It was therefore with no little interest that I detected a series of vents at four separate localities, viz. Castleton, Grange Mill, Hopton, and Kniveton Wood. I have no doubt that a more extended search will bring others to light. Those observed by me are all filled with coarse agglomerate, the blocks in which are mostly composed of different lavas, sometimes with the addition of blocks of limestone, while the matrix consists mainly of lapilli of basic devitrified glass.

The most typical examples form a group of two, possibly three, vents which rise into two isolated, smooth, grassy dome-shaped hills at Grange Mill, five miles west from Matlock Bath.[33] In external form and colour, these eminences present a contrast to the scarped slopes of limestone around them. They at once recall the contours of many of the volcanic necks in Central Scotland. On examination it is found that the material composing them is a dull green agglomerate, the matrix of which is a compact substance weathering spheroidally, and full of small lapilli of minutely vesicular diabase. The larger stones consist, for the most part, of various vesicular dolerites or diabases, together with some pieces of limestone and occasionally large blocks of the latter rock, altered into a saccharoid condition. Two dykes of dolerite or basalt traverse the margin of the larger vent.

The steep sides of these agglomerate domes rise from the low ground around them to a height of 100 to 180 feet, their summits being a little more than 900 feet above the sea. The smaller neck is nearly circular, and measures about 1000 feet in diameter. The larger mass is less regular in shape, and is prolonged into such a bulge on the south-east as to suggest that its prolongation in that direction may really mark the position of a third and much smaller vent contiguous to it. The longer diameter of the larger mass is 2300 and the shorter 1300 feet.

On the south and west sides, the surrounding limestone can be traced up to within a few feet of the edge of the agglomerate, and its strata are there found to be much jumbled and broken, while their texture is rather more crystalline than usual, though not saccharoid. The two necks are separated by a narrow valley in which no rock is visible. Their opposite declivities meet at the bottom of this hollow, and are so definitely marked off that, even in the absence of proof that they are disjoined by intervening limestone, there can be little hesitation in regarding each hill as marking a distinct vent. A wider valley extends along the eastern base of the necks, and slopes upward on its east side until it is crowned by a long escarpment of limestone, which reaches a height of 1000 feet above the sea, or about 100 feet above the valley from which it rises. Unfortunately, the bottom and slopes of this depression are thickly covered with soil, but at one or two places debris of fine tuff may be observed, and at the northern and southern ends of the hollow well-bedded green and reddish tuff appears, dipping gently below the limestone escarpment. This band of volcanic detritus evidently underlies the limestone, and forms most of the gentle slope on the east side of the valley. It may be from 70 to 100 feet thick. That it was discharged from one or both of the necks seems tolerably clear. Its material resembles that forming the matrix of the agglomerate. The general arrangement of the rocks at this interesting locality is represented in Fig. 179, which is reduced from my survey on the scale of six inches to a mile. A section across the smaller vent would show the structure represented in Fig. 180.

This group of vents lies in the southern of the two tracts of the volcanic district. In the northern tract a mass of agglomerate pierces the base of the limestone escarpment about a quarter of a mile west from the entrance to the Peak Cavern at Castleton.[34] It is rudely semicircular in area, stretching down the slope until its northern extension is lost under the lower ground. The agglomerate is not well exposed, but it can be seen to be a green, granular crumbling rock, made up in great part of minutely vesicular lapilli, enclosing blocks of various diabases two feet long or more. From the abrupt way in which this agglomerate rises through the limestone, there can be little doubt that it marks the position of one of the volcanic vents of the time. As it stands on the extreme northern verge of the limestone area, the ground further north being covered with the Yoredale rocks and Millstone Grit, it is the most northerly of the whole volcanic district.

Along the southern margin of the limestone country a group of agglomerate masses probably marks another chain of vents. These are specially interesting, inasmuch as they abut on the Yoredale series, and may thus be looked upon as among the latest of the volcanic chimneys. One of them is seen at Hopton,[35] where along the side of the road a good section is exposed of coarse tumultuous agglomerate, having a dull green matrix, composed of green, brown, and black, minutely cellular, basic, devitrified, glassy lapilli, showing under the microscope abundant microlites and crystals or calcareous pseudomorphs of olivine, augite, and felspar, and much magnetite dust. Through this matrix are distributed blocks of slaggy basalt and dolerite. An interesting feature of this mass is the occurrence in it of some veins, two or three inches broad, of a compact black porphyritic basalt. I did not trace the relations of this agglomerate to the stratified rocks around it. But its internal structure and composition mark it out as a true neck. It extends, according to the Geological Survey map, for about half a mile along the edge of the limestone, and is represented as being separated by two faults from the Yoredale series immediately to the south. So long as the belief is entertained that the toadstones are contemporaneous outflows of lava lying on certain definite horizons, far below the summit of the limestones, the position of the Hopton agglomerate is only explicable on the assumption of some dislocation by which the Yoredale shales have been brought down against it. But when we realize that the rock is an unstratified agglomerate, probably marking the place of a volcanic vent, and therefore rising transgressively through the surrounding strata, the necessity for a fault is removed, or if a fault is inserted its existence should be justified on other evidence than the relations of the igneous rock to the surrounding strata.

Four miles to the south-west of Hopton, on the slope of the hill at Kniveton Wood, another remarkable mass of agglomerate forms a rounded ridge between the two forks of a small stream.[36] Its granular matrix, like that of the other necks, consists of lapilli of minutely vesicular basic glassy lava or pumice, and encloses large and small rounded blocks of finely cellular basalt and pieces of limestone. The rock is unstratified, and in all respects resembles that of ordinary Carboniferous necks in Scotland. Its relations to the Yoredale rocks are laid bare in the channels of the streamlets. There the shales and thin limestones may be seen much broken and plicated, their curved and fractured ends striking directly at the agglomerate. They may be traced to within a yard of the agglomerate. On the Geological Survey map the igneous rock is represented as bounded by two parallel faults. But I hardly think that this explanation suffices to account for the relations of the rocks and their remarkable boundary-line, which seems to me to be undoubtedly the wall of a volcanic vent. To the east of the streams, another mass of agglomerate may mark another neck, while to the north, a third detached area of the same kind of rock, rising among the limestones, may be regarded as likewise a distinct mass. At this locality, therefore, there are two, possibly three, vents. One of these, from the way in which it cuts across the Yoredale shales and limestones, is to be assigned to a time later than the older part of the Yoredale series, and thus, like the Hopton mass, it indicates that in the south of the volcanic area eruptions did not cease with the close of the deposition of the thick limestones, but were prolonged even into the time of the Yoredale rocks.

A further proof of the late age of these southern patches of volcanic material is shown by two bands of vesicular toadstone in the Yoredale series, a little south from the village of Kniveton. These rocks are traced on the Survey Map, and are shown in a diagram in the Memoir, where their position is sought to be explained by a system of parallel faulting.[37] I was able to trace the actual contact of the western band with the strata underneath it, and satisfied myself that there is no fault at the junction. The igneous material is regularly bedded with the Yoredale shales and limestones. Either, therefore, these bands are intercalated lava-streams or intrusive sills. If mere vesicular structure were enough to distinguish true outflowing lavas, then there could be no doubt about these Kniveton rocks. But this structure is found in so many Carboniferous sills, particularly in those thin sheets which have been injected into coals and black shales, that its presence is far from decisive. The vesicles in the Kniveton rocks are small and pea-like, tolerably uniform in size and shape, and crowded together. They are thus not at all like the irregular cavities in the ordinary cellular and scoriaceous lavas of the toadstone series.

Whether or not the question of their true relations be ever satisfactorily settled, these Kniveton bands are certainly younger than the lower portion of the Yoredale group. Their evidence thus agrees with that of the southern agglomerates in showing that the volcanic activity of this region was continued even after the thick calcareous masses of the Carboniferous Limestone series had ceased to be deposited.

Besides the six necks to which I have referred, a rock in Ember Lane, above Bonsall, probably belongs to another vent.[38] It is particularly interesting from the great preponderance of limestone fragments in it. The volcanic explosions at this locality broke up the already solidified limestones on the floor of the Carboniferous Limestone sea, and strewed them around, mingled with volcanic blocks and dust of the prevailing type.

When the district has been more carefully searched, other centres of eruption will no doubt be discovered. It may then be possible to depict the distribution of the active vents, and to connect with them the outflow of the bedded lavas. So far as I have been able to ascertain, there are no necks of dolerite or basalt, though, as I have shown, dykes or veins of molten rock are occasionally to be found in the agglomerates of the necks.

4. THE LAVAS AND TUFFS.—I have referred to the opinion of De la Beche that the toadstones of Derbyshire were poured out as lava-streams without any accompanying fragmentary discharges, and to the correction of this opinion by the subsequent observations of Jukes and of the Geological Survey. But though the existence of interbedded tuffs has long been known, it was not until Mr. Bemrose's more careful scrutiny that the relative importance of the tuffs among the lavas was first indicated. He has shown that a number of the bands mapped as "toadstone" are tuffs, and he has discovered other bands of tuff which have not yet been placed on any published map.

In examining the outcrops of the various toadstones of Derbyshire we learn that some of them are lavas without tuffs, probably including a number of bands, which are really sills; that others are formed of both lavas and tuffs, and that a third type shows only bedded tuff. Each of these developments will deserve separate description. But before entering into details, we may take note of the varying thicknesses of the different toadstones which have been determined by observation at the surface or by measurement underneath in mining operations. In some cases a distinct band of toadstone, separated by many feet or yards of limestone from the next band, and therefore serving to mark a separate volcanic discharge, may not exceed a yard or two in total thickness, and from that minimum may swell out to 100 feet. The majority of the bands probably range between 50 and 100 feet in thickness. In one exceptional case at Snitterton, a mass of "blackstone" is said to have been proved to be 240 feet thick, but this rock may not improbably have been a sill.[39] The true contemporaneous intercalations seem to be generally less than 100 feet in thickness.

(a) Lavas without Tuffs.—Examples occur of sheets of toadstone which consist entirely of contemporaneously ejected diabase, basalt or dolerite. This rock is then dull green or brown in colour, more or less earthy in texture, and irregularly amygdaloidal. The vesicles are extremely varied in size, form and distribution, sometimes expanding until the rock becomes a slaggy mass. A central more solid portion between a scoriaceous bottom and top may sometimes be observed, as at the Great Rocks Quarry, Peak Forest Limeworks (Fig. 181). In this, as in other examples, a remarkably hummocky and uneven surface of limestone lies below the igneous band, the calcareous rock presenting knobs and ridges, separated by cauldron-shaped cavities and clefts, some of which are several yards deep. These inequalities are filled in and covered over with a soft yellow and brown clay, varying up to three or four feet thickness, and passing upwards into the more solid toadstone. There can hardly be any doubt that this singularly uneven limestone surface is due to the solvent action of water lying between the limestone and the somewhat impervious toadstone above, and that the clay represents partly the insoluble residue of the calcareous rock, but chiefly the result of the action of the infiltrating water on the bottom of the igneous band.[40]

Junctions of the upper surfaces of the lava-sheets with the overlying limestone show that the igneous material sometimes assumed hummocky forms, which the calcareous deposits gradually overspread and covered.[41] A good example of this kind may be observed by the roadside at the foot of Raven's Tor, Millersdale. As shown in the subjoined figure, the limestone has here been worn into a cave, the floor of which is formed by the toadstone. The latter rock, of the usual dull green, slaggy and amygdaloidal character, is covered immediately by the limestone, but I did not observe any fragments of the toadstone, nor any trace of ashy materials in the overlying calcareous strata. This section shows that after the outflow of the lava, the sedimentation of the limestone was quietly resumed, and the igneous interruption was entirely buried.

In some cases there is evidence of more than one outflow of lava in the same band of toadstone. Jukes believed that each band "was the result, not of one simultaneous ejection of igneous matter, but of several, proceeding from different foci uniting together to form one band," and he found that near Buxton, two solid beds of toadstone could be seen to have proceeded from opposite quarters towards each other without overlapping.[42]

In Millersdale the authors of the Geological Survey Memoir on North Derbyshire observed that a band of toadstone about 100 feet thick showed six distinct divisions, which they were disposed to regard as marking so many separate beds.[43] In Tideswell Dale, on the west side of the valley, immediately to the south of the old toadstone quarry, two bands of toadstone are seen to be separated by a few yards of limestone.

(b) Lavas with Tuffs.—It will probably be found that in many, if not in most cases, the outflow of lava was preceded, accompanied or followed by fragmental discharges. As far back as 1861, Jukes noticed that a toadstone band, about 50 feet thick, near Buxton consisted of two solid beds of lava "with beds of purple and green ash, greatly decomposed into clay, both above and below each bed and between the two."[44]

An interesting section, showing this intercalation of the two kinds of material is exposed at the lime-kilns beyond the southern end of the railway viaduct at Millersdale Station. Over a mass of solid blue limestone (1 in Fig. 183) lies a band of bright yellow and brown clay (2), varying from six inches to two feet in thickness. This may be compared with the clay found above the limestone at Peak Forest (Fig. 181). But it is probably a layer of highly decomposed tuff. It is succeeded by a thin band of greenish limestone (3) containing an admixture of fine volcanic detritus, and partially cut out by an irregular bed, four to eight feet thick, of a highly slaggy, greenish, decomposing, spheroidal and amygdaloidal diabase (4). This unmistakable lava-sheet is followed by a bed of green granular tuff (5), which in some places reaches a thickness of three feet, but rapidly dies out. Over a space several yards in breadth, the succeeding strata are concealed, and the next visible rock is a dark, compact dolerite which weathers spheroidally (6).

(c) Tuffs without Lavas.—Mr. Bemrose has shown that some of the bands of toadstone consist entirely of bedded tuff. In these cases, so far as the present visible outcrops allow us to judge, no outflow of lava accompanied the eruption of fragmentary materials. But that the ejection of these materials was not the result of a sudden spasmodic explosion, but of a continued series of discharges varying in duration and intensity, is indicated by the well-bedded character of the tuff and the alternation of finer and coarser layers. Large blocks of lava, two feet or more in diameter, may mark some of the more vigorous paroxysms of the vents, while the usual fine granular nature of the tuff may point to the prevailing uniformity and less violent character of the eruptions. Bands of tuff 70 feet or more in thickness, without the intercalation of any limestone or other non-volcanic intercalation, point to episodes of such continued volcanic activity that the ordinary sedimentation of the sea-bottom was interrupted, or at least masked, by the abundant fall of dust and stones.

One of the best exposures of such intercalations of bedded tuffs was pointed out to me by Mr. Bemrose, immediately to the east of the village of Litton. The matrix is crowded with the usual minutely vesicular glassy lapilli, and encloses fragments of diabase of all sizes, up to blocks more than a foot in diameter. The rock is well stratified, and the layers of coarse and fine detritus pass beneath a group of limestone beds. The actual junction is concealed under the roadway, but only two or three feet of rock cannot be seen. The lowest visible layer of limestone is nodular and contains decayed bluish fragments which may be volcanic lapilli. Immediately above the lower limestones the calcareous bands become richly fossiliferous. Some of their layers consist mainly of large bunches of coral; others are crowded with cup-corals, or are made up mainly of crinoids with abundant brachiopods, polyzoa, lamellibranchs, gasteropods and occasional fish-teeth. This remarkable profusion of marine life is interesting inasmuch as it succeeds immediately the band of volcanic ash.

Another well-marked zone of tuff, with no traceable accompaniment of lava, has already been referred to as connected with the Grangemill vents. In this case also, the limestone that lies directly upon the volcanic material is rather impure and nodular in character. The tuff itself is well bedded, perhaps from 70 to 100 feet thick and dips underneath an overlying series of marine limestones.

I did not observe thin partings of tuff and disseminated volcanic lapilli among the limestones, such as are so marked in the Lower Carboniferous formations of West Lothian, and in the Limerick basin, to be described in the following chapter. But a diligent search might discover examples of them, and thus prove that, besides the more prolonged and continuous eruptions that produced the thick bands of tuff, there were occasional feeble and intermittent explosions during the accumulation of the thick sheets of limestone. Some of the layers of "red clay" observed in shafts sunk for mining purposes may perhaps represent such spasmodic discharges of fine fragmental material.

5. THE SILLS.—No attempt has yet been made to determine whether and to what extent the toadstone bands include true intrusive sheets. My own brief examination of the ground does not warrant me in making any positive statement on this subject. I can hardly doubt, however, that some, perhaps not a few, of the toadstone bands are really sills. In the accounts of these rocks contained in the mining records a distinction, as already remarked, appears to have been generally drawn between "toadstone" and "blackstone." The latter term is applied to the black, fresh, more coarsely crystalline, and generally non-amygdaloidal rocks, which, so far as I have been able to examine them, have the general external and many of the internal characters of the Carboniferous sills of Central Scotland. At Snitterton near Matlock one of these "blackstones," as already mentioned, is said to have been found to be 240 feet thick.[45]

It is stated that the toadstones, though subject to great variations in thickness, are never seen to cut across the limestones.[46] But I suspect that proofs of intrusion and transgression will be found when diligently sought for. It appeared to me that the dark, compact, crystalline dolerite, which was formerly quarried in the middle of Tideswell Dale, may be separated from the vesicular toadstone of that valley, which is undoubtedly a true lava-flow, and that it does not always occupy the same horizon there, being sometimes below and sometimes above the amygdaloid. Where it rests on a band of red clay the latter rock has been made columnar to a depth of nine feet.[47] Alteration of this kind is very rare among the Carboniferous bedded lavas, but is by no means infrequent in the case of sills. But the most important proof of alteration which I have myself observed occurs at Dale Farm near the village of Peak Forest, where the limestone above a coarsely crystalline dolerite has been converted into a white saccharoid marble for about two yards from the junction.

3. THE ISLE OF MAN

Rising from the middle of the Irish Sea, within sight of each of the three kingdoms, with a history and associations so distinct, yet so intimately linked with those of the rest of Britain, this interesting island presents in its geological structure features that connect it alike with England, Scotland and Ireland, while at the same time it retains a marked individuality in regard to some of the rocks that form its framework. Its great central ridge of grits and slates, which still rises 2000 feet above the sea in the summit of Snaefell, must have formed a tract of dry land in Carboniferous time, until it sank under sea-level, and was buried beneath the Carboniferous and later formations. Along the southern margin of this ancient land, a relic of the floor of the Carboniferous sea has been preserved in a small basin of Carboniferous Limestone which covers about seven or eight square miles. This remnant has a special interest in geological history, for it has preserved the records of a series of volcanic eruptions which took place contemporaneously with the deposition of the Carboniferous Limestone.

The geology of the Isle of Man was sketched in outline by J. F. Berger,[48] J. Macculloch,[49] and J. S. Henslow,[50] and was afterwards more fully illustrated by J. G. Cumming.[51] To the last-named observer we owe the recognition of true intercalated volcanic rocks among the calcareous formations of the southern end of the island. These rocks have subsequently been studied in greater detail by a number of geologists. An excellent general account of them was published in 1874 by Mr. John Horne, of the Geological Survey.[52] A few years later some further observations on them were prepared by J. Clifton Ward.[53] More recently their petrography has been studied by Messrs. E. Dickson, P. Holland and F. Rutley,[54] and in more detail by Mr. B. Hobson.[55] To some of the observations of these writers reference will be made in the succeeding pages. During the progress of the Geological Survey in the Isle of Man, the rocks in question have been mapped in detail by Mr. A. Strahan and Mr. G. W. Lamplugh, and I have had an opportunity of examining the coast-sections with the last-named geologist. The following description of these sections is taken mainly from my field note-book. The full details will appear in the official Memoirs.

It may be remarked at the outset that the last outcrop of the plateau-lavas of the Solway basin occurs only 60 miles from the south end of the Isle of Man, at the foot of the hills of Galloway, the blue outline of which can be seen from that island. The distance from the Manx volcanoes to the nearest of the puys of Liddesdale is about 100 miles. Though the fragment which has been left of the ejections is too small to warrant any confident parallelism, there appears to be reason to believe that, alike in geological age and in manner of activity, the Manx volcanoes may be classed with the type of the puys.

The Carboniferous strata of the Isle of Man lie in a small trough at the south end of the island. The lowest members of the series consist of red conglomerates and sandstones, which pass upward into dark limestones full of the characteristic fossils of the Carboniferous Limestone. As the bottom of the basin is on the whole inclined seawards, the highest strata occur along the extreme southern coast. It is there that the volcanic rocks are displayed. They occupy a narrow strip less than two miles in length, which is almost entirely confined to the range of cliffs and the ledges of the foreshore. Yet though thus extremely limited in area, they have been so admirably dissected along the coast, that they furnish a singularly ample body of evidence bearing on the history of Carboniferous volcanic action.

Unfortunately the bottom of the volcanic group is nowhere visible. At the east or lower end of the series, exposed on the shore, an agglomerate with its dykes appears to truncate the Castletown Limestones. No trace of any tuff has been noticed among these lower limestones. We may infer that the volcanic activity began after they were deposited. The highest accessible portions of the volcanic group, as Mr. Horne showed, are clearly exposed on the coast at Poyll Vaaish, intercalated in and overlying the dark limestones of that locality (Fig. 184), which have been assigned, from their fossil contents, to the upper part of the Carboniferous Limestone series.[56] The Manx volcanoes may therefore be regarded as having probably been in eruption during the later portion of the Carboniferous Limestone period.

Owing to irregularities of inclination, the thickness of the volcanic group can only be approximately estimated. It is probably not less than 200 or 300 feet. But as merely the edge of the group lies on the land, the volcanic rocks may reach a considerably greater extent and thickness under the sea.

The volcanic materials consist mainly of bedded tuffs, but include also several necks of agglomerate and a number of dykes and sills. So far as I have observed, they comprise no true lava-streams.[57] These Manx tuffs present many of the familiar features of those belonging to the puy-eruptions of Central Scotland, but with some peculiarities worthy of attention. They are on the whole distinctly bedded, and as their inclination is generally in a westerly direction, an ascending order can be traced in them from the eastern end of the section to the highest parts of the group associated with the Poyll Vaaish limestones. Their colour is the usual dull yellowish-green, varying slightly in tint with changes in the texture of the materials, the palest bands consisting of the finest dust or volcanic mud. Great differences in the size of their fragmentary constituents may be observed in successive beds, coarse and fine bands rapidly alternating, with no admixture of non-volcanic sediment, though occasional layers of fine ash or mudstone, showing distinct current-bedding, may be noticed.

Pauses in the succession of eruptions are marked by the intercalation of seams of limestone or groups of limestone, shale and black impure chert. Such interstratifications are sometimes curiously local and interrupted. They may be observed to die out rapidly, thereby allowing the tuff above and below them to unite into one continuous mass. They seem to have been accumulated in hollows of the tuff during somewhat prolonged intervals of volcanic quiescence, and to have been suddenly brought to an end by a renewal of the eruptions. There are some four or five such intercalated groups of calcareous strata in the thick series of tuffs, and we may regard them as marking the chief pauses in the continuity or energy of the volcanic explosions.

An attentive examination of these interpolated sedimentary deposits affords some interesting information as to the submarine conditions in which the eruptions took place. The intercalations, sometimes 12 feet or more in thickness, consist mainly of dark limestones, enclosing the usual Carboniferous Limestone fossils; black shales, sometimes showing very fragmentary and much macerated remains of ferns and other land-plants; and black impure argillaceous chert or flint, arranged in bands interposed between the other strata, and also in detached lumps and strings. The dark flaggy limestones and black shales may be paralleled lithologically with those of Castletown and Poyll Vaaish. Indeed, there seems to be little doubt that they represent the contemporaneous type of marine sediment that was gathering on the sea-floor outside the volcanic area, and which during intervals of quiescence or feeble eruptivity spread more or less continuously into that area. The thick mass of tuff must thus have been strictly contemporaneous with a group of calcareous muddy and siliceous deposits which gathered over the bottom beyond the limits of the showers of ashes.

One of the most singular features of these sedimentary intercalations is the occurrence of the black cherty material. It may generally be observed best developed at the bottom and top of each group of included strata. Looking at the lumps of this substance scattered through the adjoining tuffs, we might at first take them for ejected fragments, and such no doubt may have been the derivation of some of them. But further examination will show that, as a rule, they are of a concretionary nature, and were formed in situ contemporaneously with or subsequent to the deposition of the tuffs. The accompanying section (Fig. 185) represents the manner in which the chert is distributed through two or three square yards of tuff overlying one of the calcareous groups. The material has been segregated not only into lumps, but into veins and bands, which, though on the whole parallel with the general stratification-planes of the deposits, sometimes run irregularly in tongues or strings across these planes, as shown in Fig. 186, where the dark chert band which overlies the limestones and shales sends a tongue upwards for several inches into the overlying tuff.

That these interstratified calcareous and muddy strata were laid down in water of some considerable depth may be inferred from their general lithological characters. The dark carbonaceous aspect of the limestones points to the probable intermingling of much decayed vegetation with the remains of the calcareous organisms of which these strata chiefly consist. The thin unimportant bands or partings of dark shale show that only the finest muddy sediment reached the quiet depths in which the strata were deposited, while the macerated fern-fragments suggest a long flotation and ultimate entombment of terrestrial vegetation borne seawards from some neighbouring land.

The cherty bands and nodules, like the flints of the chalk, bear their testimony to the quiet character of the sedimentation in rather deep water beyond the limits within which the sediment from the land was mainly accumulated on the sea-bottom. The origin of these siliceous parts of the series of deposits has still to be investigated. Whether or not they are to be referred to organic causes like chalk-flints, and the radiolarian cherts of the Lower Silurian system, they furnish a fresh example of the remarkable association of such siliceous material with volcanic phenomena, which has now been observed in many widely separated areas all over the world.

If we next turn to the stratification of the tuffs, we obtain further evidence of undisturbed conditions of deposition on the sea-floor. The bedding of these volcanic masses, though distinct, appears for the most part to be due rather to the eruption and settlement of alternately finer and coarser detritus than to any marked drifting and rearrangement of these materials by current-action into different layers. Throughout the series of tuffs, indeed, there is, on the whole, a notable absence of any structure suggestive of strong currents or of wave-action in the dispersal and reassortment of the volcanic detritus. The ashes and stones were discharged in such a way as to gather irregularly over the sea-floor into ridges and hollows. There does not seem to have been sufficient movement in the bottom water to level down these inequalities of surface, for we find that they remained long enough to allow twelve feet or more of calcareous and siliceous ooze to gather in the hollows, while the intervening ridges still stood uneffaced until buried under the next fall of ashes. At rare intervals some transient current or deeper wave may have reached the bottom and spread out the volcanic detritus lying there. Such exceptional disturbances of the still water are not improbably indicated by occasional well-defined stratification, and even by distinct false-bedding, in certain finer layers of tuff.

The materials of the tuffs are remarkably uniform in character and conspicuously volcanic in origin. With the exception of occasional blocks of limestone, which range up to masses several feet, and occasionally several yards, in diameter, the dust, lapilli and included stones consist entirely of fragmentary basic lava, so persistent in its lithological features that we may regard its slightly different varieties as merely marking different conditions of the same rock. The accumulation of pumiceous ash in this southern coast of the Isle of Man is one of the most remarkable in Britain. As Mr. Hobson has well shown, the matrix of this tuff consists of irregular lapilli, representing what may have been various conditions of solidification in one original volcanic magma. This magma he has described as an "augite-porphyrite" or olivine-basalt. Some of the lapilli, as he noted, consist of a pumice "crowded with vesicles which occupy more space than the solid part"; others show nearly as many vesicles, but the glass is made brown by the number of its fine dust-like inclusions; a third type presents the cells and cell-walls in nearly equal proportions. The same observer found that where the substance is most cellular the vesicles, fairly uniform in size, measure about a tenth of a millimetre in longest diameter.

An interesting feature of the tuffs is the abundant occurrence of loose felspar crystals throughout the whole group up to the highest visible strata. These crystals, sometimes nearly an inch in length, appear conspicuously as white spots on weathered surfaces of the rock. They are so much decayed, however, that it is difficult to extract them entire. On the most cursory inspection they are observed to enclose blebs of a greenish substance like the material that fills up the vesicles in the pumiceous fragments and in the pieces of cellular lava.

I have not ascertained the original source of these scattered felspars. In one of the dykes on the north side of the agglomerate at Scarlet Point, as was pointed out by Mr. Hobson, large crystals of plagioclase occur in the melaphyre, but the felspars in the tuffs and agglomerates differ so much from these that we cannot suppose them to have come from the explosion of such a rock. I failed to detect any other mineral in detached crystals in the tuffs, but a more diligent search might reveal such, and afford some grounds for speculating on the probable nature of the magma from the explosion of which the scattered crystals were derived. It is at least certain that this magma must have included a large proportion of plagioclase crystals.

Between the lapilli and the minute pumice-dust that constitute the matrix of this tuff much calcite may be detected. Though this mineral may have been partly derived from the decay of the felspar in the lava-fragments, I believe that it is mainly to be attributed to the intermingling of fine calcareous ooze with the ash accumulated on the sea-floor. A more remarkable association of the same kind will be described in later pages from King's County in Ireland. That abundant calcareous organisms peopled the sea in which the Manx Carboniferous volcanoes were active is shown by the contemporaneously deposited limestones. The tuffs themselves are occasionally fossiliferous. Species of Spirifer, Productus and other brachiopods, together with broken stems of encrinites, may be found in them, and doubtless the diffused calcite, though now crystalline, as in the limestones, and showing no organic structure, owes its presence to the detritus of once living organisms.

The stones imbedded in the tuff consist almost exclusively of slightly different varieties of the same pale, always vesicular rock, and sometimes pass into a coarse slag. They vary up to six feet or more in length. In many cases, they appear to have been derived from the disruption of already solidified lava, for their vesicles are not elongated or arranged with reference to the form of the block, but have been broken across and appear in section on the outer surface. In other instances, however, the cavities are large and irregular in the centre of the block, while on the outside they are smaller and are drawn out round the rudely spherical shape of the mass, as in true volcanic bombs.

The limestone fragments enclosed in the tuff include pieces of the dark carbonaceous and of the pale encrinal varieties. In no case did I observe any sensible alteration of these fragments. They seem to have been derived from material disrupted and ejected during the opening of successive vents, and not to have been exposed for any considerable time to the metamorphic influence of volcanic heat and vapours.

Narrow though the strip of volcanic material is along the south coast of the Isle of Man, it has fortunately preserved for us some of the vents from which the tuffs were ejected. A group of these vents, three or four in number, may be traced along the shore in a general W.N.W. and E.S.E. line from Scarlet Point for rather more than a mile. Their margins are in some places exceedingly well defined. The most striking example of this feature occurs in the most westerly vent, where a neck of remarkably coarse volcanic agglomerate rises vertically through well-bedded, westerly-dipping tuff (Fig. 187). In other portions of their boundaries no sharp line can be drawn between the material filling the vent and that of the surrounding tuffs. Hence it is difficult to define precisely the form and size of the vents. I am inclined to believe from this indefiniteness of outline, and from the remarkable structure of the dykes, to which I shall afterwards refer, that the presently visible parts of these necks must lie close to the mouths of the original vents, if indeed they do not actually contain parts of the craters and of their surrounding walls.

The materials that have filled up the eruptive vents consist chiefly of agglomerate, but partly also of intrusive portions of vesicular lava. The agglomerate is composed of similar materials to the tuffs. Its matrix shows the same extraordinarily abundant fine greenish-grey basic pumiceous lapilli, with the same kind of plentiful loose felspar-crystals. The large blocks of lava, too, resemble in composition and structure those of the bedded tuffs, but greatly exceed them in size and abundance.

Besides the fragments of vesicular lava, there occur also occasional blocks of limestone. Some of these are several yards in length. Messrs. Strahan and Lamplugh have mapped a large mass of limestone at the Scarlet vent, which, so far as can be observed, lies in the agglomerate—a large cake of white limestone with pebbles of quartz, which has probably been broken off from some underlying bed and carried up in the chimney of the volcano.

As a rule the agglomerate is a tumultuous, unstratified mass. But in many places it shows lines of bedding and, as already stated, passes outward into ordinary bedded tuff, the number and size of the ejected blocks rapidly diminishing. Where this transition occurs we seem to see a remnant of the base of the actual volcanic cone. Thus, in the most westerly vent already cited, while the wall of the vent has been laid bare on the side next the sea, so that the agglomerate on the beach descends vertically through the surrounding bedded tuffs, on the western side the cliffs have preserved a portion of the material that accumulated outside the orifice (Fig. 187). In this section we observe that the coarse agglomerate which fills up the main part of the vent has been left with a hummocky, uneven surface, and that a subsequent and perhaps feebler eruption of finer material has covered over these inequalities, and has extended to the left above the fine tuffs through which the agglomerate has been drilled.