least doubt that the amorphous bosses, both basic and acid, are younger than the surrounding bedded basalts, and that the acid protrusions are on the whole younger than the basic, I shall now proceed to show how these conclusions are established by the evidence of each of the areas where the several kinds of rock occur.

[Footnote 343: In two of my excursions in Mull, and once in Skye, I was accompanied by my former colleague Mr. H. M. Cadell, who rendered me great assistance in mapping those regions.]

iii. DESCRIPTIONS OF THE SEVERAL GABBRO-DISTRICTS

1. _The Gabbro of Skye_

The largest, most picturesque, and to the geologist most important area of Tertiary gabbro in Britain, is that of Skye (Map. VI.). Though, like every other portion of the Tertiary volcanic districts, it has suffered enormous denudation, and has thereby been trenched to the very core, it reveals, more conspicuously and clearly than can be seen anywhere else, the relation of the gabbro to the bedded basalts on the one hand, and to the acid protrusions on the other. Its chief portion is that which rises into the group of the Cuillin Hills, which for blackness of hue, ruggedness of surface, jaggedness of crest, and general grimness of aspect, have certainly no rivals within the limits of the British Isles (Fig. 331). It has long been known to extend eastwards into Blath Bheinn (Blaven) and its immediate northern neighbours. There is, indeed, no break whatever between the rock of the Cuillins and that of the hills on the east side of Strath na Creitheach. In Strath More the gabbro is interrupted by the granitoid mass of the Red Hills. Patches of it, however, occur further to the east, even as far as the Sound of Scalpa.

[Illustration: Fig. 331.--Scuir na Gillean, Cuillin Hills, shewing the characteristic craggy forms of the Gabbro. (From a photograph by Mr. Abraham, Keswick.)]

If we throw out of account the invading granitoid rocks, and look upon the whole tract within which the gabbro occurs as originally one connected area, we find that it covered an elliptical space measuring about nine miles from south-west to north-east and six miles from north-west to south-east, and embracing at least 40 square miles.[344] But that its original size was greater is strikingly shown more particularly on the western margin, which like that of the basalt-escarpments, has obviously been determined by denudation, for its separate beds present their truncated ends to the horizon all along the flanks of the Cuillins, from the head of Glen Brittle round to Loch Scavaig (Fig. 332), and from Strath na Creitheach round the southern flanks of Blath Bheinn to Loch Slapin and Strath More.

[Footnote 344: Though this and the other bosses are here spoken of as consisting of gabbro, it will be understood that this rock only constitutes the larger portion of their mass, which includes also dolerites and other more basic compounds, together with involved portions of the plateau-basalts and masses of agglomerate which probably mark the position of older vents.]

[Illustration: Fig. 332.--Section across Glen Brittle, to show the general relations of the Bedded Basalts (_a_) and the Gabbros (_b_).]

The first point to be ascertained in regard to the gabbro and associated basic rocks of the mountainous tract is their connection in geological structure and age with the bedded basalts of the plateau. This initial and fundamental relation, as Forbes long ago said, can be examined along the whole western and southern flank of the Cuillin Hills, from the foot of Glen Sligachan round to the mouth of Loch Scavaig. Even from a distance, the observer, who is favoured with clear weather, can readily trace the almost level sheets of basalt till they dip gently under the darker, more massive rock of the hills. Tourists, who approach Skye by way of Loch Coruisk, have an opportunity, as the steamer nears the island of Soay, of following with the eye the basalt-terraces of the promontory of Rudh' an Dunain until they disappear under the gabbro of the last spur of the Cuillins that guards the western entrance to Loch Scavaig.

What is so evident at a distance becomes still more striking when viewed from nearer ground. Nowhere can it be more impressively seen than at the head of Glen Brittle. Looking westwards, the traveller sees in front of him only the familiar level terraces and green slopes of the basalt-plateau, rising platform above platform to a height of nearly 1500 feet above the sea. But turning to the east, he beholds the dark, gloomy, cauldron-like Corry na Creiche, from which rise some of the ruggedest and loftiest crests of the Cuillins. On the hills that project from either side of this recess and half enclose it, the bedded basalts mount from the bottom of the valley, with their lines of parallel terrace dipping gently inward below the black rugged gabbro that crowns them and sweeps round to form the back or head of the corry. Down the whole length of Glen Brittle the same structure conspicuously governs the topographical features. On the right hand, the ordinary terraced basalts form the slopes; and they rise for some 500 or 600 feet up the eastern side, until they pass under the darker, more rugged, and less distinctly bedded rocks of the mountains (Fig. 332). The dip of the whole series is here at a gentle angle towards south-east, that is, into or under the main mass of the Cuillin group.

When, however, we proceed to examine the junction between the two rocks we find it to be less simple than it appears. It is not an instance of mere superposition. The gabbro unquestionably overlies the basalts, and is therefore of younger date. But it overlies them, not as they rest on each other, in regular conformable sequence of eruption, but intrusively, as a sill does upon the rocks on which it appears to follow in the unbroken order of accumulation. This important structure may be ascertained in almost any of the many sections cut by the torrents which have so deeply trenched with gullies the flanks of the hills. Starting from the ordinary bedded basalts, we observe, in mounting the slopes and approaching the gabbro, that the rocks insensibly assume that indurated shattery character, which has been referred to as characteristic of them round the margins of vents, and which will be shown to be not less so in contact with large eruptive masses of basic or acid rock.[345] Beds of dolerite make their appearance among the basalts, so distinctly crystalline, and so similar in character to the rocks of the sills, that there can be little hesitation in regarding them as intrusive. These sills increase in size and number as we ascend, though hardened amygdaloidal basalts may still be observed. True gabbros then supervene in massive beds, and at last we find ourselves entirely within the gabbro area, where, however, thin bands of highly altered basalt may still for some distance appear. One further fact will generally be noticed, viz. that before reaching the main mass of gabbro, veins and sills of basalt, as well as of various felsitic and porphyritic members of the acid group, come in abundantly, crossing and recrossing each other in the most intricate network. The base of the thick gabbro-sheets is thus another horizon on which, as on that below the plateau-basalts, intrusive masses have been especially developed. Through all these rocks numerous parallel basalt-dykes, running in a general persistent N.N.W. direction, with a later N.E. series, rise from below the sea-level up even to the very crests of the Cuillins (Fig. 333).

[Footnote 345: This indurated, altered character of the bedded basalts near the intrusive bosses and sills will be more particularly described in a later chapter in connection with the granophyre intrusions (see p. 386). The metamorphism induced by the basic rocks has generally been less pronounced than that effected by the acid masses.]

The sections on the western side of the gabbro area of Skye thus prove that this rock inosculates with the bedded basalts by sending into them, between their bedding planes, sheets which vary in texture from fine dolerites at the outside into coarse gabbros further towards the central mass, and that this intrusion has been accompanied by a certain amount of induration of the older rocks.

[Illustration: Fig. 333.--View of the crest of the Cuillin Hills, showing the weathering of the gabbro along its joints, and of a compound basic dyke which rises through it. (From a photograph by Mr. Abraham, Keswick.)]

On the eastern side, the same structure can be even more distinctly seen, for it is not only exposed in gullies and steep declivities, but can be traced outward into the basalt-plateau. In the promontory of Strathaird, Jurassic sandstones and shales, which form the coast-line and lower grounds, are surmounted by the bedded basalts. Denudation has cut the plateau into two parts. The smaller of these makes the outlier that rises into Ben Meabost (1128 feet). The larger stretches continuously from Glen Scaladal and Strathaird House northward into Blath Bheinn. Hence from the ordinary terraced basalts, with their amygdaloids, thin tuffs, red partings, and seams of lignite, every step can be followed into the huge gabbro mountain. Starting from the black Jurassic shales on which the lowest basalt lies, we walk over the successive terraces up into the projecting ridge of An da Bheinn. But as we ascend, sheets of dolerite and gabbro make their appearance between the basalts, which gradually assume the altered aspect already noticed. The dip of the whole series is at a low angle northwards, and the beds can be followed round the head of the Glen nan Leac into the southern slopes of Blath Bheinn. Seen from the eastern side of this valley, the bedded character of that mountain is remarkably distinct, but it becomes less marked towards the upper part of the ridge where the gabbros preponderate. One of the most striking features of the locality is the number and persistence of the dykes, which strike across from the ordinary unaltered basalts of the plateau up into the highest gabbros of the range. Where less durable than the intractable gabbro, they have weathered out on the face of the precipices, thereby causing the vertical rifts and gashes and the deep notches on the crest that form so marked a feature in the scenery. On the other hand, they are often less destructible than the plateau-basalts, and hence in the Glen nan Leac they may be seen projecting as low dams across the stream which throws itself over them in picturesque waterfalls. The youngest dykes in the Blath Bheinn group of hills, have been found by Mr. Harker to have a north-easterly trend, and a north-westerly hade of about 40°, and to give a stratified appearance to the gabbro when viewed from a distance.

The deep dark hollow of the Coire Uaigneich has been cut out of the very core of Blath Bheinn, and lays bare the structure of the east part of the mountain in the most impressive as well as instructive way (Fig. 334). By ascending into this recess from Loch Slapin, we pass over the whole series of rocks, and can examine them in an almost continuous section in the bed of the stream and on the bare rocky slopes on either side. Sandstones and shales of the Jurassic series extend up the Allt na Dunaiche for nearly a mile, much veined with basalt and quartz-porphyry, by which the sandstones are locally indurated into quartzite. At last these strata are overlapped by the basalts of the Strathaird plateau, which with a marked inclination to N.N.W., here dip towards the mountains. But by the time these rocks have reached the valley, they have already lost their usual brown colour and crumbling surfaces, and have assumed the indurated splintery character, though still showing their amygdaloidal structure. They are much traversed by felsitic veins and strings which proceed from a broad band of fine-grained hornblende-granite that runs up the bottom of the Coire Uaigneich and, ascending the col, crosses it south-westwards into the Glen nan Leac. On the left or south-eastern side of this intrusive mass, a portion of Lias shales and limestone (here and there altered into white marble) is traceable for several hundred yards up the stream.[346]

[Footnote 346: This limestone was formerly identified by me with the Cambrian strata of the district. It was noticed by Von Oyenhausen and Von Dechen, who, as Mr. Harker has recently ascertained, correctly believed it to be a portion of the Lias torn off and carried upward by the eruptive rocks (Karsten's _Archiv_, i. p. 79).]

The bedded basalts of Strathaird, after dipping down towards the N.N.W., bend up where they are interbanded with dolerites and gabbros, and form the prominence called An Stac, which rises as the eastern boundary of the Coire Uaigneich. Their steep dip away from the mountain is well seen from the east side, and their outward inclination is continued into the ridge to the southward. Similar rocks appear on the other flank of the band of granite, and form the base of Blath Bheinn. They are likewise continued in the mountains further north called Sgurr nan Each and Belig, where they dip in a northerly direction away from Blath Bheinn, which seems to be the centre of uprise, with the gabbro-sheets dipping away from it. The bedded basalts have been traced by Mr. Harker up to a height of well over 2000 feet on the Blath Bheinn range. They are of the usual altered, indurated, and splintery character. The intrusive sheets interposed between them become thicker and more abundant higher up, until they constitute the main mass of the mountain. But that they are in separate sheets, and not in one amorphous mass, can be recognized by the parallel lines that mark their boundaries. The junction of the gabbro sills and the lavas is a very irregular one, portions of the latter rocks being enveloped in the intrusive sills.

The granite which sends out veins into the surrounding rocks is obviously the youngest protrusion of the locality, except of course the basalt-dykes which cross it, and which are nowhere seen in a more imposing display than round the flanks of Blath Bheinn. A section across the corry shows the structure represented in Fig. 334.

It is thus demonstrable that when its line of junction with the surrounding plateau-basalts is traced in some detail, the gabbro is found to overlie them as a whole, but also to be intercalated with them in innumerable beds, bands, or veins which rapidly die out as they recede outwards from the main central mass; that these interposed beds are intrusive sheets or sills from that mass which have cut off and enveloped portions of the basalts, and that the contiguous bedded basalts show more or less marked metamorphism.

We have now to consider the structure of the interior of the gabbro area of the Cuillin Hills. The first impression of the geologist who visits that wild district is that the main mass of rock is as thoroughly amorphous as a core of granite. Yet a little further examination will reveal to him many varieties of texture, sometimes graduating into, sometimes sharply marked off from, each other, and suggesting that the rock is not the product of one single protrusion. He will notice further indications of successive discharges or extravasations of crystalline material during probably a protracted period of time, and in the intricate network of veins crossing each other and the general body of the rock in every direction, as well as in the system of basalt-dykes that traverse all the other rocks, he will recognize the completion of the evidence of repeated renewals of subterranean energy.

[Illustration: Fig. 334.--Section across the Coire Uaigneich, Skye.

_a_, _b_, Jurassic sandstones and shales; _c_ _c_, bedded basalts and dolerites; _d_ _d_, gabbros and dolerites with indurated basalts; _e_, fine-grained hornblende-granite sending veins into surrounding rocks; _f_ _f_, basalt-dykes running through all the other rocks. ]

But the observer will be struck with the absence of the more usual proofs of volcanic activity in such forms as vesicular lavas and abundant masses of slag, bombs and tuffs, which are commonly associated with the idea of the centre of a volcanic orifice, though he will meet with isolated masses of coarse volcanic agglomerate within the gabbro area and along some parts of its junction with the granophyre. The general characters of the rocks around him suggest that he stands, as it were, far beneath that upper part of the earth's crust which is familiar to us in the phenomena of modern volcanoes; that he has been admitted into the heart of one of the deeper layers, where he can study the operations that go on at the very roots of an active vent.

When the geologist begins a more leisurely and systematic examination of the interior of the gabbro area of Skye he soon sees reason to modify the impression he may at first have received that this rugged region presents the characters of one single eruptive mass. The more he climbs among the hills the more will he meet with evidence of long-continued and oft-repeated extravasation, one portion having solidified before another broke through it, and both having been subsequently disrupted by still later protrusions.

But if by chance he should begin his examination of the ground upon some of the more typically banded varieties of rock, he may for a time almost refuse to admit that these can be either of volcanic origin or of Tertiary age.[347] He will find among them such startling counterparts of the structure of the ancient Lewisian gneiss of the North-West of Scotland that he may well be pardoned if for a time he seeks for evidence that they really do belong to that primeval formation, and have only been accidentally involved among the Tertiary volcanic rocks. If, for instance, he should land in Loch Scavaig, and first set foot upon the gabbros as they appear around Loch Coruisk, he would find himself upon masses of grey coarsely crystalline, rudely banded rock, like much of the old gneiss of Sutherland and Ross. Ascending over the ice-worn domes, he would notice that the banding becomes here and there more definitely marked by strong differences in texture and colour, while elsewhere it disappears and is replaced by a granitoid arrangement of the crystals, which are often as large as walnuts.

[Footnote 347: See _Quart. Journ. Geol. Soc._ vol. l. pp. 217, 657, and a paper by the author, "Sur la Structure rubannée des plus anciens Gneiss et des Gabbros Tertiaires," _Compt. rend. Cong. Géol. Internat._ 1894, p. 139.]

[Illustration: Fig. 335.--Banded and puckered gabbro, Druim an Eidhne, Glen Sligachan, Skye.]

Nowhere is the gneissoid banding more beautifully developed than on the east side of the Cuillin group near the head of Glen Sligachan along the ridge of Druim an Eidhne. It was at this locality that the four typical structures were observed which have already been referred to (p. 329). The varieties of colour and composition depend upon the exceedingly irregular distribution of the component minerals. The paler bands, rich in felspar, lie parallel with dark brown bands full of pyroxene, olivine and magnetite, in which, moreover, thin ribs of glistening black consist in large part of the iron ore. These layers vary in thickness from mere pasteboard-like laminæ to beds a yard or more in thickness. Within a space of a few square yards their parallelism reminds one of stratified deposits (Fig. 336), but traced over a wider space they are found to be more or less irregular in thickness and lenticular in form.

[Illustration: Fig. 336.--Banded structure in the Gabbro, from the ridge of Druim an Eidhne between Loch Coruisk and Glen Sligachan.]

The resemblance to gneisses, and sometimes to the flow-structure of coarse rhyolites, is still further sustained by occasional undulations or minute puckerings (Fig. 335). Still more extraordinary are the examples of the actual plication of a group of successive bands, as shown in Fig. 337, wherein such a group about ten feet thick is shown to have been doubly folded between parallel bands above and below. This structure is not due to any deformation of the gabbro long subsequent to the consolidation of the mass. It belongs to the phenomena of protrusion and solidification. An examination of thin slices of these rocks under the microscope reveals no evidence of crushing. On the contrary, the minerals of one band interlock with those of the band adjoining, in such a manner as to prove that the differences of composition cannot be due to crushing and shearing or to successive intrusion, but must have been present before the final consolidation of the whole rock.[348]

[Footnote 348: Mr. J. J. H. Teall and A. G., _Quart. Journ. Geol. Soc._ vol. 1. (1894), p. 652.]

The conclusion which seems most consonant with the facts is that the magma which supplied the visible masses of gabbro in Skye existed below in a heterogeneous condition, that portions of it, differing considerably from each other in composition, were simultaneously intruded, and that by the deformation of these portions during their intrusion their present plicated structures were produced. A careful study of these banded gabbros offers many suggestive points of comparison with the gneisses and anorthosite (Norian) rocks of pre-Cambrian age. It seems in the highest degree probable that the banded structures and peculiar mineral aggregation in these ancient rocks arose under conditions closely analogous to, if not identical with, those in which the Tertiary gabbros of Skye originated.[349]

[Footnote 349: Consult the Memoirs cited in the footnote on p. 342.]

Similar structures are found to be widely developed through the gabbros of the Cuillin Hills. Not only are these rocks disposed in distinct beds, but many of the beds display the most perfect banding. Thus the mountains that surround the head of Loch Scavaig and sweep round Loch Coruisk up to the great splintered crests of Sgurr na Banachdich display on their bare black crags a distinct bedded structure. On the east side of Loch Scavaig the rock presents a rudely-banded character, the bands or beds being piled over each other from the sea-level up to the summits of the rugged precipices, and dipping into the hill at angles of 25° to 35°. Abundant dykes and veins of various basic, intermediate and acid rocks cut this structure. The individual layers here show sometimes the wavy and puckered condition already referred to.

Even from a distance the alternating lighter and darker bands can readily be seen, so that this structure, with the variations in its inclination, can be followed from hill to hill (Fig. 338). The regularity of the arrangement, however, is often less pronounced on closer inspection. While the gabbro is rudely disposed in thick beds, indicative of different intrusive sheets or sills, with which the banding is generally parallel, considerable irregularities may be observed in the arrangement of the structure of individual sheets. These sheets may be parallel to each other, and yet, while in some the banding is tolerably regular in the direction of the planes of the sheets, in others it is much twisted or inclined at various angles.

[Illustration: Fig. 337.--Banded and doubly-folded Gabbro, Druim an Eidhne, 10 feet broad.]

On the west side of the Coruisk river the banding is vertical; southward from that stream it inclines slightly towards the south, but soon again becomes vertical, and continues conspicuously so at the junction of the gabbro with the Torridon sandstones and plateau-basalts on the west side of Loch Scavaig.

Thus, instead of being one great eruptive boss, the gabbro of this district is in reality an exceedingly complicated network of sills, veins and dykes. While the general inclination of the bedding sometimes continues uniform in direction and amount from one ridge to another, it is apt to change rapidly, as if the complex assemblage of intruded masses had been disrupted and had subsided in different directions. For example, after overlying the bedded basalts of the plateau all the way from Glen Brittle to the west side of Loch Scavaig, the gabbro descends abruptly across these basalts and also across the Torridon sandstones, on which they unconformably rest. These two groups of rocks are not only truncated by the gabbro, but are traversed by the intricate system of sills, dykes and veins already referred to. Where it abuts against the sandstones and basalts in Loch Scavaig, the gabbro is arranged in vertical bands of different mineral composition and texture. Much of it is remarkably coarse, some bands displaying pyroxene crystals more than an inch in length. There is no fine-grained selvage here, indicative of more rapid cooling. So coarse, indeed, is the rock close up against the sandstone, that the junction-line can hardly be supposed to be the normal contact of the intrusive rock. This inference is confirmed by the existence of a singular kind of breccia between the gabbro and the sandstones. It is a tumultuous mass of fragments of coarse and fine gabbro, Torridon sandstone and shale, and plateau-basalts, embedded in a pale crystalline matrix of fine granular granophyre; veins from this acid intrusion run off into the gabbro on the one side as well as into the Torridon sandstones on the other. It would seem that this junction-line has been one of great movement, that the gabbro-sheets have subsided against a fault-wall of plateau-basalt and Torridon sandstone, and that subsequently an intrusion of finely granular granophyre has come up the fissure, involving in its ascent fragments of all the materials around.

The rocks for a considerable distance to the south of the gabbro are intensely altered. The Torridon sandstone has been so indurated as to pass into a bleached white quartzite, while the shales interstratified with it have been converted into a kind of porcellanite. But the most interesting alterations are those to be observed in the plateau-basalts, which at a height of about 300 feet above the sea, are to be seen in nearly horizontal sheets that lie immediately on the upturned edges of the Torridon sandstones. These lavas have suffered great metamorphism, to which more particular reference will be made in

##