xlvii. Some of the acid dykes and sills of the Inner Hebrides are

varieties of rhyolite. No undoubted example has yet been observed of a superficial rhyolite-lava, though such not improbably appeared in the interval between the lower and upper basalts of Antrim.

ii. STRATIGRAPHICAL POSITION.--ANALOGIES FROM CENTRAL FRANCE

In the history of opinion regarding the relative position of the Tertiary eruptive rocks, no feature is so remarkable as the universal acceptance of the misconception regarding the place of the acid protrusions. In tracing this mistake to its source, we find that it probably arose from the fact that along their line of junction the granitoid masses generally underlie the basic. This order of superposition, which would usually suffice to fix the age of two groups of stratified rocks, is obviously not of itself enough to settle the relative epochs of two groups of intrusive rocks. Yet it has been assumed as adequate for this purpose, and hence what can be proved to be really the youngest has been placed as the oldest part of the Tertiary volcanic series.

Macculloch, who showed that his "syenites" and "porphyries" had invaded the Secondary strata of the Inner Hebrides, and must therefore be of younger date than these, left their relations to the other igneous rocks of the region in a curiously indefinite position. He was disposed to regard them all as merely parts of one great series; and seems to have thought that they graduate into each other, and that any attempt to discriminate between them as to relative age is superfluous. Yet he evidently felt that the contrasts of topography which he described could hardly fail to raise the question of whether rocks so distinct in outward form did not differ also in relative antiquity. But he dismissed the question without answering it, remarking that if there is any difference of age between the two kinds of rock, "there appears no great prospect of discovering it."[373] He records an instance of a vein of "syenite" traversing the "hypersthene rock" in the valley of Coruisk. "If this vein," he says, "could be traced to the mass of syenite, it might be held a sufficient ground of judgment, but under the present circumstances it is incapable of affording any assistance in solving the difficulty."[374] Instead, however, of being a solitary instance, it is only one of hundreds of similar intrusions which can be connected with the general body of granitic and granophyric masses, and which put the relative ages of the several groups of rock beyond any further doubt.

[Footnote 373: _Western Islands_, i. p. 368; see also pp. 488, 575, 578.]

[Footnote 374: _Op. cit._ p. 370.]

Boué, who knew the geology of some of the extinct volcanic regions of Europe, recognized the similarity of the Scottish masses to those of the Continent, and classed the acid rocks as "trachytes." He saw in each of the volcanic areas of the West of Scotland a trachytic centre, and supposed that the more granitoid parts might represent the centres in the European trachytic masses. He traced in imagination the flow of the lava-streams from these foci of volcanic activity, distinguishing them as products of different epochs of eruption, among the last of which he thought that the trachytic porphyries might have been discharged. He admitted, however, that his restoration could not be based on the few available data without recourse to theoretical notions drawn from the analogy of other regions.[375]

[Footnote 375: _Essai Géologique sur l'Écosse_, pp. 291, 322, 327.]

In the careful exploration of the central region of Skye made by Von Oeynhausen and Von Dechen, these able observers traced the boundary between the "syenite" and the "hypersthene rock"; and as they found the former lying underneath the latter, they seem naturally to have considered it to be the older protrusion of the two.[376] Principal Forbes came to a similar conclusion from the fact that he found the dark gabbro always overlying the light-coloured felspathic masses.[377] Professor Zirkel also observed the same relative position, and adopted the same inference as to the relative age of the rocks.[378] Professor Judd followed these writers in placing the acid rocks before the basic. He supposed the granitoid masses to form the cores of volcanic piles probably of Eocene age, through and over which the protrusions of gabbro and the eruptions of the plateau-basalts took place.[379]

[Footnote 376: Karsten's _Archiv_, i. p. 82. It will be shown in later pages that the apparent infraposition of the granophyre is often deceptive, the real junction being vertical.]

[Footnote 377: _Edin. New Phil. Jour._ xl. (1846) p. 84.]

[Footnote 378: _Zeitsch. Deutsch. Geol. Gesellsch._ xxiii. (1871) pp. 90, 95. He says that the gabbro seems to be the younger rock, so far as their relations to each other can be seen.]

[Footnote 379: _Quart. Jour. Geol. Soc._ xxx. (1874) p. 255.]

The evidence for the posteriority of the acid rocks will be fully detailed in later pages. Before entering upon its consideration, however, I would remark that the uprise of the British granophyres presents so many points of resemblance to that of the trachytes and phonolites among the basalt-plateaux of Auvergne and the Velay in Central France, that a brief account of the acid protrusions of these regions may be suitably given here as an introduction to the account of those of the Inner Hebrides. A succession of stages in the progress of denudation allows us to follow the gradual isolation and dissection of the French volcanic groups. The youngest examples occur in the chain of cones and craters, in the region of the Puy de Dôme. These may be of Pleistocene, or even of more recent date. Older and more deeply eroded than these are the numerous domes and cones in the territory of Haute Loire. Yet more ancient and still more stupendously denuded come the bosses, sills and dykes of Britain. Nevertheless, the geologist, by the methods so admirably devised by Desmarest, may follow the chain of relationship through these different regions and trace a remarkable continuity of structure. The younger rocks serve to illustrate the original condition of the more ancient, while the latter, by their extensive denudation, permit points of structure to be seen which in the former are still concealed.

No feature in the interesting volcanic district of Auvergne has attracted more attention than the trachytic protrusions.[380] Rising conspicuously along the chain of puys, they claim notice even from a distance owing to the topographical contrast which their pale rounded domes offer to the truncated, crater-bearing cones of dark cinders around them. They consist of masses of a pale variety of trachyte (domite), which in ground-plan present a circular or somewhat elliptical outline. They vary in size from the nearly circular dome of the Grand Sarcoui, which measures about 400 yards in diameter, to the largest mass of all--that of the Puy de Dôme, which extends for some 1500 yards from north to south with a breadth varying from 500 to 800 yards. They are likewise prominent from their height; in the Puy de Dôme they form the highest elevation of the whole region (1465 metres), and even in the less conspicuous hills they rise from 500 to 600 feet above the surrounding plateau.

[Footnote 380: The admirable Map and Memoirs of Desmarest on Auvergne are classics in geology. Scrope's work, vol. i. p. 45, gives still the best published account of this district. See also the work of Lecoq (_ibid._). The results of more detailed petrographical research regarding the rocks will be found in the essays of M. Michel Lévy (_Bull. Soc. Géol. France_, 1890, p. 688) and in the Clermont sheet of the Geological Survey Map of France (Feuille, 166). A bibliography of the district up to the year 1890 is given in the volume of the _Bull. Soc. Géol. France_ just cited, p. 674.]

Five such dome-shaped protrusions of trachyte have made their appearance among the cinder-cones in a space of about five English miles in length by about two miles in extreme breadth. Though opinions have varied as to the mode of formation of these domes, there has been a general agreement that their present topographic contours cannot be far from the original outlines assumed by the masses at the time of their production. The position of the trachyte bosses among the puys serves to show that they were not deep-seated masses which have been entirely uncovered by denudation, but were essentially superficial, and were protruded to the surface at various points along the plateau in the midst of already existing cinder-cones. In some cases, they have risen on or near the position of the vents of these cones. Thus the Puy de Chopine is half encircled by the crater of the Puy de la Goutte, and the Grand Sarcoui stands in a similar relation to the fragmentary crater-wall of the Petit Sarcoui.

M. Michel Lévy, in pointing out the superficial character of the domitic protrusions, has forcibly dwelt on the evidence that these rocks have undergone a comparatively trifling denudation, and that they could never have extended much beyond their present limits.[381] As Scrope pointed out, they were obviously protruded in a pasty condition, not flowing out in streams like the other lavas of the district, but consolidating within their chimneys and rising from these in rounded domes.

[Footnote 381: _Op. cit._ p. 711.]

Undoubtedly denudation, cannot have left them altogether unaffected, but must have removed some amount of material from their surface. There is reason to believe that the material so removed may have been in large part of a fragmental character, and that it was under a covering of loose pyroclastic debris that the upward termination of the trachyte column assumed its typical dome-form. Thus in the crater-wall of the Puy de la Goutte, layers of buff-coloured trachytic tuff dip gently away from the central domite mass of the Puy de Chopine. That this material was thrown out from the vent previous to the uprise of the domite may be inferred from the way in which the latter rock has obliterated the northern half of the crater. The relations of the rocks are somewhat obscured by talus and herbage, but when I last visited the locality in the spring of 1895 the structure seemed to me to be as expressed in the accompanying diagram (Fig. 344).[382]

[Footnote 382: Compare M. Michel Lévy, _ibid._]

The relative date of the protrusion of the trachytic domes cannot be very precisely defined. There can, indeed, be no doubt that it belongs to a late phase of the volcanic history. It came long after the outpouring of the older basaltic plateaux, of which large fragments emerge from beyond the limits of the younger lavas on both sides of the great ridge of the puys, and not only long after that outpouring, but even after the widespread sheets of basalt had been deeply trenched by valleys and isolated into outliers capping the hill-tops. Yet there is good evidence also that the uprise of the comparatively acid trachytes was not the last volcanic episode of the district. The abundance of dark slags and fragments of basalt lying on the domite hills shows that discharges of more basic detritus occurred after these hills had taken their place in the landscape.

Since the latest eruptions, a gradual alteration of the topographical features by denudation has been slowly but continuously going on. The Grand Sarcoui, possibly from having originally had a considerable covering of fragmentary material, shows least the effects of this waste. Its remarkably regular form, like that of an inverted cauldron (the "Chaudron," as it is called in the district), presents, in a distant view, a smooth grassy surface which slopes steeply down into the great volcanic plain. But on a nearer examination these declivities are found to be seamed with trenches which the rain-storms of centuries have dug out. The covering of loose debris has been largely washed away, though many fragments of dark slag are still strewn over the slopes, and the scars are now being cut into the domite below. A more advanced stage of decay may be seen on the Puy de Dôme, where, from greater elevation and exposure, the domite is already deeply gashed by gullies and ravines, while the slopes below are strewn with its detritus.

The region of the Velay displays on a far more extensive scale the protrusion of trachytic and phonolitic bosses, but as its volcanic history goes back beyond the time of the Puys of Auvergne, its volcanic monuments have consequently been more extensively affected by denudation.[383] A series of basaltic eruptions forming extensive sheets can there be traced, the oldest dating from Miocene time, the youngest coming down to the age of the mammoth, cave-bear and early man. During this prolonged outpouring of basic lavas there were several intervals during which materials of a more acid nature--trachytes and phonolites--were erupted. These rocks occur partly as extensive tracts, covering five or six square miles, like those of the Mezenc, the Megal, the Pic de Lizieux, and the Rand, and partly in isolated conical or dome-shaped prominences, sometimes only a few hundred feet in diameter. Upwards of one hundred distinct eruptions of phonolite have been observed in the Velay. Even in the tracts where they cover the largest space, several prominent eminences may usually be observed, not unlike in general shape the isolated cones and domes of Auvergne. In these wider areas there appears to be evidence of the outcome of the lava from one or more vents, either as superficial streams or as underground intrusive sheets. M. Boule has expressed his opinion that most of the masses of trachyte and phonolite have been the result of local and limited eruptions, the pasty rock having risen in and accumulated around its pipe, without flowing far in any direction. A section across one of these masses would present a somewhat mushroom-shaped form.[384]

[Footnote 383: In addition to the work of Scrope, the student of this important volcanic district will find an invaluable guide in the Le Puy Sheet (No. 186) of the Geological Survey Map of France, and in the _Bulletins_ of the Survey, particularly those by MM. Termier and Boule, No. 13 (1890) and No. 28 (1892).]

[Footnote 384: _Bull. Carte. Géol. France_, No. 28 (tome iv.) p. 125.]

That fragmentary ejections accompanied the protrusion of these rocks, though probably on a very limited scale, is shown by the occasional survival of portions of trachyte tuff around them. One of the most notable of these deposits occurs in the hollow between the Suc du Pertuis and the next dome to the south. It consists of fine and coarse, trachytic detritus, which in one place is rudely bedded and appears to dip away from the phonolite dome behind it at an angle of 30°. This material and its inclination are what might be expected to occur round an eruptive vent, and may be compared with those of the crater-wall of the Puy de la Goutte in relation to the domite boss of the Puy de Chopine.

The denudation of Velay has undoubtedly advanced considerably further than that of the Puys of Auvergne. The pyroclastic material which may have originally covered the domes of trachyte and phonolite has been in great part swept away. The surrounding rocks, too, both aqueous and igneous, have been extensively removed from around the necks of more enduring material. Hence the trachyte and phonolite bosses stand out with so striking a prominence as to arrest the eye even for a distance of many miles.

There cannot be any doubt that these necks have pierced the older basalts, and therefore belong to a later epoch in the volcanic history. The approximately horizontal sheets of basalt have been deeply eroded and reduced to mere fragments, and in some instances their existing portions owe their survival to the protection afforded to them by the immense protrusions of more acid material. But there is here, as well as in Auvergne, evidence of the uprise of a later more basic magma, for sheets of basalt are found overlying some parts of the trachytes and phonolites.

While the external forms of these Velay necks recall with singular vividness the features of many more ancient necks in Britain, an examination of the internal structure of some of them affords some further interesting points of resemblance. The slabs into which, by means of weathering along the joints, the rock is apt to split up are sometimes arranged with a general dip outwards from the centre of the hill, so that their flat surfaces roughly coincide with the hillslopes. In other cases the peculiar platy structure, so characteristic of phonolite, is disposed vertically or dips at a steep angle into the hill, so that the edges of the slabs are presented to the declivities, which consequently become more abrupt and rugged.

Though none of the volcanic series in Auvergne or the Velay is so acid in composition as the more acid members of the Tertiary volcanic series of Britain, the manner in which the trachytes and phonolites of the French region make their appearance presents some suggestive analogies to that of the corresponding rocks in this country. We see that they were erupted long after the outpouring of extensive basaltic plateaux, that they belonged to successive epochs of volcanic activity, that they were protruded in a pasty condition to the surface, where, more or less covered with fragmentary ejections, they terminated in dome-shaped hills or spread out to a limited distance around the vents, and lastly, that they were succeeded by a still later series of more basic eruptions, which completed the long volcanic history. We shall see in the following pages how closely the various stages in this complex record of volcanic activity may be paralleled in the geological records of Tertiary time in Britain.[385]

[Footnote 385: The phonolite necks of Bohemia, which form so prominent a feature in the Tertiary geology of that country, might likewise be cited here in illustration of the acid domes and bosses of the British Isles.]