Fig. 222.—Radiating columnar dyke in the tuff of a volcanic vent, Rock and Spindle, two and a half miles east from St. Andrews.

A columnar arrangement may often be observed among the basalt dykes. When the vein or dyke is vertical, the columns of course seem piled in horizontal layers one above the other. The exposed side of the dyke then reveals a wall of rock, seemingly built up of hexagonal or polygonal, neatly fitting blocks of masonry, as in the Lower Carboniferous vent of the Binn of Burntisland (Figs. 166, 168). An inclination of the dyke from the vertical throws up the columns to a proportional departure from the horizontal. Sometimes a beautiful fan-shaped grouping of the prisms has taken place. Of this structure the Rock and Spindle, near St. Andrews, presents a familiar example (Fig. 222). Much more striking, however, though less known, is the magnificent basalt mass of Kincraig, to the west of Elie, where the columns sweep from summit to base of the cliff, a height of fully 150 feet, like the Orgues d'Expailly, near Le Puy in Auvergne. The general position of this basalt in the vent is represented in the section (B, Fig. 218). The curvature of the basalt is shown in Fig. 223, which is taken from the Elie side looking westward, beyond the intrusions, to the picturesque cliffs of tuff. The details of the cliff are given in Fig. 225.

That many of the dykes served as lines of escape for the basalt to the outer slopes of the cones is highly probable, though denudation has usually destroyed the proofs of such an outflow. A distinct radiation of the dykes from the centre of a neck is still sometimes traceable. This structure is most marked on the south cone of Largo Law, where a number of hard ribs of basalt project from the slopes of the hill. Their general trend is such that if prolonged they would meet somewhere in the centre of the cone. On the south-east side of the hill a minor eminence, termed the Craig Rock, stands out prominently (Fig. 209). It is oblong in shape, and, like the dykes, points towards the centre of the cone. It consists of a compact columnar basalt, the columns converging from the sides towards the top of the ridge. It looks like the fragment of a lava-current which flowed down a gully on the outer slope of the cone (B' in Fig. 226).

Veins of basalt are not confined to the necks, but may be seen running across the surrounding rocks. The shore at St. Monans furnishes some instructive examples of this character. As the veins thin away from the main mass of basalt they become more close-grained and lighter in colour, and when they enter dark shales or other carbonaceous rocks they pass, as usual, into the white earthy clay-like "white-trap." The influence of carbonaceous strata in thus altering basic dykes and sills may be instructively studied along the shore of the East of Fife. A good instance occurs near St. Monans Church (Fig. 227), where a vein of "white-trap" traverses black shales which have been somewhat jumbled.

Fig. 223.—View of part of the shore front of the great vent at Kincraig, looking westward, with the columnar basalt in front.

In a modern volcano no opportunity is afforded of examining the contact of the erupted material with the rocks through which the vent has been opened. But in the basin of the Firth of Forth, within the area now under description, a numerous series of coast-sections lays bare this relation in the most satisfactory manner. The superincumbent cones of tuff have been swept away, and we can examine, as it were, the very roots of the old volcanoes. The margin of a neck or volcanic vent is thus found to be almost always sharply defined. The rocks through which the funnel has been drilled have been cut across, as if a huge auger had been sunk through them. This is well displayed in the beautifully perfect neck already cited at Newark Castle, near St. Monans (Fig. 224). The strata through which this neck rises consist of shales, sandstones, thin coal and encrinal limestones, dipping in a westerly direction at angles ranging from 25° to 60°. At the south end of the neck they are sharply truncated, as if by a fault. Elsewhere they are much jumbled, slender vein-like portions of the tuff being insinuated among the projecting strata. A large vertical bed of sandstone, 24 yards long by 7 yards broad, stands up as a sinuous reef on the east side of the vent (s). It is a portion of some of the surrounding strata, but, so far as can be seen at the surface, is entirely surrounded with agglomerate. Here and there the shales have been excessively crumpled, and at the north end have been invaded by a vein of basalt which, where it runs through them, assumes the usual clay-like character. The strata have been blown out, and their place has been occupied by a corresponding mass of volcanic agglomerate. But their remaining truncated edges round the margin of the orifice have undergone comparatively little alteration. In some places they have been hardened, but their usual texture and structure remain unaffected.

Fig. 224.—Plan of volcanic neck on beach near St. Monans.

T, Neck of tuff enclosing a mass of sandstone (s), and piercing sandstones and shales With beds of limestone, (l l), and a thin seam of coal (c); B, Basalt "white-trap" dyke. The arrows show the dip of the strata.

In a few examples, the progress of denudation has not advanced so far that the cone cannot still be partially made out amidst its surrounding masses of tuff. One of the most interesting of these is Largo Law, of which an outline has been given in Fig. 209. The accompanying section (Fig. 226) represents what appears to me to be the structure of this hill. Each of the two now conjoined cones was probably in succession the vent of the volcano. The southern and rather lower eminence, as already mentioned, is traversed by rib-like dykes of basalt, which point towards its top, where there is a bed of the same rock underlying a capping of tuff. On its eastern declivity lies the basalt stream already described (p. 87). The higher cone is surmounted by a cake of basalt which, as I have above suggested, may have solidified at the bottom of the latest crater. Of course all trace of the crater has disappeared, but the general conical form of the volcanic mass remains. Doubtless, still more of the old volcano would have been removed by denudation but for the protection afforded to the tuff by the intrusion of the basalt. The upper dotted lines in the figure are inserted merely to indicate hypothetically how the cone may originally have stood. On the west side the sheets of tuff which were thrown out over the surrounding country have been almost entirely removed, but on the east and south they still cover an extensive area. (See Fig. 208).

Fig. 225.—Columnar basalt in the neck of Kincraig, Elie, seen from the west.

Fig. 226.—Section across Largo Law.

l l, Lower Carboniferous strata; t, tuff of cones; t', tuff of plain beyond the cones; B B, basalt ascending vents and sending out veins: B', basalt which has probably flowed out at the surface. The dotted lines are suggestive of the original outline of the hill.

(2) Sills.—In the Clyde coal-field and in the basin of the Firth of Forth, among the vast number of sills which there traverse the Carboniferous formations, it is possible that some belong to the Permian volcanic period (see vol. i. p. 474). Where the sheets have been intruded along horizons that lie below the upper stratigraphical limit of the puy eruptions, they may not unnaturally be held to belong to these manifestations of volcanic energy, though it is obviously quite conceivable that some of them may be of much later date. But where they lie above the highest platforms of Carboniferous lavas and tuffs, they may be assigned to a younger volcanic period. We know as yet of only two such periods after the deposition of the Carboniferous Limestone series in Scotland—Permian and older Tertiary. Unless, therefore, these higher sills formed part of some other display of subterranean activity which is not known to have culminated in eruptions at the surface, they must be looked upon as probably either Permian or Tertiary.

In the great coal-field of Stirlingshire and Lanarkshire, among the large sills that break into the Millstone Grit and the Coal-measures, one lies entirely in the Coal-measures, and covers about six square miles of ground, stretching from near Caldercruix Station, a little east of Airdrie, to near Kirk of Shotts, a distance of about four miles. A group of smaller sheets, possibly connected with the larger mass, runs for four miles further west to beyond New Monkland. Another chain of sills, which may also be part of the same great intrusion, extends from the Cant Hills, near the Kirk of Shotts, for more than eight miles in a north-easterly direction. The largest mass in this chain stretches from Blackridge, west of Bathgate, for upwards of three miles, covering an area of about three square miles and terminating on the north at the line of dislocation which has been followed by one of the east and west dykes. Another large sill, which appears nearly two miles further east on the north side of that dyke, lies on a lower stratigraphical horizon, for it cuts the Carboniferous Limestone series, and does not reach the top of the Millstone Grit. This sill is cut through by two of the later dykes.

Fig. 227.—Vein of "white-trap" cutting black carbonaceous shales, a little west from St. Monans Church.

That these great intrusions took place later than the deposition of the Coal-measures is obvious. There is no satisfactory evidence to enable us to decide to which of the two post-Carboniferous volcanic periods they may with most probability be assigned. As one of them is distinctly cut by dykes that have been referred to the Tertiary series, it might be plausibly argued that it at least is of pre-Tertiary date, and therefore probably Permian. On the other hand, as will be shown in a later chapter, some portion of the sills appears to be connected with the younger or Tertiary dykes. This problem must for the present remain unsolved.

In Ayrshire where, as already described, basic sills traverse the Permian volcanic series, other large intrusive sheets are found around the Permian basin. On the north side an important group of them, passing through the Coal-measures into the Carboniferous Limestone series, runs from Troon eastward for more than eight miles to beyond Craigie. On the south side a much more extensive series may be traced from the River Ayr southwards into the Dalmellington coal-field, and thence north-eastwards in a wide semicircular sweep into the coal-field of New Cumnock and Airds Moss. That some of these sills proceed from the Permian necks has been definitely ascertained, and this fact has been already alluded to in connection with the vents. I have little doubt that the great majority, if not the whole, of these intrusive sheets are to be referred to the Permian period.

Some of the sills must be later than a part of the Permian volcanic eruptions, for they are found in at least three places intercalated in the zone of lavas and tuffs. But no instance has been observed of their traversing the basin of Permian sandstone which overlies that zone, though a few dykes, possibly of Tertiary age, do cut this sandstone.