It is one of the most interesting points in the Tertiary volcanic history that, in spite of the enormous geological revolutions that have passed since they became extinct, the sites of many scattered vents can still be recognized. A far greater number must lie buried under the basalts, and of others the positions are concealed by the sea, which now covers so large an area of the old lava-fields. Nevertheless, partly within the area of the plateaux, but still more on the surrounding tracts from which the basalts have been removed by denudation, the traces of unmistakable vents of discharge may be recognized amid the general wreck.

In Britain and the Faroe Isles, it is chiefly along the coast-line that the process of denudation has revealed the volcanic vents of Tertiary time. The interior of the country is often loaded with peat, covered with herbage, or strewn with glacial detritus: and even where indications of the vents are to be detected, it is not always possible to ascertain their true limits and connections. But where the structure of the plateaux has been laid bare along ranges of rocky precipice, the vents have sometimes been so admirably dissected by the sea that every feature of their arrangements can be satisfactorily determined.

As the actual physical connexion of these volcanic orifices with the plateaux has been in most cases removed by denudation, we can usually only by inference place them in what was probably their true relation to the plateau-eruptions. Those which project from the surface of the plateaux must, of course, be younger than the basalts through which they rise; how much younger we cannot tell. They may possibly be later than any of the plateau-sheets; they may even belong to a subsequent and waning condition of volcanic action. On the other hand, the vents which can now be traced outside of the present limits of the edges of the plateaux may, like those just mentioned, be younger than the basalt-sheets, or, on the contrary, they may be records of a period of eruptivity anterior to the emission of any of the rocks of the plateaux, and may have been deeply buried under a mass of basalt-beds subsequently removed. Positive demonstration is, from the nature of the case, impossible in these instances. But examples will be cited from the Western Isles and from Faroe, where the vents can be proved to belong to the time of the plateau-eruptions, for they are seen to have broken through some of the basalt-sheets and to have been buried under others. With this clear evidence of relationship in some cases, there need be little hesitation in believing that in other instances where no such positive connexion can be found, but where the vents are obviously such as the general structure of the plateaux would have led us to expect, they may be confidently regarded as part of the phenomena of the plateau-eruptions.

Sometimes the vents can be linked with lines of fissures or dykes. This is especially the case where they are small in size. More usually, however, no such relation can be demonstrated. It will be remembered that among the modern Icelandic eruptions, some eruptive vents, like the later cinder-cones of Laki, are ranged in a linear direction along the great fissure, while others, of an older series in the same district, almost engulphed amidst the more recent lavas, are clustered irregularly in groups. A similar diversity of arrangement has been observed among the volcanic cones of the Velay in Central France.

Considering as a whole the volcanic necks or eruptive vents which rise from the older rocks around the Tertiary basalt-plateaux, and sometimes even from the surface of these plateaux themselves, we may conveniently follow the same classification as was adopted in dealing with those of Palæozoic age, and, according to the nature of the material that now fills them, arrange them in two series: (1) Those occupied by some form of crystalline eruptive rock, and (2) those filled with volcanic agglomerate.

i. VENTS FILLED WITH DOLERITE, BASALT, ETC.

These, as the composition of the plateaux would lead us to anticipate, are numerous. They perhaps attain their most conspicuous development in Antrim, either on the tableland or among the underlying rocks round its edges. The finest example in that district is undoubtedly furnished by the lofty eminence called Slemish, which rises above the surrounding basalt-terrace, to a height of 1437 feet above the sea (Fig. 294). It is elliptical in ground-plan, measuring some 4000 feet in length by 1000 in breadth. Seen from the north, it appears as a nearly perfect cone. The material of which it consists is a coarsely crystalline olivine-dolerite, presenting under the microscope a nearly holocrystalline aggregate, in which the lath-shaped felspars penetrate the augite, with abundant fresh olivine, and wedge-shaped patches of interstitial matter. The rock is massive and amorphous, except that it is divided by parallel joints into large quadrangular blocks like a granitic rock, and wholly different from the character of the surrounding basalts. The latter, which possess the ordinary characters of the rocks of the plateaux, can be followed to within 80 yards of this neck, which rises steeply from them, but their actual junction with it is concealed under the depth of talus.

At the nearest point to which the two rocks are traceable, the basalts appear somewhat indurated, break with a peculiar splintery fracture, and weather with a white crust. These characters are still better shown on abundant fragments which may be picked up among the debris further up the slope. There can be no doubt, I think, that a ring of flinty basalt, differing considerably in texture from the usual aspect of that rock in the district, surrounds the neck. The meaning of this ring will be more clearly seen from the description of another example in Mull. About four miles to the north-east of Slemish, a smaller and less conspicuous neck rises out of the plateau-basalts. The rock of which it consists is less coarsely crystalline than that of Slemish, but its relations to the surrounding volcanic rocks are obviously the same. On the west side of Belfast Lough a boss of similar rock, about 1200 feet in diameter, rises at the very edge of the basalt escarpment into the eminence known as Carnmony Hill (Fig. 295). On its northern side it presents along its wall a mass of interposed volcanic agglomerate.[295] On visiting with Mr. M'Henry the quarry opened on the eastern face of this vent, I was much struck with the remarkable cellular structure of some parts of the dolerite. Many of the vesicles are lined with a thin pellicle of black glass, and the same substance occurs in minute patches in the body of the rock. A thin slice exhibiting this structure was found by Mr. Watts to possess the following characters:—"The rock is an ophitic dolerite consisting of plagioclase, augite, and iron ores, without olivine, enclosing one or two patches of finer basalt. The vesicles in the latter, and certain angular spaces between the crystals of the former, have been wholly or partially filled with brown glass, the outer part of which has been converted into radiating crystals of a brown mineral." The occurrence of patches of glass which seem to have been squeezed into vesicles or cracks in the body of a dolerite or andesite has been noticed in some of the Tertiary dykes. But in the present case the glass occurs as a mere coating on the walls of the larger spheroidal vesicles, the interior of which generally remains empty.

Of the other doleritic necks scattered over the surface of the Antrim plateau, I will refer to only one which occurs on the hillslopes between Glenarm and Larne. It forms a prominence known as the Scawt Hill, and consists of a boss of basalt, which, in rising through a vent in the plateau-sheets, has carried up with it and converted into marble a large mass of chalk which is now exposed along its eastern wall (Fig. 296).

As examples of similar necks which have been exposed by denudation outside the present limits of the same plateau, I may allude to those which rise through the Cretaceous and other Secondary strata on the northern coast near Ballintoy. One of the most striking of these may be seen at Bendoo, where a plug of basalt, measuring about 1400 feet in one diameter and 800 feet in another, rises through the Chalk, and alters it around the line of contact (Fig. 297). Another remarkably picturesque example is to be seen near Cushendall, where a prominent doleritic cone rises out of the platform of Old Red Sandstone, some distance to the north of the present edge of the volcanic escarpment (Fig. 298).

The greater coarseness of grain of the material filling these pipes, compared with that of the sheets in the terraces, is only what the very different conditions of cooling and consolidation would lead us to expect. There is no essential difference of composition between the two rocks. Where the erupted material has been poured out at the surface, it has assumed a finely crystalline texture, while, where it has slowly solidified within a volcanic pipe at some depth beneath the surface, and where consequently its component crystals have had more time for development, the resulting structure is much more largely crystalline, with a more or less complete development of the ophitic structure.

In the island of Mull, another instance of the same kind of vent has been observed and described by Professor Judd.[296] It rises in the conspicuous hill, 'S Airde Beinne (Sarta Beinn), about two miles south-west from Tobermory, and consists of a coarsely crystalline dolerite, which becomes finer in grain towards the outer margin (Fig. 299). No bedding, or structure of any kind beyond jointing, is perceptible in it. Examined in thin sections under the microscope, this rock is found to be another typical ophitic dolerite, consisting of lath-shaped felspars embedded in augite, with here and there wedge-shaped portions of interstitial matter and grains of olivine. Dr. Hatch found the felspars to contain spherical inclusions of devitrified glass, filled with black granules and trichites, and he observed that, under a high power, the interstitial matter is seen to consist mainly of a greenish-brown isotropic substance, in which are inclosed small crystals of augite, skeleton-forms and microlites of felspar, sometimes in stellate aggregates, as well as club-shaped, cruciform, arrow-headed and often crested microlites of magnetite.

Towering prominently above the flat basalt sheets, this neck has an oval form, measuring about half a mile in length by a quarter of a mile in breadth. Its central portion, however, instead of rising into a rugged hill-top, as is usually the case, sinks into a deep hollow, which is filled with water, and reminds one of a true crater-lake (Figs. 299, 300). The middle of the neck is thus concealed from view, and we can only examine the hard prominent ring of dolerite that surrounds the tarn. The material occupying the hollow may be softer than that of the ring, and may have been scooped out by denudation. What we now see may not be the original surface, but may have been exposed after the removal of possibly hundreds of feet of overlying material. On the other hand, it is conceivable that the hollow is really a crater-lake which was filled up with detritus and may have been overspread with basalt, since removed. It may be suggestively compared with the crater-hollows revealed by denudation on the cliffs of Stromö and Portree Harbour, which will be described in a later part of this chapter. Possibly some more easily removable agglomerate, representing an eruption later than that of the dolerite, may occupy the centre of the volcanic pipe.

One of the most interesting features of this vent is to be found in its relation to the surrounding basalts. The marginal parts of the rock along the line of contact are much finer in grain than the rest, and have obviously cooled more rapidly. The contrast between them and the ordinary dolerite nearer the centre, however, cannot be properly understood, except in thin sections under the microscope. Dr. Hatch, to whom I submitted my specimens, observed that, in place of the structure above described, the marginal parts show an absence of the ophitic grouping except in small isolated patches. Instead of occurring in large grains or plates enveloping the felspars, the augite is found in numerous small roundish grains, together with grains of magnetite, in equal abundance and of similar size. The felspars are speckled over with opaque particles; olivine has not been detected.

For miles around the vent, the plateau-rocks are of the usual type—black, compact, sometimes amygdaloidal, alternating with more coarsely crystalline decomposing bands, the separation between different sheets being often marked by the ordinary red ferruginous partings. But around the margin of the neck, they have undergone a remarkable metamorphism. The portions of them which adhere to the outer wall of the neck have lost their distinct bedding, and have been, as it were, welded together into an indurated compact, black to dull-grey rock, so shattery and jointed that fresh hand-specimens, three or four inches in length, are not easily obtainable. Especially marked is one set of joints which, running approximately parallel, cause the rock to split into plates or slabs. These joints are sometimes curved. Yet, in spite of the alteration from its normal character, the basalt retains in places some of its more usual external features, such, for instance, as its amygdaloidal structure, the amygdales consisting of calcite, finely acicular mesotype, and other minerals.

Examined under the microscope, this altered basalt presents "a confused aggregate of colourless microlites (felspar?) and innumerable minute granules of magnetite, these two constituents being very unequally distributed. Sometimes the colourless portions preponderate, in other places the opaque granules are heaped together in black patches, which may possibly mark the position of fused augites."[297]

In the zone of contact-metamorphism around some of the volcanic pipes in the plateaux, we see changes analogous to, but less developed than, those which have been superinduced on so large a scale round the great eruptive bosses of gabbro, granophyre, etc., that have broken up the terraced basalts along the west coast of Scotland. I shall accordingly return to this subject in connection with phenomena presented by these younger rocks (p. 386).

ii. VENTS FILLED WITH AGGLOMERATE

While the necks of dolerite or basalt cannot always be satisfactorily discriminated from bosses which may never have established a connection with the surface, there is no room for any doubt in this respect in the case of those filled with fragmentary materials. As has been already pointed out, the occurrence of true volcanic agglomerate may be accepted as evidence of the existence of an eruptive vent communicating with the surface of the earth. The agglomerate in the vents associated with the basalt-plateaux, like that of the Palæozoic vents, is generally exceedingly coarse, and without any trace of structure. Blocks of all sizes up to masses some yards in length, and of the most diversified materials, both volcanic and non-volcanic, are dispersed confusedly through a granular paste of similar miscellaneous composition.

An instructive example of the general characteristics of agglomerate-vents, and of the relation of these vents to the surrounding tuffs and basalts, is to be found at the island of Carrick-a-raide, on the north coast of Antrim, and on the opposite mainland. The visible mass of this neck is about 1000 feet in diameter, but the boundaries, except on the land side, are concealed by the sea. The material filling up the vent is a coarse agglomerate, in which blocks and bombs of basalt, with pieces of chalk and flint, are stuck at all angles in a dull dirty-green granular tuff. Some large and small intrusions of basalt rise through it. Owing partly to these intrusions, and partly to the grass-covered slope that separates it from the line of cliff, the actual contact of this neck with the volcanic beds of the escarpment cannot be seen. I have no doubt, however, that the tuff, which has already been referred to as so conspicuous a member of the series here, was discharged from this vent.[298] The materials are as usual coarser in the pipe than beyond it, but the finer portion or matrix of the agglomerate is similar to many bands of the tuff. The structure of the locality may be diagrammatically represented as in Fig. 301. The bedded tuff is thickest in the neighbourhood of the vent, and gradually dies away on either side of it.

But another important inference may be drawn from this locality. I have already pointed out that the lower basalts here reach their minimum thickness. Their basement beds thin away towards the vent as markedly as the tuff thickens. Obviously they cannot have proceeded from that point of eruption. Yet, that they had begun to be poured out before the discharge of the tuff is shown by their underlying as well as overlying that rock, though westward, owing to the thinning away of the undermost basalts, the tuff comes to lie directly on the Chalk. Hence, we may legitimately infer that in this neighbourhood one or more other vents supplied the sheets of the lower basalts.

In the island of Mull a number of detached bosses or patches of agglomerate much obscured by invasions of granophyre probably mark the sites of volcanic vents. They will be more particularly noticed in Chapter xlvii. One of their most interesting features is the large number of fragments of felsitic or rhyolitic rocks which they contain.

In the promontory of Ardnamurchan, where the basalt-plateau has been invaded and displaced by later intrusions of crystalline rocks, and has likewise been reduced to such a fragmentary condition by denudation, some interesting examples of agglomerate necks have been laid bare. One of the largest of these occurs on the north shore at Faskadale. Cut open by the sea for more than a quarter of a mile, this neck is seen to be filled with a coarse agglomerate, composed mainly of basalt-blocks and debris, but crowded also with angular and subangular pieces of different close-grained andesitic, felsitic and porphyritic rocks belonging to the acid series to be afterwards described.[299] Some of these stones exhibit a very perfect flow-structure, and closely resemble certain fine-grained, flinty, intrusive rocks in Mull, to which allusion will subsequently be made. The matrix of the agglomerate is of the usual dull dirty-green colour, but is so intensely indurated that on a fresh fracture it can hardly be distinguished from some of the crystalline rocks of the locality. The neck is pierced in all directions with dykes and veins of basalt, dolerite, andesite, gabbro, and felsitic rocks. Similar intrusions continue and increase in numbers farther west until the cliffs become a labyrinth of dykes and veins running through a mass of rocks which appears to consist mainly of dull dolerites and fine gabbros. Though the relations of this vent to the plateau-basalts are not quite plain, the agglomerate seemed to me to rise out of these rocks. At least the basalts extend from Achateny to Faskadale, but, as they are followed westwards, they are more and more invaded by eruptive sheets, and assume the indurated character to which I have already referred.

On the south side of the peninsula of Ardnamurchan, another agglomerate, noticed by Professor Judd,[300] rises into the bold headland of Maclean's Nose, at the mouth of Loch Sunart, and affords better evidence of its relation to the bedded basalts. It measures about 1000 yards in length by 300 in breadth, and its summit rises more that 900 feet above the sea, which washes the base of its southern front. It is filled with an agglomerate even coarser than that on the northern coast. The blocks are of all sizes, up to eight or ten feet in diameter. By far the largest proportion of them consists of varieties of basalt and andesite, slaggy and vesicular structures being especially conspicuous. There are also large blocks of different andesitic porphyries and felsitic rocks like those just referred to, a porphyry with felspar crystals two inches long being particularly abundant. All the stones are more or less rounded, and are wrapped up in a dull-green compact matrix of basalt-debris. There is no stratification or structure of any kind in the mass. Numerous dykes or veins of basalt, of andesite, and of a porphyry, resembling that of Craignure, in Mull, traverse the agglomerate. Some of the narrow basalt-dykes cut through the others.

The position of the vent, with reference to the surrounding rocks, will be understood from the accompanying section (Fig. 302). On the eastern side, the agglomerate can be seen to abut against the truncated ends of the flat beds of the plateau-basalts, which are of the usual bedded compact and amygdaloidal character. There can be no doubt, therefore, that the vent has been opened through these basalts. But it will be observed that the latter belong to the lower part of the volcanic series. These lowest sheets are exposed on the slope, resting upon yellowish and spotted grey sandstone, with seams of jet and a reddish breccia, which, lying in hollows of the quartzites, quartz-schists, and mica-schists, form no doubt the local base of the Jurassic rocks of the district. Hence, the vent, though younger than the older sheets of the plateau, may quite well be contemporaneous with some of the later sheets.[301]

An interesting feature at this locality is the peculiar grouping of some of the large dykes in the area around the agglomerate. They run in the direction of the vent, and one or other of them may represent the fissure or fissures on which the volcanic orifice was blown open to the surface. Another notable element in the geological structure of the ground is the vast amount of intrusive material, both in dykes and sheets, which has been erupted. The intrusive sheets of Ben Hiant form the most prominent eminence in this part of Ardnamurchan. Reserving them for description in the following Chapter (p. 318), I will only remark here that they partly overlie the agglomerate, and are therefore, to some extent at least, younger than the vent. They belong to that late stage in the history of the basalt-plateaux when the molten material, no longer getting ready egress to the surface, forced its way among the rocks about the base of the bedded basalts, and more especially on the sites of older vents, which were doubtless weak places, where it could more easily find relief.

The large neck now described is only one of a group scattered around it in the ground to the north. Two of these may be seen rising through a detached area of Jurassic limestones and shales at the northern base of Ben Hiant. A third, almost obliterated by the intrusive sheets, may be traced at the western end of that mountain above Coiremhuilinn. Two others rising through the schists on either side of Beinn na h-Urchrach, have been much invaded by the sills of that eminence (Fig. 326). It is doubtless owing to the extensive denudation of the basalt-plateau, and the consequent uncovering of the rocks underneath it, that this series of vents has been laid bare.[302]

By far the largest mass of agglomerate in any of the Tertiary volcanic areas of Britain is that which occurs on the north side of the main valley of Strath, in Skye.[303] Unfortunately, it has been so seriously invaded by the eruptive rocks of the Red Hills, that its original dimensions and its relations to the surrounding rocks, especially to the bedded basalts, are much obscured (see Fig. 348). It can be followed continuously from the lower end of Loch Kilchrist along the southern slopes of Beinn Dearg Bheag round to the western roots of Beinn Dearg Mhor—a distance of more than two miles in a straight line, and from Kilbride to the flank of Beinn na Caillich above Coire-chat-achan—a direct distance of two miles and a quarter. A similar rock, possibly a portion of the same mass, appears in Creagan Dubha, on the north side of the Red Hills. If the whole of this agglomerate forms part of one originally continuous mass, it must have been upwards of two miles in diameter. There may, however, have been two or three closely adjacent vents. The Beinn na Caillich patch, for example, appears to belong to a different area, and that of Creagan Dubha is also probably distinct. But there seems no reason to doubt that the mass which forms Cnoc nam Fitheach, and all the long declivity on the southern flank of Beinn Dearg Bheag, occupies part of the site of a single volcano. Owing to the absence of sufficient sections, it is hardly possible to determine how much of this fragmentary material should be assigned to the actual chimney. The diameter of the whole mass is almost two miles. But possibly a considerable proportion of this accumulation belongs to the external cone which gathered round the vent, so that the eruptive pipe might thus be of much smaller dimensions than the superficial area of the agglomerate. The subsequent invasion of so much granophyre, not only that of the Red Hills, but that of numerous smaller intrusions, has indurated the agglomerate and made the investigation of its structure somewhat unsatisfactory.

It might be supposed that the mere existence of intrusive bosses and veins rather furnishes an argument in favour of considering the visible agglomerate to belong to a deeper-seated part of the erupted material than the external cone. But, as will be afterwards shown, there is some reason to regard the present conical or dome-shaped outlines of the granophyre hills as not far from their original forms, and to believe that, like the trachytic Puys of Auvergne, they were much more superficial than plutonic eruptions. A study of the cinder cones of Central France shows that even these superficial accumulations have been invaded not only by bosses but by dykes.[304]

The agglomerate of the great Strath vent is a coarse tumultuous assemblage of blocks and bombs, imbedded in the usual dull, dirty-green matrix. Among the stones, grit and sandstone, together with scoriaceous, vesicular and amygdaloidal basalts are specially abundant; also pieces of various quartz-porphyries and granophyres, among which a black felsite like that of Mull may often be recognised. In some places, large masses of altered limestone and quartzite (Cambrian) are included; in others, pieces of yellow sandstone and dark shale (Jurassic), or of the bedded lavas. Some of these masses may be 100 yards or more in length. Occasionally a breccia, mainly made up of acid materials—granophyre or granite,—has been noticed by Mr. Harker along the north side of the Red Hills, which he thinks may rather be of the nature of a crush-breccia than a part of the true agglomerate.

The agglomerate of this district is wholly without stratification or structure of any kind. On the north-west side of Loch Kilchrist, indeed, it weathers into large tabular forms, the parallel surfaces of which dip to south-west; but this is probably due only to jointing. Here and there, dykes of basalt cut the rock in a general north-westerly direction, but their number is remarkably small when compared with the prodigious quantity of them in the limestone at the bottom and opposite side of the valley, some of which may possibly mark the fissure on which the vent was placed. More abundant and extensive are the masses of granophyre that rise particularly along the outer margin of the agglomerate near Loch Kilchrist. These may be connected with the great boss that forms the Red Hills, of which further details will be given in Chapter xlvi.[305]

The important question of the relation of this agglomerate to the plateau-basalts does not admit of satisfactory treatment, owing to destruction of the evidence by the intrusion of the granophyre, and likewise to enormous denudation. Nevertheless, some traces still remain to indicate that the basalts once stretched over the site of the vent, which probably rose through them. Looking westward from the Hanks of Beinn Dearg Bheag to the other side of Loch Slapin, the geologist sees the bold basalt-escarpment of Strathaird presenting its truncated beds to him at a distance of only two miles. That these lavas were once prolonged eastwards beyond their present limits is obvious, and that they stretched at least over these two intervening miles can hardly be doubted. But we can still detect relics of them on the flanks of Beinn Dearg. As we follow the agglomerate round the margin of the granophyre that mounts steeply from it, we lose it here and there under beds of amygdaloidal basalt. The rocks next the great eruptive mass of the mountain are so indurated and shattered that it is difficult to separate them from each other and determine their relative positions. But, so far as I could ascertain, these basalts are fragments of beds that overlie the agglomerate (Fig. 303). This is not the only place along the flanks of the Red Hills where portions of the bedded basalts have survived. Other localities will be subsequently alluded to.

The Strath vent has been drilled through the Cambrian limestone, and as the result of protracted denudation it now towers steeply 500 or 600 feet above that formation on the floor of the valley. Of the material discharged from it over the surrounding country no certain trace now remains. We may infer from the nature of the rock which fills it that towards the end, if not from the beginning of its activity, its discharges consisted mainly of dust and stones. A cone, of which the remains are two miles in diameter, must surely have sent its fragmentary materials far and wide over the surrounding region. But on the bare platform of older rocks to the south, beyond the bottom of the agglomerate declivities, not a vestige of these erupted materials can now be found. Westward the escarpment of Strathaird remains to assure us that no thick showers of ashes fell at even so short a distance as two miles, either before or during the outpouring of the successive basalt sheets still remaining there. We may therefore conclude with some confidence that here, as at Ardnamurchan, the vent is younger than at least the older parts of the basalt-plateau. Unfortunately the uprise of the large bosses of granophyre that stretch from the Red Hills to Loch Sligachan has entirely destroyed the vent and its connections in that direction. There is no certain proof that any molten rock ever issued from this orifice, unless we suppose the fragmentary patches of amygdaloid on the southern flank of Beinn Dearg Bheag to be portions of flows that proceeded from this centre of eruption. The basalt-plateau which still remains in Strathaird no doubt formerly extended eastwards over Strath and northwards across the site of the Red Hills and Cuillins, joining on to the continuous tableland north of Lochs Brittle and Sligachan. How much of the plateau had been built up here before the outburst of the vent cannot be ascertained. The agglomerate may possibly, of course, belong to the very latest period of the plateau-eruptions, or even to a still younger phase of Tertiary volcanic history. The impression, however, made on my mind by a study of the evidence from the Western and Faroe Isles is that the necks of agglomerate, like those of dolerite and basalt, really belong to different epochs of the plateau period itself; and mark some of the vents from which the materials of the plateaux were successively emitted.

The example of Carrick-a-raide (p. 277) is peculiarly suggestive when we regard it in connexion with the great Strath vent. Already the progress of denudation has removed at least half of the layer of dust and stones which, thrown out from that little orifice, fell over the bare chalk-wolds and black basalt-fields of Antrim. The neck that marks the position of the volcanic funnel has been largely cut away by the waves, and is almost entirely isolated among them. The vents at Canna, Portree and the Faroe Isles, to be afterwards described, unquestionably belong to the eruptions of the plateau-period, for their connection with the basalts can be clearly established. At the Strath vent, however, the march of destruction has been greater. The connexion between this vent and the materials ejected from it has been entirely removed, and we can only guess from the size of the remaining neck what may have been the area covered by the discharges from this largest of all the volcanic cones of the Inner Hebrides.

Other masses of similar agglomerate are observable in the same region of Skye, where they not improbably mark the sites of other vents. Unfortunately their original limits and relations to the rocks through which the eruptive orifices were drilled have been much obscured by the uprise of the great masses of gabbro and granophyre of the Cuillin Hills. Several of these isolated intrusions occur in the midst of the gabbro, as in Harta Corry and on the west side of the Blaven ridge. Another mass is interposed between the gabbro and granophyre on Druim an Eidhne and at the base of the lavas between Druim an Eidhne and the Camasunary valley. Mr. Harker has found a huge mass of agglomerate underlying the bedded basalts to the north and west of Belig, one of the hills on the west side of the large valley that runs from the head of Loch Slapin to Loch Aynort. This mass has its bottom concealed by the granophyre which underlies it; but it reaches a maximum thickness of perhaps 1000 feet, rapidly thinning out and disappearing. It generally resembles the Strath agglomerate, but is distinguished by including a large proportion of fragments of gabbro. Mr. Harker remarks that "a study of these agglomerates points to the existence of both gabbros and granophyres older than the volcanic series, and therefore distinct from the gabbros and granophyres now exposed at the surface."

It is a suggestive fact that so many detached masses of agglomerate should occur around and within the areas of the great eruptive bosses of gabbro and granophyre. They seem to indicate the former existence of groups of volcanic vents in these tracts, and may thus account for the uprise of such large bodies of intrusive material through what must have been a weakened part of the terrestrial crust.

Further north in Skye a much smaller but more perfectly preserved vent has been laid open by denudation on the south side of Portree Bay—a deep inlet which has been cut out of the plateau-basalts and their underlying platform of Jurassic sandstones and shales. The great escarpment of the basalts has, at the recess of Camas Garbh, been trenched by a small rivulet, aided by the presence of two dykes. The gully thus formed exposes a section of a neck of agglomerate that underlies the basalts of the upper half of the cliff. This neck is connected with a thick deposit of volcanic conglomerate and tuff which, lying between the basalts, extends from the neck to a considerable distance on either hand. The general relations of the rocks at this locality are represented in Fig. 304.