Chapter XXXVII…

Chapter XXXVII

THE SEVERAL BASALT-PLATEAUX AND THEIR GEOLOGICAL HISTORY, ANTRIM, MULL, MORVEN AND ARDNAMURCHAN

There are five districts in North-western Europe where the original widespread Tertiary lava-fields have been less extensively eroded than elsewhere, or at least where they have survived in larger and thicker masses. Whether or not each of them was an isolated area of volcanic activity cannot now be determined. Their several outflows of lava within the area of the British Isles may have united into one continuous volcanic tract, and their present isolation there may be due entirely to subterranean movements and denudation. There is a certain convenience, however, in treating the districts separately. They are--1. Antrim; 2. Mull, Morven and Ardnamurchan; 3. Small Isles; 4. Skye; 5. The Faroe Islands.

i. ANTRIM[230]

[Footnote 230: The basalts of Antrim are the subject of an abundant literature. I may refer particularly to the papers of Berger and Conybeare (Trans. Geol. Soc. iii.), the Geological Report of Portlock, and the Explanations of the Sheets of the Geological Survey of Ireland. Other papers will be afterwards cited. The general features of the Antrim plateau are shown on Map VII.]

The largest of the basalt-plateaux of Britain is that which forms so prominent a feature in the scenery and geology of the North of Ireland, stretching from Lough Foyle to Belfast Lough, and from Rathlin Island to beyond the southern margin of Lough Neagh. Its area may be roughly computed at about 2000 square miles. But, as its truncated strata rise high along its borders, and look far over the surrounding low grounds, it must be regarded as a mere fragment of the original volcanic plain. It may be described as an undulating tableland, which almost everywhere terminates in a range of bold cliffs, but which, towards the centre and south, sinks gently into the basin of Lough Neagh. The marginal line of escarpment, however, presents considerable irregularity both in height and form, besides being liable to frequent local interruptions. It is highest on the west side, one of its crests reaching at Mullaghmore, in County Londonderry, a height of 1825 feet. It sinks down into the valley of the Bann, east of which it gradually ascends, forming the well-known range of cliffs from the Giant's Causeway and Bengore Head to Ballycastle. It then strikes inland, and making a wide curve in which it reaches a height of more than 1300 feet, comes to the sea again at Garron Point. From that headland the cliffs of basalt form a belt of picturesque ground southwards beyond Belfast, interrupted only by valleys that convey the drainage of the interior of the plateau to the North Channel. Above the valley of the Lagan the crest of the plateau rises to a height of more than 1500 feet.

Throughout most of its extent the basalt-escarpment rests on the white limestone or Chalk of Antrim, beneath which lie soft Lias shales and Triassic marls. Here and there, where the substratum of Chalk is thin, the action of underground water on the crumbling shales and marls below it has given rise to landslips. The slopes beneath the base of the basalt are strewn with slipped masses of that rock, almost all the way from Cushendall to Larne, some of the detached portions being so large as to be readily taken for parts of the unmoved rock. On the west side also, a group of huge landslips cumbers the declivities beneath the mural front of Benevenagh.

I have found some difficulty in the attempt to ascertain what was the probable form of surface over which the volcanic rocks of this plateau began to be poured out. The Chalk sinks below the sea-level on the north coast, but, in the outlier of Slieve Gallion, three miles beyond the western base of the escarpment, it rises to a height of 1500 feet above the sea. On the east side also, it shows remarkable differences of level. Thus, below the White Head at the mouth of Belfast Lough, it passes under the sea-level, but only 16 miles to the south, where it crops out from under the basalt, its surface is about 1000 feet above that level. If these variations in height existed at the time of the outpouring of the basalt, the surface of the ground over which the eruptions took place was so irregular that some hundreds of feet of lava must have accumulated before the higher chalk hills were buried under the volcanic discharges. But it seems to me that much of this inequality in the height of the upper surface of the Chalk is to be attributed to unequal movements since the volcanic period, which involved the basalt in their effects, as well as the platform of Chalk below it. Had the present undulations of that platform been older than the volcanic discharges, it is obvious that upper portions of the basalt-series would have overlapped lower, and would have come to rest directly on the Chalk. But this arrangement, so far as I am aware, never occurs, except on a trifling scale. Wherever the Chalk appears, it is covered by sheets of the lower and not of the upper of the two groups into which the Antrim basalts are divisible. We have actual proof of considerable terrestrial disturbance, subsequent to the date of the formation of the volcanic plateau. Thus, near Ballycastle, a fault lets down the basalt and its Chalk platform against the crystalline schists of that district. On the east side of the fault, the Chalk is found far up the slope, circling round the base of the beautiful cone of Knocklayd--an outlier of the basalt which reaches a height of 1695 feet (Fig. 263). The amount of vertical displacement of the volcanic sheets is here 700 feet.[231] Many other displacements, as shown by the mapping of my colleagues in the Geological Survey, have shifted the base of the escarpment from a few inches up to several hundred feet. Besides actual dislocations, the Antrim plateau has undergone some marked subsidences of which the most notable is that of Lough Neagh.[232]

[Footnote 231: Explanatory Memoir of Sheets 7 and 8, Geological Survey, Ireland, by Messrs. Symes, Egan, and M'Henry (1888), p. 37.]

[Footnote 232: These inequalities in the level of the base of the Antrim plateau will be more particularly discussed in Chapter xlix., in connection with the subsidences and dislocations which have affected the region since the close of the volcanic period.]

It is evident, therefore, that the present position of the Chalk platform is far from agreeing with that which it presented to the outflow of the sheets of basalt. But, on the other hand, there can be no doubt that its surface at the beginning of the volcanic outbursts was not a level plain. It was probably a rolling country of low bare chalk-downs, like parts of the South-east of England. The Irish Chalk attains its maximum thickness of perhaps 250 feet at Ballintoy. But it is liable to rapid diminution. On the shore at Ballycastle about 150 feet of it can be seen, its base being concealed; but only two and a half miles to the south, on the outlier of Knocklayd, the thickness is not quite half so much. On the west side of the plateau also, there are rapid changes in the thickness of the Chalk. Such variations appear to be mainly attributable to unequal erosion before the overflow of the basalts. So great indeed had been the denudation of the Cretaceous and underlying Secondary formations previous to the beginning of the volcanic eruptions, that in some places the whole of these strata had been stripped off the country, so that the older platform of Palæozoic or still more ancient masses was laid bare. Thus, on the west side of the escarpment, the basalt steals across the Chalk and comes to rest directly upon Lower Carboniferous rocks.

The authors who have described the junction of the Chalk and basalts in Antrim have generally referred to the uneven surface of the former rock as exposed in any given section. The floor on which the basalt lies is remarkably irregular, rising into ridges and sinking into hollows or trenches, but almost everywhere presenting a layer of earthy rubbish made of brown ferruginous clays, mixed with pieces of flint, chalk, and even basalt.[233] The flints are generally reddened and shattery. The chalk itself has been described as indurated, and its flints as partially burned by the influence of the overlying basalt. But I have not noticed, at any locality, evidence of alteration of the solid chalk, except where dykes or intrusive sheets have penetrated it.[234] There can be no doubt that the hardness of the rock is an original peculiarity, due to the circumstances of its formation. The irregular earthy rubble, that almost always intervenes between the chalk and the base of the basalt, like the "clay with flints" so general over the Chalk of Southern England, no doubt represents long-continued subærial weathering previous to the outflow of the basalt. Even, therefore, if there were no other evidence, we might infer with some confidence from this layer of rubble, that the surface over which the lavas were poured was a terrestrial one. Here and there, too, we may detect traces of the subsidence of the basalt into swallow-holes dissolved in the chalk subsequent to the outflow of the basalt-sheets.

[Footnote 233: Portlock, Report on Geology of Londonderry, etc. (Geological Survey), p. 117.]

[Footnote 234: See Portlock, op. cit. p. 116.]

The Antrim plateau is not only the largest in the British Islands, it is also the most continuous and regular. It may be regarded, indeed, as one unbroken sheet of volcanic material, not disrupted by any such mountainous masses of intrusive rock as in the other plateaux interrupt the continuity of the horizontal or gently inclined sheets of basalt. Around its margin, indeed, a few outliers tower above the plains, and serve as impressive memorials of its losses by denudation. Of these, by much the most picturesque and imposing, though not the loftiest, is Knocklayd already referred to, which forms so striking a feature in the north-east of Antrim (Fig. 263).

The total thickness of volcanic rocks in the Antrim plateau exceeds 1000 feet; but, as the upper part of the series has been removed by denudation, the whole depth of lava originally poured out cannot now be told. A well-marked group of tuffs and clays, traceable throughout a large part of Antrim, forms a good horizon in the midst of the basalts, which are thus divisible into a lower and upper group (Fig 264).

The Lower Basalts have a thickness of from 400 to 500 feet. But, as already mentioned (p. 194), they die out in about six miles to no more than 40 feet at Ballintoy. They are distinguished by their generally cellular and amygdaloidal character, and less frequently columnar structure. The successive flows, each averaging perhaps above 15 feet in thickness, are often separated by thin red ferruginous clayey partings, sometimes by bands of green or brown fine gravelly tuff. The most extensive of these tuff-bands occurs in the lower part of the group at Ballintoy, and can be traced along the coast for about five miles. In the middle of its course, near the picturesque Carrick-a-raide, it reaches a maximum thickness of about 100 feet and gradually dies out to east and west. The neck of coarse agglomerate at Carrick-a-raide, is doubtless the vent from which this mass of tuff was discharged (see Fig. 301). Owing to the thinning out of the sheets of basalts, as they approach the vent, the tuff comes to rest directly on the Chalk, and for some distance westwards forms the actual base of the volcanic series.[235] Occasional seams of carbonaceous clays, or of lignite, appear in different horizons among the basalts. Beneath the whole mass of basalt, indeed, remains of terrestrial vegetation here and there occur. Thus, near Banbridge, County Down, a patch of lignite, four feet ten inches thick, underlies the basalt, and rests directly on Silurian rocks. Such fragmentary records are an interesting memorial of the wooded land-surface over which the earliest outflows of basalt spread.

[Footnote 235: See Explanation of Sheets 7 and 8 of the Geological Survey of Ireland (1888), p. 23.]

In looking at the great basalt-escarpments of Antrim, the Inner Hebrides or the Faroe Islands, and in following with the eye the successive sheets of lava in orderly sequence of level bands from the breaking waves at the base to the beetling crest above, we are apt to take note only of the proofs of regularity and repetition in the outflows of molten rock and to miss the evidence that these outflows did not always rapidly follow each other, but were separated by intervals of varying, sometimes even of long duration. One of the most frequent and conspicuous proofs of such intervals is to be found in the red layers or partings above referred to which, throughout all the basalt-plateaux, so commonly intervene between successive sheets of basalt. These red streaks cannot fail to arrest the eye on the coast-precipices where by their brilliant contrast of colour, they help to emphasize the bedded character of the whole volcanic series.

Examined more closely, they are found to consist of clay or bole which shades into the decomposed top of the bed whereon it lies, and is usually somewhat sharply marked off from that which covers it. This layer has long, and I think correctly, been regarded as due to the atmospheric disintegration of the surface of the basalt on which it rests, before the eruption of the overlying flow. It varies in thickness from a mere line up to a foot or more, and it passes into the tuffs and clays which are sometimes interposed between the sheets of basalts. It may be looked upon as probably furnishing evidence of the lapse of an interval sufficiently extended to permit a considerable subserial decay of the surface of a lava-sheet before the outflow of the next lava. But an attentive study of the plateaux discloses other and even more remarkable indications that the pauses between the consecutive basalt-beds were frequently so prolonged as to allow extensive topographical changes to be made in a district. Nowhere is the long duration of some of these intervals more impressively taught than in the central zone of sedimentary strata in Antrim.

This persistent group of tuffs, clays, and iron-ore is generally from 30 to 40 and sometimes as much as 70 feet thick. From the occurrence of the ore in it, it has been explored more diligently in recent years than any other group of rocks in the district, and its outcrop is now known over most of the plateau. The iron-ore bed varies from less than an inch up to 18 inches in thickness, and consists of pisolitic concretions of hæmatite, from the size of a pea to that of a hazel nut, wrapped up in a soft ochreous clayey matrix.[236] Where it is absent, its place is sometimes taken by an aluminous clay, worked as "bauxite," which has yielded stumps of trees and numerous leaves and cones. Beneath the iron-ore or its representative, lies what is called the "pavement,"--a ferruginous tuff, 8 to 10 feet thick, resting on "lithomarge,"--a lilac or violet mottled aluminous earth sometimes full of rounded blocks or bombs of basalt. The well-known horizon for fossil plants at Ballypallidy is a red tuff in this zone. The section of strata between the two basalt-groups at this locality may serve as an illustration of the nature and arrangement of the deposits.[237]

[Footnote 236: Consult a good essay on the Iron-ore and Basalts of North-east Ireland by Messrs. Tate and Holden, Quart. Journ. Geol. Soc. xxvi. (1870), p. 151. In this paper the nature, composition and modes of origin of the iron-ore and its associated strata are fully discussed.]

[Footnote 237: A. M'Henry, Geol. Mag. (1895), p. 263.]

Upper Basalt, compact and often columnar sheets.

Brown laminated tuff and volcanic clays.

Laminated brown impure earthy lignite, 2 feet 3 inches.

Brown and red variegated clays, tuffs and sandy layers, with irregular seams of coarse conglomerate composed of rounded and subangular fragments of rhyolite and basalt, 3 feet 4 inches.

Brown, red and yellowish laminated tuffs, mudstones, and bole, with occasional layers of fine conglomerate (rhyolitic and basaltic), pisolitic iron-ore band and plant-beds, 8 feet 10 inches.

Lower basalt, amygdaloidal.

In some of the Ballypallidy tuffs the most frequent lapilli are pieces of green and brown glass, which Mr. Watts compares with the pitchstone of Sandy Braes, though rarely containing phenocrysts as that rock does. He has found also in these strata a smaller proportion of lithoidal rhyolites and occasionally fragments of basic rock.

The pale and coloured clays that occur in this marked sedimentary intercalation have doubtless been produced by the decomposition of the volcanic rocks and the washing of their fine detritus by water. Possibly this decay may have been in part the result of solfataric action. From true bauxite or aluminium-hydrate, the sediments vary in composition and specific gravity and pass into aluminous silicates and iron-ores. They seem to indicate a prolonged interval of volcanic quiescence when the lavas and tuffs already erupted were denuded and decomposed.[238]

[Footnote 238: See a note on Bauxite by Professor G. A. Cole, Scientif. Trans. Royal Dublin Soc. vol. vi. series ii. (1896), p. 105.]

The area over which this interesting series of stratified deposits now extends is obviously much less than it was originally. It has indeed been so reduced by denudation into mere scattered patches that it probably does not exceed 170 square miles. But the group can be traced from Divis Hill, near Belfast, to Rathlin Island, a distance of 50 miles, and from the valley of the Bann to the coast above Glenarm, more than 20 miles. There can be little doubt that it was once continuous over all that area, and that it probably extended some way further on each side. If the so-called Pliocene clays of Lough Neagh be regarded as parts of this group of strata, its extent will be still further increased. Hence the original area over which the iron-ore and its accompanying tuffs and clays were laid down can hardly have been less than 1000 square miles. This extensive tract was evidently the site of a lake during the volcanic period, formed by a subsidence of the floor of the lower basalts. The salts of iron contained in solution in the water, whether derived from the decay of the surrounding lavas or from the discharges of chalybeate springs, were precipitated as peroxide in pisolitic form, as similar ores are now being formed on lake-bottoms in Sweden. For a long interval, quiet sedimentation went on in this lake, the only sign of volcanic energy during that time being the dust and stones that were thrown out and fell over the water-basin, or were washed into it by rains from the cones of the lava-slopes around.

It may here be remarked that the tendency to subsidence in the Antrim plateau seems to have characterized this region since an early part of the volcanic period. The lake in which the deposits now described accumulated was entirely effaced and overspread by the thick group of upper basalts. But long after the eruptions had ceased, a renewed sinking of the ground gave rise to the sheet of water which now forms Lough Neagh.[239]

[Footnote 239: This subject will be discussed in Chapter xlix.]

Nowhere else among the Tertiary basalt-plateaux of Britain has any trace been found of so marked and prolonged a pause in the volcanic activity as is indicated by the Antrim zone of tuffs and clays. Throughout the Inner Hebrides, indeed, numerous intercalations of sedimentary material occur among the basalts, but these consist mainly of tuffs and volcanic conglomerates with less frequent shales and coal-seams, and they never suggest so distinct and lengthened an interval as is indicated by the Antrim deposit.

It is not improbable that this interval was marked by the outbreak of rhyolitic eruptions somewhere in the region. The abundance of rhyolite fragments in some of the tuffs is striking evidence that acid rocks were in one way or other brought to the surface at this time. At Glenarm one of the members of the stratified series is a marked rhyolitic conglomerate, composed of rounded pebbles of a rock not unlike the well-known rhyolite of Tardree and Carnearny. These fragments, obviously of local origin, must either have been derived from a surface of acid rock laid bare by denudation, or from rhyolite ejected in lapilli or poured out in streams. I formerly believed that all the Antrim rhyolites had been injected into the basalts after the close of the plateau-period. But the proved abundance and wide extent of the rhyolitic detritus among the sediments associated with the iron-ore point to a possible outflow of acid lavas with accompanying tuffs during the sedimentary interval between the two groups of basalt. The characters of the Antrim rhyolites, however, will be more particularly discussed in Chapter xlvii., in connection with the acid rocks of the Tertiary volcanic series.

Immediately above the iron-ore of Antrim, or separated from it in places by only a few inches of tuff, comes the group of Upper Basalts, which varies up to 600 feet in thickness, though as the upper portion has been everywhere removed by denudation, no measure remains of what may have been the original depth of the group. The general character of these basalts is more frequently columnar, black and compact, and with fewer examples of a strongly amygdaloidal structure than in the lower group. But this distinction is less marked in the south than in the north of Antrim, so that where the intervening zone of tuffs and iron-ore disappears, no satisfactory line of division can be traced between the two groups of basalt. The occurrence of that zone, however, by giving rise to a hollow or slope, from which the upper basalts rise as a steep bank or cliff, furnishes a convenient topographical feature for mapping the boundary of these rocks. Among the upper basalts, also, there is perhaps a less frequent occurrence of those thin red partings of bole between successive flows, so conspicuous in the lower group. But the flows are not less distinctly marked off from each other. Nowhere can their characteristic features be better seen than along the magnificent range of cliffs from the Giant's Causeway eastwards. The columnar bed that forms the Causeway is the lowest sheet of the upper group, and may be seen resting directly on the zone of grey and red tuffs. It is about 60 or 70 feet thick; and, while perfectly regular in its columnar structure at the Causeway and the "Organ," assumes further eastward the confusedly starch-like arrangement of prisms already referred to. But in the great cliff section of the "Amphitheatre," the more regular structure is resumed, the bed swells out to about 80 feet in thickness, and columns of that length run up the face of the precipice, weathering out at the top into separate pillars, which, perched on the crest of an outstanding ridge, are known as the "Chimneys." The basalt-beds that succeed the lowest one are each only about 10 to 15 feet thick (Fig. 265).

Between the successive sheets of the Upper Basalts thin seams of red ferruginous clay though, as I have said, less frequent perhaps than in the lower group, continue to show that the intervals between successive eruptions were of sufficient duration to admit of some subærial decay of the surface of a lava before the outflow of the next bed. Occasional thin layers of tuff also, and even of pisolitic iron-ore, have been observed among these higher basalts. But the most interesting and important intercalations are inconstant seams of lignite. One of the most conspicuous of these lies immediately above the basalt of the "Causeway," where it was long worked for fuel, and was found to be more than six feet thick. But it is quite local, as may be seen at the "Organ" over which it lies, having a thickness of only 12 inches and rapidly dying out so as to allow the basalts above and below it to come together. The removal of the upper portion of the basalts by denudation has destroyed the records of the latest part of the volcanic history of the Irish plateaux.

It is obvious that nowhere in Antrim does any trace exist of a central vent or cone from which the volcanic materials were discharged. There is no perceptible thickening of the individual basalt-sheets, nor of the whole series in one general direction, in such a manner as to point to the site of some chief focus of eruption. Nor can we place reliance on the inclination of the several parts of the plateau. I have pointed out that the varying dip of the beds must be attributed mainly to post-volcanic movements, or at least to movements which, if not later than all the phases of volcanic action, must have succeeded the outpouring of the plateau-basalts. There has been a general subsidence towards the central and southern tracts now occupied by the valley of the Bann and Lough Neagh. But nowhere in the depression is there any trace of the ruins of a central cone or focus of discharge.

The Antrim plateau, in these respects, resembles the others. But as has already been remarked, it differs from them in one important particular. It has nowhere been disrupted by huge bosses of younger rocks, such as have broken up the continuity of the old lava-fields further north. Yet it also is not without its memorials of younger protrusions. It contains not a few excellent examples of true volcanic vents, and, as above stated, it includes some small acid bosses that may represent the great protrusions of the Inner Hebrides, and may have been connected with superficial outflows of rhyolitic lava and showers of rhyolitic tuff.

ii. MULL, MORVEN AND ARDNAMURCHAN

This plateau covers nearly the whole of the island of Mull, embraces a portion of Morven on the Argyleshire mainland, and, stretching across Loch Sunart, includes the western part of the peninsula of Ardnamurchan (Map VI.). That these now disconnected areas were once united into a continuous lava-field which extended far beyond its present limits is impressively indicated by their margin of cliffs and fringe of scattered islands and outliers. The plateau went west, at least, as far as the Treshnish Isles, which are composed of basalt. On its eastern border, a capping of basalt on the top of Beinn Iadain (1873 feet) in Morven, and others further north, prove that its volcanic sheets once spread into the interior of Argyleshire (Fig. 266). On the south, its fine range of lofty cliffs, with their horizontal bars of basalt, bear witness to the diminution which it has undergone on that side; while, on the north, similar sea-walls tell the same tale. Not only has it suffered by waste along its margin, it has also been deeply trenched by the excavation of glens and arms of the sea. The Sound of Mull cuts it in two, and the mainland portion is further bisected by Loch Sunart, and again by Loch Aline. The island of Mull is so penetrated by sea-lochs and divided by deep valleys that a comparatively slight depression would turn it into a group of islands. But, besides its enormous denudation, this plateau has been subjected to disruption, and perhaps also to subsidence, from subterranean movements. In the southern portion of the island of Mull it has been broken up by the intrusion of large bosses and sheets of gabbro, and by masses as well as innumerable veins of various granitoid and felsitic rocks. In Ardnamurchan, it has suffered so much disturbance from the same cause that its original structure has been almost obliterated over a considerable area. Moreover, it has been dislocated by many faults, by which different portions have been greatly shifted in level. The most important of these breaks is one noticed by Professor Judd, and visible to every tourist who sails up the Sound of Mull. It traverses the cliffs on the Morven side, opposite Craignure, bringing the basalts against the crystalline schists, and strikes thence inland, wheeling round into the long valley in which Lochs Arienas and Teacus lie. On its western side, the base of the basalt-series is almost at the sea-level; on its eastern side, that platform rises high into the outliers of Beinn na h-Uamha (1521 feet) and Beinn Iadain. The amount of displacement is probably more than 1000 feet. Many other minor faults in the same district show how much the crust of the earth has been fractured here since older Tertiary time.

[Footnote 240: There are no fewer than three faults in the basalt-capping on the summit of Beinn Iadain. By bringing the basalts and schists into juxtaposition, they have given rise to topographical features that can be seen even from a distance.]

A little to the west of Mull, and belonging originally to the same plateau, lies the isle of Staffa, the famous columnar basalts of which first attracted the attention of travellers, and gave to the Tertiary volcanic rocks of Scotland their celebrity (Fig. 266_a_).

In spite of the extent to which it has suffered from denudation and subterranean disturbance, and indeed in consequence thereof, the Mull plateau presents clear sections of many features in the history of the basalt-outflows and of the subsequent phases of Tertiary volcanic action which cannot be seen in the more regular and continuous tableland of Antrim. Moreover, it still possesses in its highest mountain, Ben More (3169 feet), a greater thickness, and probably a higher series, of lavas than can now be seen in any other of the plateaux.

The difficulties, already referred to in regard to Antrim, of tracing the probable form of ground on which the volcanic eruptions began, are even greater in the case of the Mull plateau. We can dimly perceive that the depression in the crystalline rocks of the Highlands which had, from at least the older part of the Jurassic period, stretched in a N.N.W. direction along what is now the western margin of Argyleshire, lay beneath the sea in Jurassic time, and was then more or less filled up with sedimentary deposits. The hollow appears thereafter to have become a land-valley, whence the Jurassic strata were to a large extent cleared out by denudation before its subsequent submergence under the sea in which the upper Cretaceous deposits accumulated. Professor Judd has shown that relics of these Cretaceous strata appear on both sides of the plateau from under the protecting cover of basalt-sheets. But, before the volcanic eruptions began, the area had once again been raised into land, and the youngest Secondary formations had been extensively eroded.

In their general aspect the basalts of Mull agree with those of Antrim, and the circumstances under which they were erupted were no doubt essentially the same. But considerable differences in detail are observable between the succession of rocks in the two areas. When I first visited the island in 1866, the only available maps, with any pretensions to accuracy, were the Admiralty charts; but, as these do not give the interior except in a generalized way, it was difficult to plot sections from them, and to arrive at satisfactory conclusions as to the thickness of different groups of rock. Accordingly, as the successive nearly flat flows of basalt can be traced from the sea-level up to the top of Ben More, I contented myself with the fact that the total depth of lava-beds in Mull was at least equal to the height of that mountain, or 3169 feet. The publication of the Ordnance Survey Maps now enables us to make a nearer approximation to the truth. From the western base of the magnificent headland of Gribon, the basalts in almost horizontal beds rise in one vast sweep of precipice and terraced slope to a height of over 1600 feet, and then stretch eastwards to pass under the higher part of Ben More, at a distance of some eight miles. They have a slight easterly inclination, so that the basement sheets seen at the sea-level, at the mouth of Loch Scridain, gradually sink below that level as they go eastward. It is not easy to get a measurement of dip among these basalts, except from a distance. If we take the inclination at only 1°, the beds which are at the base of the cliff on the west, must be about 700 feet below the sea on the line of Ben More, which would give a total thickness of nearly 3900 feet of bedded lava below the top of that mountain. We shall not probably overestimate the thickness of the Mull plateau if we put it at 3500 feet.

The base of the volcanic series of Mull can best be seen on the south coast at Carsaig, and at the foot of the precipices of Gribon. As already stated, it is there found resting above Cretaceous and Jurassic rocks. The lowest beds are basalt-tuffs, of the usual dull green colour. They are in places much intermingled with sandy and gravelly sediment, as if the volcanic debris had fallen into water where such sediment was in course of deposition. One of the most interesting features, indeed, in this basement part of the series, is the occurrence of bands of non-volcanic material which accumulated after the tuffs and some of the lavas had been erupted, but before the main mass of basalts. Those at Carsaig include a lenticular bed, 25 feet thick, of rolled flints, which, with some associated sandy bands, lies between sheets of basalt. On the opposite side of the promontory is the well-known locality of Ardtun, from which the first land-plants in the volcanic series were determined. The actual base of the basalts is not there seen, being covered by the sea. The "leaf-beds," with their accompanying sandstones, gravels, and limestone, lie upon a sheet of basalt, which in some parts is exceedingly slaggy on the top, passing down into a black compact structure, and assuming at the base of the cliff a columnar arrangement, with the prisms curved and built up endways towards each other. Some of the gravels exceed 30 feet in thickness, and consist of rolled flints, bits of chalk, and pieces of basalt and other basic igneous rocks. But some of their most interesting ingredients are pebbles of sanidine lavas, which have been recognized in them by Prof. G. Cole.[241] No known protrusions of such lavas occur anywhere beneath or interstratified with the plateau-basalts of this district. As will be afterwards shown, all the visible acid rocks, the geological relations of which can be ascertained, are here of younger date than these basalts. I am disposed to regard the fragments found in the Ardtun conglomerates as probably derived from some of the basalt-conglomerates of the plateau, in which fragments of siliceous igneous rocks do occur. Though there is no evidence that any lavas of that nature were here poured out at the surface before or during the emission of the basalts, the contents of these fragmental volcanic accumulations suggest that such lavas, already consolidated, lay at some depth beneath the surface, and that fragments were torn off from them during the explosions that threw out the materials of the basalt-conglomerates to the surface.

[Footnote 241: Quart. Jour. Geol. Soc. xliii. (1887) p. 277.]

The succession of strata at the Ardtun headland varies considerably in a short distance, some of the sedimentary deposits rapidly increasing or diminishing in thickness. The section as measured by Mr. Starkie Gardner is as follows[242]:--

Columnar basalt, 40 feet. Position of first leaf-bed, obscured by grass, about 2 feet. Gravel varying from about 25 feet to a maximum of nearly 40 feet. Black or second leaf-bed, 2-1/2 feet. Gravel about 7 feet. Grey clay, 2 feet. Laminated sandstone, 6 inches, with 3 inches of fine limestone, containing leaves at the base. Clay, with leaves at base, 1 foot. Clunch, with rootlets, 7 inches. Amorphous basalt, becoming columnar at base, about 60 feet.

[Footnote 242: Op. cit. p. 280.]

Mr. Starkie Gardner has called attention to the extraordinarily fresh condition of the vegetation in some of the layers of the Ardtun section. One of the leaf-beds he has found to be made up for an inch or two of a pressed mass of leaves, lying layer upon layer, and retaining almost the colours of dead vegetation. Among the plants represented is a large purple Ginkgo and a fine Platanites, one leaf measuring 15-1/2 inches long by 10-1/2 broad. The characteristic dicotyledonous leaves at this locality possessed relatively large foliage.[243]

[Footnote 243: For fuller local details regarding the Ardtun leaf-beds, I may refer to the original paper by the Duke of Argyll (Quart. Jour. Geol. Soc. vii. p. 89), and to the memoir by Mr. Starkie Gardner (op. cit. xliii. (1887), p. 270).]

To the early observations of Macculloch we are indebted for the record of an interesting fact in connection with the vegetation of the land-surface over which the first lava-flows spread. He figured a vertical tree trunk, imbedded in prismatic basalt, and rightly referred it to some species of fir.[244] This relic may still be seen under the basalt precipices of Gribon. Mr. Gardner found it to be "a large trunk of a coniferous tree, five feet in diameter, perhaps Podocarpus, which has been enveloped, as it stood, in one of the flows of trap to the height of 40 feet. Its solidity and girth evidently enabled it to resist the fire, but it had decayed before the next flow passed over it, for its trunk is a hollow cylinder filled with debris, and lined with the charred wood. A limb of another, or perhaps the same tree, is in a fissure not far off."[245]

[Footnote 244: Western Islands, vol. i. p. 568, and plate xxi. Fig. 1.]

[Footnote 245: Quart. Jour. Geol. Soc. xliii. p. 283.]

At different levels in the volcanic series of Mull, beds of lignite and even true coal are observable. These seem to be always mere lenticular patches, only a few square yards in extent. The best example I have met with lies among the basalts near Carsaig. It is in part a black glossy coal, and partly dull and shaly. Some years ago it was between two and three feet thick, but now, owing to its having been dug away by the shepherds, only some six or eight inches are to be seen. It lies between two basalt-flows, and rapidly disappears on either side.

More frequent than these inconstant layers of fossil vegetation are the thin partings of tuff and layers of red clay, sometimes containing iron-ore, which occur at intervals throughout the series between different flows of basalt. But even such intercalations are of trifling thickness, and only of limited extent. The magnificent precipices of M'Gorry's Head and Gribon expose a succession of beds of columnar amorphous and amygdaloidal basalt, which must attain a thickness of at least 2500 feet, before they are overlain by the higher group of pale lavas in Ben More. On the east side of the island, thin tuffs and bands of basalt-conglomerate occur on different horizons among the bedded basalts, from near the sea-level up to the summit of the ridge which culminates in Beinn Meadhon (2087 feet), Dùn-da-Ghaoithe (2512 feet), and Mainnir-nam-Fiadh (2483 feet). Reference has already been made to the remarkably coarse character of some of the breccias intercalated among the basalts in this part of Mull, and to the enormous dimensions of some of the masses of mica-schist and quartzite which have been carried up from a depth of 2000 feet or more by volcanic agency (see ante, p. 196, and Fig. 262).

Above the ordinary compact and amygdaloidal basalt comes the higher group of pale lavas already referred to as forming the uppermost part of Ben More, whence it stretches continuously along the pointed ridge of A'Chioch, and thence northwards into Beinn Fhada. The same lavas are likewise found in two outliers, capping Beinn a' Chraig, a mile further north, and I have found fragments of them on some of the loftier ridges to the south-east. This highest and youngest group of lavas in the plateaux has been reduced to mere isolated patches, and a little further denudation will remove it altogether. Yet it is not less than about 800 feet thick, and consists of bedded andesitic or trachytic lavas, which alternate with and follow continuously and conformably upon the top of the ordinary plateau-basalts. These dull, finely crystalline or compact, light-grey rocks weather with a characteristic platy form, which has been mistaken for the bedding of tuffs. The fissility, however, has none of the regularity or parallelism of true bedding, and may be observed to run sometimes parallel with the bedding of the sheets, sometimes obliquely or even at right angles to it. Even where this structure is best developed, the truly crystalline nature of the rocks can readily be detected. Some of them are porphyritic and amygdaloidal, the very topmost bed of the mountain being a coarse amygdaloid. Intercalated with these curious rocks there are others in which the ordinary characters of the dolerites and basalts of the plateaux can be recognised. The amygdaloids are often full of delicate prisms of epidote.

In Mull, as in the other areas of terraced basalts, we everywhere meet with gently inclined sheets, which do not thicken out individually or collectively in any given direction, except as the result of unequal denudations. So far as I have been able to discover, they afford no evidence of any great volcanic cone from which they proceeded. Their present inclinations are unquestionably due, as in Ireland, to movements subsequent to the formation of the plateau. In Loch-na-keal they dip gently to the E.N.E.; in Ulva and the north-west coast to N.N.E.; near Salen to W.S.W. on the one side, and N.W. on the other. Round the southern and eastern margins of the mountainous tract of the island, they dip generally inwards to the high grounds.

The Mull plateau presents a striking contrast to that of Antrim, in the extraordinary extent to which it has been disrupted by later protrusions of massive basic and acid rocks over a rudely circular area, extending from the head of Loch Scridain to the Sound of Mull, and from Loch-na-keal to Loch Buy. The bedded basalts have been invaded by masses of dolerite, gabbro, and granophyre, with various allied kinds of rock. They have not only been disturbed in their continuity, but have undergone considerable metamorphism.

Again, further to the north, in the promontory of Ardnamurchan, the plateau has been disrupted in a similar way, and only a few recognisable fragments of it have been left. These changes will be more appropriately discussed in connection with similar phenomena in the other plateaux further north.

Spotted a typo or a problem in this chapter? Tell us

Chapter XXXVIII

THE BASALT-PLATEAU OF THE PARISH OF SMALL ISLES--RIVERS OF THE VOLCANIC PERIOD

iii. PARISH OF SMALL ISLES PLATEAU

The parish of Small Isles includes the islands of Eigg, Rum, Canna, Sanday and Muck (Map VI.). The fragmentary basalt-plateau which it contains, although the smallest of the whole series, is surpassed by none in the variety and interest of its geology. It contains by far the most complete records of the rivers which, during the volcanic period, flowed across the lava plains. And it alone has preserved a relic of the latest lava which, after the basalt-plateau had been built up and had been greatly eroded, flowed over the denuded surface in streams of volcanic-glass that found their way into a river-channel and sealed it up.

That the fragments of the basaltic plateau preserved in each member of the group of the Small Isles were once connected as a continuous volcanic plain can hardly be doubted. Indeed, as already stated, they were not improbably united with the plateau of Skye on the north, and with that of Mull, Morven and Ardnamurchan oh the south. Taking the whole space of land and sea within which the basalt of Small Isles is now confined, we may compute it at not much less than 200 square miles. In Eigg, Muck, Canna and Sanday the basalts retain their almost horizontal position, and from underneath them the Jurassic strata emerge in the first of these islands. The central part of the plateau in the island of Rum has suffered greatly from denudation. It now consists of four small outliers of basalt, which lie at levels of 1200 feet and upwards, on the western slope. The basalt is underlain by a thick mass of red Torridon Sandstone, which, with some gneisses and schists, forms the general underlying platform of this island. These rocks are doubtless a continuation of the red sandstone and schists of Sleat, in Skye, and like them have been subjected to those post-Cambrian convolutions and metamorphism whereby the Lewisian Gneiss and Torridon Sandstone have been brought above younger rocks, and have been crushed and rolled out so as to assume a new schistose arrangement. Before the time when volcanic action began, a mass of high ground, consisting of these ancient rocks, stood where the island of Rum is now situated. The streams of basalt spread around it, not only covering the surrounding low tracts of Jurassic rocks, but gradually accumulating against the hills, and thus reducing them both in area and in height above the plain.[246] Viewed from Canna the western coast of Rum presents a striking picture of the general relations of the volcanic masses of the Inner Hebrides and of the enormous denudation which they have undergone (Fig. 267). The Torridon Sandstones are there seen to mount into ranges of hills, capped with outliers of the basalt-plateau, while behind rise the great eruptive bosses of gabbro and granophyre. The edges of the sheets that form the outliers would, if prolonged, cover the northern or lower half of the island, where pre-Cambrian rocks form the surface. In the southern half, the continuity of the basalt has been partly obscured and partly destroyed by the protrusion of the great masses of gabbro that form the singularly picturesque mountain group to which this island owes its prominence as a landmark far and wide along the West Coast of Scotland.

[Footnote 246: That the lava-fields did not completely bury this nucleus of older rocks has been supposed to be shown by the fragments of red sandstone found in the ancient river-bed of Eigg, which was scooped out of the basalt-plateau and sealed up under pitchstone. But I am disposed to think that these fragments, together with those of Jurassic sandstone, came, not from Rum, but from some district more to the north and east, as will be explained in a later page. At Canna, a few miles to the west, fragments of red sandstone not improbably derived from Rum are abundant in the conglomerates between the basalts.]

The most varied and interesting of the fragments of the basaltic plateau in the area of the Small Isles is that which forms the island of Canna, with its appendage Sanday. Canna measures five miles in length by from half a mile to a mile in breadth, and consists entirely of the rocks of the plateau and their accompaniments. The basalts are exposed along the north coast in a range of mural precipices rising to a height of about 600 feet above the sea. From the top of that escarpment the ground falls by successive rocky terraces and grassy slopes to the southern shore-line. Sanday, connected with the large island by a shoal and foot-bridge, is two miles long and 220 to about 1200 yards broad. Its highest cliffs range along its southern shore to a height of 193 feet, whence they slope gently northward into the hollow between the two islands. This peculiar topography accounts for the manner in which the geological sections of most interest are distributed.

The first, and still the best, account of the geology of these islands is that of Macculloch. He showed that the rocks all belong to the series of the plateau-basalts, and he described the presence among them of a "trap-conglomerate." He noticed the occurrence also of trap-tuff and the occasional appearance of carbonized wood in these deposits. Reasoning upon these observations in his characteristically vague and verbose manner, "bewildered in the regions of conjecture," he concludes that the basalts instead of belonging to "one general formation" have been successively deposited on the same spot, "since lapse of time is evidently implied in the formation of a conglomerate." He inclines to believe that they have been discharged by ancient volcanoes from which in the course of time all traces of their original outline have been more or less completely removed, the existing basalts being merely fragments of once more extensive masses.[247]

[Footnote 247: Western Isles, vol. i. pp. 448-459, and pl. xix. Figs. 2, 3 and 4. See also Jameson's Mineralogy of the Scottish Isles.]

Macculloch regarded the intercalated-conglomerates as having been arranged under water and as marking pauses in the deposition of the sheets of "trap." He gave two diagrams in illustration of the relations of these detrital deposits, but he expressed no definite opinion as to their origin, though from one passage it would seem that he inclined towards the belief that they were formed in the sea.[248] Since his time, so far as I am aware, no fresh light has been thrown upon the subject.

[Footnote 248: Op. cit. pp. 449, 457, pl. xix. Figs. 2 and 3.]

During a yachting cruise in the summer of 1894 I visited Canna for the first time and found so much that was new to me in regard to the history of Tertiary volcanic action, and which demanded a careful survey, that I returned to the locality the following summer and remained in the island until I had mapped it and its dependencies upon the Ordnance Survey sheets on the scale of six inches to a mile. The following narrative is the result of the observations then made.

As far back as the year 1865 I published an account of an ancient river-channel which, during the volcanic period, had been eroded on the surface of the basalt-plateau, and of which a small portion had been preserved under a stream of pitchstone-lava that had flowed into and buried it.[249] This water-course, now marked by the picturesque ridge of the Scuir of Eigg, was shown to have been excavated by a stream which came from the north-east or east, and to be younger, not only than the plateau-basalts of the district, but than even the dykes which cut these basalts. Yet that it belonged to the volcanic period was proved by the manner in which it had been sealed up and preserved under the black glassy lava of the Scuir. Its history and the data from which this history is compiled will be narrated in a later part of this chapter.

[Footnote 249: Scenery of Scotland (1865); Quart. Journ. Geo. Soc. vol. xxvii. (1871), p. 303.]

My examination of the islands of Canna and Sanday, however, brought to light other and more abundant evidence of river-action in the same region of the Inner Hebrides, but belonging to an earlier part of the volcanic period. This evidence reveals that a powerful river, flowing westwards from the Highland mountains, swept over the volcanic plain, while the sheets of basalt were still being poured forth, and while volcanic eruptions were taking place from cones of slag.

The basalt-plateau of Canna resembles in all essential particulars those of the other Western Isles. Its base is everywhere concealed under the sea, but from the fragments of Torridon Sandstone in its agglomerates we may infer that it probably rests on that formation, like the volcanic outliers in Rum. It is formed of successive sheets of different basalts including the usual banded, amygdaloidal and columnar forms. Some of them towards the west are specially marked by the great abundance and large size of their porphyritic felspars. The magnetic properties of the basalts at the east end of the island have long been known, and have given rise to various modern myths regarding their influence on the compasses of passing vessels.

But it is in its conglomerates, tuffs and agglomerates and the light they cast on some aspects of the volcanic period, elsewhere hardly recorded, that the geology of Canna possesses a special importance. To these, therefore, we may at once turn.

The conglomerates are best developed at the eastern end of the island, where the cliffs present the structure represented in Fig. 268. At the base, and passing under the level of the sea, lies the agglomerate (a) of a vent which will be described in Chapter xli., together with other eruptive orifices of the various plateaux (p. 288). This rock has a somewhat uneven upper surface which rises in places about 150 feet above high tide-mark. Here and there it shades off upward into the conglomerate that overlies it; water-worn pebbles appear among its contents, and rude traces of bedding begin to show themselves, until, within the course of a few feet, we pass upward into an undoubted conglomerate. Elsewhere, however, and particularly along the precipices west of Compass Hill, the two deposits are more distinctly marked off from each other. The agglomerate has there a hummocky, irregular upper surface, as if it had been thrown down in heaps. The hollows between these protuberances have been filled up with conglomerate and sandstone, forming the base of the thick overlying deposit.

It is thus clear that the loose materials of the vent were directly exposed at the surface when the conglomerate was accumulated, and, indeed, that these materials served to supply some of the detritus of which the conglomerate consists. The absence of any trace of a cone and crater at the vent may perhaps be explicable on the supposition that their incoherent material was washed down by the currents that swept along and deposited the conglomerate.

The mass of sedimentary material (b) which overlies the agglomerate of the vent forms a conspicuous feature along the lower half of the precipices at the eastern end of Canna. It rises to a height of 250 to 300 feet above sea-level, and must reach a maximum thickness of probably not less than 100 to 150 feet. It gradually descends in a westward direction, both along the northern cliffs and in the lower ground round Canna Harbour, insomuch that in about a mile, owing to the gentle westerly dip of the whole volcanic series, combined with the effect of a number of small faults, it passes under the level of the sea.

Great variation in the character of the detritus composing this thick group of strata may be observed as it is followed westward. On the cliffs below Compass Hill, as represented in Fig. 268, the coarse conglomerate with water-worn stones, hardly to be distinguished from the volcanic agglomerate of the vent, shows more or less distinct bedding, or at least a succession of coarser and finer bands. Towards its base it encloses numerous pieces of Torridon Sandstone, sometimes subangular, but often so well and smoothly rounded as to show that they must have been long subjected to the action of moving water. It is further observable that, while in the agglomerate the volcanic stones have rough surfaces, those in the conglomerate begin to show increasing evidence of attrition, until, as the deposit is traced upwards, they become almost as well rounded and water-worn as the non-volcanic stones.

Yet amidst and overlying these proofs of transport from some little distance lie abundant huge slags and blocks of amygdaloidal lava, sometimes closely aggregated, sometimes scattered through a volcanic tuff or ashy sandstone. The composition and structure of these stones, and the manner of their dispersion through the deposit, leave little doubt that they were ejected from the vent. We are thus confronted with the interesting fact that, while the materials of the volcanic cone were being washed down by running water, eruptions were still taking place. But by degrees these indications of contemporaneous volcanic activity diminish. The detrital materials become coarser and more distinctly water-rolled until they pass into greenish sandstones and fine conglomerates. Yet the matrix even of these higher sediments is largely composed of fine volcanic detritus, and probably points to occasional discharges of dust and ashes.

Various sills or intrusive sheets have been injected into this sedimentary group along the precipices at the east end of Canna, and form there lenticular bands. One of these (c) is shown in Fig. 268.

Immediately above the massive greenish pebbly sandstone (d) which caps the stratified series lies a group of basalts (e), composed of several distinct beds, having a united thickness of from 80 to 100 feet. The lowest of these has a regular columnar structure, while those overlying it exhibit the confused starch-like grouping of curved and rather indistinctly-formed prisms.

The next band in upward succession is one of conglomerate (f), which runs as a continuous and conspicuous feature along the upper part of the cliff. This rock presents in many respects a strong contrast to the conglomerates underneath. It is dull-green to yellow in colour, and is well stratified, being marked by the interstratification of finer layers, and passing down into a band of pebbly sandstone, which rests immediately on the basalt (e). Its component stones are thoroughly water-worn, ranging up to six inches or even more in length. But its most distinctive character lies in the nature of its pebbles. Instead of consisting mainly of volcanic materials, these stones have almost all been transported for some distance. They include abundant fragments of Torridon Sandstone, gneiss, schists, grits, and other rocks like those in Rum and Western Inverness-shire. No such rocks exist in situ in Canna. The nearest tract of Torridon Sandstone is in Rum, about four miles to the eastward. But the pieces of schist and epidotic grit like the rocks of the Western Highlands, have probably travelled at least 30 miles.

It is important to observe that all these transported stones indicate a derivation from some source lying to the eastward of Canna. The evidence in this respect agrees with that furnished by the ancient river-gravel under the pitchstone of the Scuir of Eigg. It is clear that the waters which found their way across the lava-fields of this part of the Inner Hebrides took their rise somewhere to the eastward, probably among the mountains of Inverness-shire.

The conglomerate now described is from 40 to 50 feet thick. It can be followed along the face of the cliffs for more than a mile on the north side of Canna. Less persistent on the south side, its outcrop strikes from the edge of the precipice inland, keeping to the south of the top of Compass Hill. It is well seen in the ravine above the Coroghon, but cannot be followed further westward among the basalt-terraces. Yet, though this stratified intercalation is not traceable far as a band of conglomerate, the same stratigraphical horizon is probably indicated elsewhere by other kinds of sedimentary deposits, to which further reference will be made in the sequel.

The section now described establishes the existence of at least two successive platforms of conglomerate in the volcanic series. Following these platforms along their outcrop, we obtain additional light on their origin, and on the topographical conditions under which they were deposited, and we learn further that other prolonged intervals, which were likewise marked by intercalations of sedimentary material, occurred in the outpouring of the basalts.

Taking first the lower conglomerate of Compass Hill and tracing it westward, we find it to form the depression in which the sheltered inlet of Canna Harbour lies. It is exposed along the shores and also in the islands enclosed within the same bay. But it is not traceable further west, possibly because it seems to sink beneath the level of the sea. To the south-east, though it is there likewise for the most part concealed under the waves, it rises above them in one or two parts of the coast-line of Sanday, particularly at the Uamh Ruadh or Red Cave, and likewise on a surf-beaten skerry off Ceann an Eilein, the highest part of the Sanday cliffs--a distance of about a mile and a half from Compass Hill. Throughout this space it retains its remarkably coarse character and is mainly made up of volcanic material.

The numerous sections exposed in Canna Harbour enable us to study the composition and local variations of this curious deposit. On the north side of the basin, while the lower part of the sedimentary series continues to be an exceedingly coarse volcanic conglomerate, it passes upward into finer conglomerates, tuffs, and shales. In front of Canna House the imbedded blocks are of large size, occasionally as much as three or four feet in diameter. They are still more gigantic on the island of Eilean a' Bhaird, where I found one to contain 150 cubic feet in the exposed part, the rest being still concealed in the matrix. As they are generally somewhat rounded, here and there markedly so, most of these stones have probably undergone a certain amount of attrition in water. The great majority of them, and certainly all those of larger size, are pieces of basalt, dolerite, andesite, etc. Among them huge blocks of amygdaloid and coarsely vesicular lava are specially abundant. Some of these look like pieces of slag torn from the upper surface of lava-streams. Others, displaying a highly vesicular centre and a close-grained outer crust, are suggestive of bombs. It is interesting to note here again that the amygdaloidal blocks present their zeolitic infiltrations so precisely like those of the amygdaloids of the plateau that it seems reasonable to suppose the carbonate of lime, zeolites, etc. to have been introduced before the blocks were imbedded in the conglomerate.

The whole aspect of this deposit is eminently volcanic. It looks like a vast sheet of lava-fragments swept away from one or more cones of slags and cinders, or from the scoriaceous surface of a lava-stream. Where the vesicles were still empty, the large boulders could be more easily swept along by moving water. But a powerful current must have been needed to transport and wear down into more or less rounded forms blocks of basic lava, many of which must weigh several tons. The large block on Eilean a' Bhaird probably exceeds 12 tons in weight.

Besides the obviously volcanic contents of the conglomerate there occur here also, as in the Compass Hill cliffs, abundant pieces of Torridon Sandstone. These stones are notably smaller in size and more perfectly water-worn and even polished than the blocks of lava. Obviously they have travelled further and have undergone more prolonged attrition.

The matrix of the rock consists essentially of the fine detritus of basic lavas, probably mingled with true volcanic dust. The coarser parts display only the feeblest indication of stratification; indeed, in a limited exposure the rock might be regarded as a tumultuous agglomerate. But the manner in which the deposit is intercalated with, and sometimes overlies, green tuffs and shales, together with the water-worn condition of its stones, shows that it has not been accumulated in a volcanic chimney, but has been thrown down by some powerful body of water, with probably the co-operation of volcanic discharges.

While the composition of the conglomerate suffices to indicate that this deposit was formed at a time when some volcano was active in the immediate neighbourhood, singularly convincing proofs of the work of this vent are to be seen in the form of intercalated sheets of lava. Thus on Eilean a' Bhaird the boulders of the conglomerate are overlain and wrapped round by a sheet of rudely prismatic basalt, with lines of vesicles arranged in the direction of the bedding. A similar relation can be traced along the beach between Canna House and the wooden pier, where successive sheets of basalt have flowed over the conglomerate (Fig. 269).

But, besides coarse volcanic detritus, the sedimentary platform represented by the lower conglomerate of Compass Hill includes other deposits of which good sections may be examined all round Canna Harbour. Beds of fine well-stratified dull-green tuff pass by an admixture of pebbles into fine ashy conglomerate or pebbly sandstone, and by an increase in the proportion of their fine detritus into volcanic mudstone and fine shales. The shales vary from a pale grey or white tone into blackish grey, brown, and black. They are well stratified and are frequently interleaved with layers of fine tuff. The darker bands are carbonaceous, and are not infrequently full of ill-preserved vegetation. Indeed, leaves and stems in a rather macerated condition are of common occurrence in all the shaly layers. Here and there, especially in some ashy shales in front of Canna House, I observed a recognisable Sequoia. The mudstones are dull green, close-grained shattery rocks composed of fine volcanic detritus, and pass both laterally and vertically into shales, tuffs, and conglomerates. They suggest showers of fine dust or streams of volcanic mud. They, too, contain fragmentary plants.

It is a noteworthy fact that the sedimentary intercalations among the Canna basalts generally end upward in carbonaceous shales or coaly layers. The strong currents and overflows of water, which rolled and spread out the coarse materials of the conglomerates, gave way to quieter conditions that allowed silt and mud to gather over the water-bottom, while leaves and other fragments of vegetation, blown or washed into these quiet reaches, were the last of the suspended materials to sink to the bottom. Good illustrations of this sequence in the case of the lower conglomerate zone of Canna may be studied along the shores of Sanday, from the Catholic Chapel eastwards. The fine pebbly sandstones, tuffs, and shales, which there overlie the coarse conglomerate, are surmounted by dark brown or black carbonaceous shale, with lenticles of matted vegetation that pass into impure coal. Immediately overlying this coaly layer comes a sheet of prismatic vesicular basalt, followed by another with an exceedingly slaggy texture.

Lenticles of shale and mudstone likewise occur in the heart of the finer parts of the conglomerate, especially towards the top, as may be seen in the section exposed beneath the basalt behind the first cottage west from Canna House. One of the most interesting layers in this section is a seam of tuff, varying up to about two inches in thickness, which lies at the top of the lenticular band of tuffs and shales, and immediately beneath the band of basalt-conglomerate, on which a basalt, carrying a vesicular band near its bottom, rests. Traced laterally, the dark brown tuff of this seam gradually passes into a series of rounded bodies and flattened shells composed of a colourless mineral which has evidently been developed in situ after the deposition of the tuff. Mr. Harker's notes on thin slices made from this band are as follows:--

"This is a rusty-brown, dull-looking rock, rather soft and seemingly light, but too absorbent to permit of its specific gravity being tested. The dark brown mass is in great part studded with little spheroidal bodies, 1/50 to 1/10 inch in diameter, of paler colour, but the larger ones having a dark nucleus. In other parts larger flat bodies have been formed, as if by the coalescence of the spheroids, extending as inconstant bands in the direction of lamination for perhaps 1/2 inch, with a thickness of 1/10 inch or less. The appearance is that of a spherulitic rather than an oolitic structure.

"A slice [6658 A] shows the general mass of the rock to be of an extremely finely divided but coherent substance of brown colour, which can scarcely be other than a fine volcanic dust, composed of minute particles of basic glass or 'palagonite' compacted together. Scattered through this are fragments of crystals recognizable as triclinic and perhaps monoclinic felspars, green hornblende, augite, olivine (?), and magnetite, usually quite fresh.

"The curious spheroidal and elongated growths already mentioned are better seen in another slide [6658 B], where they occupy the larger part of the field, leaving only an interstitial framework of the brown matrix. The substance of the little spheroids is clear, colourless, and apparently structureless. The centre is often occupied by an irregularly stellate patch of brown colour, and sometimes cracks tend to run in radiating fashion, but these are the only indications of radial structure. The outer boundary is sharply defined, and where the slice is shattered the spheroids have separated from the matrix. The matrix is darker than in the normal rock, being obscured by iron-oxide which we may conceive as having been expelled from the spaces occupied by the spheroids. The little crystal-fragments are enclosed in the spheroids as well as in the matrix, but there is no appearance of their having served as starting-points for radiate growths. The flat elongated bodies are like the spheroids, with merely the modifications implied in their different shape.

"The identity of the clear colourless substance seems to be rather doubtful. It is sensibly isotropic and of refractive power distinctly lower than that of felspar. These characters would agree with analcime, which is not unknown as a contact-mineral; but it is difficult to understand how analcime, even a lime-bearing variety like that of Plas Newydd,[250] could be formed in abundance from palagonitic material. An alternative supposition, perhaps more probable, is that the clear substance is a glass, modified from its former nature, especially by the expulsion of the iron-oxide into the remaining matrix. A comparison is at once suggested with certain types of 'Knotenschiefer,' but respecting the thermal metamorphism of fine volcanic tuffs there seems to be little or no direct information."

[Footnote 250: Henslow, Trans. Camb. Phil. Soc. (1821), vol. i. p. 408; Mr. Harker, Geol. Mag. (1887), p. 414. Mr. W. W. Watts suggests a comparison with the hexagonal bodies figured by Mr. Monckton in an altered limestone from Stirlingshire: Quart. Journ. Geol. Soc., vol. li. p. 487.]

Lenticular interstratifications of shale and mudstone make their appearance even in the coarser parts of the conglomerate, as may be observed on the beach below Canna House where, as shown in Fig. 269, some shales and tuffs (a) full of ill-defined leaves are surmounted by a conglomerate (b). The deposition of this overlying bed of boulders has given rise to some scooping-out of the finer strata underneath. Subsequently both the conglomerate and shales have been overspread by a stream of dolerite (c), the slaggy bottom of which has ploughed its way through them.

Before discussing the probable conditions under which the group of sedimentary deposits now described was formed, we may conveniently follow the upper conglomerate band of the Compass Hill, and note the variations in structure and composition which its outcrop presents.

This yellowish conglomerate can be traced along the cliffs for more than a mile, when it descends below the sea-level at the solitary stack of Bod an Stòl. A few hundred yards further west, what is probably the same band appears again at the base of the precipice overlain by prismatic basalts. But the conglomerate, here only 12 feet thick, is made of much finer detritus which, largely composed of volcanic material, includes small well-rounded and polished pebbles of Torridon Sandstone. Beneath it lies a bed of dark shale, with remains of plants, resting immediately on a zeolitic amygdaloid which plunges into the sea. The chief interest of this locality is to be found in the shale which, instead of being at the top of the sedimentary group, lies at the bottom. I was informed by Mr. A. Thom that leaves had been obtained from this shale; but I was not successful in my search for them. The locality is only accessible by boat, and, as the coast is fully exposed to the Atlantic swell, landing at the place is usually difficult and often impossible.

About a mile and a half still further west, where a foreshore fronts the precipice of Earnagream at the Camas Tharbernish, a band of intercalated sedimentary material underlies the great escarpment of basalts and rests upon the slaggy sheet with the singular surface already referred to (p. 187). This band not improbably occupies the same platform as the upper conglomerate of Compass Hill. It is only about seven feet thick, the lower four feet consisting of a dull green pebbly tuff or ashy sandstone, with small rounded pieces of Torridon Sandstone, while the upper three feet are formed of dark shale with crowded but indistinct remains of plants. Here the more usual order in the sequence of deposition is restored. The shale is indurated and shattery, so that no slabs can be extracted without the use of quarrying tools.

Rather less than half a mile towards the south, on the roadside at the gully of Cùl nam Marbh, the basalts enclose a sedimentary interstratification which not improbably lies on the same horizon as those just described along the northern shore. The relations of the rocks at this locality are shown in Fig. 270. A remarkable slaggy basalt (a) rises into a hummock, against which have been deposited some fine granular tuffs (b) whereof only a few inches are visible, that pass up into a thin band of dark shale (c), including a layer of pebbly ferruginous tuff, with small rounded pea-like pieces of basalt, basic pumice, bole, limonite, etc. At the top of this shale an irregular parting of coaly material (d) lies immediately under the slaggy base of the succeeding basalt (e). It will be observed that this upper lava cuts out the shale and thus comes to rest directly upon the lower sheet. At the point where it begins to descend it has caught up and enclosed a small tree-stump (d´) which stands upright on the coaly parting and shale. This stump, at the time of my visit, measured five inches in height by three inches in breadth; it had been thoroughly charred and was crumbling away on exposure, but among the pieces which I took from it sufficient trace of structure can be detected with the microscope to show the tree to have been a conifer.

We have here another instance of the deposition of volcanic dust and fine mud in a pool that filled a hollow in the lava-field. Again we see that the closing act of sedimentation was the subsidence of vegetable matter in the pool, which was finally buried under another outflow of basalt.

It is on the southern coast of the isle of Sanday that the higher intercalations of sedimentary material among the basalts are most instructively displayed. At the eastern end of this island, as already stated, the lowest and coarsest conglomerate is visible on a skerry immediately to the south of the headland Ceann an Eilein. It doubtless underlies the Sanday cliffs, but is not there visible, for the basalts descend below sea-level. These volcanic sheets have a slight inclination westward; hence in that direction we gradually pass into higher parts of the series. In the Creag nam Faoileann (Seamews' Crag) and the gully that cuts its eastern end, likewise in the two singularly picturesque stacks of Dùn Mòr and Dùn Beag (Big and Little Gull Rocks), which here rise from the foreshore, two distinct platforms of detrital material may be noticed among the basalts. Both of these can be well seen on Dùn Mòr, about 100 feet high, which is represented in Fig. 271. The lower band, four or live feet thick, is here a rather coarse conglomerate which lies upon a sheet of scoriaceous basalt that extends up to the base of the Creag nam Faoileann. It is directly overlain by another basalt, about 30 feet thick, which dips seawards and forms a broad shelving platform, whereon the tides rise and fall. On this stack a second coarse conglomerate, about 10 feet thick, forms a conspicuous band about a third of the height from the bottom; it is composed mainly of well-rounded blocks of various lavas up to 18 inches or more in diameter, but it contains also pieces of Torridon Sandstone. It is covered by about 60 feet of basalt, which towards the base is somewhat regularly columnar, but passes upward into the wavy, starch-like, prismatic structure.

If now we trace these two intercalated zones of conglomerate along the shore, we find them both rapidly to change their characters and to disappear. The lower, though formed of coarse detritus under the Dùn Mòr, passes on the opposite cliff in a space of not more than 60 yards, into fine tuff and shale, about six feet thick, which become carbonaceous at the top where they are overlain by the next basalt. A hundred yards to the east, the band likewise consists of tuffs and ashy shales, which underlie the basalts on the Dùn Beag, and again show the usual coaly layers at the top. On the east side of the gully in the coast, about 160 yards to the north-east of Dùn Mòr, the same band is reduced to not more than three feet in thickness, consisting chiefly of fine conglomerate, wherein well water-worn pebbles of Torridon Sandstone and epidotic grit appear among the predominant volcanic detritus. This conglomerate is surmounted by a few inches of dark carbonaceous mudstone or shale. Rough slaggy basalts lie above and below the band.

The upper conglomerate dies out, both towards the east and the west, in the cliff opposite to Dùn Mòr, dwindling down at last to merely a few pebbles between the basalts. It lies in a kind of channel or hollow among these lavas. This depression, in an east and west direction, cannot be more than about 65 yards broad.

Probably still higher in the series of basalts is another intercalation of sedimentary layers which may be seen in the little bay to the east of Tallabric, rather more than a mile to the west of the Creag nam Faoileann. It rests upon a coarsely slaggy amygdaloid, and is from six to ten feet in thickness. The lower and larger part of the deposit consists of greenish pebbly sandstone and fine conglomerate, largely composed of basaltic detritus, but including abundant well-smoothed and polished pebbles of Torridon Sandstone, green grit, quartzite, etc. The stones vary from mere pea-like pebbles up to pieces two or three inches long, the largest being generally fragments of slag and amygdaloid which are less water-worn than the sandstones and other foreign ingredients. The uppermost two or three feet of the intercalation consist of dark carbonaceous mudstone or shale, made up in large measure of volcanic detritus, which may have been derived partly from eruptions of fine dust, partly from subærial disintegration of the basalt-sheets. Some layers of these finer strata are full of remains of much macerated plants.

Other thin coaly intercalations have been observed among the basalts of Canna, some of which may possibly mark still higher horizons than those now described. But, confining our attention to the regular sequence of intercalations exposed along the Sanday coast, we find at least four distinct platforms of interstratified sediment among the plateau-basalts of this district. Each of these marks a longer or shorter interval in the outflow of lava, and points to the action of moving-water over the surface of the lava-fields.

We may now consider the probable conditions under which this intervention of aqueous action took place. The idea that the sea had anything to do with these conglomerates, sandstones, and shales may be summarily dismissed from consideration. The evidence that the basalt-eruptions took place on a terrestrial surface is entirely convincing, and geologists are now agreed upon this question.

Excluding marine action, we have to choose among forms of fresh water--between lakes on the one hand and rivers on the other. That the agency concerned in the transport and deposition of these strata was that of a river may be confidently concluded on the following grounds:--

1. The large size and rolled shape of the boulders in the conglomerates. To move blocks several tons in weight, and not only to move them but to wear them into more or less rounded forms, must have required the operation of strong currents of water. The coarse detritus intercalated among the basalts is quite comparable to the shingle of a modern river, which descends with rapidity and in ample volume from a range of hills.

2. The evidence that the materials of the conglomerates are not entirely local, but include a marked proportion of foreign stones. The proofs of transport are admirably exhibited by pieces of Torridon Sandstone, epidotic grit, quartzite, and other hard rocks none of which occur in situ except at some distance from Canna. These stones are often not merely rounded, but so well smoothed and polished as to show that they must have been rolled along for some considerable time in water.

3. The lenticular character and rapid lithological variations of the strata, both laterally and vertically. The coarse conglomerates die out as they are followed along their outcrop and pass into finer sediment. They seem to occur in irregular banks, which may not be more than 200 feet broad, like the shingle-banks of a river. The coarser sediment generally lies in the lower part of the sedimentary group. But cases may be observed, such as that shown in Fig. 269, where fine sediment, laid down upon the bottom conglomerate, has subsequently been overspread by another inroad of coarse shingle. Such alternations are not difficult to understand if they are looked upon as indicating the successive floods and quieter intervals of a river.

For these reasons I regard the platforms of sedimentary materials intercalated among the basalts of Canna and Sanday as the successive flood-plains of a river which, like the rivers that traverse the lava-deserts of Iceland, flowed perhaps in many separate channels across the basalt-fields of the Inner Hebrides, and was liable to have its course shifted from time to time by fresh volcanic eruptions. That this river came from the east or north-east and had its source among the Western Highlands of Inverness-shire, may be inferred from the nature of the stones which it has carried for 30 miles or more along its bed. And that it crossed in its course the tract of Torridon Sandstone, of which a portion still remains in Rum, is manifest from the abundance of the fragments of that formation in the conglomerates.

With the remarkable exception of the section on Dùn Beag, to be immediately referred to, no trace of any eroded channel of this river through the lavas of the great volcanic plain has been preserved. Possibly frequent invasions of its bed by streams of basalt from different vents hindered it from remaining long enough in one course to erode anything like a gorge or canon. But, in any case, the main channel of the river probably lay rather to the east of the present islands of Canna and Sanday, on ground which is now covered by the sea. The banks or sheets of boulder-conglomerate undoubtedly show where its current swept with great force over the lava-plain, but the manner in which these coarser materials are so often covered with fine silt suggests that the sedimentary materials now visible were rather deposited on the low grounds over which the steam rushed in times of flood. Pools of water would often be left after such inundations, and in these depressions silt would gradually accumulate, partly carried in suspension by the river, partly washed in by rain, while drift-wood that found its way into these eddies, and leaves blown into them from the trees and shrubs of the surrounding country, would remain for some time afloat and would be the last of the detritus to sink to the bottom. Hence, no doubt, the carbonaceous character of the hardened silt in the upper part of each intercalation of sediment.

If we were to look upon the volcanic materials in the conglomerates as derived from the subærial disintegration of the fields of basalt, we should be compelled to admit a very large amount of erosion of the surface of the volcanic plain during the period when the river flowed over that tract. It would be necessary to suppose not only that there was a considerable rainfall, but that the differences of temperature, either from day to night, or from summer to winter, were so great as to split up the lavas at the surface, in order to provide the river with the blocks which it has rolled into rounded boulders. I do not think, however, that such a deduction would be sound. If we compare the materials that have filled up the large eruptive vent at the east end of Canna (to be afterwards described) with the great majority of the blocks in the coarse conglomerates, we cannot fail to note their strong resemblance. The abundance of lumps of slaggy lava in the river-shingle corresponds with their predominance in the agglomerate of the vent. The boulders of basalt, dolerite, and andesite which crowd the conglomerates need not have been derived from the action of atmospheric waste on the lava-fields, but might quite well have been mainly supplied by the demolition of volcanic cones of fragmental materials.

That such has really been the chief source of the blocks in the conglomerates I cannot doubt. At the east end of Canna we actually detect a volcanic cone, partly washed down and overlain by a pile of river-shingle. There were probably many such mounds of slag and stones along lines of fissure all over the lava-fields. The river in its winding course might come upon one cone after another, and during times of flood, or when its waters burst through any temporary barrier created by volcanic operations it would attack the slopes of loose material and sweep their detritus onward. At the same time, the current would carry forward its own natural burden of far-transported sediment, and hence on its flood-plains, buried and preserved under sheets of basalt, we find abundant pebbles of the old Highland rocks which it had borne across the whole breadth of the basaltic lowland.

But the destruction of volcanic cones was probably not the only source of the basaltic detritus in the conglomerates of Canna and Sanday. I have shown that these conglomerates pass laterally into tuffs, and are sometimes underlain, sometimes overlain, with similar material. It is quite obvious that their deposition was contemporaneous with volcanic action in the immediate neighbourhood, and that at least part of their finer sediment was obtained directly from volcanic explosions. In wandering over the coast-sections of these coarse deposits, I have been impressed with the enormous size of many of the stones, their resemblance to the ejected blocks of the agglomerate, and the distinction that may sometimes be made with more or less clearness between their rather angular forms and the more rounded and somewhat water-worn aspect of the other boulders. It seems to me not improbable that some of the remarkably coarse masses of unstratified conglomerate in Canna Harbour consist largely of ejected blocks from the adjacent vent.

The only instance which I have observed of erosion of the basalt contemporaneous with the operations of the river that spread out this conglomerate is to be found in the striking stack of Dùn Beag already alluded to.[251] This extraordinary monument of geological history forms an outlying obelisk which rises from the platform of the shore to a height of about 70 feet. Seen from the south-west, it appears to consist entirely of bedded basalt resting on some stratified tuff and shale which intervene between these lavas and that of the broad platform of basalt on which the obelisk stands. On that side it presents no essential difference from the structure of the Dùn Mòr to the west, save that the lower conglomerate of that outlier is here represented by fine sediment, and the upper conglomerate is wanting. The general aspect of this south-western front of the stack is shown in Fig. 272. If, however, we approach the rock from the coast-gully to the north, we form a very different impression of its structure. It then appears to consist chiefly of conglomerate with a capping of basalt on the top (Fig. 273). Not until a close scrutiny is made of the eastern and western faces of the column do the true structure and history of this singular piece of topography become apparent.

[Footnote 251: This pinnacle of rock is referred to by Macculloch in his account of Canna, and is figured in Plate xix. Fig. 3 of his work already cited. But neither his description nor his drawing conveys any idea of the real structure of the rock.]

On the eastern front, the section represented in Fig. 274 is exposed. At the bottom, forming the pediment of the column, lies a sheet of slaggy and vesicular or amygdaloidal basalt (a), which shelves gently in a south-westerly direction into the sea. The lowest band (b) in the structure of the stack is a thin group of lilac, brown, and green shale and volcanic mudstone or tuff, which encloses pieces of coniferous wood, and becomes markedly carbonaceous in its uppermost layers. Above these strata on the south front comes the pile of bedded basalts (c) with their slaggy lower and upper surfaces. But as we follow them round the east side, we find them to be abruptly cut off by a mass of conglomerate (d). That the vertical junction-line is not a fault is speedily ascertained. The lower platform of slaggy basalt runs on unbroken under both shales and conglomerate. Moreover, the line of meeting of this conglomerate with the basalts that overlie the shales is not a clean-cut straight wall, but displays projections and recesses of the igneous rocks, round and into which the materials of the conglomerate have been deposited. The pebbles may be seen filling up little crevices, passing under overhanging ledges of the basalts, and sharply truncating lines of scoriaceous structure in these rocks. The same relations may be observed on the west front of the stack. There the ashy shales and tuffs are sharply cut out by the conglomerate, which wraps round and underlies a projecting cornice of the slaggy bottom of the basalt that rests on the stratified band (Fig. 275).

The conglomerate is rudely stratified horizontally, its bedding being best shown by occasional partings of greenish sandstone. It consists of well-rounded, polished, and water-worn stones, chiefly of members of the volcanic series--basalts, and dolerites, both compact and amygdaloidal or slaggy--but with a conspicuous admixture of Torridon Sandstone, gneiss, grey granite, grit and different schists. The coarsest part of the deposit lies toward the bottom where the volcanic blocks, some of them being six and eight feet in diameter, may have originally fallen from the basalts against which the conglomerate now reposes. The far-transported stones are also of considerable size, pieces of granite and gneiss frequently exceeding a foot in length. The well-rounded pebbles of foreign materials have been washed into the interstices between the large volcanic blocks.

It is, I think, tolerably clear that the wall of basalt against which this conglomerate has been laid down is one of erosion. The beds of basalt have here been trenched by some agent which has likewise scooped out the soft underlying shales, and even cut them away from under their protecting cover of basalt. There can be little hesitation in regarding this agent as a water-course, which for some considerable interval of time continued to dig its channel through the hard basalts. There is not room enough between the basalt-wall of Dùn Beag and the opposite cliffs of the shore (where no trace of this conglomerate is to be seen) for any large stream to have found its way. I do not therefore seek to identify this relic of an ancient waterway with the channel of the main river which deposited the conglomerate bands of Canna and Sanday. More probably it was either a mere torrential chasm, or a tributary stream draining a certain part of the volcanic plateau and allowed to retain its channel long enough to be able to erode it to a depth of nearly 50 feet. Erosion had reached down through the underlying tuffs to the slaggy basalt below, but before it had made any progress in that sheet its operations were brought to an end at this locality by the floods that swept in the coarse shingle, and by the subsequent stream of basalt of which a mere outlying fragment now forms the upper third of the stack (e, Fig. 274).

That the ravine or gully of Dùn Beag probably lay within the reach of the floods of the main river, may be inferred from the number and size of the far-transported rocks in its conglomerate. It was filled up gradually, but the conditions of deposition remained little changed during the process, except that the largest blocks of rock were swept into the chasm in the earlier part of its history, while much smaller and more water-worn shingle were introduced towards the close.

Denudation, which has performed such marvels in the topography of the West of Scotland since older Tertiary time, has here obliterated every trace of this ancient gully, save the little fragment of one of the walls which survives in the stack of Dùn Beag. When in the course of centuries this picturesque obelisk shall have yielded to the action of the elements, the last leaflet of one of the most interesting chapters in the geological history of the Inner Hebrides will have been destroyed.

The question naturally arises--What was the subsequent history of the river which has left so many records of its floods entombed among the basalts of Canna and Sanday? In particular, can any connection be traced or plausibly conjectured between it and the river-bed preserved under the Scuir of Eigg? To this question I shall return after the evidence for the existence and date of the latter stream has been laid before the reader.

In the chain of the Inner Hebrides, broken as it is in outline and varied in its types of scenery, there is no object more striking than the island of Eigg. Though only about five miles long and from a mile and a half to three miles and a half broad, and nowhere reaching a height of so much as 1300 feet, this little island, from the singularity of one feature of its surface, forms a conspicuous and familiar landmark. Viewed in the simplest way, Eigg may be regarded as consisting of an isolated part of the basaltic plateau which, instead of forming a rolling tableland or a chain of hills with terraced sides, as in Antrim, Mull and Skye, has been so tilted that, while it caps a lofty cliff about 1000 feet above the waves at the north end, it slopes gently along the length of the island to the south end. In its southern half, however, the ground rises, owing to the preservation of an upper mass of lavas, which denudation has removed from the northern half. On this thicker part of the plateau stands the distinguishing feature of the island, the strange fantastic ridge of the Scuir, which, seen from the north or south, looks like a long steep hill-crest, ending in a sharp precipice on the east. Viewed from the east, this precipice is seen to be the end of a huge mountain-wall, which rises vertically above the basalt-plateau to a height of more than 350 feet. The accompanying map (Fig. 276) shows that the ridge of the Scuir corresponds with the area occupied by a mass of pitchstone, and that while the basaltic rocks cover the whole of the rest of the southern half of the island, they gradually rise towards the north, successive members of the Jurassic series making their appearance until, at the cliffs of Dunan Thalasgair, the latter cover the greater part of the surface, and leave the volcanic rocks as a mere stripe capping the cliffs. In the section (Fig. 277) the general structure of the island is represented.

In Eigg the fragment of the basalt-plateau which has been preserved, rests unconformably on successive platforms of the Jurassic formations. Its component sheets of lava rise in cliffs around the greater part of the island. As they dip gently southwards their lower members are seen along the northern and eastern shores, while on the south-west side their higher portions are exposed in the lofty precipices which there plunge vertically into the sea. The total thickness of the volcanic series may here be about 1100 feet. The rocks consist of the usual types--black, fine-grained, columnar and amorphous basalts, more coarsely crystalline dolerites, dull earthy amygdaloids with red partings, and occasional thin bands of basalt-conglomerate or tuff. The individual beds range in thickness from 20 to 50 or 60 feet. Though they seem quite continuous when looked at from the sea, yet, on closer examination, they are found not unfrequently to die out, the place of one bed being taken by another, or even by more than one, in continuation of the same horizon. The only marked petrographical variety which occurs among them is a light-coloured band which stands out conspicuously among the darker ordinary sheets of the escarpment on the east side of the island. The microscopic characters of this rock show it to belong to the same series of highly felspathic, andesitic, or trachitic lavas as the "pale group" of Ben More, in Mull. It is strongly vesicular, and the cells are in some parts so flattened and elongated as to impart a kind of fissile texture to the rock. There can be no doubt that this band is a true lava, and that it was poured out during the accumulation of the basalt-plateau. It supplies an interesting example of the intercalation of a lighter and less basic lava among the ordinary heavy basic basalts and dolerites.

That feature of the island of Eigg which renders it so remarkable and conspicuous an object on the west coast is the long ridge of the Scuir. Rising gently from the valley which crosses the island from Laig Bay to the Harbour, the basaltic plateau ascends south-westwards in a succession of terraces, until along its upper part it forms a long crest, from 900 to 1000 feet above the sea, to which it descends on the other or south-west side, first by a sharp slope, and then by a range of precipices. Along the watershed of this crest runs, in a graceful double curve, the abrupt ridge of the Scuir, terminating on the north-west at the edge of the great sea-cliff (975 feet), and ending off on the south-east in that strange well-known mountain-wall (1272 feet high) which rises in a sheer cliff nearly 300 feet above the basalt-plateau on the one side and more than 400 feet on the other (Fig. 278). The total length of the Scuir ridge is two miles and a quarter, its greatest breadth 1520, its least breadth 350 feet. Its surface is very irregular, rising into minor hills and sinking into rock-basins, of which nine are small tarns, besides still smaller pools, while six others, also filled with water, lie partly on the ridge and partly on the basaltic plateau. No one, indeed, who looks on the Scuir from below, and notes how evenly it rests upon the basalt-plateau, would be prepared for so rugged a landscape as that which meets his eye everywhere along the top of the ridge. Two minor arms project from the east side of the ridge; one of these forms the rounded hill called Beinn Tighe (968 feet), the other the hill of A chor Bheinn.

Singular as the Scuir of Eigg is, regarded merely as one of the landmarks of the Hebrides, its geological history is not less peculiar. The natural impression which arises in the mind when this mountain comes into view for the first time is, that the huge wall is part of a great dyke or intrusive mass which has been thrust through the older rocks.[252] It was not until after some time that the influence of this first impression passed off my own mind, and the true structure of the mass became apparent.

[Footnote 252: Hay Cunningham remarks:--"In regard to the relations of the pitchstone-porphyry of the Scuir and the trap-rocks with which it is connected, it can, after a most careful examination around the whole mass, be confidently asserted that it exists as a great vein which has been erupted through the other Plutonic rocks--thus agreeing in age with all the other pitchstones of the island." Macculloch leaves us to infer that he regarded the rock of the Scuir to be regularly interstratified with the highest beds of the dolerite series (Western Isles, i. p. 522). Hugh Miller speaks of the Scuir of Eigg as "resting on the remains of a prostrate forest."--Cruise of the Betsy, p. 32.]

The ridge of the Scuir, presenting as it does so strong a topographical contrast to the green terraced slopes of the plateau-basalts on which it rests, consists of some very distinct bands of black and grey lava, long known as "pitchstone-porphyry." To the nature and history of these rocks I shall return after we have considered a remarkable bed of conglomerate which lies below them. On the lower or southern side of the ridge the bottom of the pitchstone, dipping into the hill, is exposed on the roof of a small cave where the ends of its columns form a polygonal reticulation. It is there seen to repose upon a bed of breccia or conglomerate, having a pale-yellow or grey felspathic matrix like the more decomposing parts of the grey devitrified parts of the pitchstone. Through this deposit are dispersed great numbers of angular and subangular pieces of pitchstone, some of which have a striped texture. Fragments of basalt, red (Torridon) sandstone, and other rocks are rare; and the bed suggests the idea that it is a kind of brecciated base or floor of the main pitchstone mass. A similar rock is found along the bottom of the pitchstone on both sides of the ridge (c, in Fig. 279). Here and there where this breccia is only a yard or two in thickness, it consists of subangular fragments of the various dolerites and basalts of the neighbourhood, together with pieces of red sandstone, quartzite, clay-slate, etc. The matrix is in some places a mass of hard basalt debris; in others it becomes more calcareous, passing into a sandstone or grit in which chips and angular or irregular-shaped pieces of coniferous wood are abundant.[253] A little further east, beyond the base of the Scuir, a patch of similar breccia is seen, but with the stones much more rounded and smoothed. This outlier rests against the denuded ends of the basalt-beds forming the side of the hill. Its interest arises from the evidence it affords of the prolongation of the deposit eastward, and consequently of the former extension of the precipice of the Scuir considerably beyond its present front.

[Footnote 253: The microscopic structure of this wood was briefly described by Witham (Fossil Vegetables, p. 37), and two magnified representations were given to show its coniferous character. Lindley and Hutton further described it in their "Fossil Flora," naming it Pinites eiggensis, and regarding it as belonging to the Oolitic series of the Hebrides--an inference founded perhaps on the erroneous statement of Witham to that effect. William Nicol corrected that statement by showing that the wood-fragments occurred, not among the "lias rocks," but "among the debris of the pitchstone" (Edin. New Phil. Journal, xviii. p. 154). Hay Cunningham, in the paper already cited, states that the fossil wood really lies in the pitchstone itself! The actual position of the wood, however, in the breccia and conglomerates underlying the pitchstone is beyond all dispute. I have myself dug it out of the bed. The geological horizon assigned to this conifer, on account of its supposed occurrence among Oolitic rocks, being founded on error, no greater weight can be attached to the identification of the plant with an Oolitic species. Our knowledge of the specific varieties of the microscopic structure of ancient vegetation is hardly precise enough to warrant us in definitely fixing the horizon of a plant merely from the examination of the minute texture of a fragment of its wood. From the internal organization of the Eigg pine, there is no evidence that the fossil is of Jurassic age. From the position of the wood above the dolerites and underneath the pitchstone of the Scuir it is absolutely certain that the plant is not of Jurassic but of Tertiary date.]

It is at the extreme north-western extremity of the pitchstone ridge, however, that the most remarkable exposure of this intercalated detrital band is now to be seen. Sweeping along the crest of the plateau the ridge reaches the edge of the great precipice of Bideann Boidheach, by which its end is truncated, so as to lay open a section of the gravelly deposit along which the pitchstone flowed.

The accompanying diagram (Fig. 279) represents the natural section there exposed. Rising over each other in successive beds, with a hardly perceptible southerly dip of 2°, the sheets of basalt form a mural cliff about 700 feet high. The bedded character of these rocks and their alternations of compact, columnar, amorphous and amygdaloidal beds are here strikingly seen. They are traversed by veins and dykes of an exceedingly close-grained, sometimes almost flinty, basalt. But the conspicuous feature of the cliff is the hollow which has been worn out of these rocks, and which, after being partially filled with coarse conglomerate, has been buried under the huge pitchstone mass of the Scuir. The conglomerate consists of water-worn fragments, chiefly of dolerite and basalt, but with some also of the white Jurassic sandstones, imbedded in a compacted sand derived from the waste of the older volcanic rocks. The grey devitrified bands in the pitchstone, so conspicuous at the east end of the Scuir, here disappear and leave the conglomerate covered by one huge overlying mass of glassy pitchstone.

If any doubt could arise as to the origin of the mass of detritus exposed under the pitchstone at the east end of the Scuir it would be dispelled by the section at the west end, which shows with unmistakable clearness that the conglomerate is a fluviatile deposit and lies in the actual channel of the ancient river which was eroded out of the basalt plateau, and was subsequently sealed up by streams of pitchstone-lava.

An examination of the fragments of rock found in the conglomerate affords here, as in Canna and Sanday, some indication of the direction in which the river flowed. The occurrence of pieces of red sandstone, which no one who knows West-Highland geology can fail to recognize as of Torridonian derivation, at once makes it clear that the higher grounds from which they were borne probably lay to the north or north-east. The fragments of white sandstone may also have been derived from the same quarter, for the thick Jurassic series of Eigg once extended further in that direction. The pieces of quartzite and clay-slate bear similar testimony to an eastern or north-eastern source. In short, there seems every probability that this old Tertiary river flowed through a forest-clad region, of which the red Torridon mountains of Ross-shire, the white sandstone cliffs of Raasay and Skye, and the quartzite and schist uplands of Western Inverness-shire are but fragments, that it passed over a wide and long tract of the volcanic plateau, and continued to flow long enough to be able to carve out for itself a channel on the surface of the basalt. Its course across what is now the island of Eigg took a somewhat north-westerly direction, probably guided by inequalities on the surface of the lava-plain. It is there marked by the winding ridge of the Scuir, the pitchstone of which flowed into the river-bed and sealed it up. Several minor spurs, which project from the eastern side of the main ridge, show the positions of small tributary rivulets that entered the principal channel from the slopes of the basaltic tableland. One of these, on the south-east side of the hill called Corven, must have been a gully in the basalt with a rapid or waterfall. The pitchstone has flowed into it, and some of the rounded pebbles that lay in the channel of this vanished brook may still be gathered where the degradation of the pitchstone has once more exposed them to the light. That the Eigg river here flowed in a westerly direction may be inferred from the angle at which the beds of the small tributaries meet the main stream, and also from the fact that the old river-bed at the east end of the Scuir is considerably higher than at the west end.

Several features in the geological structure of this locality serve to impress on the mind the great lapse of time represented by the erosion of the river-channel of Eigg. Thus at the narrowest point of the pitchstone ridge, near the little Loch a' Bhealaich, the bottom of the glassy lava is about 200 feet above its base on the south side, so that the valley cut out of the plateau-basalts must have been more than 200 feet deep. Even the little tributaries had cut ravines or cañons in the basalts before the ground was buried under the floods of pitchstone. In the most northerly spur of the ridge, for example, the hill of Beinn Tighe, which represents one of these tributaries, shows a considerable difference between the level of the bottom of the pitchstone on the east and west sides.

Again, all along the ridge of the Scuir, the basalt-dykes are abruptly cut off at the denuded surface on which the pitchstone rests. This feature is conspicuously displayed on the great sea-wall at the west end (Fig. 279). The truncation of the dykes demonstrates that a considerable mass of material must have been eroded before these lava-filled fissures could be laid bare at the surface. And the removal of this material shows that the denudation must have been continued for a long period of time.

The river-channel of Eigg, since it was eroded long after the cessation of the outflows of basalt in the plateau of Small Isles, must be much later in origin than those of Canna and Sanday which, as we have seen, were contemporaneous with the basalt-eruptions. But the river that excavated the channels and deposited the gravels may have been the same in both areas.

In dealing with this subject, though the evidence is admittedly scanty, we are not left wholly to conjecture. A consideration of the general topographical features of the wide region of the Inner Hebrides, from the beginning of the volcanic period onward, will convince us that, in spite of the effects of prolonged basalt-eruptions, the persistent flow of the drainage of the Western Highlands must have taken a westerly direction. It was towards the west that the low grounds lay. Though the long and broad valley which stretched northwards from Antrim, between the line of the Outer Hebrides and the West of Scotland, was gradually buried under a depth of two or three thousand feet of lava, the volcanic plain that overspread it probably remained even to the end lower than the mountainous Western Highlands. Hence the rivers, no matter how constantly they may have had their beds filled up and may have been driven into new channels, would nevertheless always seek their way westwards into the Atlantic.

On Canna and Sanday the traces of a river are preserved which poured its flood-waters across the lava-fields in that part of the volcanic region, while streams of basalt were still from time to time issuing from vents and fissures. Not more than fourteen miles to south-east stands the Scuir of Eigg, with its buried river-channel and its striking evidence that there, also, a river flowed westwards, but at a far later time, when the basalt-eruptions had ceased and the volcanic plain had been already deeply trenched by erosion, yet before the subterranean fires were finally quenched, as the pitchstone of the Scuir abundantly proves.

When one reflects upon the enormous denudation of this region, to which more special reference will be made in the sequel, one is not surprised that many connecting links should have been effaced. The astonishment rather arises that so continuous a story can still be deciphered. Even, however, had the original record been left complete, it would have been exceedingly difficult to trace the successive mutations of a river-channel during long ages of volcanic eruptions. Such a channel would have been concealed from view by each lava-stream that poured into it, and would not have been again exposed save by the very process of erosion that destroys while it reveals.

While, therefore, there is not and can never be any positive proof that in the fluviatile records of Canna, Sanday and Eigg successive phases are registered in the history of one single stream, I believe that this identity is highly probable. It was a river which seems to have risen among the mountains of Western Inverness-shire, and it had doubtless already taken its course to the sea before any volcanic eruptions began. It continued to flow westwards across the lava-floor that gradually spread over the plains. Its channel was constantly being filled up by fresh streams of basalt or deflected by the uprise of new cinder-cones. But, fed by the Atlantic rains, it maintained its seaward flow until the general subsidence which carried so much of the volcanic plain below the sea. Yet the higher part of this ancient water-course is no doubt unsubmerged, still traversing the schists of the Western Highlands as it has done since older Tertiary time. It may, perhaps, be recognized in one of the glens which carry seaward the drainage of the districts of Morar, Arisaig, or Moidart.

Let us now turn to the remarkable lava which has sealed up the river-channel of Eigg, and of which the remaining fragment stands up as the great ridge of the Scuir. This rock presents characters that strongly distinguish it from the surrounding basalts. It is not one single uniform mass, but consists of a number of distinct varieties, some of which are a volcanic glass, while others are a grey "porphyry," or devitrified pitchstone. These are arranged in somewhat irregular, but well-marked, and, in a general sense, horizontal sheets. On the great eastern terminal gable of the Scuir this bedded structure is not clearly displayed, for the cliff seems there to be built up of one homogeneous mass, save a markedly columnar band that runs obliquely up the base of the precipice (Fig. 278). If, however, the ridge is looked at from the south, the truly bedded character of its materials becomes a conspicuous feature. Along the cliffs on that side the two varieties of rock are strongly distinguished by their contrasting colour and mode of weathering, the sombre-hued pitchstone standing up in a huge precipice striped with columns, and barred horizontally with bands of the pale-grey "porphyry," which, from its greater proneness to decay, seems sunk into the face of the cliff. At the south-east end of the ridge the bedding is especially distinct. West of the precipices, to the south of the Loch a' Bhealaich, the dark pitchstone which forms the main mass is divided by two long parallel intercalations of grey rock, and two other short lenticular seams of the same material (see Figs. 280, 281). It is clear from these features, which are not seen by most travellers who pass Eigg in the tourist-steamer that the Scuir is in no sense of the word a dyke.

But although the Scuir is thus a bedded mass, the bedding is far different from the regularity and parallelism of that which obtains among the bedded basalt-rocks below. Even where no intervening "porphyry" occurs, the pitchstone can be recognized as made up of many beds, each marked by the different angle at which its columns lie. And when the "porphyry" does occur and forms so striking a division in the pitchstone, its beds die out rapidly, appearing now on one horizon, now on another, along the face of the cliffs, and thickening and thinning abruptly in short distances along the line of the same bed. Perhaps the best place for examining these features is at the Bhealaich, the only gully practicable for ascent or descent, at the south-eastern face of the ridge.

By much the larger part of the mass of the Scuir consists of vitreous material. As a rule this rock is columnar, the columns being much slimmer and shorter than those of the basalt-rocks. They rise sometimes vertically, and often obliquely, or project even horizontally from the face of the cliff. They are seldom quite straight, but have a wavy outline; and when grouped in knolls here and there along the top of the ridge they remind one of gigantic bunches of some of the Palæozoic corals, such as Lithostrotion. In other cases they slope out from a common centre, and show an arrangement not very unlike that of a Highland peat-stack.

The pitchstone of the Scuir differs considerably in petrographical character from other pitchstones of the island which occur in dykes and veins. Its base is of a velvet-black colour, and is so much less vitreous in aspect than ordinary pitchstone as to have been described by Jameson and later writers as intermediate between pitchstone and basalt.[254] A chemical analysis of the rock by Mr. Barker North,[255] gave the following composition:--

Silica 65·81 Alumina 14·01 Ferric oxide 4·43 Lime 2·01 Magnesia 0·89 Soda 4·15 Potash 6·08 Loss in ignition 2·70 ------ 100·08

[Footnote 254: Mineralogy of the Scottish Isles, vol. ii. p. 47. See also Macculloch, Western Isles, vol. i. p. 521, and Hay Cunningham, Mem. Wern. Soc. vol. viii. p. 155.]

[Footnote 255: Quart. Journ. Geol. Soc. vol. xlvi. (1890), p. 379.]

The grey devitrified bands, which occur as a subordinate part of the mass of the Scuir ridge, are usually somewhat decomposed. Where a fresh fracture is obtained, the material shows a fine-grained, sometimes almost flinty, grey felsitic base, containing clear granules of quartz, and facets of glassy felspar. In some places the rock is strongly porphyritic. Examined under the microscope it presents a more thoroughly devitrified groundmass, with the minutest depolarizing microlites, large porphyritic crystals of plagioclase and sanidine, grains of augite, and sometimes exceedingly abundant particles of magnetite.[256]

[Footnote 256: The microscopic structure of the identical pitchstone of Hysgeir is given on p. 247.]

Although the line of separation between the grey dull felsitic sheets and the more ordinary glassy pitchstone is usually well defined, the two rocks may be observed to shade into each other in such a manner as to show that the lithoid material is only a devitrified and somewhat decomposed condition of the glassy rock. This connection is particularly to be observed under the precipice at the east end of the Scuir. At that locality the pitchstone is underlain by a very hard flinty band, varying in colour from white through various shades of flesh-colour and brown into black, containing a little free quartz and crystals of glassy felspar. Where it becomes black it passes into a rock like that of the main mass of the Scuir. Such vitreous parts of the bed lie as kernels in the midst of the more lithoid and decomposed rock. The lower six feet of the "porphyry" are white and still more decomposed. The relations of this mass are represented in Fig. 282, where the basalt-rocks of the plateau (a) are shown to be cut through by basalt dykes (b b), and overlain by the "porphyry" (c) and the pitchstone (d). In the porphyry are shown several pitchstone kernels (p, p). It is deserving of remark also that in different parts of the Scuir, particularly along the north side, the bottom of the pitchstone beds passes into a dull grey earthy lithoid substance, like that now under description.

The bedded character of the rock of the Scuir and the well-marked lithological distinction between its several component sheets show the lava to have been the product of a number of separate outflows that found their way one after another into the river-valley, which was the lowest ground in the vicinity of the active vent. There can be little doubt, I think, that the lava flowed down the valley. Its successive streams are still inclined from east to west. The vent of eruption, therefore, ought to be looked for towards the east. Nowhere within the Tertiary volcanic region is there any boss of pitchstone or any mass the shape or size of which is suggestive of this vent. In the island of Eigg no boss of any kind exists, save those of granophyric porphyry to be afterwards referred to. But none of these affords any satisfactory links of connection with the rock of the Scuir. More probably the vent lay somewhere to the east on ground now overflowed by the sea. The pitchstone veins of Eigg may represent some of the subterranean extrusions from the same volcanic pipe, and if so, its site could not be far off.

The rock of the Scuir of Eigg has a special importance in the history of the volcanic plateaux. It is, so far as we know, the latest of all the superficial lavas of Britain.[257] From the basalts on which it rests it was separated by an enormous interval of time, during which these older lavas were traversed by dykes and were worn down into valleys. Its presence shows that long after the basalts of Small Isles had ceased to be erupted, a new outbreak of volcanic activity took place in this district, when lavas of a more acid composition flowed out at the surface. Whether this outburst was synchronous with the appearance of the great granophyric protrusions of the Inner Hebrides, or with the still later extravasation of pitchstone dykes, can only be surmised.

[Footnote 257: The rocks of Beinn Hiant in Ardnamurchan have been claimed by Professor Judd as superficial lavas. For reasons to be afterwards given (p. 318) I regard them as intrusive sheets. Professor Cole believes the rhyolites and pitchstones of Tardree to be probably evidence of a volcano later than the basalts of Antrim. As I have not been able to detect any actual proofs of superficial outflow there, I relegate the description of the rocks to a future chapter, in which the acid protrusions will be discussed (p. 426).]

When one scans the great precipice on the west side of Eigg, with its transverse section of the pitchstone-lava, buried river-bed and basalt-plateau underneath, there seems no chance of any further westward trace of the pitchstone being ever found. The truncated end of the Scuir looks from the top of the cliff out to sea, and the progress of denudation might have been supposed to have effectually destroyed all evidence of the continuation of the rock in a westerly direction. Some years ago, however, my friend Prof. Heddle, while cruising among the Inner Hebrides, landed upon the little uninhabited islet of Hysgeir, which, some eighteen miles to the westward of Eigg, rises out of the open sea. He at once recognized the identity of the rock composing this islet with that of the Scuir, and in the year 1892 published a brief account of this interesting discovery.[258]

[Footnote 258: Appendix C to A Vertebrate Fauna of Argyle and the Inner Hebrides, by Messrs. J. A. Harvie-Brown and Thomas E. Buckley, p. 248.]

I have myself been able to land on Hysgeir in two successive summers, and can entirely confirm Prof. Heddle's identification. The islet stands on the eastern edge of the submarine ridge which, running in a north-easterly direction, culminates in the island of Canna. Hysgeir is a mere reef or skerry, of which the top rises only 38 feet above the Ordnance datum-level. Its surface is one of bare rock, save where a short but luxuriant growth of grasses has found root on the higher parts of two or three of its ridges, and on the old storm-beach of shingle which remains on the summit. The rock undulates in long low swells, that run in a general direction 20° to 45° west of north, and are separated by narrow channels or hollows. The place is a favourite haunt of gulls, terns, eider-ducks and grey seals, and is used by the proprietor of Canna for the occasional pasturage of sheep or cattle. So numerous are the sea-fowl during the breeding-season that the geologist, intent upon his own pursuits, may often tread on their nests unawares, while he is the centre of a restless circle of white wings and anxious cries.

The pitchstone of Hysgeir, like that of Eigg, is columnar, the columns being irregularly polygonal and varying from three to ten inches in diameter. They are packed so close together that the domes of rock on which their ends appear look like rounded masses of honeycomb. They may here and there be observed to be arranged radially with their ends at right angles to the curved exterior of the ridges, as if this external surface represented the original form of the cooled pitchstone, and were not due to mere denudation. There can be no doubt, however, that the island has been well ice-worn.

At the north-west promontory a beautiful example of fan-shaped grouping of columns may be observed on a face of rock which descends vertically into the sea. Here, too, is almost the only section on which the sides of the columns may be examined, for, as a rule, it is merely their ends on the rounded domes which are to be observed, and which everywhere slip under the waves. The columns in a cliff from 15 to 20 feet high show the slightly wavy, starch-like arrangement so often to be met with among the plateau-basalts.

The rock presents a tolerably uniform texture throughout, though in some parts it is blacker, more resinous, and less charged with porphyritic enclosures than in the general body of the rock. Large fresh felspars are generally scattered through it. To the naked eye it reproduces every feature of the pitchstone of the Scuir of Eigg.

A microscopic examination completes our recognition of the identity of these two rocks. Mr. Harker has examined a thin slice prepared from the Hysgeir pitchstone, and remarks regarding it that "the large felspars are not the only porphyritic element. The microscope shows the presence also of smaller imperfect crystals of augite, very faint green in the slice, and small grains of magnetite. The felspars have been deeply corroded by the enveloping magma, and irregular included patches of the groundmass occupy nearly half the bulk of some of the crystals. This latter feature is seen especially in some of the larger crystals, which seem to be sanidine. They are, for the most part, apparently simple crystals, but in places there is a scarcely defined lamellar twinning, or, again, small patches not extinguishing with the rest; so that we are probably dealing with some perthitic intergrowth on a minute scale.[259]

[Footnote 259: Comp. Prof. Judd's remarks on the Scuir of Eigg rock, Quart. Journ. Geol. Soc. vol. xlvi. (1890), p. 380.]

"Rather smaller felspar-crystals are rounded by corrosion, but lack the inclusions of groundmass; these have albite-and sometimes pericline-lamellation, and may be referred to oligoclase-andesine. The groundmass of the rock is a brown glass with perlitic cracks, enclosing very numerous microlites of felspar about ·001 inch in length [6619]. The rock is probably to be regarded as a dacite rather than a rhyolite, and thus agrees with Mr. Barker North's analysis of the Eigg pitchstone."[260]

[Footnote 260: Op. cit. p. 379.]

There is no trace of any conglomerate in situ like that under the Scuir of Eigg, nor of any other rock, aqueous or igneous. As the pitchstone everywhere slips under the sea, its geological relations are entirely concealed.

The great variety of materials met with in the form of boulders on the island is a testimony to the transport of erratics from the neighbouring islands and the mainland during the Glacial Period. The most abundant rock in these boulders is Torridon Sandstone, derived no doubt from the hills of Rum, but there occur also various kinds of schist, gneisses, quartzites, granites, porphyries, probably from the west of Inverness-shire, as well as pieces of white sandstone, probably Jurassic, which may have come from Eigg.

That the pitchstone of Hysgeir is a continuation of that of the Scuir may be regarded as highly probable. If not a continuation, it must be another stream of the same kind, and doubtless of the same date. If it be regarded as probably a westward prolongation of the Eigg rock, and if it be about as thick as that mass at the west end of the Scuir, then its bottom lies 200 or 300 feet under the waves. The river-channel occupied by the Eigg pitchstone undoubtedly sloped from east to west. The position of Hysgeir, 18 miles further west, may indicate a further fall in the same direction at the rate of perhaps as much as 35 feet in the mile.[261] Unfortunately, however, as no trace of the river-bed can now be seen on this island, any statement in regard to it must rest on mere conjecture.

[Footnote 261: Rep. Brit. Assoc. 1894, p. 653.]

Although the question of the denudation of the basalt-plateaux since the close of the volcanic period will be the subject of a special chapter in a later part of this volume, I cannot here refrain from calling attention to the pitchstone of Eigg and Hysgeir as one of the most impressive monuments of denudation to be found within the British Isles. Though now so prominent an object in the West Highlands, this rock once occupied the bottom of a valley worn out of the basaltic tableland. Prolonged and stupendous denudation has destroyed the connection with its source, has cut down its ends into beetling precipices, has reduced the former surrounding hills into gentle slopes and undulating lowland, and has turned the bottom of the ancient valley into a long, narrow and high crest. Moreover, we see that the erosion has not been uniform. The great wall of the Scuir does not stand fairly on the crest of the basalt-plateau but on the south side of it, so that the southern half of the old valley, with all its surrounding hills, has been entirely cut away. That subsidence has also come into play in the destruction of even the youngest parts of the volcanic plateaux will be more fully discussed in a later chapter. I need only remark here that the submergence of Hysgeir probably points to extensive depression of the land-surface on which the lavas were poured out.

Spotted a typo or a problem in this chapter? Tell us

XXXIX. THE BASALT-PLATEAUX OF SKYE AND OF THE FAROE ISLES

iv. THE SKYE PLATEAU

This largest and geologically most important of all the Scottish plateaux comprises the island of Skye, at least as far south as Loch Eishort, and the southern half of Raasay, but is shown by its sills to stretch as far as the Shiant Isles on the north, and the Point of Sleat on the south (see Map VI.). It may be reckoned to embrace an area of not less than 800 square miles. The evidence that its limits, like those of the other plateaux, are now greatly less than they originally were, is abundant and impressive. The truncated edges of its basalts, rising here and there for a thousand feet as a great sea-wall above the breakers at their base, and presenting everywhere their succession of level or gently inclined bars, are among the most impressive monuments of denudation in this country. But still more striking to the geologist is the proof, furnished beyond the margins of the plateau, that the Jurassic and other older rocks there visible were originally buried deep under the basalt-sheets, which have thus been entirely stripped off that part of the country.

Throughout most of the district, wherever the base of the basalts can be seen, it is found to rest upon some member of the Jurassic series, but with a complete unconformability. The underlying sedimentary strata had been dislocated and extensively denuded before the volcanic period began. On the southern margin, however, the red (Torridon) sandstones emerge from under the basalts of Loch Scavaig, and extending into the island of Soay are prolonged under the sea into Rum. This ridge probably represents the range of the ancient high ground of the latter island already referred to.

Nowhere are the distinctive topographical features and geological structure of the basalt-plateaux better displayed than in the northern half of the island of Skye. The green terraced slopes, with their parallel bands of brown rock formed by the outcrop of the nearly flat basalt-beds, rise from the bottoms of the valleys into flat-topped ridges and truncated cones (Fig. 283). The hills everywhere present a curiously tabular form that bears witness to the horizontal sheets of rock of which they are composed.[262] And along the sea-precipices, each excessive sheet of basalt can be counted from base to summit, and followed from promontory to promontory (Figs. 284, 286). In the district of Trotternish, the basalt hills reach a height of 2360 feet. Further west, the singular flat-topped eminences, called "Macleod's Tables" (Fig. 283) ascend to 1600 feet.

[Footnote 262: These features are more fully described in my Scenery of Scotland, 2nd edit (1887), pp. 74, 145, 216.]

Along the western side of Skye, the basalts descend beneath the level of the Atlantic, save at Eist in Duirinish, where the Secondary strata, with their belt of intrusive sills, rise from underneath them, and at the Sound of Soa, where they rest on the Torridon Sandstone. Along the eastern side, their base runs on the top of the great Jurassic escarpment, whose white and yellow sandstones rise there, and on the east side of Raasay, into long lines of pale cliffs. To the south-east, the regularity of the volcanic plateau is effaced, as in Mull and Ardnamurchan, by the protrusion of extensive masses of eruptive rocks constituting the Cuillin and Red Hills, east of which the basalts have been almost entirely removed by denudation, so as to expose the older rocks which they once covered, and through which the younger eruptive bosses made their way. This is undoubtedly the most instructive district for the study of that late phase in the volcanic history of Britain comprised in the eruptive bosses of basic and acid rocks.

The magnificent plateau of this island has been so profoundly cut down into glens and arms of the sea, and its component layers are exposed along so many leagues of precipice, that its structure is perhaps more completely laid open than that of any of the other Tertiary volcanic areas in Britain. It is built up of a succession of basalts and dolerites of the usual types, which still reach a thickness of more than 2000 feet, though in this instance, also, denudation has left only a portion of them, without any evidence by which to reckon what their total original depth may have been. In rambling over Skye, the geologist is more than ever struck with the remarkable scarcity and insignificance of the interstratifications of tuff or of any other kind of sedimentary deposit between the successive lava-sheets. One of the thickest accumulations of volcanic tuff and conglomerate has already been referred to as occurring on the south side of Portree Harbour, where it attains a depth of about 200 feet. As it is in immediate connection with its parent vent, it will be more fully alluded to in Chapter xli. Here, as is so generally observable among the basalt-plateaux, traces of vegetation are plentiful among the stratified intercalations, even forming thin seams of lignite and coal, one of which was formerly worked. That volcanic eruptions, though possibly of a feebler kind, continued during the interval between the basalt-outflows at this locality, is shown by the thick accumulation of tuff and by the occurrence of abundant lapilli of fine basic pumice among the shales, even to a distance of several miles from the vent.

Another conspicuous intercalation of sedimentary materials in the Skye plateau occurs on the Talisker cliffs at the mouth of Loch Bracadale, where, on the face of the great precipice of Rudha nan Clach, some conspicuous bands of lilac and red are interspersed among the basalts. These bands were noticed by Macculloch, who described them as varieties of "iron-clay."[263] I have not had an opportunity of examining them except from the sea at a little distance. But they suggest a similarity to some of the variegated clays between the upper and lower basalt series of Antrim.

[Footnote 263: Western Islands, vol. i. p. 376.]

Though good coal is not well developed in the Tertiary volcanic plateaux of the British Isles, it has already been pointed out that coaly layers are abundant, and that as the vegetable matter may confidently be assumed always to indicate terrestrial vegetation, the presence of the carbonaceous bands may be regarded as good evidence of some lapse of time between the eruption of the basalts which they separate. I have also called attention to the fact that the vegetable material is more especially observable in the highest parts of a group of intercalated sediments between two sheets of basalt. This relation, so strikingly exhibited in the isle of Canna, as already observed, is also to be remarked in the Skye plateau. I may here cite an interesting example which occurs at the base of the lofty sea-cliff of An Ceannaich, to the south of Dunvegan Head, on the west coast of Skye (Fig. 285). At the base of the precipice, ledges of a highly cellular basalt (a) show a singularly scoriaceous and amygdaloidal structure, with abundant and beautiful zeolites, the hollows of the upper surface of the sheet being filled in with dark brown carbonaceous shale, forming a layer from one to fourteen inches thick, marked by coaly streaks and lenticles (b). A band of green and yellow sandstone (c) next supervenes, which, from its pale colour, attracts attention from a distance, and led me, while yachting along the coast, to land at the locality in the hope that it might prove to be a plant-bearing limestone. This sandy stratum is only some three or four inches thick at the north end of the section, but increases rapidly southward to a thickness of as many feet or more, when, owing to the cessation of the underlying shale, it comes to lie directly on the amygdaloid and to enclose slaggy portions of that rock. Immediately above the sandstone two or three feet of fissile shale, black with plant-remains (d), include brown layers that yield to the knife like some oil-shales. The next stratum is a seam of coal (e) about a foot thick, of remarkable purity. It is glossy, hard, and cubical, including layers that break like jet. It has been succeeded by a deposit of green sand (f), but while this material was in course of deposition another outpouring of lava (g) took place, whereby the terrestrial pool or hollow of the lava-field, in which the group of sedimentary materials accumulated, was filled up and buried. This lava is about 20 feet thick, and consists of a coarsely-crystalline, jointed dolerite with highly amygdaloidal upper and under surface. Its slaggy bottom has caught up or pushed aside the layer of green sand, so as to lie directly on the coal, and has there been converted into the earthy modification so familiar under the name of "white trap" among our coal-fields. It is interesting to find that this kind of alteration, where molten rock comes in contact with carbonaceous materials, is not confined to subterranean sills, but may show itself in lavas that have flowed over a terrestrial surface.

From the frequent intercalation of such local deposits of sedimentary material between the basalts, we may reasonably infer that during older Tertiary time the rainfall in North-Western Europe was copious enough to supply many little lakes and streams of water. As the surface of the lava-fields decayed into soil, vegetation spread over it, so that, perhaps for long intervals, some tracts remained green and forest-clad. But volcanic action still continued to show itself, now from one vent, now from another. These wooded tracts were buried under overflows of lava, and, the water-courses being filled up, their streams were driven into new channels, and other pools and lakes were formed.

In no part of the Tertiary volcanic area of Britain can the characters of the lavas and the structure of the plateaux be better seen than along the west side of Skye, north of Loch Bracadale. The precipices rise sheer out of the sea, to heights of sometimes 1000 feet, and from base to summit every individual bed may be counted. Some particulars have already been given (p. 192) regarding the average thickness of the basalt-sheets on this coast-line. The general aspect of these cliffs and the arrangement of their component lavas is shown in Fig. 286. As a further detailed illustration of the general succession of the basalts in the Skye plateau, I give a diagrammatic view of the largest of Macleod's Maidens--the three weird sea-stalks that rise so grandly in front of the storm-swept precipice at the mouth of Loch Bracadale. The height of the stack must be at least 150 feet (Figs. 284 and 287). About ten distinct sheets of igneous rock can be counted in it, which gives an average thickness of 15 feet for the individual beds. It will be observed that there is a kind of alternation between the compact, prismatic basalts and the more earthy amygdaloids, but that the former are generally thickest.[264]

[Footnote 264: A striking and illustrative contrast between the relative thickness of the beds of the two kinds of rock is supplied by the fine sections of this district. The amygdaloids range from perhaps 6 or 8 to 25 or 30 feet; but the prismatic basalts, while never so thin as the others, sometimes enormously exceed them in bulk. In the island of Wiay, for example, a bed of compact black basalt, with the confused starch-like grouping of columns, reaches a thickness of no less than 170 feet. Its bottom rests upon a red parting on the top of a dull greenish earthy amygdaloid. It is possible, however, that some of these columnar sheets of basalt are really sills.]

These features, which are repeated on cliff after cliff, may be considered typical for all the plateaux. Another characteristic point, well displayed here, is the intervening red parting between the successive beds. If the occurrence and thickness of this layer could be assumed as an indication of the relative lapse of time between the different flows of lava, it would furnish us with a rude kind of chronometer for estimating the proportionate duration of the intervals between the eruptions. It is to be noticed on the top both of the compact prismatic and of the earthy amygdaloidal sheets; but it is more frequent and generally thicker on the latter than on the former, which may only mean that the surfaces of the cellular lavas were more prone to subærial decay than those of the compact varieties. Nevertheless, I am disposed to attach some value to it, as an index of time. In the present instance, for example, it seems to me probable that the lavas in the lower half of Macleod's Maiden, where the red layers are very prominent, were poured out at longer intervals than those that form the upper half. The remarkable banded arrangement of the vesicles in one of the cellular lavas of this sea-stack has been already referred to (p. 191).

Another characteristic plateau-feature is admirably displayed in Skye--the flatness of the basalts and the continuity of their level terraces (though not of individual sheets) from cliff to cliff and hillside to hillside. This feature may be followed with almost tiresome monotony over the whole of the island, north of a line drawn from Loch Brittle to Loch Sligachan. Throughout that wide region, the regularity of the basalt-plateau is unbroken, except by minor protrusions of eruptive rock, which, as far as I have noticed, do not seriously affect the topography. But south of the line just indicated, the plateau undergoes the same remarkable change as in Rum, Ardnamurchan and Mull. Portions of it which have survived indicate with sufficient clearness that it once spread southwards and eastwards over the mountainous district, and even farther south into the low parts of the island. Its removal from that tract has been of the utmost value to geological research, for some of the subterranean aspects of volcanism have thereby been revealed, which would otherwise have remained buried under the thick cover of basalt. Denudation has likewise cut deeply into the eruptive bosses, and has carved out of them the groups of the Red Hills and the Cuillins, to whose picturesque forms Skye owes so much of its charm.

In this, as in each of the other plateaux, there is no trace of any thickening of the basalts towards a supposed central vent of eruption. The nearly level sheets may be followed up to the very edge of the great mountainous tract of eruptive rocks, retaining all the way their usual characters; they do not become thicker there either collectively or individually, nor are they more abundantly interstratified with tuffs or volcanic conglomerates. On the contrary, their very base is exposed around the mountain ground, and the thickest interstratifications of fragmentary materials are found at a distance from that area. So far as regards the structure of the remaining part of the plateau, the eruption of the gabbros and granitoid rocks might apparently have taken place as well anywhere further north.

v. THE FAROE ISLANDS[265]

[Footnote 265: For references to the recent geological literature connected with these islands see the footnote ante, p. 191.]

Though these islands lie beyond the limits of the region embraced by the present work, I wish to cite them for the singular confirmation and extension they afford to observations made among British Tertiary volcanic rocks. Over a united extent of coast-cliffs which may be roughly estimated at about 500 English miles, the nearly level sheets of basalts, with their occasional tuffs, conglomerates, leaf-beds and coals, can be followed with singular clearness. Although the Faroe Islands have been so frequently visited and so often described that their general structure is sufficiently well known, they present in their details such a mass of new material for the illustration of volcanic action that they deserve a far more minute and patient survey than they have yet received. They cannot be adequately mapped and understood by the traveller who merely sails round them. They must be laboriously explored, island by island and cliff by cliff.

While I cannot pretend to more than a mere general acquaintance with their structure, I have learnt by experience that one may sail near their precipices and yet miss some essential features of their volcanic structure. In the summer of the year 1894 I passed close to the noble range of precipices on the west side of Stromö, at the mouth of the Vaagöfjord, and sketched the sill which forms so striking a part of the geology of that district (Figs. 312, 328 and 329). But I failed to observe a much more remarkable and interesting feature at the base of the same sea-cliffs. The following summer, probably under better conditions of light, I was fortunate enough to detect with my field-glass, from the deck of the yacht, what looked like a mass of agglomerate, and found on closer examination the interesting group of volcanic vents described in Chapter xli. The magnificent precipices of Faroe, which in Myling Head reach a height of 2260 feet, present a series of natural sections altogether without a rival in the rest of Europe. They are less concealed with verdure than those of Mull and Skye, and therefore display their geological details with even greater clearness than can be found either in Scotland or in Ireland. I would especially refer to the bare precipitous sides of the long narrow islands of Kalsö and Kunö, as admirable sections wherein the characters of the plateau-basalts are revealed as in a series of gigantic diagrams. The scarcity of vegetation, and the steepness of the declivities which prevents the abundant accumulation of screes of detritus, enable the observer to trace individual beds of basalt with the eye for several miles. Thus on the west side of Kunö, one conspicuous dark sheet in the lower part of the section can be followed from opposite Mygledahl in Kalsö to the southern end of the island. There is one concealed space at the mouth of the corrie behind Kunö village, but the same, or at least a similar band of rock at the same level, emerges from the detritus on the further side, and may possibly run into the opposite promontory of Bodö. It extends in Kunö for at least six geographical miles.

These vast escarpments of naked rock show, with even greater clearness than the precipices of the Inner Hebrides, how frequently the basalts die out, now in one direction now in another. The two sides of the Kalsöfjord exhibit many examples of this structure, and some striking instances of it are to be seen on the west side of Haraldsfjord. In these cliffs, which must be about 2000 feet high, upwards of forty distinct flows can sometimes be traced from the sea-level to the crest. The average thickness of each bed is thus somewhat less than 50 feet. Such vast escarpments, with wide semicircular corries scooped out of their sides, such serrated crests and dark rifts in the precipices, such deep fjords winding through nearly horizontal basalts, of which the parallel sheets can be followed by the eye from island to island, fill the mind with a vivid conception at once of the enormous scale of the volcanic eruptions and of the stupendous denudation which this portion of North-Western Europe has undergone since Tertiary time.

As the lenticular character of the basalts, and the evidence they supply of having been discharged from many small local vents are of great importance in the comprehension of the volcanic history of the plateaux, some further illustrations of these features may with advantage be given here. Thus the traveller who skirts the western precipices of Suderö will notice some good examples to the north of the highest part of the cliffs. On Stromö he will detect other cases of the same structure. Similar features will arrest his attention on the precipices of Sandö, where, though at first sight the basalts seem to be regular and continuous, a nearer view of them reveals such sections as that shown in Fig. 288, where a group of sheets rapidly dies out towards the north against a thicker band that thins away in the opposite direction. Further north he will come upon other examples in the range of low cliffs between Kirkebonaes and Thorshaven, and more impressive still in the rugged precipices that front the Atlantic on the western front of Hestö (Fig. 289), where the disappearance is in a northerly direction.

But it is in the northern part of the Faroes, where the basalt-plateau has been so deeply trenched by parallel fjords as to be broken up into a group of long, narrow, lofty, and precipitous insular ridges, that the really local and non-persistent character of the lavas can best be seen. The eastern cliffs of Svinö present admirable examples, where in the same vertical wall of rock some of the basalts die out to the south, others to the north, while occasionally a shorter sheet may be seen to disappear in both directions as if it were the end of a stream that flowed at right angles to the others (Fig. 290).

The more the basalt-plateaux of Britain and the Faroe Islands are studied, the more certain does the conclusion become that these widespread sheets of lava never flowed from a few large central volcanoes of the type of Etna or Vesuvius, but were emitted from innumerable minor vents or from open fissures. In a later chapter an account will be given of the vents, which may still be seen under the overlying sheets of basalt, and, in particular, a remarkable group in the Faroe Islands will be described.

The occurrence of tuffs, leaf-beds and thin coals between the plateau-basalts of the Faroe Islands has long been known. These stratified deposits are well seen in the island of Suderö, where they serve to divide two distinct series of basalts, like the iron-ore and its accompaniments in Antrim. As a characteristic illustration of the same diversity of deposits observable between the lava-sheets of the basalt-plateaux of the British Isles I give here a section exposed on the east side of this island--a locality often visited and described in connexion with its coal-seams (Fig. 291). At the base lies a sheet of basalt (a) with an irregularly lumpy upper surface. It may be remarked that the lower group of basalts is marked by the occurrence of numerous columnar sheets, some of them possibly sills, and also more massive, solid, and durable basalts than the sheets above. The lowest of the intercalated sediments are light-coloured clays, passing down into dark nodular mudstone and dark shale, the whole having a thickness of at least 20 feet (b). These strata are succeeded by (c) pale clays with black plant-remains, about three feet thick. Immediately above this band comes the coal or coaly layer (d), here about six inches thick, which improves in thickness and quality further inland, where it has been occasionally worked for economic purposes. A deposit of green and brown volcanic mudstone (e), twelve feet in thickness, overlies the coal and passes under a well-bedded granular green tuff and mudstone three feet thick (f). The uppermost band is another volcanic mudstone (g) four feet in thickness, dark green in colour, and more or less distinctly stratified, with irregular concretions, and also pieces of wood. Above this layer comes another thick overlying group of basalts (h) distinguished by their abundantly amygdaloidal character, and by their weathering into globular forms which at a little distance give them a resemblance to agglomerates.

We have here an intercalated group of strata upwards of 40 feet thick, consisting partly of tuffs and partly of fine clays, which may either have been derived from volcanic explosions or from the atmospheric disintegration of basaltic lavas. Through some of these strata abundant carbonaceous streaks and other traces of plants are distributed, while among them lies a band almost wholly composed of compressed vegetation. Unfortunately none of the strata at this locality seem to have preserved the plant-remains with sufficient definiteness for identification. There can be no doubt, however, that they were terrestrial forms like those of Mull and Antrim.

This coal, with its accompanying sedimentary deposits, has been traced through Suderö, and another outcrop, possibly of the same horizon, occurs on Myggenaes, the extreme western member of the group of islands, at a distance of some 40 miles.[266]

[Footnote 266: See in particular Prof. J. Geikie, Trans. Roy. Soc. Edin. vol. xxx. (1880), p. 229.]

Spotted a typo or a problem in this chapter? Tell us

Chapter XL

THE MODERN VOLCANOES OF ICELAND AS ILLUSTRATIVE OF THE TERTIARY VOLCANIC HISTORY OF NORTH-WESTERN EUROPE

From the facts stated in the foregoing chapters concerning the structure of the basalt-plateaux of North-Western Europe, it is evident that in none of these areas have the eruptions come from one great central volcano like Etna or Vesuvius. On the contrary, in every instance there is abundant evidence that the basalt has flowed from many scattered points of eruption. The uniformity of the lava-sheets in petrographical characters, their continuity when viewed in mass, their general horizontality, and their constant thinning away in different directions, show that the eruptive vents must have been distributed over the whole plateau-areas.

The conditions under which such eruptions took place can be most readily understood by a comparison of the phenomena with those observable in modern volcanic tracts where extensive outflows of lava have taken place without the existence of any great central cones. Of these regions the most instructive is undoubtedly to be found among the recent lava-deserts of Iceland. There the parallels to the structures described from the British and Faroe plateaux are so numerous and so close that an account of the Icelandic region may appropriately be inserted here.

The evidence furnished by Iceland is of special value in our present enquiry, inasmuch as that island, besides its modern eruptions, includes vast basaltic plateaux of Tertiary age. These areas of nearly level sheets of basalt belong to the same geological period as those of the British and Faroe Islands, and display the same internal structure and external features. But they have this distinguishing peculiarity that the volcanic fires beneath them are not yet extinguished. They have been broken through again and again in recent times by volcanic eruptions which have repeated many of the characteristics of their Tertiary predecessors. The old and the new development of the same volcanic type are thus visible side by side.

The Tertiary volcanic series of Iceland reaches a thickness of upwards of 3000 metres, or nearly 10,000 English feet, but as its base is nowhere seen, it may be still thicker. Its successive sheets, piled over each other in parallel layers, form terraced hills and bold escarpments along the coast, whence they slope gently inland. The plateau, as in the Faroe Islands and in Scotland, has been extensively eroded, and has been trenched by many long valleys and fjords The composition of the basalts remains remarkably uniform over the island. The lava sheets are often decomposing, amygdaloidal, and filled with zeolites; while higher in the series compact basalts abound, the uppermost fine-grained sheets being especially constant in structure and composition. Numerous dykes traverse the plateau, and some of them cut even its highest members. The parallel with the geological structure of the Inner Hebrides is continued in Iceland by the appearance of intrusive masses of gabbro and granophyre, which represent the deeper parts of the Tertiary volcanic series, while the basalts were poured out at the surface. Thus, at Papafjord, the gabbro rises into mountainous peaks and, like the similar rock in Mull and Skye, is intersected by dykes of a coarse-grained granitoid liparite or granophyre. Large dykes and ramifying veins of the same acid material, often with a thoroughly granitic aspect, extend into the basalts.[267]

[Footnote 267: Mr. Thoroddsen, Dansk. Geografisk Tidsskrift, vol. xiii.]

A long series of eruptions has taken place in Iceland since the Glacial Period. There were likewise pre-glacial eruptions. The glaciated lava-streams are found underneath the modern lavas. So far indeed as is known, no evidence exists of any important cessation of subterranean activity there since Tertiary time.[268] The existing volcanic phenomena may with probability be regarded as the survival of those which were so widely manifested over the Icelandic area and the north-west of Europe in the older Tertiary ages. A careful study of them may therefore be expected to throw light on the history of the Tertiary basaltic plateaux; while, on the other hand, the thorough dissection of these plateaux by the denuding agencies will not improbably be found to explain some parts of the subterranean mechanism of the modern Icelandic volcanoes.

[Footnote 268: See Dr. Johnston-Lavis, Scottish Geographical Magazine, 1895, p. 442.]

In calling attention to some of the more obvious analogies which may be traced between the modern and the ancient volcanoes, I am more particularly indebted to the excellent memoirs of the resident Icelandic geologist, Mr. Th. Thoroddsen, who has examined so large a part of the island.[269] The account given by Mr. A. Holland of the Laki craters has likewise been of much service to me.[270] Among other recent observers I may cite Dr. Tempest Anderson,[271] who has made himself familiar with extensive tracts of Iceland. He was accompanied one year by Dr. Johnston-Lavis, who has published a narrative of the journey.[272]

[Footnote 269: See In particular his paper on the volcanoes of north-east Iceland (Bihang till. k. Svensk. Vet. Akad. Handl. xiv. ii. No. 5, 1888) and that on Snaefell and Faxebugt in the south-west of the island (op. cit. xvii. ii. No. 2, 1891); also papers in Dansk. Geografisk Tidsskrift, vols. xii. xiii. (1893-95); Verhand. Gesellsch. Erdkunde zu Berlin, 1894-95.]

[Footnote 270: "Lakis Kratere og Lavaströmme, Universitætsprogram," Christiania, 1885. See Mr. Thoroddsen's remarks on this paper, Verhand. Gesell. Erdkunde, 1894, p. 289.]

[Footnote 271: Brit. Assoc. Rep. 1894, p. 650.]

[Footnote 272: Dr. Johnston-Lavis, Scottish Geographical Magazine, September 1895.]

It is a mistake to suppose that the Icelandic volcanoes are generally built on the plan of such mountains as Vesuvius or Etna. Mr. Thoroddsen can evidently hardly repress his impatience to find these two Italian cones cited in almost every handbook of geology as types of modern volcanoes and their operations. The regular volcanic cone, composed of alternations of lavas and tuffs, plays a very subordinate part in Iceland.

The fundamental feature in the Icelandic eruptions is the production of fissures which reach the surface and discharge streams of lava from many points. Two systems of such fissures appear to be specially marked, one in southern Iceland running from south-west to north-east, the other, in the north part of the island, stretching from south to north.[273] Hekla and Laki belong to the former. The dislocations have often followed the boundaries of the "horsts," or solid blocks of country which have withstood terrestrial displacement. The vast outbreaks of Odádahraun and Myvatn have almost all issued from fissures of that nature.

[Footnote 273: In the Snaefell promontory they run nearly east and west. Mr. Thoroddsen, Bihang. Svensk. Akad. xvii. (ii.) No. 2, p. 91.]

The violent eruption of 1875 in Askja found its exit at the intersection of two lines of fissures. Many large fissures were opened on the surface in a nearly north and south direction, which could be followed for 80 kilometres or nearly 50 English miles. Some of them became the theatre of intense volcanic activity.[274]

[Footnote 274: Mr. Thoroddsen, op. cit. xiv. ii. No. 5, p. 63.]

Many lines of fissure are traceable at the surface as clefts or "gjás," that run nearly straight for long distances, with a width of one to three yards, and sometimes of unknown depth.[275] The most stupendous example of the structure yet discovered is probably the Eldgjá found by Dr. Thoroddsen in the year 1893, below the Mýrdalsjökull. This gigantic chasm has a length of 30 kilometres (more than 18 English miles), and a depth of 130 to 200 metres (426 to 656 feet). Over its vertical walls lofty waterfalls plunge from the crest to the bottom.

[Footnote 275: On the various modes of origin of these chasms, see Dr. Tempest Anderson, Brit. Assoc. Rep. p. 650. The gjá shown in Fig. 292 is not an eruptive fissure. For this and the following illustration I am indebted to the kindness of Dr. Tempest Anderson, who himself photographed the scenes.]

Occasionally a fissure has not been continuously opened to the surface. An interesting example of such intermittent chasms is supplied by the great rent which gave forth the enormous volume of lava in 1783. The mountain of Laki, composed of palagonite tuff, stands on the line of this dislocation, but has not been entirely ruptured. The fissure has closed up beneath the mountain, a short distance above the bottom of the slope, as is shown by the position of a couple of small craters.[276]

[Footnote 276: Mr. A. Helland, op. cit. p. 25.]

Some fissures have remained mere open chasms without any discharge of volcanic material; others have served as passages for the escape of lava and the ejection of loose slags and cinders.[277]

[Footnote 277: Mr. Thoroddsen has observed that in the Reykjanes peninsula in the south-west of Iceland, by the subsidence of one side of a fissure, a row of four craters has been cut through, leaving their segments perched upon the upper side. Globus. vol. lxix. No. 5.]

In some instances, according to Mr. Thoroddsen, lava wells out from the whole length of a fissure without giving rise to the formation of cones, the molten material issuing either from one or from both sides and flowing out tranquilly. Thus from three points on the great Eldgjá chasm lava spread out quietly without giving rise to any craters, though at the southern prolongation of the fissure, where it becomes narrower, a row of low slag-cones was formed. The three lava-streams flooded the low ground over an area of 693 square kilometres, or 270 English square miles. In the great majority of cases, however, the lava as it ascends in the fissure gives rise to long ramparts of slags and blocks of lava piled up on either side, or to a row of cones along the line of the open chasm. Thus, on the Laki fissure, which runs for about 20 miles in a north-east direction, the cones amount to some hundreds in number.

The cones consist generally of slags, cinders, and blocks of lava. They are on the whole not quite circular but oblong, their major axis coinciding with the line of the chasm on which they have been piled up, as along the marvellous line of the Laki fissure. In many places they are exceedingly irregular in form, changes in the direction of outflow of lava or of escape of steam having caused the cones partially to efface each other.

As regards their size, the cones present a wide range. Some of them are only a few yards in diameter, others several hundred yards. Generally they are comparatively low mounds. On a fissure hardly 30 feet long, Mr. Thoroddsen found a row of twelve small cones built exactly like those of largest size, but with craters less than three feet in diameter. On the Laki fissure some are only a couple of yards high; the majority are much less than 50 yards in height, and hardly one is as much as 100 yards.[278] And yet these little monticules, as Mr. Helland remarks, represent the pipes from which milliards of cubic metres of lava have issued. While other European volcanoes form conspicuous features in the landscape, the Icelandic volcanoes of the Laki district, from which the vastest floods of lava have issued in modern times, are so low that they might escape notice unless they were actually sought for.[279]

[Footnote 278: Mr. Thoroddsen, however, states that there are about 100 ranging between 20 and 100 metres in height.]

[Footnote 279: Op. cit. p. 27.]

As they have generally arisen along lines of fissure, the cones are, for the most part, grouped in rows. The hundreds of cones that mark the line of the Laki fissure present an extraordinary picture of volcanic energy of this type. In other instances the cones occur in groups, though this distribution may have arisen from the irregular uprise of scattered vents along a series of parallel fissures. Thus to the north-east of Laki a series of old cones entirely surrounded by the lavas of 1783 lie in groups, the most northerly of which consists of about 100 exceedingly small craters that have sent out streams of lava towards the N.N.E.[280]

[Footnote 280: Op. cit. p. 25. The great lava-fields of Iceland are likewise dotted over with secondary craters or "hornitos" which have no direct connection with the magma below, but arise from local causes affecting the outflowing lava. They are grouped in hundreds over a small space.]

It would appear from Mr. Helland's observations that the same fissure has sometimes been made use of at more than one period of eruption. He describes some old craters on the line of the Laki fissure, which had been active long before the outbreak of 1783.[281]

[Footnote 281: Op. cit. p. 26.]

When the lava issues from fissures it is in such a condition of plasticity that it can be drawn out into threads and spun into ropes. When the slope over which it flows is steep it often splits up into blocks on the surface. Where the ground is flat the lava spreads out uniformly on all sides, forming wide plains as level as a floor. Thus the vast lava-desert of Odádahraun covers a plain 3640 square kilometres in area, or, if the small-lava-streams north from Vatnajökull be included, 4390 square kilometres. This vast flood of lava (about 1700 English square miles in extent) would, according to Mr. Thoroddsen, cover Denmark to a depth of 16 feet. The whole of this enormous discharge has been given forth from more than twenty vents situated for the most part on parallel fissures.

Not less striking is the picture of fissure-eruption to be met with at Laki--the scene of the great lava-floods of 1783. "Conceive now," says Mr. Helland, "these hundreds of craters, or, as they are called by the Icelanders, 'borge,' lying one behind another in a long row; every one of them having sent out two or more streams of lava, now to the one side, now to the other. Understand further that these streams merge into each other, so as to flow wholly round the cones and form fields of lava miles in width, which, like vast frozen floods, flow down to the country districts, and you may form some idea of this remarkable region."[282]

[Footnote 282: Op. cit. p. 24. Mr. Helland allows an average thickness of 30 metres for the mass of lava which issued in two streams, one 80 kilometres (nearly 50 miles), the other 45 kilometres (about 28 miles) long. He estimates the total volume of lava discharged in the 1783 eruption at 27 milliards of cubic metres, equal to a block 10 kilometres (6 miles 376 yards) long, 5 kilometres (3 miles 188 yards) broad, and 540 metres (1771 feet) high; op. cit. p. 31. Mr. Thoroddsen remarks that the older estimates of the volume of lava discharged by this eruption have been greatly exaggerated. He puts the area covered by lava at 565 square kilometres and the contents at 12-1/3 cubic kilometres. Verhand. Gesell. Erdkunde Berlin, 1894, p. 296.]

The basaltic lavas have issued in a comparatively liquid state, form thin sheets and reach to great distances. The western stream from the Laki eruption of 1783 flowed for upwards of 40 miles; a prehistoric lava from Trölladyngjá in Odádahraun flowed for more than 60 miles.

In the course of time the successive streams of lava poured out upon one of these wide volcanic plains gradually increase the height of the ground, while preserving its generally level aspect. The loose slag-cones of earlier eruptions are effaced or swallowed up, as one lava-stream follows another. Eventually, when, by the operation of running water or by fissure and subsidence, transverse sections are cut through these lava-sheets, the observer can generally notice only horizontal beds of lava piled one above another, including the dykes connected with them and intercalated masses of loose slag, that remain as relics of the old craters.

In some places the lava has gradually built up enormous domes, like those of Hawaii, having a gentle inclination in every direction, as may be seen especially in the district between Floderne Skjalfanafljot and Jökulsà Most of the large volcanic piles of North Iceland are of this nature. The highest of them are 1209 and 1491 metres high by from 6 to 15 kilometres in diameter. The elliptical crater of the highest of these eminences measures 1100 by 380 metres.[283]

[Footnote 283: Mr. Thoroddsen, op. cit. xiv. ii. No. 5, pp. 10, 23.]

Large conical volcanoes of the Vesuvian type built up of alternating lavas and tuffs are not common in Iceland, but some occur and rise into lofty glacier-covered mountains, such as Öræfajökull (6241 feet), Eyjafjallajökull (5432), and Snaefellsjökull (4577). Hekla (4961) also is similarly composed of sheets of lava and tuffs, but has not been built as a cone. It forms an oblong ridge which has been fissured in the direction of its length and bears a row of craters along the fissure.[284]

[Footnote 284: Mr. Thoroddsen, Dansk. Geograf. Tidsskrift, vol. xiii.]

Explosion-craters likewise occur among the modern volcanic phenomena of Iceland. One of these was formed by a violent explosion at Askja on 29th March 1875. It has a diameter of only about 280 feet, yet so great was the vigour of the outburst that pumiceous stones were spread over an area of more than 100 Danish (468 English) square miles, and the dust was carried as far as Norway and Sweden. Nine years later Mr. Thoroddsen found the bottom of this crater filled with bluish-green boiling mud, which will probably in the end become a sheet of still water. The borders of these Icelandic explosion-craters seem to be very little higher than the ground around them. Most of the ejected material is expelled with such force and to such a distance that only a small fraction of it falls down around the orifice of eruption.[285]

[Footnote 285: Mr. Thoroddsen, op. cit.]

There is still another feature of the Icelandic volcanic regions which may be cited as an interesting parallel to the sequence of eruptive discharges among the Inner Hebrides. While the lavas are as a rule more or less basic--many of them being true basalts--they have been at different times pierced by much more acid liparites and obsidians. Examples of these rocks of post-Glacial age have recently been traced on the ground by Mr. Thoroddsen,[286] and their petrographical characters have been studied by Mr. Bäckström.[287] The wide distribution of such rocks all over the island, their occurrence in isolated bosses among the more basic lavas, and their remarkable internal structures have been noted by several observers.[288] The liparites and obsidians are contrasted with the basalt by the colours and forms of their streams. Some of them are so black as to look like heaps of coal, though their surfaces pass into grey pumice. They have flowed out in a much less liquid condition than the basalts, and have consequently formed short, thick and irregular sheets. The liparites and basalts appear to have been nearly contemporaneous. They certainly belong to the same volcanic cycle and their vents lie close to each other. Though none of the acid eruptions are known to have occurred in modern times, some of the liparites are crusted with sulphur and from the connected fissures steam still rises.

[Footnote 286: Geol. Fören. Stockholm Förhandl. xiii. (1891), p. 609; Bihang. Svensk. Vet. Akad. Handl. xvii. ii. p. 21 (1891); _Dansk. Geograf. Tidsskrift_, xiii. (1895).]

[Footnote 287: Geol. Fören. Stockholm Förhandl. xiii. (1891), p. 637.]

[Footnote 288: See in particular C. W. Schmidt, Zeitsch. Deutsch. Geol. Gesellsch. xxxvii. (1885), p. 737.]

It will thus be seen how entirely the modern volcanic eruptions of Iceland agree with the phenomena presented by our Tertiary basalt-plateaux. It is, therefore, to the Icelandic type of fissure-eruptions, and not to great central composite cones like Vesuvius or Etna that we must look for the modern analogies that will best serve as commentary and explanation for the latest chapter in the long volcanic history of the British Isles.[289]

[Footnote 289: In his memoir of 1874, Professor Judd announced his conclusion that there were formerly five great volcanoes amongst the Western Isles, and that the lavas of the plateaux had issued from these. He subsequently reiterated this view (Quart. Journ. Geol. Soc. xlv., 1890, p. 187), and ridiculed the explanation of fissure-eruptions. The evidence adduced by me in a paper published in 1896 (same journal, vol. lii. p. 331) and reprinted with additions in this chapter, will, I trust, be regarded by geologists as having finally settled this question.]

As a further but more ancient illustration of the type of volcanic action which appears to have been prevalent during the formation of the Tertiary volcanic plateaux of Britain, I may again refer to the vast basalt-fields of Western America. The basalt of Idaho stretches out as an apparently limitless plain. Along its northern boundary, this sea of black lava runs up the valleys and round the promontories of the older trachytic hills with almost the flatness of a sheet of water. It has been deeply trenched, however, by the streams that wind across it, and especially by the Snake River, which has cut out a gorge some 700 feet deep, on the walls of which the successive beds of basalt lie horizontally one upon another, winding along the curving face of the precipice exactly as those of Antrim and the Inner Hebrides do along their sea-worn escarpments. Here and there, a low cinder-cone on the surface of the plain marks the site of a late outflow. One is struck, however, with the singular absence of tuffs and volcanic conglomerates. The basalts appear to have flowed out stream after stream with few fragmentary discharges.

These characteristic features of one distinctive type of volcanic action have been repeated over a vast region, or rather a whole series of regions, in Western America, the united area of which must equal that of a considerable part of Europe. From Idaho, the basalt-fields may be followed southwards interruptedly into Utah and Nevada, and across the great plateau-country of the cañons into Arizona and New Mexico, northwards into Montana, and westwards into Oregon. The tract which has as yet been most carefully traversed and described is probably that of the high plateaux of Utah and Arizona. Thus on the Uinkaret plateau, which measures some 45 to 50 miles in length by 8 to 12 in breadth, a thick covering of basalt has been spread composed of many successive flows. Between 160 and 170 separate cones have been counted on this area, most of them quite small, mere low mounds of scoriæ, though a few reach a height of 700 or 800 feet, with a diameter of a mile. From three to seven or eight may be found in a row, as if springing from a single line of fissure. But generally the grouping is quite irregular.[290] My friend Captain C. E. Dutton, from whose admirable memoir these details are quoted, remarks further that among the Utah plateaux no trace of a cone is to be found at or near some of the most recent basalt-fields, and that the most extensive outpours are most frequently without cones. "The lavas," he adds, "appear to have reached the surface and overflowed like water from a spring, spreading out immediately and deluging a broad surface around the orifice."[291] The deep gorges cut by the rivers through these thick accumulations of horizontal or nearly horizontal basalts, have here and there revealed parallel dykes that traverse the rocks, and in at least one case have shown the dyke running for half a mile up a cliff and actually communicating with a crater of scoriæ at the top.[292] Again, in New Mexico, Captain Dutton noticed vast tracts of younger basalt, about which "a striking fact is the entire absence of all distinguishable traces of the vents from which they came. Some of them, however, indicate unmistakably their sources in small depressed cones of very flat profiles. No fragmental ejecta (scoriæ, lapilli, etc.) have been found in connection with these young eruptions."[293] Such I believe to have been the general conditions under which the basalts of the Tertiary plateaux of the British Isles were also erupted.[294]

[Footnote 290: Captain C. E. Dutton, "Tertiary History of the Grand Cañon District," U.S. Geol. Survey (1882), p. 104.]

[Footnote 291: Captain C. E. Dutton, "Geology of the High Plateaux of Utah," U.S. Geol. Survey of the Rocky Mountain Region (1880), pp. 198, 200. See also pp. 232, 234, 276 of the same Monograph for additional examples.]

[Footnote 292: Tertiary History of the Grand Cañon, etc., p. 95.]

[Footnote 293: Nature, xxxi. (1884), p. 49.]

[Footnote 294: I may again refer to Hopkins's Researches in Physical Geology, where the conditions of the problem here discussed have been distinctly realized. Speaking of the ejection of lava from a number of fissures, he remarks that the imperfect fluidity of the melted material "would seem to require a number of points or lines of ejection as a necessary condition." "If there were only a single centre of eruption, a bed of such matter approximating to uniformity of thickness, could only be produced on a surface of a conical form." "Where no such tendency to this conical structure can be traced, it would probably be in vain to look for any single centre of eruption. On the supposition, too, of ejection through continued fissures, or from a number of points, that minor unevenness of surface which must probably have existed under all circumstances during the formation of the earth's crust, would not necessarily destroy the continuity of a comparatively thin extensive bed of the ejected matter, in the same degree in which it would inevitably produce that effect in the case of central ejection" (Cambridge Phil. Trans. vi. 1835, p. 71).]

Although we may be convinced, from their general structure and relations, that the stratified lavas of these plateaux have been poured out from fissures and not from great central cones, it must obviously be difficult to obtain demonstrative evidence of this origin from any single section. Of the thousands of dykes which traverse the British plateaux and the ground around them, I am not aware of a single one which can be actually seen to have ever communicated with the surface. The very process of denudation which has revealed these dykes has at the same time removed all trace of any former connection they may have had with the surface. The only places where we may hopefully search for the missing evidence are the fronts of the escarpments. On these precipices dykes may sometimes be seen to end off at some particular platform among the basalt-sheets, but I have never found a case which could be confidently cited as an example of lava rising in a fissure and spreading out as a superficial sheet. That this connection may eventually be found when a more detailed survey is made of these great sea-walls I fully anticipate.

In recently mapping the basalt-plateau of Strathaird in Skye, Mr. Harker has made some interesting observations regarding the probable connection of the dykes with the plateau basalts. He has noticed that the flanks of Slat Bheinn, a portion of the plateau, are abundantly traversed by dykes containing numerous enclosed pieces of gabbro, while the basalt on the summit of the plateau is full of similar fragments--an occurrence not observed elsewhere. It is conceivable that the gabbro-bearing basalt-sheets are sills, but Mr. Harker has found no proof that they are so, the evidence so far as it has been collected being rather in favour of the view that these sheets are superficial lavas, and that they have been supplied from the dyke-fissures.

Various considerations suffice to assure us that actual instances of the outflow of the basalt from its parent fissures should be expected to be exceptional. The absence or scarcity of beds of scoriæ among the basalt-plateaux may be taken as an indication that the lava as a rule flowed out without the formation of cinder-cones, and therefore that these conspicuous monuments of the eruptive vents were probably always rare in Britain. If the lava was poured out tranquilly from one or two points along a fissure which were subsequently buried under floods of similar lava issuing from other fissures, the chances that such points of emission should be laid open along the front of any escarpment are small. And, even when so exposed, it might be difficult to feel sure that the dyke below was really the feeder of the basalt above, unless the cliff were accessible and the rocks could be scrutinized foot by foot. These elements of uncertainty are happily removed where the volcanic energy has drilled well-marked funnels of discharge and left them filled with the erupted materials, as will be narrated in the next chapter.

Spotted a typo or a problem in this chapter? Tell us