The older volcanic series of this hill has been broken through by the agglomerate which occupies a true neck, and is abruptly marked off from all the rocks older than itself. There is no trace of any of the older lavas or tuffs thickening towards this vent. On the contrary they are completely truncated by it, and their outcrops on the north side reappear from under the agglomerate on the south side. Their escarpments are wrapped round by the agglomerate which likewise fills the head of the hollow that had been previously worn by denudation out of the stratified deposits between the oldest lavas. There is thus a violent unconformability between the later and the older volcanic rocks of Arthur Seat.

The length of time indicated by this stratigraphical break must be great. There is no known discordance in the Carboniferous system of the Lothians, yet the Coal-measures, Millstone Grit, Carboniferous Limestone series and much of the Calciferous Sandstones were stripped from this hill before the eruption of the agglomerate. It will be shown in the sequel that a nearly similar amount of denudation preceded some of the probably Permian eruptions of Fife.

The agglomerate contains abundant fragments of the older volcanic series. Its matrix is a dull red gravelly detritus, crowded with blocks of all sizes up to a yard or more in diameter. It is pierced by a column or plug of basalt, which sends veins into it, and rises to the apex of the hill. A beautiful olivine-basalt forms the lateral mass of the Lion's Haunch, which rests on the agglomerate.

In general characters the agglomerate of Arthur Seat resembles that of some of the younger vents of Fife which pierce the Coal-measures and are connected with tuffs that lie unconformably on the Carboniferous Limestone. On these various grounds I think that it may be reasonably assigned to the same geological period.

That a new vent should be opened, after the lapse of one or more geological periods, on or near the site of more ancient volcanic orifices is an incident of which, as we have seen, the geological history of the British Isles furnishes a number of examples. It will be remembered that little more than a mile to the south of Arthur Seat lies the great vent of the Braid Hills, which in the time of the Lower Old Red Sandstone gave forth such a huge pile of lavas and tuffs. Volcanic energy thereafter entirely died away, and in this district was succeeded by a prolonged period of quiescence, during which the Lower Old Red Sandstone was upraised and extensively denuded, while the Upper Old Red Sandstone was deposited. At length, in the immediate neighbourhood, from one or more vents, the exact site of which is not certainly known, the older lavas and tuffs of Arthur Seat, Calton Hill and Craiglockhart Hill were erupted. Again, after another vast interval, a new volcano appeared, and the agglomerate and younger basalts of Arthur Seat were ejected from it. This is one of the most striking examples in this country of the remarkable persistence of volcanic energy in the same locality.

There is no evidence at Arthur Seat itself to fix the geological date of the last volcanic activity of the hill. If the group of younger rocks stood alone, with no other trace of post-Carboniferous eruptions in the surrounding district, a plausible conjecture as to its age would not be easily offered. But in reality it is not a solitary example of such rocks; for within sight, on the opposite side of the Firth of Forth, its counterparts may be seen. To the description of these numerous and clearer illustrations I now proceed.

The East of Fife is remarkable for a large assemblage of volcanic vents, which, unlike those in Ayrshire and Nithsdale, stand alone, their superficial ejections having been removed by denudation, and no connection being traceable between them and any Permian sandstones. The vents filled up with agglomerate and pierced with plugs and veins of basalt, rise through the Carboniferous rocks, but have left no record for precisely defining their geological age. On the one hand, it is quite certain that in this district volcanic eruptions took place during the earlier half of the Carboniferous period. To the north of Largo, and still more distinctly to the north-east of Leven, sections occur to show the contemporaneous outpouring of volcanic rocks during the time of the Carboniferous Limestone. The Leven section, seen in a ravine a little to the north-east of the town, is specially important. It presents a succession of red and green fine sandy tuffs, interstratified with fire-clays and sandstones, and containing a zone of basalt in the centre. These rocks lie not far from the top of the Carboniferous Limestone series.

On the other hand, there is equally clear proof of far later eruptions. From St. Andrews to Elie a chain of necks may be traced, having the same general characters, and piercing alike the Calciferous Sandstones, and the older part of the Carboniferous Limestone series. That these vents must in many cases be long posterior to the rocks among which they rise, is indicated by some curious and interesting kinds of evidence. They are often replete with angular fragments of shale, sandstone and limestone, of precisely the same mineral characters as the surrounding strata, and containing the same organic remains in an identical state of fossilization. It is clear that these strata must have had very much their present lithological aspect before the vents were opened through them. Again, the necks may often be observed to rise among much contorted strata, as, for example, along the crest of a sharp anticlinal arch, or across a synclinal basin. The Carboniferous rocks must thus have been considerably plicated before the time of the volcanic eruptions. In the next place, the vents often occur on lines of dislocation without being affected thereby. They must be posterior, however, not only to these dislocations, but also to much subsequent denudation, inasmuch as their materials overspread the rocks on each side of a fault without displacement. Hence we conclude with confidence, that a great deal of volcanic activity in the East of Fife must have been posterior to most, if not all, of the Carboniferous period.

In the neighbourhood of Largo, further important evidence is presented, confirming and extending this conclusion. The highest member of the Upper Coal-measures, consisting of various red sandstones, with red and purple clays, shales, thin coals and ironstones, is prolonged from the Fife coal-field in a tongue which extends eastward beyond the village of Lower Largo. It is well displayed on the shore, where every bed may be followed in succession along the beach for a space of nearly two miles. Two volcanic necks, presenting the same features as those which pierce the older portions of the Carboniferous system to the east, rise through these red rocks. We are thus carried not only beyond the time of the Carboniferous Limestone, but beyond the close of the very latest stage of the Carboniferous period in Central Scotland. Connected with these and other vents farther north, there is a large area of tuff which has been thrown out upon the faulted and greatly denuded Carboniferous rocks. It may be traced passing from the red Upper Coal-measures across the large fault which here separates that formation from the Carboniferous Limestone, and extending inland athwart different horizons of the latter series. Outlying fragmentary cakes of it may be seen resting on the upturned edges of the sandstones, shales and coal-seams, even at a distance of some miles towards the north-west, proving that the fragmentary materials discharged from the vents spread over a considerable area. The accompanying section (Fig. 208) may serve as an illustration of the relation between this sheet of bedded tuff and the underlying rocks.

Though interstratified volcanic rocks occur in the Carboniferous system of the East of Fife, no connection has been traced between them and any of the vents now referred to. While none of these vents can be proved to be of Carboniferous age, it is of course possible that such may be the true date of some of them. Others, nevertheless, and probably much the largest number, judged from the data just given, may be regarded as probably post-Carboniferous. Those which happen to rise through the uppermost Coal-measures do not appear to be distinguishable by any essential characters from those which pierce indifferently the Carboniferous Limestone series and Calciferous Sandstones of the East of Fife. They seem to be all one connected aggregate, resembling each other alike in their external characters, internal structure and component materials, and the limit of their age must be determined by the geological horizon of the youngest formation which they traverse. By this process of reasoning I reach the conclusion that this remarkable series of old volcanoes in the East of Scotland not improbably dates from the same time as that of Ayrshire and Nithsdale, already described.

Some idea of the importance and interest of the volcanic area of Eastern Fife may be gathered from the fact that in a space of about 70 square miles no fewer than 60 necks may be counted, and others are probably concealed below the drift-deposits which cover so much of the interior of the country. The area of this remarkable display extends from St. Andrews Bay and the Vale of the Eden southwards to the coast of the Firth of Forth between Lundin Links and St. Monans. All over the inland tract the necks form more or less marked eminences, of which the largest are conspicuous landmarks from the southern side of the Firth. But the distinguishing characteristic of the area is the display of the necks along the coast, where, in a series of natural dissections, their form, composition, internal structure and relations to the surrounding rocks have been laid open in such clearness and variety as have been met with in the volcanic records of no other geological period within the compass of these islands. As this district thus possesses a singular interest and value for the study of volcanic vents, I shall enter in some detail into the description of the sections so admirably laid bare.

As in Ayrshire, the necks in the East of Fife generally rise as isolated conical or dome-shaped hills, with smooth grassy slopes, but where a dyke or boss of basalt occurs in them, it usually stands out as a crag or knoll. Largo Law (Fig. 209) may be taken as a singularly perfect example of the cone-shaped neck. This hill, however, comprises more than one vent. The mass of tuff of which it consists probably includes at least three distinct funnels of discharge, and surrounding it there still remains a good deal of the fragmental material that gathered around these vents and is now seen to lie unconformably upon the Carboniferous formations (Fig. 208). There must be a total area of not much less than four square miles over which tuff occupies the surface of the ground.

While the Fife necks possess the great advantage of having been laid bare by the sea, their frequent small size on the coast allows their whole area to be examined. As illustrations of these little vents, two plates are here given from the coast-line to the east of St. Andrews, where a number of small necks of agglomerate have been planted among the plicated Calciferous Sandstones. In Fig. 210 the abrupt truncation of the sandstones by the volcanic rock is well shown. The strata on the right have been broken through, and the sea has indented a small gully along the wall of the old volcanic funnel. The sandstones in front, however, still adhere firmly to the agglomerate, which rises above them as a rugged mass of rock.

In Fig. 212 the edge of the vent can be traced partly in section and partly in plan for about half of its circumference. On the right hand, the actual wall of the funnel is visible where the false-bedded sandstones are sharply cut off by the agglomerate. In front the strata appear in plan on the beach, and their ledges can be seen to the left striking at the margin of the neck.

The shape of the Fife vents is, as usual, generally circular or oval; but is subject to considerable irregularity. The coast-section between Largo and St. Monans exposes many ground-plans of them, and permits their irregularities to be closely examined. The accompanying figure (Fig. 211) exhibits some characteristic forms. Eccentricities of outline no doubt arose from the irregular way in which the rocks yielded to the forces of explosion during the piercing of a volcanic orifice. This is often well shown by the veins and nests of tuff or agglomerate which have been forced into the rents or sinuosities of the orifices. In other cases, however, it is probable that, as among the Ayrshire necks, and those of Carboniferous age already cited, what appears now as one volcanic neck was the result of a shifting of the actual funnel of discharge, so that the neck really represents several closely adjacent vents. The case of Largo Law has been already noticed. The necks at Kellie Law (Fig. 213) show clearly the same structure, the Law itself (1) probably consisting of two contiguous vents, while a third (2) forms a smaller cone immediately to the east. Such a slight lateral displacement of the vent has been noticed at many Tertiary and recent volcanic orifices. In the island or peninsula of Volcanello, for example, three craters indicate successive shiftings of the vent, the most perfect of them marking the latest and diminishing phase of volcanic activity (Fig. 214, compare Fig. 29, vol. i., p. 70).

The Fife necks vary from only a few yards up to perhaps 4000 feet in diameter. One of the smallest and most completely exposed occurs on the shore at Newark Castle, near St. Monans. It measures only 60 yards in length by about 37 yards in breadth. A ground-plan of it is given in Fig 224. Still smaller is the neck at Buddo Ness, on the coast east of St. Andrews, which measures only 20 yards across.

From the way in which the vents have been dissected by the sea along the Fife coast, the geologist is enabled to study in minute detail the effects of the volcanic operations upon the strata through which the funnels have been drilled. Considerable variation may be observed in the nature and amount of change. Sometimes the orifice has been made without any noticeable alteration of the sandstones, shales and limestones, which retain their dip and strike up to the very wall of the chimney. Usually there is more or less jumbling and crushing of the stratification, and often a considerable amount of induration. As a typical example of these effects I give a section from the margin of the neck of tuff on the east side of Elie Harbour (Fig. 215). Here the sandstones and shales (a) have been doubled over and dragged down against the tuff (b). They have likewise been hardened into a kind of quartzite, and this alteration extends for about 20 to 30 feet from the edge of the neck.

The material which has filled up the vents is almost entirely fragmental, varying from a coarse agglomerate to a fine volcanic tuff. Some minor necks have been completely or in great part filled with angular debris of the ordinary rocks of the neighbourhood. In the western neck on the Largo shore, for example, which rises through the red rocks of the Upper Coal-measures, the material consists largely of fragments of red sandstone, clay and shale. Between Elie and St. Monans, some of the necks are filled almost wholly with debris of black shale and encrinal limestone.

There does not appear to be any relation between the diameter of a funnel and the size of the blocks that now fill it. Some of the larger necks, for example, consist of comparatively fine tuff. The Buddo Ness, on the other hand, though so small a vent, is packed with blocks of shale six feet long, while the sandstone through which the orifice has been drilled passes, as usual, into quartzite for several yards away from the edge. As an example of the general aspect presented by one of the coarse agglomerates in the necks of the Fife coast, a view is given in Fig. 216 of a portion of the neck at Ardross, about two miles east from Elie. This thoroughly volcanic accumulation is here shown to consist of blocks of all sizes heaped together without any definite arrangement.

Since the first stage in the history of the vents has been the perforation of the solid crust by explosion, and the consequent production of debris from the disrupted rocks, we may hope to detect underneath the pile of thoroughly volcanic ejections, traces of the first explosions. I have been much struck with the fact that in the East of Fife such traces may frequently be found here and there within the outer border of the vents. At Largo, and again between Elie and St. Monans, it may be noticed that the mass of material adhering to the wall of a neck, exposed in ground-plan upon the beach, often consists largely, or even wholly, of debris of sandstone, shale and limestone, while the central and chief mass is made up of green tuff or agglomerate, with occasional pieces of the surrounding stratified rocks scattered through it. It seems probable, therefore, that the sections of these Fife necks, laid bare on the present shore, do not lie far below the original crater-bottoms.

Some light might be expected to be thrown upon the phenomena in an active volcanic chimney by the condition of the fragments of recognizable sedimentary rocks imbedded in the ejected debris which has filled up the orifice. But the assistance from this source is neither so full nor so reliable as could be wished. In some of the Fife vents, indeed, the fragments of shale, sandstone and other sedimentary strata are so unchanged that they cannot on a fresh fracture be distinguished from the adjacent parent strata. The Spirifers, Lingulæ, crinoids, cyprid-cases, ganoid scales and other fossils are often as fresh and perfect in the fragments of rock imbedded in tuff as they are in the rock in situ. In some cases, however, distinct, and occasionally even extreme, metamorphism may be detected, varying in intensity from mere induration to the production of a crystalline texture. The amount of alteration has depended not merely upon the heat of the volcanic vent, but also in great measure upon the susceptibility of the fragments to undergo change and the duration of their exposure to it.

Dr. Heddle has computed the temperature to which fragments of shale, etc., in tuff-necks of the Fife coast have been subjected. He found that the bituminous shales have lost all their illuminants, and of organic matter have retained only some black carbonaceous particles; that the encrinal limestones have become granular and crystalline; that the sandstones present themselves as quartzite, and that black carbonaceous clays show every stage of a passage into Lydian-stone. He inferred from the slight depth to which the alteration has penetrated the larger calcareous fragments, that the heat to which they were exposed must have been but of short continuance. As the result of his experiments, he concluded that the temperature at which the fragments were finally ejected from the volcanic vents probably lay between 660° and 900° Fahr.[98]

It may be perhaps legitimate to infer that, while the fragments that fell back into the volcanic funnel, or which were detached from the sides of the vent, after having been exposed for some time to intense heat under considerable pressure, would suffer more or less metamorphism, those, on the other hand, which were discharged by the æriform explosions from the cool upper crust, on the first outburst of a vent, would not exhibit any trace of such a change. Where, therefore, we meet with a neck full of fragments of unaltered stratified rocks, we may suppose it to have been that of a short-lived volcano; where, on the other hand, the fragments are few and much altered, they may mark the site of a vent which continued longer active. The metamorphism of the fragmentary contents of a volcanic funnel by the action of ascending vapours has already been described in the case of one of the vents of the Carboniferous plateaux (vol. i. p. 404).

One of the most curious and puzzling features in the contents of the tuff necks of the Fife coast is the occurrence there of crystals and fragments of minerals, often of considerable size, which do not bear evidence of having-been formed in situ, but have undoubtedly been ejected with the other detritus. Dr. Heddle has noticed the fact, and has described some of the minerals which occur in this way. The following list comprises the species which he and I have found:—

Hornblende, in rounded fragments of a glassy black cleavable variety.
Augite, sometimes in small crystals, elsewhere in rounded fragments of an augitic glass.
Orthoclase (Sanidine), abundant in worn twin crystals in the tuffs of the east of Fife.
Plagioclase.
Biotite.
Pyrope, in the tuffs (and more rarely in the basalts) of Elie.
Nigrine, common in some of the dykes, more rarely in the tuffs of Elie.
Saponite, Delessite and other decomposition products.
Semi-opal, one specimen found in the later (Permian?) agglomerate of Arthur's Seat.
Asphalt, abundant at Kincraig, near Elie.
Fragments of wood, with structure well preserved, may be included here.

Dr. Heddle has described from the neck of tuff at Kinkell, near St. Andrews, large twin crystals of a glassy orthoclase, which are invariably much worn, and preserve only rudely the form of crystals. He justly remarks that they have no connection with drusy cavity, exfiltration vein, or with any other mineral, and look as if a portion of their substance has been dissolved away. Internally, however, they are quite fresh and brilliant in lustre, though sometimes much fissured.[99]

The tuffs at Elie are full of similar crystals. I obtained from one of the necks east of that village a specimen which measures 4 inches in length, 3¹/₂ in breadth, and 2¹/₄ in thickness, and weighs about 2 lbs. It is, however, a well-striated felspar. From the same tuff I procured an orthoclase twin in the Carlsbad form. All the felspar pieces, though fresh and brilliant internally, have the same rounded and abraded external appearance.

The fragments of hornblende form a characteristic feature in several of the Elie dykes (to be afterwards described), and in the neighbourhood of these intrusive rocks occur more sparingly in the tuff. It is a glossy-black cleavable mineral, in rounded pieces of all sizes, up to that of a small egg. Dr. Heddle obtained a cleavage angle of 124° 19', and found on analysis that the mineral was hornblende.[100]

Augite occurs sparingly in two forms among the rocks. I have obtained small crystals from the red agglomerate on the south side of Arthur Seat, recalling in their general appearance those of Somma. Lumps of an augitic glass have been found by Dr. Heddle, sometimes as large as a pigeon's egg, in two of the dykes at Elie, and in the tuff at the Kinkell neck, near St. Andrews. He observed the same substance at the Giant's Causeway, both in the basalt and scattered through one of the interstratified beds of red bole. Much larger rounded masses of a similar augitic glass, but with a distinct trace of cleavage, have already been referred to as occurring in a volcanic vent of Upper Old Red Sandstone age at John o' Groat's House.[101]

Biotite is not a rare mineral in some tuffs. It may be obtained in Lower Carboniferous tuffs of Dunbar, in plates nearly an inch broad; but the largest specimen I have obtained is one from the same Elie vent which yielded the large felspar fragment. It measures 2¹/₂ × 2 × ¹/₂ inches. These mica tables, like the other minerals, are abraded specimens.

That these various minerals were ejected as fragments, and have not been formed in situ, is the conclusion forced upon the observer who examines carefully their mode of occurrence. Some of them were carried up to the surface by liquid volcanic mud, and appear in dykes of that material like plums in a cake. But even there they present the same evidence of attrition. They assuredly have not been formed in the dykes any more than in the surrounding tuff. In both cases they are extraneous objects which have been accidentally involved in the volcanic rocks. Dr. Heddle remarks that the occurrence of the worn pieces of orthoclase in the tuff is an enigma to him. I have been as unable to frame any satisfactory explanation of it.[102]

It might have been thought that within the throat of a volcano, if in any circumstances, loose materials should have taken an indefinite amorphous aggregation. And, as has been shown in the foregoing chapters, this is usually the case where the materials are coarse and the vent small. Oblong blocks are found stuck on end, while small and large are all mixed confusedly together. But in numerous cases where the tuff is more gravelly in texture, and sometimes even where it is coarse, traces of stratification may be observed. Layers of coarse and fine material succeed each other, as they are seen to do among the ordinary interstratified tuffs. The stratification is usually at high angles of inclination, often vertical. So distinctly do the lines of deposit appear amid the confused and jumbled masses, that an observer may be tempted to explain the problem by supposing the tuff to belong not to a neck, but to an interbedded deposit which has somehow been broken up by dislocations. That the stratification, however, belongs to the original volcanic vents themselves is made exceedingly clear by some of the coast-sections in the East of Fife. On both sides of Elie, examples occur in which a distinct circular disposition of the bedding can be traced corresponding to the general form of the neck. The accompanying ground-plan (Fig. 217) represents this structure as seen in the neck which forms the headland at Elie harbour. Alternations of coarse and fine tuff with bands of coarse agglomerate, dipping at angles of 60° and upwards, may be traced round about half of the circle. The incomplete part may have been destroyed by the formation of another contiguous neck immediately to the east. To the west of Earlsferry another large, but also imperfect, circle may be traced in one of the shore necks. A quarter of a mile farther west rises the great cliff-line of Kincraig, where a large neck has been cut open into a range of precipices 200 feet high, as well as into a tide-washed platform more than half a mile long. The inward dip and high angles of the tuff are admirably laid bare along that portion of the coast-line. The section in which almost every bed can be seen, and where, therefore, there is no need for hypothetical restoration, is as shown in Fig. 218.

I have already referred to the frequently abundant pieces of stratified tuff, found as ejected blocks in vents filled with tuff, and to the derivation of these blocks from tuff originally deposited within the crater. There can, I think, be little hesitation in regarding the stratification of these Fife vents as a record of successive deposits of volcanic detritus inside the vents. The general dip inwards from the outer rim of the vent strikingly recalls that of some modern volcanoes. By way of illustration, I give here a section of part of the outer rim of the crater of the Island of Volcano, sketched by myself in the year 1870 while ascending the mountain from the north side (Fig. 220). The crater wall at this point consists of two distinct parts—an older tuff (a), which may have been in great measure cleared out of the crater before the ejection of the newer tuff (b). The latter lies on the outer slope of the cone at the usual angle of 30°. It folds over the crest of the rim, and dips down to the flat tuff-covered crater bottom, at an angle of 37°. These are its natural angles of repose.

Applying modern analogies of this kind, I have been led to conclude that the stratification so conspicuous in the tuff of the vents in the east of Fife and in the Carboniferous series of the Lothians belongs to the interior of the crater and the upper part of the volcanic funnel.[103] These stratified tuffs, on this view of their origin, must be regarded as remains of the beds of dust and stones which gathered within the crater and volcanic orifice, and which, on the cessation of volcanic action, sometimes remained in their original position, or were dislocated and slipped down into the cavity beneath. That the tuffs consolidated on slopes, perhaps quite as steep as those of Volcano, is now and then indicated by an interesting structure. The larger stones imbedded in the layers of tuff may be observed to have on their fronts in one direction a small heap of coarse gravelly debris, while fine tuff is heaped up against their opposite side. This arrangement doubtless points to deposit on a slope of loose debris, from which the larger blocks protruded so as to arrest the smaller stones, and allow the fine dust to gather behind.

If the inference be correct, that the stratification here described belongs to the old craters or the upper parts of the funnels, it furnishes additional evidence of the wide interval of time that elapsed between the deposition of the Carboniferous strata and the outbreak of these vents. During that interval prolonged denudation reduced the upturned Carboniferous Limestone series to nearly its present form of surface, and any materials discharged from the vents over the surrounding ground would obviously lie with a violent unconformability on the rocks below.

The frequent great disturbance in the bedding of the tuff within the vents may be connected with some kind of collapse, subsidence or shrinkage of the materials in the funnel below. That a movement of this nature did take place is shown by the remarkable bending down of the strata round the margins of the vents, which has been already described.

The minor vents for the most part contain only fragmentary materials; but those of larger size usually present masses of lava in some characteristic forms. In not a few cases, the lava has risen in the central pipe and has hardened there into a column of solid rock. Subsequent denudation, by removing most of the cone, has left the top of this thick column projecting as a round knoll upon the hill-top. Arthur Seat presents a good example of this structure. Where the denudation has not proceeded so far, we may still meet with a remnant of the cake of lava which sometimes overflowed the bottom of a crater. The summit of Largo Law affords indications of this arrangement, the cone of tuff being there capped with basalt, evidently the product of successive streams, which welling out irregularly covered the crater bottom with hummocks and hollows (Fig. 226). The knolls are beautifully columnar, and sometimes show a divergent arrangement of the prisms.

But the most frequent form assumed by the lava in the necks is that of veins or dykes running as wall-like bands through the tuff or agglomerate. Many admirable examples may be cited from the shore between Largo and St. Monans. Two illustrations of them are given in Figs. 219 and 221. In Fig. 219 the dyke is about four feet broad, and is seen to present the common transverse jointing as it pursues its way through the tuff. White veins of calcite along its margin serve to define its limits. Its position in reference to the general body of the neck is shown in the ground-plan Fig. 224. The second instance (Fig. 221) is that of a dyke of basalt only one foot wide, which runs like a wall up the agglomerate of the Kincraig neck near Elie. It is seen at the bottom of the cliff projecting from the agglomerate; but higher up it has decayed, leaving its fissure as a gaping chasm. Here the general character of the pyroclastic material is well brought out. One or two large blocks may be seen imbedded in it, the largest lying above where the dyke bends away to the left.

The intruded masses vary in breadth from mere threadlike veins up to dykes several yards in breadth, which sometimes expand into large irregular lumps. They generally consist of some form of basalt; now and then, as at Ruddon Point, near Elie, they are amygdaloidal; and it may be observed among them, as among dykes in general, that where the amygdaloidal texture is developed, it is apt to occur most markedly in the central part of the vein, the amygdales running there in one or more lines parallel with the general trend of the mass.

That the basalt of these veins and dykes was sometimes injected in an extremely liquid condition is shown by its frequently exceedingly close homogeneous texture. Within the neck on the shore to the west of Largo, the basalt assumes in places an almost flinty character, which here and there passes into a thin external varnish of basalt-glass. A farther indication of the liquidity of the original rock seems to be furnished by the great number of included extraneous fragments here and there to be observed in the basalt.

But besides basalt, other materials may more rarely be detected assuming the form of dykes or veins within the necks. Thus, at the Largo neck just referred to, strings of an exceedingly horny quartz-felsite accompany the basalt—a remarkable conjunction of acid and basic rock within the same volcanic chimney. To the east of Elie some dykes, which stand out prominently on the beach from a platform worn by the sea in a neck, consist of an extremely compact volcanic mudstone, stuck full of the worn twin crystals of orthoclase and pieces of hornblende and biotite already noticed. So like is this rock to one of the decomposing basalts that its true fragmental nature may easily escape notice, and it might be classed confidently as a somewhat decayed basalt. A considerable amount of a similar fine compact mudstone is to be seen round the edges of some of the Elie vents. This material must have been injected into open fissures, where it solidified. There is further evidence of the presence of "mud-lava" in some of the vents of East Fife, where these orifices contain a remarkable compact volcanic sandstone, composed of the usual detritus, but weathering into spheroidal crusts, so as externally to be readily mistaken for some form of basalt.