We now come to the examination of another distinct phase of volcanic action during Tertiary time in Britain. The igneous rocks that have been under consideration in the foregoing chapters, whether poured out at the surface or injected below ground, have been chiefly of basic, partly indeed, like the peridotites, of ultra-basic character. Some, however, have shown an andesitic or intermediate composition. Reference has also been made to the probable eruption of acid rhyolites in the long interval between the outflow of the lower and the upper basalts in Antrim. But we now encounter a great series, decidedly acid in composition, in the more largely crystalline members of which the excess of silica is visible to the eye in the form of free quartz. While there is a strong contrast in chemical composition between this series and the rocks hitherto under discussion, there are also marked differences in structure and mode of occurrence. Like the gabbros, all the masses of acid rock now visible appear to be intrusive. They have been injected beneath the surface, and therefore record for us subterranean rather than superficial manifestations of volcanic action.

The existence of rocks of this class in the midst of the basic masses has long been recognized. They were noticed by Jameson, who described the hills between Loch Sligachan and Broadford as composed of "a compound of felspar and quartz, or what may be called a granitel, with occasional veins of pitchstone."[359] Macculloch gave a fuller account of the same region, and classed the rocks as chiefly "syenite" and "porphyry."[360] In Antrim, also, even in the midst of the basalt-tableland, masses of "pitchstone-porphyry "pearlstone-porphyry," "clay-porphyry," and "greystone" were observed and described.[361] In more recent years Professor Zirkel has given a brief account of the so-called "syenite and porphyry" of Mull and Skye,[362] and the late Professor Von Lasaulx fully described the "trachyte" or rhyolite of Antrim.[363]

This interesting series of rocks embraces a greater variety of petrographical characters than any other portion of the British Tertiary volcanic rocks. On the one hand, it presents thoroughly vitreous masses, some of which in their colour, lustre and microscopic structure remind us of recent obsidians. On the other hand, it affords coarsely crystalline compounds, to which no other name than granite can be assigned, and which, did we not know their geological position, might almost be classed with some of the most ancient eruptive rocks. Between these two extremes abundant gradations may be found, including beautiful spherulitic rocks, felsites and rhyolites.

In dealing with such a series of intrusive rocks, we again encounter the difficulty of reaching certainty as to their relative dates of eruption, since in each case all that can usually be affirmed is that the intrusive mass is younger than that into which it is injected. It is quite possible that protrusions of acid rocks occurred at intervals during the accumulation of the basic masses, as may perhaps be inferred from the rhyolite-tuffs and conglomerates of Antrim and from the occurrence of fragments of siliceous lavas in the gravels near the base of the basalt-plateau of Mull, and in the agglomerates of that island as well as of other districts.[364] It is probable, therefore, that at the time when the basalts of the plateaux were emitted, there existed, within reach of volcanic explosions, masses of granophyric, felsitic or rhyolitic rocks, fragments from which were shot up the funnels of discharge. That portions of these rocks were actually intruded into the basalt-sheets before the building up of the plateaux was completed appears to be proved in Antrim. Elsewhere, however, no evidence has yet been obtained of any such intrusion until after the close of the plateau-period. On the contrary, in every case where the relative ages of the rocks can be fixed, the acid are younger than the basic protrusions.

The only known exceptions to this rule are the latest basalt-dykes. Hence, while amid the large and varied series of acid rocks, which no doubt represents a wide interval of time, some may belong to comparatively early epochs in the protracted volcanic period, the actual available evidence places the emission of these rocks, as a whole, towards the end of the volcanic history. This evidence I shall bring, forward in full detail, since it necessitates an abandonment of what has been the general belief in regard to the relative ages of the rocks.

i. PETROGRAPHY OF THE ACID ROCKS

The classification of the rocks which best harmonizes the field-evidence and the detailed study of their mineralogical composition, is one that arranges these volcanic protrusions into two series. In the one, the orthoclase is sanidine, and the rocks range from the most vitreous pitchstone through perlitic and spherulitic varieties to rhyolite ("quartz-trachyte"). In the other series, which embraces by far the largest proportion of the whole, the orthoclase is always turbid, and in this respect as well as in many others the rocks remind us rather of ancient eruptive masses than of those which have appeared in Tertiary time. They range from flinty felsitic varieties, which are obviously devitrified glasses, through different textures of quartz-porphyry into granophyre, and finally into granite. As I have been unable to recognize any essential difference of structure and composition between these acid Tertiary rocks and those of far earlier geological time, I give them the names which no petrographer would hesitate to apply to them if they were of Palæozoic age. It has long appeared to me that these rocks furnish conclusive evidence of the misleading artificiality of any petrographical nomenclature in which relative antiquity is made an essential element of discrimination.

Granite.—That true granites form part of the Tertiary volcanic series of the British Isles has now been completely established. They occur as bosses and sills which have been intruded into the gabbros and all older rocks. They are thus proved not only to belong to the Tertiary period, but to one of the latest phases of its volcanic history. But besides these granites, the relative age of which can be definitely fixed, there occur others which, standing alone and at some distance from the basaltic plateaux, can only be inferentially classed in the Tertiary series. To this group belong the granite masses of the Isle of Arran and the Mourne Mountains in north-eastern Ireland.

Taking first the unquestionably Tertiary granites which occur as bosses and intrusive sheets, we have to note that the more coarsely crystalline granophyres are hardly to be distinguished externally from granite. As the dark ferro-magnesian constituent of these rocks was generally believed to be hornblende, they were called by the older petrographers "syenite"; that is, granite with hornblende instead of mica. The peculiar micropegmatitic groundmass, which constitutes the distinguishing feature of the granophyres, may occasionally be observed so reduced in amount as only to appear here and there between the other minerals, which are grouped in a granitic structure. From this condition, one step further carries us into a true granite, from which all trace of the granophyric character has disappeared. Such gradations may be traced even within short distances in the same boss of rock. Thus, in the hornblende-biotite-granite boss of Beinn-an-Dubhaich, Skye, a thoroughly granitic arrangement of the component minerals is observable in the centre, while a specimen taken from near the edge on the shore of Camas Malag shows the development of a granophyric groundmass. But, though the large bosses are usually somewhat coarsely crystalline in the centre, and tend to assume finer felsitic textures around their borders, as was observed long ago by Oeynhausen and Von Dechen,[365] the granitic structure is sometimes exhibited even at the very edge, and not only so, but in the dykes that protrude from the bosses into the surrounding rocks. Thus the Beinn-an-Dubhaich mass, at its margin in Camas Malag, sends a vein into the surrounding limestone, but though more close-grained than the main body of the rock, this vein is neither felsitic nor granophyric, but truly granitic in structure.

So far as I have observed, the true granites contain a brown mica and also a little hornblende, both visible to the naked eye, but generally somewhat decomposed. These rocks are thus hornblende-biotite-granites (amphibole-granitites of Rosenbusch). They may be defined as medium-grained aggregates of quartz, orthoclase (also plagioclase), biotite and hornblende, with sometimes magnetite, apatite, epidote and zircon. Dr. Hatch found that in some instances (Beinn-an-Dubhaich) the quartz contains minute inclusions (glass?), bearing immovable bubbles with strongly-marked contours; while in others (Beinn-na-Chro, Skye) this mineral is full of liquid inclusions with bubbles, sometimes vibratile, sometimes fixed. He remarked that the quartz and felspar have consolidated almost simultaneously, but that in some instances (Marsco, Glen Sligachan) there are isolated roughly idiomorphic crystals, of a white, less turbid orthoclase, which belong to a slightly earlier consolidation than that of the more kaolinized felspar of the rest of the rock.

The granite of the island of Arran, in the Birth of Clyde, which is here included in the Tertiary volcanic series, has long been recognized as consisting of two distinct portions, an eastern or coarse-grained, and a western or fine-grained variety. The latter sends veins into the former. These granites contain orthoclase, plagioclase, quartz and dark mica, the quartz being often idiomorphic with respect to the felspar, and a tendency towards a micropegmatitic structure being sometimes observable. A distinguishing characteristic of the Arran granite is the cavernous or drusy structure which it presents, the cavities being often lined with well-crystallized orthoclase and smoky quartz.[366] The granite of the Mourne Mountains in Ireland closely resembles that of Arran. Its druses, with their beautifully terminated minerals, have long been well known.

Microgranite.—This term is applied to certain intrusive masses, which megascopically may be classed with the quartz-porphyries and felsites, but which microscopically are found to possess a holocrystalline granitic groundmass of quartz and orthoclase, through which are scattered porphyritic crystals of the same two minerals, sometimes also with plagioclase, augite, magnetite or apatite. Rocks of this type do not appear to be abundant. They occur as dykes and bosses, but occasionally also as sheets. I have collected them from Skye, Rum and Ardnamurchan.

Granophyre.—Under this name may be grouped the large majority of the acid rocks which play an important part in the geology of the West of Scotland. They are typically developed in the islands of Mull and Skye. Generally pale grey or buff in colour, they range in texture from the true granites, into which, as above stated, they graduate, to exceedingly close-grained varieties like the felsites of Palæozoic formations. In the great majority of them the micrographic intergrowth of quartz and felspar, known as micropegmatite, is their conspicuous structure, and even constitutes most of their substance. They may thus be classed generally as granophyres, in the sense in which this term is employed by Rosenbusch, but without his limitation of it to pre-Tertiary rocks.

The specific gravity of these rocks has been determined from a series of specimens by Mr. A. Harker to range from about 2·3 among the felsites to 2·7 among the granites. No chemical analyses of these rocks have yet been made, but they have been subjected to microscopical examination, and their general structure and composition are now known.

The typical granophyre of the Inner Hebrides outwardly closely resembles an ordinary granite of medium grain, in which the component dull felspar and clear quartz can be readily distinguished by the naked eye. Throughout all the varieties of texture there is a strong tendency to the development of minute irregularly-shaped drusy (miarolitic) cavities, which here and there give a carious aspect to the rock. That these cavities, however, are part of the original structure of the rock, and are not due to mere weathering, is shown by the well-terminated crystals of quartz and felspar which project into them. On a small scale, it is the same structure so characteristic of the granite of the Mourne Mountains and of parts of that of Arran.

Examined under the microscope, a normal specimen of the granophyre of the Western Isles presents a holocrystalline groundmass, which fills all the interspaces between the crystals of earlier consolidation. This groundmass consists of an aggregate of clear quartz and turbid orthoclase, arranged as micropegmatite, but also in more or less idiomorphic crystals. In some parts, the two dominant minerals are grouped in alternate parallel fibres, diverging from the surface of the enclosed crystals, which are thus more or less completely surrounded by a radially fibrous mass. The felspathic portion of the micropegmatite which usually surrounds the orthoclase crystals, when viewed between crossed Nicols, is found to extinguish simultaneously with the central crystal.[367] In other parts, the felspar forms a kind of network, the meshes of which are filled up with quartz. Through the groundmass, besides the clear quartz and dull orthoclase, some ferro-magnesian or other additional constituent is generally distributed, but usually somewhat decomposed. In certain varieties Dr. Hatch found an abundant brown mica, as in the rock at Camas Malag, Skye. In others, a pyroxene occurs, which he observed in minute greenish grains, sometimes completely enclosed in the quartz. In a third variety, the dark constituent is hornblende, the most remarkable example of which is one to be seen at Ishriff, in the Glen More of Mull, where the ferro-magnesian mineral takes the form of long dirty-green needles, conspicuous on a weathered surface of the rock. A fourth variety is distinguished by containing plagioclase in addition to or instead of orthoclase. In the rock of the sheet forming Cnoc Carnach, near Heast, in Skye, Dr. Hatch observed both orthoclase and plagioclase scattered through a fine micropegmatitic groundmass, and in a part of the boss at Ishriff he found the rock to be composed mainly of plagioclase, in a micropegmatitic groundmass of quartz and felspar, with a few scattered grains of a pale brown augite and grains of magnetite. A fifth variety is marked by the prominence of the crystals of quartz and felspar of earlier consolidation, and by the fineness of grain in the surrounding micropegmatitic groundmass, whereby a distinct porphyritic structure is developed. Rocks of this kind are megascopically like ordinary quartz-porphyries. Still another variety has been detected by Mr. Teall in the rock of Meall Dearg, at the head of Glen Sligachan, Skye, in which, besides irregular patches which may represent decayed biotite, and others which are possibly ilmenite, the rare mineral riebeckite is present.[368]

Felsite.—The close-grained rocks into which the ordinary granophyres frequently graduate may be conveniently grouped under the general name of Felsite. They differ in no essential feature from the felsites of the Palæozoic formations. They are more particularly developed, as might be expected, in those places where the conditions have been most favourable for rapid cooling, while the more coarsely crystalline granophyres occur where the material may be supposed to have consolidated most slowly. Where the acid magma has been injected into chinks and fissures so as to take the form of veins or dykes, it is sometimes felsitic, sometimes granophyric, in texture. Along the margin of large bosses, like those of Mull and Skye, it frequently though not invariably has assumed a fine texture, with even spherulitic and flow-structures. But in the centre of large bosses it usually appears as coarse granophyre or as granite.

The felsites vary in texture from flinty or horny to dull finely-granular, and in colour from white through shades of grey, buff and lilac, to black, generally with porphyritic felspars and blebs of quartz. Where these porphyritic enclosures increase in size and number, the rocks cannot be distinguished externally from ancient quartz-porphyries. In general the groundmass of these rocks has been completely devitrified. But in some dykes enough of the glassy base remains to show their original vitreous condition. A gradation can thus be traced from thoroughly glassy pitchstone into completely lithoid felsites and crystalline granophyres.

A characteristic feature of the felsitic varieties of acid rock is their flow-structure, which they often display in great perfection. Sometimes, indeed, this structure has been so strongly developed as to cause the rock to weather along the planes of flow and to break up into thin slabs.

Many of these rocks also present admirably developed spherulitic structures, varying from microscopic minuteness up to large round or egg-shaped balls nearly two inches in diameter, and often distributed in lines along those of flow-structure. They likewise exhibit a frequent development of micropegmatite. No line indeed can be drawn between these felsites and the granitoid varieties, for the same characteristic granophyric intergrowth of felspar and quartz runs through them all.

Pitchstone.—This name is applied to the glassy varieties apart from their chemical composition, and specially denotes the possession of a vitreous structure. Some of the rocks to which it has been applied are probably glassy varieties of andesite, others are dacites, while some may be as acid as the most acid felsites and granophyres. The pitchstones are found in veins or dykes which traverse different geological formations up to and including the great granophyre bosses of the Inner Hebrides. They vary in colour from a deep jet-black or raven-black to a pale bottle-green, and in lustre from an almost glassy obsidian-like to a dull resinous aspect. Occasionally they assume a felsitic texture, owing to devitrification, and also a finely spherulitic structure. Some varieties appear to the naked eye to be perfectly homogeneous, others become porphyritic by the appearance of abundant sanidine crystals.

The microscopic structure of the British pitchstones has not yet been fully worked out. The beautiful feathery microlites of the Arran dykes, first made known by David Forbes, and subsequently described by Zirkel, Allport and others, are well known objects to geological collectors. Dr. Hatch, in whose hands I placed my tolerably large collection of specimens and their thin slides, furnished me with some preliminary notes on the slides, from which the following generalized summary is compiled.

At the one end of the pitchstone group we have a nearly pure glass, with no microlites, and only a few scattered crystals of sanidine, quartz, augite or magetite. The glass in thin slices is almost colourless, but generally inclines to yellow, sometimes to dark-grey. Some varieties of the rock are crowded with microlites, in others these bodies are gathered into groups, the glass between which is nearly free from them. Among the minerals that have been observed in this microlitic form are sanidine, augite, hornblende (forming the beautiful green feathery or fern-like aggregates in the Arran pitchstones, Fig. 3) and magnetite. Sometimes the rudimentary forms appear as globulites, or as belonites, but more commonly as dark trichites. Among the more definite mineral forms are grains of sanidine, quartz and augite. The porphyritic crystals are chiefly sanidine, augite and magnetite, but plagioclase occasionally occurs. The development of spherulites is well seen in a few of the slides, and occasionally perlitic structure makes its appearance.

The interesting rhyolitic areas of Antrim include several varieties of pitchstone. One of these is described by Professor Cole as "a glassy pyroxene-rhyolite, on the verge of the rhyolitic andesites." Another is a blue-black porphyritic obsidian.[369]

Rhyolite (Quartz-Trachyte).—This rock has been abundantly erupted in north-east Ireland, where it rises in occasional bosses among the plateau-basalts.[370] It is best exposed at the Tardree and Carnearny Hills, where it has long been quarried. Its petrographical characters at that locality were described by Von Lasaulx as those of a typical quartz-trachyte rich in tridymite, and containing large crystals of glassy sanidine, isolated narrow laths of plagioclase (probably andesine), grains of smoky-grey quartz, partly bounded by dihexahedral faces, and a few scattered flakes of a dark-coloured mica. The groundmass is microgranitic, and under a high power is resolvable into a confused aggregate of minute microlites of felspar, with interstitial quartz-granules.[371] More recently a detailed investigation of the petrography of the Antrim rhyolites has been conducted by Professor Cole, who has called attention to their remarkable varieties of structure, ranging from perfect volcanic glass to a thoroughly lithoidal texture, and exhibiting flow, perlitic and spherulitic structures.[372]

Intrusive masses of rhyolite are also found in the Carlingford region. One of these, seen at Forkhill, is a velvet-black almost resinous rock with abundant quartz and felspar, and sometimes displaying beautiful flow-structure. It will be more particularly described in Chapter 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.

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]

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]

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.

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.

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]

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]

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.