CHAPTER XXVII. Geological Structure of the Carboniferous Puys of Scotland

GEOLOGICAL STRUCTURE OF THE CARBONIFEROUS PUYS OF SCOTLAND

1. Vents: Relation of the Necks to the Rocks through which they rise--Evidence of the probable Subærial Character of some of the Cones or Puys of Tuff--Entombment of the Volcanic Cones and their Relation to the Superficial Ejections. 2. Bedded Tuffs and Lavas--Effects of Subsequent Dislocations. 3. Sills, Bosses and Dykes.

The puy-type of volcanic hill differs widely in one respect from those which we have hitherto been considering. In the earlier epochs of volcanism within the British area, it is the masses of material discharged from the vent, rather than the vents themselves which arrest attention. Indeed, so copiously have these masses been erupted that the vents are often buried, or their positions have been rendered doubtful, by the uprise in and around them of sills and bosses of molten rock. But among the Carboniferous puys the vent is often the only record that remains of the volcanic activity. In some cases we know that it never ejected any igneous material to the surface. In others, though it may be filled with volcanic agglomerate or tuff, there is no record of any shower of such detritus having been discharged from it. In yet a third class of examples, we see that lava rose in the vent, but no evidence remains as to whether or not it ever flowed out above ground. Other cases occur where beds of lava or of tuff, or of both together, have been intercalated in a group of strata, but with no trace now visible of the vent from which they came. The most complete chronicle, preserving at once a record of the outflow of lava, of the showering forth of ashes and bombs, and of the necks that mark the vents of eruption, is only to be found in some of the districts.

I shall therefore, in the present instance, reverse the order of arrangement followed in the previous chapters, and treat first of the vents, then of the materials emitted from them, and lastly of the sills and dykes.

i. VENTS

A large number of vents rise through the Carboniferous rocks of Scotland. Some of these are not associated with any interbedded volcanic material, so that their geological age cannot be more precisely defined than by saying that they must be later than the particular formations which they pierce. Some of them, as I shall endeavour to show, are in all probability of Permian age. But many, from their position with reference to the nearest intercalated lavas and tuffs, are to be regarded as almost certainly belonging to the Carboniferous period. Those which are immediately surrounded by sheets of lava and tuff, similar in character to the materials in the vents themselves, may without hesitation be connected with these sheets as marking the orifices of discharge.

The vents of the puys are in general much less than those of the plateaux. Their smallest examples measure only a few yards in diameter, their largest seldom much exceed half a mile.

The dislocations of the Carboniferous system are probably on the whole later than its volcanic phenomena. It is at least certain that the lavas and tuffs of the puys have been extensively faulted, like the surrounding sedimentary strata, and the vents seldom show any apparent relation to faults. It may sometimes be observed, however, that the vents are arranged in lines suggestive of fissures underneath. A remarkable instance of the linear distribution is furnished by the chain of necks which extends from the Vale of the Tweed at Melrose south-westwards across the watershed and down Liddesdale. The most notable part of this line lies among the uplands to the east of the Old Mosspaul Inn at the head of the Ewes Water. A string of masses of agglomerate has there solidified in a fissure among the Silurian greywackes and shales, running in a north-easterly direction for several miles. The largest connected mass of agglomerate is 4700 feet long, and from 350 to 600 feet broad (see No. 1 in Fig. 22). That this curious vent, or connected line of vents along a great fissure, belongs to the puy-eruptions of Liddesdale is shown by the abundant fragments of yellow sandstone and cement-stone which occur in the agglomerate.

Most frequently the vents are distributed irregularly in groups. As examples of this arrangement I may cite those of the west of Fife, of West Lothian and of the north of Ayrshire.

A convenient classification of the vents may be made by dividing them into four groups according to the nature of the material that now fills them: 1st, Necks of non-volcanic debris; 2nd, Necks of tuff and agglomerate; 3rd, Necks of similar materials, but with a central plug of basalt; 4th, Bosses of basalt or other lava, without agglomerate or tuff.

1. Necks of Non-volcanic Debris.--In a few instances the orifices of eruption have been filled up entirely with non-volcanic debris. They have served as funnels for the discharge of explosive vapours only, without the expulsion of any solid volcanic materials. At least no trace of fragmentary lavas is met with in them, nor are any beds of tuff or lava intercalated among the surrounding strata. Some interesting examples of this kind were laid bare in the open ironstone-workings near Carluke in Lanarkshire. They were circular in ground-plan, descended vertically into the strata, and were somewhat wider at the top of the quarry than at the bottom. They were filled with angular pieces of Carboniferous sandstone, shale, limestone, ironstone and other rocks, these materials being rudely arranged with a dip towards the centre of the neck, where the blocks were largest in size. Though no fragments of igneous rocks were observed among the debris, a few string-like veins of "white trap," or altered basalt, were seen to traverse the agglomerate here and there. The necks and the strata surrounding them were highly impregnated with pyrites and sulphate of lime.

1. Three feet coal; 2. Ontake coal; 3. Upper and Lower Black-band Ironstones; 4. Index Limestone; 5. Gas Coal and Janet Peat Coal; 6. Calmy Limestone; 7. Neck.]

A vent of the same nature, but on a much larger scale, has been mapped by Mr. Peach in the south of Perthshire, near East Grange, where it rises through the higher coal-bearing part of the Carboniferous Limestone series (Fig. 143). It has been encountered in the mining of coal and ironstone, and its cross-section has been ascertained in the underground workings which have been carried up to its margin. It measures 1500 feet in diameter from east to west and 2000 feet from north to south. It does not appear ever to have emitted any ashes or lava. Mr. Peach found it filled with dark sandy crumbling clays, full of fragments of sandstone, shale and coal. These sediments are arranged in layers that dip in the same general direction as the strata surrounding the vent. They contain abundant calcareous nodules of all sizes from that of a hazel-nut up to concretions 18 feet in diameter. The clays likewise include many of the common shells and crinoids of the Carboniferous Limestone sea, and the same fossils are enclosed in the nodules. A remarkable feature in this vent is the occurrence of abundant vertical rents, which have been filled partly with the same material that forms the nodules, and partly with sandstone.

The formation of the neck took place after the deposition of the Index Limestone, and probably about the time of the accumulation of the next limestone, which lies immediately to the west somewhat higher in the series. It would appear that the eruption which produced this funnel gave forth only gaseous explosions, and occurred on the sea-floor; that the low crater-walls were washed down to such an extent that the sea entered and carried some of its characteristic organisms into the lagoon or maar within; further, that as the silt gathered inside, successive subsidences occurred, whereby the sediment was rent by cracks into which sand and calcareous mud were washed from above.

Many necks occur wherein non-volcanic materials, though not forming the whole of the agglomerate, make up by far the larger part, with only a slight admixture of volcanic tuff between them. Among the Burntisland necks of Fife, for instance, abundant fragments of the well-marked cyprid limestone and shale may be observed, while at Niddry in Linlithgowshire blocks several yards in length, and consisting of different layers of shale and cement-stone still adhering to each other, may be seen imbedded at all angles in the tuff.

Where only the debris of non-volcanic rocks occupies a vent, we may infer that the volcanic action was limited to the explosion of steam, whereby the rocks were dislocated, and an orifice communicating with the surface was drilled through them, and that while no true volcanic rock in such a case appeared, the pipe was filled up to perhaps not far from the surface by the falling back of the shattered detritus. A little greater intensity or farther prolongation of the volcanic action would bring the column of lava up the funnel, and allow its upper part to be blown out as dust and lapilli; while still more vigorous activity would be marked by the rise of the lava into rents of the cone or its actual outflow at the surface. Every gradation in this scale of progress may be detected among the Carboniferous volcanoes of the basin of the Firth of Forth.

2. Necks of Tuff and Agglomerate.--The majority of the necks connected with the puys consist of tuff or agglomerate. Externally they generally appear as smooth rounded grassy hills that rise disconnected from other eminences. In some districts their materials consist of a greenish granular often stratified tuff, enclosing rounded balls of various basic lavas and pieces of sandstone, shale, limestone or other strata through which they have been drilled. This is their usual character in the Forth region. But in some cases, the tuff becomes a coarse agglomerate, made up partly of large blocks of basalt and other volcanic rocks and partly of the sedimentary strata around them, of which large masses, many cubic yards in bulk, may be seen. Among the enclosed fragments it is not unusual to find pieces of older stratified tuff. These resemble in general petrographical character parts of the tuff among which they are imbedded. Sometimes they have been

As a good illustration of the variety and relative proportions of the ejected blocks in the green tuff of the puy-vents, I may cite the following table of percentages which I took many years ago in the tuff which rises through the Cement-stone group on the beach at the Heads of Ayr.

Diabase and basalt 57 per cent. Older tuff 3 " Andesite (probably from Old Red Sandstone volcanic series of the neighbourhood) 14 " Limestone (cement-stone, etc.) 20 " Shale 3 " Sandstone 2 " Fossil wood 1 " ---- 100

While many examples might be cited where no molten rock of any kind has risen in the vents, or where at least all the visible materials are of a fragmentary character, yet small veins and dykes of basalt have not infrequently been injected into the tuff or agglomerate. These seldom run far, and usually present a more or less tortuous course. Thus, on the south front of the Binn of Burntisland (Figs. 166, 168) a number of basalt-dykes, which vary in breadth from five or six feet to scarcely so many inches, bifurcate and rapidly disappear in the tuff, one of them ascending tortuously to near the top of the cliff. They at once recall the appearance of the well-known dykes in the great crater wall of Somma.

Though not by any means the largest or most perfect of the vents in the basin of the Firth of Forth, the Binn of Burntisland, of which a view is given in Fig. 144, may be cited in illustration of their general characters. It presents in detail some of the most strikingly volcanic aspects of scenery anywhere to be seen in that region. Consisting of a dull green granular volcanic tuff, it rises abruptly out of the Lower Carboniferous formations to a height of 631 feet above the sea. The southern edge of this neck has been so extensively denuded, that it presents steep craggy slopes and rugged precipices, which descend from the very summit of the cone to the plain below--a vertical distance of nearly 500 feet. Here and there the action of atmospheric waste has hollowed out huge crater-like chasms in the crumbling tuff. Standing in one of these, the geologist can realize what must have been the aspect of the interior of these ancient Carboniferous volcanic cones. The scene at once reminds him of the crater-walls of a modern or not long extinct volcano. The dull-green rudely stratified tuff rises around in verdureless crumbling sheets of naked rock, roughened by the innumerable blocks of lava, which form so conspicuous an element in the composition of the mass. The ribs or veins of columnar basalt run up the declivities as black shattered walls. The frosts and rains of many centuries have restored to the tuff its original loose gravelly character. It disintegrates rapidly, and rolls down the slopes in long grey lines of volcanic sand, precisely as it no doubt did at the time of its ejection, when it fell on the outer and inner declivities of the original cone. Some of these features may be partly realized from Fig. 145, which represents a portion of the south front of the hill. Sections of this neck are given in Figs. 149 and 159.

(3) Necks of Tuff or Agglomerate with a Central Plug of Basalt or other Lava.--It has often happened that, after the explosions in a vent have begun to decrease in vigour, or have at last ceased, lava has risen in the chimney and finally sealed it up. In such cases the main mass of the rock may consist of tuff or agglomerate, which the enfeebled volcanic activity has been unable to expel from the orifice, while a plug of basalt, dolerite, or even more basic material, of much smaller dimensions, may have risen up the pipe in the centre or towards one side. Binns Hill, West Lothian, the Beath and Saline Hills of Fife, and Tinnis Hill in Liddesdale are good examples of this structure. (See Figs. 26, 148, 149 and 174).

(4) Necks of Basalt, Dolerite, etc.--In other cases no fragmental material is present in the vent, or possibly traces of it may be seen here and there adhering to the walls of the funnel, the prevailing rock being some form of lava. Necks of this kind are much less frequent in the puy- than in the plateau-type. But examples may be found in several districts. The most striking with which I am acquainted are those which form so picturesque a group of isolated cones around the volcanic basin of Limerick, to be afterwards described (Figs. 195, 196). The vents there have been filled by the uprise of much more acid rocks than the lavas of the basin, for, as I have already stated, they include even quartziferous trachytes. In the basin of the Firth of Forth some prominent bosses of basalt probably mark the sites of former vents, such as Dunearn Hill in Fife, the Castle Rock of Edinburgh, and Galabraes Hill near Bathgate. Some striking vents which occur in the Jedburgh district, in the debateable land between the plateau series on the east and the puy-series on the west, show the nearly complete usurpation of the funnel by basalt, but with portions of the tuff still remaining visible.

Relation of the Necks to the Rocks through which they rise.--A remarkable feature among the Carboniferous and Permian vents of central Scotland is presented by the effect which has been produced on the strata immediately surrounding them. In the interior of the country this effect is often

I have not been able to find an altogether satisfactory explanation of this inward dip of the strata around vents. Taking it in connection with the metamorphism, I am inclined to believe that it arose after the close of the long-continued volcanic action which had hardened the rocks around the volcanic pipe, and as the result of some kind of subsidence within the vent. The outpouring of so much tuff and lava as escaped from many of the volcanoes would doubtless often be apt to produce cavities underneath them, and on the decay of volcanic energy there might be a tendency in the solid or cavernous column filling up the funnel, to settle down by mere gravitation. So firmly, however, did much of it cohere to the sides of the pipe, that if it sank at all, it could hardly fail to drag down a portion of these sides. So general is this evidence of downward movement in all the volcanic districts of Scotland where the necks have been adequately exposed, that the structure may be regarded as normal to these volcanic vents. It has been observed among the shore-sections of the volcanoes of the Auckland district, New Zealand. Mr. C. Heaphy, in an interesting paper upon that district, gives a drawing of a crater and lava-stream abutting on the edge of a cliff where the strata bend down towards the point of eruption, as in the numerous cases in Scotland.

Evidence for the probable subærial Character of some of the Cones or Puys of Tuff.--From the stratigraphical data furnished by the basin of the Firth of Forth, it is certain that this region, during a great part of the Carboniferous period, existed as a wide shallow lagoon, sometimes overspread with sea-water deep enough to allow of the growth of corals, crinoids, and brachiopods; at other times, shoaled to such an extent with sand and mud as to be covered with wide jungles of a lepidodendroid and calamitoid vegetation. As volcanic action went on interruptedly during a vast section of that period, the vents, though generally submarine, may occasionally have been subærial. Indeed, we may suppose that the same vent might begin as a subaqueous orifice and continue to eject volcanic materials, until, as these rose above the level of the water, the vent became subærial. An instance of a submarine vent has been cited from the Perthshire coal-field (p. 426).

Among the evidence which may be collected to show that some Carboniferous volcanoes probably rose as insular cones of tuff above the surrounding waters, the structure of the tuff in many necks may be cited, for it suggests subærial rather than subaqueous stratification. The way in which the stones, large and small, are grouped together in lenticular seams may be paralleled on the slopes of many a modern volcano. Another indication of this mode of origin is supplied by the traces of wood to be met with in some of the tuff-necks. The vents of Fife and Linlithgowshire contain these traces sometimes in great abundance. The specimens are always angular fragments, and are frequently encrusted with calcite. Sometimes they present the glossy fracture and clear ligneous structure shown by sticks of well-made wood charcoal. In a neck at St. Magdalen's, near Linlithgow, the wood fragments occur as numerous black chips. So far as can be ascertained from the slices already prepared for the microscope, the wood is always coniferous. These woody fragments seldom occur in the interstratified tuffs or in the associated strata where Stigmaria, Lepidodendron, etc., are common. They are specially characteristic of the necks and adjacent tuffs. The parent trees may have grown on the volcanic cones, which as dry insular spots would support a different vegetation from the club-mosses and reeds of the surrounding swamps. As the fragments occur in the tuffs which, on the grounds already stated, may be held to have been deposited within the crater, they seem to point to intervals of volcanic quiescence, when the dormant or extinct craters were filled with a terrestrial flora, as Vesuvius was between the years 1500 and 1631, when no eruptions took place. Some of the cones, such as Saline Hill and the Binn of Burntisland, may have risen several hundred feet above the water. Clothed with dark pine woods, they would form a notable feature in the otherwise monotonous scenery of central Scotland during the Carboniferous period.

Entombment of the Volcanic Cones and their relation to the bedded Lavas and Tuffs.--From the facts above detailed, it is evident that in most cases the necks represent, as it were, the mere denuded stumps of the volcanoes. As the puys took their rise in areas which, on the whole, were undergoing a movement of subsidence, they were eventually submerged and buried under sedimentary accumulations. Their loose ashes would be apt to be washed down and strewn over the sea-bottom, so that only the lower and inner part of a cone might remain. We can hardly hope to discover any of the actual craters among these volcanic relics. The cones having been submerged and buried under many hundred feet of sediment, their present position at the surface is due to subsequent elevation and prolonged denudation. It is obvious that there must still be many buried cones which the progress of denudation has not yet reached. Some of these have been revealed in the course of mining operations. Valuable seams of coal, ironstone and oil-shale in the Scottish Carboniferous Limestone and Calciferous Sandstone series are extensively worked, and in the underground operations many illustrations of former volcanic action have been met with. The most remarkable instances of the discovery of buried volcanoes have occurred in the Dalry coal-field in the north of Ayrshire. In one pit-shaft about a mile and a half to the south-west of the village of Dalry, a thickness of 115 fathoms of tuff was passed through, and in another pit 90 fathoms of similar tuff were sunk into before the position of the black-band ironstone of that mineral field was reached by driving levels through the tuff into the sedimentary strata outside of it. Only a short distance from these thick piles of tuff, their place is entirely taken up by the ordinary sedimentary strata of the district. The working-plans of the mines show the tuff to occur in irregular patches and strips, between which the ironstone is workable. From these data we perceive that the shafts have in some cases been sunk directly upon the tops of puys of tuff, which were, in one case, nearly 700 feet, and in another instance, 540 high (Fig. 146).

Fig. 146.--Diagram of buried volcanic cone near Dalry, Ayrshire. Constructed from information obtained in mining operations.

1. Hurlet Limestone. 2. Clayband Ironstone. 3. Black-band Ironstone. 4. Borestone Coal. 5. Wee Coal. 6. Highfield Limestone. 7 and 8. Thin Limestones. 9. Linn Limestone. 10. Volcanic neck and cone of tuff. ]

It is obvious that from the condition of a completely buried and concealed cone every stage may be expected to occur up to the deeply worn-down neck representing merely the stump of the volcanic column. The subjoined diagram (Fig. 147) may serve to illustrate this process of gradual re-emergence.

1, Neck, still buried under the succeeding sedimentary accumulations; 2, Neck uncovered and denuded.]

When, in the progress of denudation, a volcanic cone began to show itself from under the cover of removed strata, it would still for a time maintain its connection with the sheets of tuff or of lava which, when active, it had erupted. A number of examples of this structure may be observed in the basin of the Firth of Forth, where the degradation of the surface has not yet proceeded so far as to isolate the column of agglomerate or tuff from the sheets of tuff that were strewn around the old volcano. In such cases, the actual limits of the vent are still more or less concealed, or at least no sharp line can be drawn between the vent and its ejections. As an illustration of this connection of a volcanic pipe with the materials ejected from it over the surrounding country I would cite Saline Hill in the west of Fife. That eminence rises to a height of 1178 feet above the sea, out of a band of tuff which can be traced across the country for fully three miles. Numerous sections in the water-courses show that this tuff is regularly interbedded in the Carboniferous Limestone series, so that the relative geological date of its eruption can be precisely fixed. On the south of Saline Hill, coal and ironstone, worked under the tuff, prove that this portion of the mass belongs to the general sheet of loose ashes and dust, extending outwards from the original cone over the floor of the sheet of water in which the Carboniferous Limestone series of strata was being deposited. But the central portion of the hill is occupied by one or more volcanic pipes. A section across the eminence from north-west to south-east would probably show the structure represented in Fig. 148. Immediately to the east of the Saline Hill lies another eminence, known as the Knock Hill, which marks the site of another eruptive vent. A coal-seam (the Little Parrot or Gas Coal) is worked along its southern base, and is found to plunge down steeply towards the volcanic rocks. This seam, however, is not the same as that worked under the Saline Hill, but lies some 600 feet below it. Probably the whole of the Knock Hill occupies the place of a former vent.

The thick parallel black lines mark the position of seams of coal and ironstone, some of which are worked under Saline Hill. T, Tuff of the necks; t, Tuff at a little distance from the cone, interstratified with the ordinary sedimentary beds; B, Basalt. The larger eminence is Saline Hill, the lower is Knock Hill. ]

A further stage of decay and denudation brings before us the entire severance of the volcanic column from the materials that were ejected from it. An excellent example of this isolation of the neck in the midst of surrounding masses of tuff and lava which proceeded from it is presented by the Binn of Burntisland, to which I have already alluded. A section across that eminence gives the geological structure represented in Fig. 149. The dip of the rocks away from the volcanic pipe at this locality has been produced long after the volcanic phenomena had ceased. The arch here shown is really the prolongation and final disappearance of the great anticlinal fold of which the Pentland Hills form the axis on the opposite side of the Firth. But if we restore the rocks to a horizontal, or approximately horizontal position, we find the Binn of Burntisland rising among them in one or more necks, which doubtless mark centres of volcanic activity in that district. A series of smaller neck-like eminences runs for two miles westward.

Striking as the forms of many of the necks are, and much as their present conical forms resemble those of active and extinct volcanoes, the evidence of extensive denudation proves that these contours are not the original outlines of the Carboniferous vents, but are in every case the result of prolonged waste. What we now see is a section of the volcanic chimney, and the conical form is due to the way in which the materials filling the chimney have yielded to the forces of denudation.

1, Sandstones; 2, Limestone (Burdiehouse); 3, Shales, etc.; b, b, Interstratified basalts; t t, Bedded tuff, etc.; T, Tuff of the great neck of Burntisland; B, Basalt veins.]

ii. BEDDED TUFFS AND LAVAS

During at least the earlier part of the period of the puys, in some districts or from certain vents, such as those of East Fife, Western Midlothian, Eastern Linlithgowshire, Northern Ayrshire, Heads of Ayr and Lower Eskdale, only fine tuff seems to have been thrown out, which we now find intercalated among the surrounding strata. These eruptions, neither so vigorous nor so long-continued as those of the plateaux, never gave forth such thick and widespread sheets of fragmentary materials as those associated with the plateaux in East Lothian and the north-east of Ayrshire. A single discharge of ashes seems in many cases to have been the sole achievement of one of those little volcanoes; at least only one thin band of tuff may be discoverable to mark its activity.

The tuff of these solitary bands is seldom coarse in texture. It usually consists of the ordinary dull green paste, with dust and lapilli of basic pumice. The local variations in the tuffs of the puys generally arise mainly from differences in the composition, size and numbers of the included ejected blocks. Generally the most abundant stones are pieces of different diabases, or basalts; then come fragments from the surrounding Carboniferous strata, from older tuffs and rarely from rocks of much deeper-seated origin.

Now and then the eruptions of tuff have consisted of extremely fine volcanic dust, which, mingling with water, took the form of a compact mudstone, as in the case of the Houston Marls (p. 423), which remind one of a volcanic mud. But in most localities the discharge of tuff, though for a time it may have completely obscured the ordinary contemporaneous sedimentation, was intermittent, so that in the intervals between successive showers of detritus, the deposition of non-volcanic sediment went on as usual. Hence it is that bands of tuff, whether they lie among lavas or among sedimentary formations, are apt to contain interstratifications of sandstone, shale, limestone or other detrital deposit, and to pass insensibly into these. The extremely gentle gradation from volcanic into non-volcanic sediment, and the occasional reappearance of thin partings of tuff bring vividly before the mind the slow dying out of volcanic energy among the Carboniferous lagoons.

Fig. 150.--Section in old quarry, west of Wester Ochiltree, Linlithgowshire. Calciferous Sandstone series. ]

The comparatively quiet character of the volcanic explosions, and the contemporaneous undisturbed deposition of sediment during the earlier part of the puy period, are exemplified in many sections throughout the areas above enumerated, as will be more fully illustrated in subsequent pages. Two typical examples may suffice for this general statement of the characters of the discharges of tuff in the puy-eruptions. In the Linlithgowshire quarry represented in Fig. 150, where about ten feet of strata have been exposed, a black shale (1) of the usual carbonaceous character, so common in the Oil-shale series of this region, may be seen at the bottom of the section. It is covered by a bed of nodular bluish-grey tuff (2) containing black shale fragments. A second black shale (3) is succeeded by a second thin band of fine pale yellowish tuff (4). Black shale (5) again supervenes, containing rounded fragments of tuff, perhaps ejected lapilli, and passing up into a layer of tuff (6). It is evident that we have here a continuous deposit of black shale which was three times interrupted by showers of volcanic dust and stones. At the close of the third interruption, the deposition of the shale was renewed and continued, with sufficient slowness to permit of the segregation of thin seams and nodules of clay ironstone round the decomposing organic remains of the muddy bottom (7). A fourth volcanic interlude now took place, and the floor of the water was once more covered with tuff (8). But the old conditions of deposit were immediately afterwards resumed (9); the muddy bottom was abundantly peopled with ostracod crustaceans, while many fishes, whose coprolites have been left in the mud, haunted the locality. At last, however, a much more serious volcanic explosion took place. A coarse agglomeratic tuff (10), with blocks sometimes nearly three feet in diameter, was then thrown out, and overspread the lagoon.

The second illustration may be taken from the admirable coast-section between Burntisland and Kinghorn, where the number of intercalations of tuff is very great. Besides thicker well-marked bands, successive innumerable thin layers occur there among the associated zones of sedimentary strata which separate the sheets of basalt. The character of these tuff-seams may be inferred from the following details of less than two feet of rock at Pettycur Point:--

Tuff 1·5 inch. Limestone 0·2 " Tuff 0·5 " Shale 0·2 " Tuff 0·1 " Shale and tuff 1·0 " Shale 0·2 " Limestone 0·5 " Shale full of volcanic dust 3·5 " Shaly limestone 1·5 " Laminated tufaceous limestone 2·0 " Limestone in thin bands, with thin laminæ of tuff 0·8 " Granular tuff 0·6 " Argillaceous limestone, with diffused tuff 0·9 " Fine granular tuff 0·7 " Argillaceous limestone, with diffused tuff 1·5 " Laminated limestone 0·1 " Limestone, with parting of granular tuff in middle 0·9 " Tufaceous shale 2·0 " Limestone 0·4 " Shaly tuff 1·25 " Laminated limestone 0·1 " Tuff 1·2 " ----- 21·65 inches.

1. Brown shaly fire-clay with rootlets, about five inches; 2. Impure coal, five or six inches, pressed down in its upper layers by the impact and weight of the stone; 3. Green crumbling ashy fire-clay, one foot, with its lower layers pressed down by the stone while the upper layers rise over it, showing that the stone fell at the time when half this seam was deposited. The fire-clay passes up into dark greenish and black ashy shale (4) about six inches thick and containing plant-remains. The stone is a pale diabase weighing about six or eight pounds. ]

Such a section as this brings vividly before the mind a long-continued intermittent feeble volcanic action during pauses between successive outbursts of lava. In such intervals of quiescence, the ordinary sediment of the lagoons accumulated, and was mixed up with the debris supplied by occasional showers of volcanic dust. In this Fife volcanic series, thin layers of sandstone, streaked with remains of the Carboniferous vegetation; beds of shale full of cyprid-cases, ganoid scales, and fragmentary ferns; thin beds of limestone, and bands of fire-clay supporting seams of coal, are interleaved with strata of tuff and sheets of basalt. Now and then a sharp discharge of larger stones is seen to have taken place, as in the case of the block many years ago described by me as having fallen and crushed down a still soft bed of coal (Fig. 151).

The Fife coast-section from which these details are taken supplies almost endless instances of the varying characters of the pyroclastic materials of the puy-eruptions. The very same cliff, bank or reef will show at one point an accumulation of excessively coarse volcanic debris and at another thin laminæ of the finest dust and lapilli. These rapid gradations are illustrated in Fig. 152, which is taken from the east end of the Kingswood Craig. The lower part of the declivity is a coarse agglomerate which passes upward into finer tuff.

Besides the thin partings and thicker layers of tuff which, intercalated among the sedimentary strata of the Carboniferous system, mark a comparatively feeble and intermittent volcanic activity, we meet in some localities with examples where the puys have piled up much thicker accumulations of fragmentary material without any intercalated streams of lava, or interstratified sandstone, shale or limestone. Thus the widespread Houston marls above described reach a thickness of some 200 feet. The vents of the Saline Hills in Fife covered the sea-floor with volcanic ashes to a depth of several hundred feet. In the north of Ayrshire the first eruptions of the puys have formed a continuous band of fine tuff traceable for some 15 miles, and in places at least 200 feet thick.

Where volcanic energy reached its highest intensity during the time of the puys, not only tuffs but sheets of lava were emitted, which, gathering round the vents, formed cones or long, connected banks and ridges. Of these there are four conspicuous examples in Scotland--the hills of the Burntisland district, the Bathgate Hills, the ground between Dalry and Galston in north Ayrshire, and a broken tract in Liddesdale. Nowhere in the volcanic history of this country have even the minutest details of that history been more admirably preserved than among the materials erupted from puys in these respective districts.

Lava-cones, answering to solitary tuff-cones among the fragmental eruptions, do not appear to have existed, or, like some of those in the great lava-fields of Northern Iceland and Western America, must have been mere small heaps of slag and cinders at the top of the lava-column, which were washed down and effaced during the subsidence and entombment of the volcanic materials. The lavas never occur without traces of fragmentary discharges. Two successive streams of basalt may indeed be found at a given locality without any visible intercalation of tuff, but proofs of the eruption of fragmental material will generally be observed to occur somewhere in the neighbourhood, associated with one or both of them, or with other lavas above or below them.

Where the phenomena of the puys have been most typically developed, lavas and tuffs succeed each other in rapid succession, with numerous or occasional interstratifications of ordinary sediment. Perhaps the most complete and interesting example of this association is to be found on the coast between Burntisland and Kirkcaldy, where, out of a total thickness of rock which may be computed to be between 1500 and 2000 feet, it will probably be a fair estimate to say that the igneous materials constitute four-fifths, or from

The general bedded character of the volcanic series is well shown in Fig. 153, which represents the alternations of lavas and tuffs in the Kingswood Craig two miles to the east of Burntisland. The harder basalts will be seen to project as bold crags while the tuffs and other stratified deposits between them give rise to grassy slopes and hollows. A nearer view of the alternation of lavas and tuffs with non-volcanic sedimentary deposits is supplied in Fig. 170, which is taken from a part of the Fife coast a little further to the east than the last illustration. Here one of the limestones of the Carboniferous Limestone series is overlain with shale and tuff, which, being easily disintegrated, have been cut away by the waves, leaving the lava above to overhang and fall off in blocks. The columnar structure of some of the basalts of this coast is well brought out in Fig. 171, which shows further how the columns sometimes merge into an amorphous part of the same sheet.

These Fife basalts illustrate admirably the peculiarities of the sheets of lava which are intercalated among the Carboniferous strata. They show how easy it generally is to discriminate between such sheets and intrusive sills. The true lavas are never so largely crystalline, nor spread out in such thick sheets as the sills; they are frequently slaggy and amygdaloidal, especially towards the top and bottom, the central portion being generally more fine-grained and sometimes porphyritic. Where most highly cellular they often decompose into a dull, earthy, dirty-green rock. Where they form a thick mass they are usually composed of different beds of varying texture. Except the differences between the more compact centre and the slaggy layer above and below, the bedded lavas do not present any marked variation in composition or structure within the same sheet. A striking exception to this rule, however, is furnished by the Bathgate "leckstone" already described. This mass forms a continuation of the great basaltic ridge of the Bathgate Hills, and though its exact relations to the surrounding strata are concealed, it appears to be an interbedded and not an intrusive sheet. The remarkable separation of its constituent minerals into an upper, lighter felspathic layer, and a lower, heavier layer, rich in olivine, augite and iron-ores, is a structure which might be more naturally expected to occur in a sill. An instance of its development in an undoubted sill will be described further on. Nevertheless, if we follow the trend of the volcanic band of the Bathgate Hills southward for only two miles beyond the picrite quarry, we find in the Skolie Burn a rock in many respects similar, and quarried for the same purpose of building oven-soles. This "leckstone" is there seen to be surmounted by a group of calcareous shales and thin limestones. The section laid bare in the stream is represented in Fig. 154. Immediately above the diabase, which is highly cellular, lies a green felspathic sandstone or shale containing detached fragments of the amygdaloid together with Lingulæ and other shells. There seems no reason to doubt that this is a true interstratified lava.

1. Slaggy diabase; 2. Green sandy shale and shaly sandstone containing Lingulæ, also pieces of slag from the underlying lava, which are completely wrapped round in the sediment; 3. Yellow calcareous shelly sandstone; 4. Dark shale with Spiriferæ, etc.; 5. Bed of blue crinoidal limestone; 6. Clays and thin coal; 7. Black and blue calcareous shales and thin limestones. ]

Where the puys attained their greatest development in Scotland, they rose in the shallow lagoons, and here and there from deeper parts of the sea-bottom, until by their successive discharges of lavas and tuffs they gradually built up piles of material, which, in the Linlithgow and Bathgate district, may have been nearly 2000 feet in thickness. It must be remembered, however, that the eruptions took place in a subsiding area, and that even the thickest volcanic ejections, if the downward movement kept pace with the volcanic activity, need not have grown into a lofty volcanic hill. Indeed, largely as the lavas and tuffs bulk in the geology of some parts of Central Scotland, their eruption does not seem to have seriously interfered with the broader physical changes that were in progress over the whole region. Thus the subsidence which led to the spread of a marine and limestone-making fauna over much of Central Scotland included also the volcanic districts. The limestones, formed of crinoids, corals and other marine organisms, extended over the submerged lavas and tuffs, and were even interstratified with them.

While the volcanic materials are found to replace locally the ordinary Carboniferous sedimentary strata, it is interesting in this regard to note that, during pauses in the volcanic activity, while the subsidence doubtless was still going on, some groups of sandstones, shales or limestones extended themselves across the volcanic ridges so as to interpose, on more than one platform, a mass of ordinary sediment between the lavas or tuffs already erupted and those of succeeding discharges, and thus to furnish valuable geological chronometers by which to define the stratigraphical horizons of the successive phases of volcanic energy.

1. Houston Coal; 2. Houston Marls and tuffs; 3. Interstratified sheets of basic lavas with occasional tuffs and intercalations of shale, sandstone, etc.; 4. Tartraven Limestone; 5. Hurlet Limestone with tuffs, shales and sandstones above and below; 6. Wardlaw Limestone; 7. Index Limestone; 8. Highest band of tuff--upward limit of the volcanic series; 9 9. Volcanic necks; 10. Sill of basalt; 11. Levenseat or Castlecary Limestone; 12. Millstone Grit; 13. Base of Coal-measures; 14. Thick doleritic sill; 15. Dolerite dyke (? Tertiary). ]

The volcanic banks or ridges not improbably emerged as islets out of the water, and were sometimes ten miles or more in length. Their materials were supplied from many separate vents along their surface, but probably never attained to anything approaching the elevation which they would have reached had they been poured out upon a stable platform. This feature in the history of the volcanic ridges is admirably shown by the fact just referred to, that recognizable stratigraphical horizons can sometimes be traced right through the heart of the thickest volcanic accumulations. One of the largest areas of basalts and tuffs connected with the puys is that of the Linlithgow and Bathgate Hills, where, as already remarked, a depth of some 2000 feet of igneous rocks has been piled up. Yet several well-known seams of stone can be traced through it, such as the Hurlet Limestone and the Index Limestone (Fig. 155). Only at the north end, where the volcanic mass is thickest and the surface-exposures of rock are not continuous, has it been impossible to subdivide the mass by mapping intercalations of sedimentary strata across it. It would thus seem that, even where the amplest accumulations gathered round the puys, they formed low flat domes, rather than prominent hills, which, as subsidence went on and the tuff-cones were washed down, gradually sank under water, and were buried under the accumulating silt of the sea-floor.

As a detailed illustration of the manner in which the growth of organically-formed limestones and the deposit of ordinary sediment took place concurrently with the occasional outflow of lava-streams over the sea-bottom, I may cite the section presented in another Linlithgowshire quarry (Fig. 156). At the bottom of the group of strata there exposed, a pale amygdaloidal, somewhat altered basalt (A) marks the upper surface of one of the submarine lavas of the period. Directly over it comes a bed of limestone (B) 15 feet thick, the lower layers of which are made up of a dense growth of the thin-stemmed coral Lithostrotion irregulare. The next stratum is a band of dark shale (C) about two feet thick, followed by about the same thickness of an impure limestone with shale seams (D). The conditions for coral and crinoid growth were evidently not favourable, for this argillaceous limestone was eventually arrested first by the deposit of a dark mud, now to be seen in the form of three or four inches of a black pyritous shale (E), and next by the inroad of a large quantity of dark sandy mud and drift vegetation, which has been preserved as a sandy shale (F), containing Calamites, Producti, ganoid scales and other traces of the life of the time. Finally, a great sheet of lava, represented by the uppermost amygdaloid (G), overspread the area, and sealed up these records of Palæozoic history.

Among the phenomena associated with the Carboniferous volcanoes mention may, in conclusion, be made of the evidence for the former existence of thermal springs and saline sublimations or incrustations. Among the plateau-tuffs of North Berwick, as has been already pointed out (p. 390), a fœtid limestone has been quarried, which bears indications of having been deposited by springs, probably in connection with the volcanic action of the district. The lower limestones of Bathgate furnish abundant laminæ of silica interleaved with calcareous matter, the whole probably due to the action of siliceous and calcareous springs connected with the active puys of that district. Some portions of the limestone are full of cellular spaces, lined with chalcedony. A saline water has been met with among the volcanic rocks to the west of Linlithgow, in a bore which was sunk to a depth of 348 feet in these rocks without reaching their bottom. The water that rose from the bore-hole was found to contain as much as 135 grains of chloride of sodium in the gallon. It is not improbable that this salt was originally produced by incrustations on the Carboniferous lavas immediately after their eruption, as has happened so often in recent times at Vesuvius, and that it was then buried under succeeding showers of tuff and streams of lava.

Subsequent Dislocation of Bedded Lavas and Tuffs.--As the interstratified volcanic materials were laid down in sheets at the surface, they necessarily behave like the ordinary sedimentary strata, and have undergone with them the various curvatures and fractures which have occurred since Carboniferous times. Notwithstanding their volcanic nature, they can be traced and mapped precisely as if they had been limestones or sandstones. This perfect conformability with the associated stratified rocks is strikingly seen in the case of the sheets of lava which lie imbedded in the heart of the great volcanic ridge of Linlithgowshire. The overlying strata having been removed from their surface for some distance, and the ground having been broken by faults, these volcanic rocks might at first be taken for irregular intrusive bosses, but their true character is that shown in Fig. 157, where by a succession of faults, with a throw in the same direction, the upper basalts of Bonnytoun Hill are gradually brought down to the level of the Firth of Forth.

iii. SILLS, BOSSES AND DYKES

One of the characteristic features of Central Scotland is the great number, and often the large size and extraordinary persistence, of the masses of eruptive, more or less basic material, which have been injected among the Carboniferous strata. The precise geological age of these intrusions cannot, of course, be more exactly defined than by stating that they are younger than the rocks which they traverse, though in many cases their association with the necks, lavas and tuffs is such as to show that they must be regarded as part of the Carboniferous volcanic phenomena.

Sills.--With regard to the sills I have been led, for the following reasons, to connect the great majority of them with the puys, though some are certainly of far later date, while others should possibly be assigned to the plateaux.

In the first place, the sills obviously connected with the plateaux are in great measure intermediate, or even somewhat acid rocks, while those of the puy series are much more basic. It is hardly possible, however, in all cases to decide to which series a particular sill should be assigned. This difficulty is particularly manifest in the western part of Midlothian, where the plateau of that district exhibits such frequent interruption, and where it often consists only of a single basaltic sheet. To the west of it lie the abundant puys with their lavas and tuffs, and between the two volcanic areas numerous sills of dolerite and diabase make their appearance. In the difficulty of deciding to which series these sills should be referred, it will be convenient to consider them with those of the puys.

1. Upper Old Red Sandstone; 2. "Ballagan Beds"; 3. Tuffs; 4. Lavas of the Campsie district of the Clyde plateau; 5 5. Necks belonging to the plateau volcanic series; 6. Trachytic sills belonging to the plateau; 7. Carboniferous Limestone series; 8. Dolerite sills cutting the Carboniferous Limestone series. f, Fault.]

A remarkable illustration of the contrast in petrographical character between the typical sills of the plateaux and those of the puys is furnished by the chain of the Campsie Fells, where, on the north side, among the Calciferous Sandstones which emerge from under the andesitic lavas of the Clyde plateau, many intrusive sheets and bosses of trachytic material may be seen, while on the southern side come the great basic sills which, from Milngavie by Kilsyth to Stirling, run in the Carboniferous Limestone series (Fig. 158). A similar contrast may be observed in Renfrewshire between the trachytic sills below the plateau-lavas south of Greenock and the basic sills above these lavas in the Carboniferous Limestone series around Johnstone and Paisley.

In the second place, the more basic sills, as a rule, appear on platforms higher in stratigraphical position than the plateaux, and wherever this is their position there cannot be any hesitation in deciding against their association with the older phase of volcanic activity.

In the third place, the basic sills often occur in obvious connection with the vents or bedded lavas and tuffs of the puy series. A conspicuous example of this dependence is supplied by the intrusive sheets of Burntisland, underlying the basalts and tuffs of that district in the immediate neighbourhood of some of the vents from which these bedded rocks were erupted (Fig. 159).

In the fourth place, even where no visible vents appear now at the surface near the sills, the latter generally occupy horizons within the stratigraphical range indicated by the interbedded volcanic rocks. It must be remembered that all the Carboniferous vents were deeply buried under sedimentary deposits, and that large as is the number of them which has been exposed by denudation, it is probably much smaller than the number still concealed from our view. The sills are to be regarded as deep-seated parts of the volcanic protrusions, and they more especially appear at the surface where the strata between which they were injected crop out from under some of the higher members of the Carboniferous system. Thus the remarkable group of sills between Kilsyth and Stirling (Fig. 158) may quite possibly be connected with a group of vents lying not far to the eastward, but now buried under the higher parts of the Carboniferous Limestone, Millstone Grit and Coal-measures. Again, the great series of sills that gives rise to such a conspicuous range of hills in the north and middle of Fife may have depended for its origin upon the efforts of a line of vents running east and west through the centre of the county, but now buried under the Coal-measures. Some vents, indeed, have been laid bare in that district, such as the conspicuous groups of the Saline Hills and the Hill of Beath, but many more may be concealed under higher Carboniferous strata further east.

1. Calciferous Sandstone series; 2. Burdiehouse Limestone; 3. Sandstones, shales and tuffs; 4. Basalts and tuffs, with intercalations of sandstone, shale and limestone; 5. Agglomerate of the Binn of Burntisland neck; 6. Basalt dyke; 7. Dyke and sill; 8 8 8. Three sills. ]

In the fifth place, the materials of which the sills consist link them in petrographical character with those that proceeded from the puys. The rocks of the intrusive sheets in West Lothian, Midlothian and Fife are very much what an examination of the bedded lavas of the puys in the same region would lead us to expect. There is, of course, the marked textural difference between masses of molten rock which have cooled very slowly within the crust of the earth and those which have solidified with rapidity at the surface, the sills being for the most part much more coarsely crystalline than the lavas, and more uniform in texture throughout, though generally finer at the margins than at the centre. There is likewise the further contrast arising from differences in the composition of the volcanic magma at widely-separated periods of its extravasation. At the time when the streams of basalt flowed out from the puys its constitution was comparatively basic, in some localities even extremely basic. Any sills dating from that time may be expected to show an equal proportion of bases. But those which were injected at a long subsequent stage in the volcanic period may well have been considerably more acid.

In actual fact the petrographical range of the sills reasonably referable to the puy-eruptions varies from picrite or limburgite to dolerite without olivine. The great majority of these sheets in the basin of the Firth of Forth, where they are chiefly displayed, are dolerites (diabases), sometimes with, but more frequently without, olivine. They include all the more coarsely crystalline rocks of the region, though occasionally they are ordinary close-grained basalts. Their texture may be observed to bear some relation to their mass, so far at least as that, where they occur in beds only two or three feet or yards in thickness, they are almost invariably closer-grained. A cellular or amygdaloidal texture is seldom to be observed among them, and never where they are largely crystalline. This texture is most often to be found in thin sills which have been injected among carbonaceous shales or coals. These intrusive sheets are generally finely cellular, and more or less decayed ("white trap").

Differences of texture may often be observed within short distances in the same sill, and likewise considerable varieties in colour and composition. The most finely crystalline portions are, as usual, those along the junction with the stratified rocks, the most crystalline occurring in the central parts of the mass. A diminution in the size of the crystalline constituents may be traced not only at the base, but also at the top of a sheet, or at any intermediate portion which has come in contact with a large mass of the surrounding rock. A good illustration is supplied by the intrusive sheet at Hound Point (Fig. 160), to the east of South Queensferry, where some layers of shale have been involved in the igneous rock, which becomes remarkably close-grained along the junction. This change in texture and absence of cellular structure form a well-marked distinction between these sheets and those which have flowed out at the surface as true lava-streams.

Some of the larger doleritic sills display a somewhat coarsely crystalline texture in their central portions, and occasionally present a notable micropegmatitic aggregate, which plays the part of interstitial substance enclosing the other minerals. Mr. Teall has referred to the frequent occurrence of this structure in the coarser parts of the Whin Sill of the north of England. It occurs also in a marked degree in the Ratho sill and in some portions of the great doleritic sill of which the crags of Stirling form a part.

But beside the differences in texture, mainly due to varying rates of cooling, the sills sometimes exhibit striking varieties of composition in the same mass of rock. These variations are more especially noticeable among the larger sills, and particularly where the material is most markedly basic. The special type of differentiation, so noticeable in the Bathgate diabase and picrite mass already alluded to, is likewise well exhibited in an intrusive sheet or group of sheets, recently exposed at Barnton, in the cutting of a railway from Edinburgh to Cramond (Fig. 161). The intrusive nature of the several bands of igneous rock which occur here is made quite evident by the alteration they have produced upon the shales with which they have come in contact. It is the uppermost and most extensive of these sills which specially deserves notice, for the differentiation of its constituents. It stretches along the cutting for several hundred yards at an angle of dip of about 15°. At the western or upper part of the mass its actual contact with the superincumbent sedimentary strata is not visible, but as the igneous rock is there a good deal finer in grain than elsewhere, its upper surface cannot be many feet distant. The upper visible portion is a light well-crystallized dolerite with a rudely bedded structure, the planes dipping westwards at 15°. About 20 or 30 feet below the upper visible termination of the mass, the dark ferro-magnesian minerals begin rapidly to increase in relative proportion to the pale felspar, and the rock consequently becomes dark-greenish brown. The change is particularly noticeable in certain bands which run parallel with the general dip. There is no definite line between the pale and dark body of the rock, the two graduating into each other and the darker part becoming deeper in colour, heavier and more decomposing, until it becomes a true typical picrite. Even in this ultra-basic portion the same rude bedding or banding may be observed.

1. Baked shale; 2. Sill of very felspathic dolerite about, nine feet thick; 3. Baked shale, eight inches; 4. Dolerite showing chilled fine-grained edge and adhering firmly to the shale below; it rapidly passes up into (5) Picrite with white felspathic veins (6); 7. Junction of picrite and dolerite with a similar vein along the line of contact; 8. Large globular body of dolerite enclosing a mass of picrite. ]

Veins in which felspar predominates over the darker minerals traverse the rock, sometimes parallel with the bedding, sometimes across it. They vary from less than an inch to a foot in width, sometimes dividing and enclosing parts of the surrounding mass. But that they are on the whole contemporaneous with the sill itself, and not long subsequent injections, is shown by the way in which the dark ferro-magnesian minerals project from the picrite into the veins and lock the two together.

But besides these injections, which doubtless represent the last and more acid portions of the magma injected into the basic parts before the final consolidation of the whole, there are to be observed irregular concretionary patches, of similar character to the veins, distributed through the picrite. On the other hand, towards its base the sill becomes a coarse dolerite round which the picrite is wrapped, and which encloses a detached portion of that rock.

It is deserving of note that while the ultra-basic portion descends almost to the very bottom of the sill, the lowest five feet show the same change as occurs at the top of the mass. There the felspar rapidly begins to predominate over the darker minerals, and the dolerite into which the rock passes shows a fine-grained margin adhering firmly to the shales on which it rests. This lower doleritic band, showing as it does the effect of chilling upon its under surface, may be due to more rapid cooling and crystallization, while in the overlying parts the mass remained sufficiently mobile to allow of a separation of the heavier minerals from the felspars, which appear in predominant quantity towards the top. It must be frankly admitted, however, that we are still very ignorant of the causes which led to this separation of ingredients in a few sills, while they were entirely absent or non-efficient in most of them.

The intrusive character of the Carboniferous sills of Central Scotland and their contact-metamorphism have been fully described, and some of them have become, as it were, "household words" in geology. Exposed in so many fine natural sections in the vicinity of Edinburgh, they early attracted the notice of geologists, and furnished a battle-ground on which many a conflict took place between the Plutonist and Neptunist champions at the beginning of the present century.

As the sills frequently lie in even sheets perfectly parallel with the bedding of the strata between which they have been injected, care is required in some cases to establish that they are of intrusive origin. One of the most obvious tests for this purpose is furnished by the alteration they produce among the strata through which they have made their way, whether these lie above or below them. The strata are sometimes crumpled up in such a manner as to indicate considerable pressure. They are occasionally broken into fragments, though this may have been due rather to the effects of gaseous explosions than to the actual protrusion of melted rock. But the most frequent change superinduced upon them is an induration which varies greatly in amount even along the edge of the same intrusive sheet. Sandstones are hardened into quartzite, breaking with a smooth clear glistening fracture. Coals are converted into a soft sooty substance, sometimes into anthracite. Limestones acquire a crystalline saccharoid structure. Shales pass generally into a kind of porcellanite, but occasionally exhibit other types of contact-metamorphism. Thus below the thick picrite sill at Barnton, near Edinburgh, the shales have assumed a finely concretionary structure by the appearance in them of spherical pea-like aggregates.

Another proof of intrusion is to be found in the manner in which sills catch up and completely enclose portions of the overlying strata. The well-known examples on Salisbury Crags (Fig. 162) are paralleled by scores of other instances in different parts of the same region.

Moreover, sills do not always remain on the same horizon; that is, between the same strata. They may be observed to steal across or break through the beds, so as to lie successively between different layers. No more instructive example of this relation on a small scale could be cited than that of the intrusive sheet which has been laid open in the Dodhead Limestone Quarry, near Burntisland. As shown in the accompanying figure (Fig. 163), this rock breaks through the limestone and then spreads out among the overlying shales, across which it passes obliquely.

Among the larger sills this transgressive character is seen to be sometimes manifested on a great scale. Thus, along the important belt of intrusive rocks that runs from Kilsyth to Stirling, the Hurlet Limestone lies in one place below, in another above, the invading mass, but in the intervening ground has been engulphed in it. Similar evidence of the widely separate horizons occupied by different parts of the same sill is supplied at Kilsyth, where the intrusive sheet lies about 70 or 80 fathoms below the Index Limestone, while at Croy, in the same neighbourhood, it actually passes above that seam.

Other interesting evidence of the intrusive nature of the Carboniferous dolerite sills of Central Scotland is supplied by the internal modifications which the eruptive rock has undergone by contact with the strata between which it has been thrust. These alterations, though partly visible to the naked eye, are best studied in thin slices with the aid of the microscope. Tracing the variations of an intrusive dolerite outwards in the direction of the rocks which it has invaded, we perceive change first in the augite. The large crystals and kernels of that mineral grow smaller until they pass into a granulated form like that characteristic of basalts. The large plates and amorphous patches of titaniferous iron or magnetite give place to minute particles, which tend to group themselves into long club-shaped bodies. The labradorite continues but little affected, except that its prisms, though as defined, may not be quite so large. The interstitial glassy groundmass remains in much the same condition and relative amount as in the centre of the rock.

Along the line of contact, while the dolerite becomes exceedingly close-grained, its felspar crystals are still quite distinct even up to the very edge. But they become fewer in relative number, and still smaller in size, though an occasional prism two or three millimetres in length may occur. They retain also their sharpness of outline, and their comparative freedom from enclosures of any kind. They tend to range themselves parallel with the surface of the contact-rock. The augite exists as a finely granular pale green substance, which might at first be taken for a glass, but it gives the characteristic action of augite with polarized light. It is intimately mixed through the clear glass of the groundmass, which it far exceeds in quantity. The iron oxides now appear as a fine granular dust, which is frequently aggregated into elongated club-shaped objects, as if round some inner pellucid or translucent microlite. In patches throughout the field, however, the oxides take the form of a geometrically perfect network of interlacing rods. This beautiful structure, described and figured by Zirkel and others, is never to be seen in any of the dolerites, except close to the line of contact with the surrounding rocks. It occurs also in some of the dykes. I have not succeeded in detecting any microlites in the sandstones at the edge of a dolerite sheet, though I have had many slices prepared for the purpose.

Where one of the dolerite sills has invaded sandstone, there is usually a tolerably sharp line of demarcation between the two rocks, though it is seldom easy to procure a hand-specimen showing the actual contact, for the stone is apt to break along the junction-line. Where, however, the rock traversed by the igneous mass is argillaceous shale, we may find a thorough welding of the two substances into each other. In such cases the dolerite at the actual contact becomes a dark opaque rock, which in thin slices under the microscope is found to be formed of a mottled or curdled segregation of exceedingly minute black grains and hairs in a clear glassy matrix, in which the augite and felspar are not individualized. But even in this tachylyte-like rock perfectly formed and very sharply defined crystals of triclinic felspar may be observed ranging themselves as usual parallel to the bounding surfaces of the rock. These characters are well seen in the contact of the intrusive sheet of dolerite with shale and sandstone at Hound Point (Fig. 160).

Another instructive example is furnished by the small threads which proceed from the dolerite of Salisbury Crags, and traverse enclosed fragments of shale (Fig. 162). Some of these miniature dykes are not more than one-eighth of an inch in diameter, and may therefore easily be included, together with part of the surrounding rock, in the field of the microscope. The dolerite in these ramifications assumes an exceedingly fine texture. The felspar is the only mineral distinctly formed into definite crystals. It occurs in prisms of an early consolidation, sometimes one-fifth of an inch long, and therefore readily recognizable by the naked eye. These prisms are perfectly shaped, contain abundant twin lamellæ, and show enclosures of the iron of the base. They had been already completely formed at the time of injection; for occasionally they may be observed projecting beyond the wall of the vein into the adjacent shale or sandstone, and they have ranged themselves parallel to the sides of the vein. The black ground, from which these large well-defined crystals stand out prominently, consists of a devitrified glass, rendered dark by the multitude of its enclosed black opaque microlites. These are very minute grains and rudely feathered rods, with a tendency to group themselves here and there into forms like portions of the rhombohedral skeletons of titaniferous iron. So thoroughly fused and liquid has the dolerite been at the time of its injection, that little threads of it, less than 1/100 of an inch in diameter, consisting of the same dark base, with well-defined felspars, may be seen isolated within the surrounding sedimentary rock. Minute grains and rounded portions of the latter may also be noticed in the marginal parts of the dolerite.

It is thus evident that specimens taken from the edge of an intrusive sheet, where the rock has rapidly chilled and solidified, represent to us an earlier stage in the history of the whole mass than specimens taken from its central portions. In fact, a series of samples collected at short intervals from the outer contact to the inner mass shows, as it were, the successive stages in the consolidation of the molten rock.

From the observations just described, it appears that the triclinic felspars began to assume the shape of large definite crystals before any of the other minerals. These felspars already existed when the molten mass forced its way among the shales, for they can be seen lying with their long axes parallel to the surface of shale, precisely as, in the well-known flow-structures, they behave round a large crystal embedded in the heart of a rock. A few feet from where the consolidation was not so rapid, the iron oxides have grouped themselves into incipient crystalline forms and skeleton crystals; the felspar crystals abundantly occur, and the augite has been left in the finely granular condition. Still further towards the interior of the mass, the normal character of the dolerite is gradually assumed.

Where a sill has been injected among carbonaceous shales and coals it has undergone great alteration along the contact, and if the sheet is only a few inches or feet thick, the change extends throughout its whole mass. Black basalts and dolerites, in such circumstances, pass into a substance like a white or pale yellow clay, which at first might be mistaken for some band of fire-clay intercalated among the other sediments. But evidence of actual intrusion may usually be found, as where the igneous rock has caught up or broken through the adjacent strata, besides altering them. Such "white traps," as they have been called, have long been familiar in the coal-fields of Scotland and Central England.

1. Hurlet Limestone; 2. Black shales; 3 3. Two sills of "White Trap"; 4. Columnar Basalts. ]

As a good illustration of the behaviour of such thin sills among carbonaceous shales I give here a section (Fig. 165) exposed in the old limestone quarry of Hillhouse, south of Linlithgow. At the bottom lies the Hurlet Limestone which has once been extensively mined at this locality. Above it comes a group of black shales which in turn are surmounted by a sheet of beautifully columnar basalt. The shales seem at first sight to include two layers of pale fire-clay, each only a few inches in thickness. Closer inspection, however, will show that these two thin intercalations are really sills which, though on the whole parallel with the bedding of the shale, may be seen to cut across it, and even at one place to send a finger into it. The upper example may also be observed to diminish rapidly in thickness in one direction.

The dimensions of the sills vary within tolerably wide limits. Although here and there the injected material dwindles down to an inch or less in thickness, running away even into threads, it more usually forms sheets of considerable depth. The rock of Salisbury Crags, for example, is fully 150 feet thick at its maximum. That of Corstorphine Hill is probably about 350 feet. The great sheet which runs among the lower limestones from Kilsyth by Denny to Stirling has been bored through to a depth of 276 feet, but as the bore started on the rock, and not in overlying strata, some addition may need to be made to that thickness.

The spheroidal weathering so characteristic of basic eruptive rocks is nowhere more characteristically displayed than among the great doleritic sills of the basin of the Firth of Forth. As an illustration of this structure an example is taken here from the large sheet at North Queensferry (Fig. 164).

While one is struck with the great size and extent of some of the sills connected with the puys, as compared with the small and local sheets underneath the plateaux, there is a further fact regarding them that

It would seem from these facts that the extent to which the crust of the earth round a volcanic orifice is injected with molten rock, in the form of intrusive sheets between the strata, does not depend upon the energy of the volcano as gauged by its superficial outpourings, but on other considerations not quite apparent. Possibly, the more effectively volcanic energy succeeded in expelling materials from the vent, the less opportunity was afforded for subterranean injections. And if the protrusion of the sills took place after the vents were solidly sealed up with agglomerate or lava, it would doubtless often be easier for the impelled magma to open a way for itself laterally between the bedding-planes of the strata than vertically through the thick solid crust. The size and extent of the sills may thus be a record of the intensity of this latest phase of the volcanic eruptions.

1. Burdiehouse Limestone; 2. Shales; 3. Diabase.]

* * * * *

Bosses.--The rounded, oval or irregularly shaped masses of igneous rock included under this head are found in some cases to be only denuded domes of sills, as, for example, in the apparently isolated patches in the oil-shale district of Linlithgowshire, which have been found to unite under ground. (Compare Fig. 157). In other instances, bosses possibly, or almost certainly, mark the position of volcanic funnels, as at the Castle Rock of Edinburgh, Dunearn Hill, Burntisland, and Galabraes, near Bathgate. But many examples occur which can only be regarded as the exposed ends of irregular bodies of molten material which has been protruded upwards into the

The connection between bosses and intrusive sheets is instructively exhibited in a railway cutting to the west of Edinburgh, where the section shown in Fig. 167 may be seen. In the space of a few yards no fewer than four distinct bands of diabase traverse the shale, thickening rapidly in one direction and uniting with a large boss of more coarsely crystalline material. Such connections must exist in all sills, for the material injected as a sheet between stratified formations cannot but be united to some column, dyke or irregular protrusion which descends to the parent magma in the interior. But it is very rarely that the geologist is permitted to see them.

* * * * *

Dykes take a comparatively unimportant place in the eruptive phenomena of the puys. They occur in some numbers, but on a small scale, among the tuff vents, and there they can without much hesitation be set down as part of the phenomena of eruption through these pipes. The Binn of Burntisland, which has been so often referred to in this Chapter, may again be cited as a typical vent for the display of this series of dykes (Figs. 149 and 159). Two additional illustrations from this locality are here given. In Fig. 166 a dyke of compact black basalt is seen on the right hand running up the steep slopes of the agglomerate. Some of these dykes are distinctly columnar, the columns diverging from the walls on each side. Where the encasing agglomerate has been removed by the weather, the side of the dyke presents a reticulated network of prism-ends. A narrow basalt-dyke of this character near the top of the Binn vent is represented in Fig. 168.

But dykes also occur apart from vents and without any apparent relation to these. They are sometimes associated with sills and bosses in such a manner as to suggest that the whole of these forms of injected material belong to one connected series of intrusions. Among the Bathgate Hills, for example, from which I have already cited instances of sills and a boss, the section represented in Fig. 169 occurs. Yet in this same district there is a group of large east and west dykes which cut all the other rocks including the bedded lavas and tuffs, and must be of later date than the highest part of the Coal-measures (Fig. 155).

It is difficult to ascertain the age of the dykes which rise through the Carboniferous formations at a distance from any interbedded sheets of lava and tuff, or from any recognizable vent. The south-east and north-west dykes, increasing in number as they go westward, and attaining a prodigious development in the great volcanic area of Antrim and the Inner Hebrides, are probably of Tertiary date. Others may possibly be Permian, while a certain number may reasonably be looked upon as Carboniferous. In petrographical characters the latter resemble the dolerites and basalts (diabases) of the finer-grained sills.