Volcanic Rocks of the Upper Miocene Period. — Madeira. — Grand Canary. — Azores. — Lower Miocene Volcanic Rocks. — Isle of Mull. — Staffa and Antrim. — The Eifel. — Upper and Lower Miocene Volcanic Rocks of Auvergne. — Hill of Gergovia. — Eocene Volcanic Rocks of Monte Bolca. — Trap of Cretaceous Period. — Oolitic Period. — Triassic Period. — Permian Period. — Carboniferous Period. — Erect Trees buried in Volcanic Ash in the Island of Arran. — Old Red Sandstone Period. — Silurian Period. — Cambrian Period. — Laurentian Volcanic Rocks.

Volcanic Rocks of the Upper Miocene Period.—Madeira.—The greater part of the volcanic eruptions of Madeira, as we have already seen (p. 532), belong to the Pliocene Period, but the most ancient of them are of Upper Miocene date, as shown by the fossil shells included in the marine tuffs which have been upraised at San Vicente, in the northern part of the island, to the height of 1300 feet above the level of the sea. A similar marine and volcanic formation constitutes the fundamental portion of the neighbouring island of Porto Santo, forty miles distant from Madeira, and is there elevated to an equal height, and covered, as in Madeira, with lavas of supra-marine origin.

The largest number of fossils have been collected from the tuffs and conglomerates and some beds of limestone in the island of Baixo, off the southern extremity of Porto Santo. They amount in this single locality to more than sixty in number, of which about fifty are mollusca, but many of these are only casts. Some of the shells probably lived on the spot during the intervals between eruptions, and some may have been cast up into the water or air together with muddy ejections, and, falling down again, have been deposited on the bottom of the sea. The hollows in some of the fragments of vesicular lava of which the breccias and conglomerates are composed are partially filled with calc-sinter, being thus half converted into amygdaloids. Among the fossil shells common to Madeira and Porto Santo, large cones, strombs, and cowries are conspicuous among the univalves, and Cardium, Spondylus, and Lithodomus among the lamellibranchiate bivalves, and among the Echinoderms the large Clypeaster called C. altus, an extinct European Miocene fossil.

[ 537 ]

The largest list of fossils has been published by Mr. Karl Meyer, in Hartung’s “Madeira;” but in the collection made by myself, and in a still larger one formed by Mr. J. Yate Johnson, several remarkable forms not in Meyer’s list occur, as, for example, Pholadomya, and a large Terebra. Mr. Johnson also found a fine specimen of Nautilus (Atruria) ziczac (Fig. 211), a well-known Falunian fossil of Europe; and in the same volcanic tuff of Baixo, the Echinoderm Brissus Scillæ, a living Mediterranean species, found fossil in the Miocene strata of Malta. Mr. Meyer identifies one-third of the Madeira shells with known European Miocene (or Falunian) forms. The huge Strombus of San Vicente and Porto Santo, S. Italicus, is an extinct shell of the Sub-apennine or Older Pliocene formations. The mollusca already obtained from various localities of Madeira and Porto Santo are not less than one hundred in number, and, according to the late Dr. S. P. Woodward, rather more than a third are of species still living, but many of these are not now inhabitants of the neighbouring sea.

It has been remarked (p. 212), that in the Older Pliocene and Upper Miocene deposits of Europe many forms occur of a more southern aspect than those now inhabiting the nearest sea. In like manner the fossil corals, or Zoantharia, six in number, which I obtained from Madeira, of the genera Astræa, Sarcinula, Hydnophora, were pronounced by Mr. Lonsdale to be forms foreign to the adjacent coasts, and agreeing with the fauna of a sea warmer than that now separating Madeira from the nearest part of the African coast. We learn, indeed, from the observations made in 1859, by the Reverend R. T. Lowe, that more than one-half, or fifty-three in ninety, of the marine mollusks collected by him from the sandy beach of Mogador are common British species, although Mogador is 18½ degrees south of the nearest shores of England. The living shells of Madeira and Porto Santo are in like manner those of a temperate climate, although in great part differing specifically from those of Mogador.*

Grand Canary.—In the Canaries, especially in the Grand Canary, the same marine Upper Miocene formation is found. Stratified tuffs, with intercalated conglomerates and lavas, are there seen in nearly horizontal layers in sea-cliffs about 300 feet high, near Las Palmas. Mr. Hartung and I were unable to find marine shells in these tuffs at a greater elevation than 400 feet above the sea; but as the deposit to which they belong reaches to the height of 1100 feet or more in the interior, we conceive that an upheaval of at least that amount has

* Linnean Proceedings; Zoology, 1860.

[ 538 ]

taken place. The Clypeaster altus, Spondylus gæderopus, Pectunculus pilosus, Cardita calyculata, and several other shells, serve to identify this formation with that of the Madeiras, and Ancillaria glandiformis, which is not rare, and some other fossils, remind us of the faluns of Touraine.

The sixty-two Miocene species which I collected in the Grand Canary were referred by the late Dr. S. P. Woodward to forty-seven genera, ten of which are no longer represented in the neighbouring sea, namely Corbis, an African form, Hinnites, now living in Oregon, Thecidium (T. Mediterranean, identical with the Miocene fossil of St. Juvat, in Brittany), Calyptræa, Hipponyx, Nerita, Erato, Oliva, Ancillaria, and Fasciolaria.

These tuffs of the southern shores of the Grand Canary, containing the Upper Miocene shells, appear to be about the same age as the most ancient volcanic rocks of the island, composed of slaty diabase, phonolite, and trachyte. Over the marine lavas and tuffs trachytic and basaltic products of subaërial volcanic origin, between 4000 and 5000 feet in thickness, have been piled, the central parts of the Grand Canary reaching the height of about 6000 feet above the level of the sea. A large portion of this mass is of Pliocene date, and some of the latest lavas have been poured out since the time when the valleys were already excavated to within a few feet of their present depth.

On the whole, the rocks of the Grand Canary, an island of a nearly circular shape, and 6½ geographical miles diameter, exhibit proofs of a long series of eruptions beginning like those of Madeira, Porto Santo, and the Azores, in the Upper Miocene period, and continued to the Post-Pliocene. The building up of the Grand Canary by subaërial eruptions, several thousand feet thick, went on simultaneously with the gradual upheaval of the earliest products of submarine eruptions, in the same manner as the Pliocene marine strata of the oldest parts of Vesuvius and Etna have been upraised during eruptions of Post-tertiary date.

In proof that movements of elevation have actually continued down to Post-tertiary times, I may remark that I found raised beaches containing shells of the Recent Period in the Grand Canary, Teneriffe, and Porto Santo. The most remarkable raised beach which I observed in the Grand Canary, in the study of which I was assisted by Don Pedro Maffiotte, is situated in the north-eastern part of the island at San Catalina, about a quarter of a mile north of Las Palmas. It intervenes between the base of the high cliff formed of the tuffs with Miocene shells and the sea-shore. From

[ 539 ]

this beach, at an elevation of twenty-five feet above high-water mark, and at a distance of about 150 feet from the present shore, I obtained more than fifty species of living marine shells. Many of them, according to Dr. S. P. Woodward, are no longer inhabitants of the contiguous sea, as, for example, Strombus bubonius, which is still living on the West Coast of Africa, and Cerithium procerum, found at Mozambique; others are Mediterranean species, as Pecten Jacobæus and P. polymorphus. Some of these testacea, such as Cardita squamosa, are inhabitants of deep water, and the deposit on the whole seems to indicate a depth of water exceeding a hundred feet.

Azores.—In the island of St. Mary’s, one of the Azores, marine fossil shells have long been known. They are found on the north-east coast on a small projecting promontory called Ponta do Papagaio (or Point-Parrot), chiefly in a limestone about twenty feet thick, which rests upon, and is again covered by, basaltic lavas, scoriæ, and conglomerates. The pebbles in the conglomerate are cemented together with carbonate of lime.

Mr. Hartung, in his account of the Azores, published in 1860, describes twenty-three shells from St. Mary’s,* of which eight perhaps are identical with living species, and twelve are with more or less certainty referred to European Tertiary forms, chiefly Upper Miocene. One of the most characteristic and abundant of the new species, Cardium Hartungi, not known as fossil in Europe, is very common in Porto Santo and Baixo, and serves to connect the Miocene fauna of the Azores and the Madeiras. In some of the Azores, as well as in the Canary islands, the volcanic fires are not yet extinct, as the recorded eruptions of Lanzerote, Teneriffe, Palma, St. Michael’s, and others, attest.

Lower Miocene Volcanic Rocks.—Isle of Mull and Antrim.—I may refer the reader to the account already given (p. 247) of leaf-beds at Ardtun, in the Isle of Mull in the Hebrides, which bear a relation to the associated volcanic rocks of Lower Miocene date analogous to that which the Madeira leaf-bed, above described (p. 532), bears to the Pliocene lavas of that island. Mr. Geikie has shown that the volcanic rocks in Mull are above 3000 feet in thickness. There seems little doubt that the well-known columnar basalt of Staffa, as well as that of Antrim in Ireland, are of the same age, and not of higher antiquity, as once suspected.

The Eifel.—A large portion of the volcanic rocks of the

* Hartung, Die Azoren, 1860; also Insel Gran Canaria, Madeira und Porto Santo, 1864, Leipsig.

[ 540 ]

Lower Rhine and the Eifel are coeval with the Lower Miocene deposits to which most of the “Brown-Coal” of Germany belongs. The Tertiary strata of that age are seen on both sides of the Rhine, in the neighbourhood of Bonn, resting unconformably on highly inclined and vertical strata of Silurian and Devonian rocks. The Brown-Coal formation of that region consists of beds of loose sand, sandstone, and conglomerate, clay with nodules of clay-iron-stone, and occasionally silex. Layers of light brown and sometimes black lignite are interstratified with the clays and sands, and often irregularly diffused through them. They contain numerous impressions of leaves and stems of trees, and are extensively worked for fuel, whence the name of the formation. In several places layers of trachytic tuff are interstratified, and in these tuffs are leaves of plants identical with those found in the brown-coal, showing that, during the period of the accumulation of the latter, some volcanic products were ejected. The igneous rocks of the Westerwald, and of the mountains called the Siebengebirge, consist partly of basaltic and partly of trachytic lavas, the latter being in general the more ancient of the two. There are many varieties of trachyte, some of which are highly crystalline, resembling a coarse-grained granite, with large separate crystals of feldspar. Trachytic tuff is also very abundant.

M. Von Dechen, in his work on the Siebengebirge,* has given a copious list of the animal and vegetable remains of the fresh-water strata associated with the brown-coal of that part of Germany. Plants of the genera Flabellaria, Ceanothus, and Daphnogene, including D. cinnamomifolia (Fig. 155), occur in these beds, with nearly 150 other plants. The fishes of the brown-coal near Bonn are found in a bituminous shale, called paper-coal, from being divisible into extremely thin leaves. The individuals are very numerous; but they appear to belong to a small number of species, some of which were referred by Agassiz to the genera Leuciscus, Aspius, and Perca. The remains of frogs also, of extinct species, have been discovered in the paper-coal; and a complete series may be seen in the museum at Bonn, from the most imperfect state of the tadpole to that of the full-grown animal. With these a salamander, scarcely distinguishable from the recent species, has been found, and the remains of many insects.

Upper and Lower Miocene Volcanic Rocks of Auvergne.—The extinct volcanoes of Auvergne and Cantal, in central France, seem to have commenced their eruptions in the Lower

* Geognost. Beschreib. des Siebengebirges am Rhein. Bonn, 1852.

[ 541 ]

Miocene period, but to have been most active during the Upper Miocene and Pliocene eras. I have already alluded to the grand succession of events of which there is evidence in Auvergne since the last retreat of the sea (see p. 527).

The earliest monuments of the Tertiary Period in that region are lacustrine deposits of great thickness, in the lowest conglomerates of which are rounded pebbles of quartz, mica-schist, granite, and other non-volcanic rocks, without the slightest intermixture of igneous products. To these conglomerates succeed argillaceous and calcareous marls and limestones, containing Lower Miocene shells and bones of mammalia, the higher beds of which sometimes alternate with volcanic tuff of contemporaneous origin. After the filling up or drainage of the ancient lakes, huge piles of trachytic and basaltic rocks, with volcanic breccias, accumulated to a thickness of several thousand feet, and were superimposed upon granite, or the contiguous lacustrine strata. The greater portion of these igneous rocks appear to have originated during the Upper Miocene and Pliocene periods; and extinct quadrupeds of those eras, belonging to the genera Mastodon, Rhinoceros, and others, were buried in ashes and beds of alluvial sand and gravel, which owe their preservation to overspreading sheets of lava.

In Auvergne, the most ancient and conspicuous of the volcanic masses is Mont Dor, which rests immediately on the granitic rocks standing apart from the fresh-water strata. This great mountain rises suddenly to the height of several thousand feet above the surrounding platform, and retains the shape of a flattened and somewhat irregular cone, the slope of which is gradually lost in the high plain around. This cone is composed of layers of scoriæ, pumice-stones, and their fine detritus, with interposed beds of trachyte and basalt, which descend often in uninterrupted sheets until they reach and spread themselves round the base of the mountain.* Conglomerates, also, composed of angular and rounded fragments of igneous rocks, are observed to alternate with the above; and the various masses are seen to dip off from the central axis, and to lie parallel to the sloping flanks of the mountain. The summit of Mont Dor terminates in seven or eight rocky peaks, where no regular crater can now be traced, but where we may easily imagine one to have existed, which may have been shattered by earthquakes, and have suffered degradation by aqueous agents. Originally, perhaps, like the highest crater of Etna, it may have formed

* Scrope’s Central France, p. 98.

[ 542 ]

an insignificant feature in the great pile, and, like it, may frequently have been destroyed and renovated.

Respecting the age of the great mass of Mont Dor, we can not come at present to any positive decision, because no organic remains have yet been found in the tuffs, except impressions of the leaves of trees of species not yet determined. It has already been stated (p. 234) that the earliest eruptions must have been posterior in origin to those grits and conglomerates of the fresh-water formation of the Limagne which contain no pebbles of volcanic rocks. But there is evidence at a few points, as in the hill of Gergovia, presently to be mentioned, that some eruptions took place before the great lakes were drained, while others occurred after the desiccation of those lakes, and when deep valleys had already been excavated through fresh-water strata.

The valley in which the cone of Tartaret, above-mentioned (p. 527), is situated affords an impressive monument of the very different dates at which the igneous eruptions of Auvergne have happened; for while the cone itself is of Post-Pliocene date, the valley is bounded by lofty precipices composed of sheets of ancient columnar trachyte and basalt, which once flowed from the summit of Mont Dor in some part of the Miocene period. These Miocene lavas had accumulated to a thickness of nearly 1000 feet before the ravine was cut down to the level of the river Couze, a river which was at length dammed up by the modern cone and the upper part of its course transformed into a lake.

Gergovia.—It has been supposed by some observers that there is an alternation of a contemporaneous sheet of lava with fresh-water strata in the hill of Gergovia, near Clermont.

[ 543 ]

But this idea has arisen from the intrusion of the dike represented in Fig. 604, which has altered the green and white marls both above and below. Nevertheless, there is a real alternation of volcanic tuff with strata containing Lower Miocene fresh-water shells, among others a Melania allied to M. inquinata (Fig. 217), with a Melanopsis and a Unio; there can, therefore, be no doubt that in Auvergne some volcanic explosions took place before the drainage of the lakes, and at a time when the Lower Miocene species of animals and plants still flourished.

Eocene Volcanic Rocks.—Monte Bolca.—The fissile limestone of Monte Bolca, near Verona, has for many centuries been celebrated in Italy for the number of perfect Ichthyolites which it contains. Agassiz has described no less than 133 species of fossil fish from this single deposit, and the multitude of individuals by which many of the species are represented is attested by the variety of specimens treasured up in the principal museums of Europe. They have been all obtained from quarries worked exclusively by lovers of natural history, for the sake of the fossils. Had the lithographic stone of Solenhofen, now regarded as so rich in fossils, been in like manner quarried solely for scientific objects, it would have remained almost a sealed book to palæontologists, so sparsely are the organic remains scattered through it. When I visited Monte Bolca, in company with Sir Roderick Murchison, in 1828, we ascertained that the fish-bearing beds were of Eocene date, containing well-known species of Nummulites, and that a long series of submarine volcanic eruptions, evidently contemporaneous, had produced beds of tuff, which are cut through by dikes of basalt. There is evidence here of a long series of submarine volcanic eruptions of Eocene date, and during some of them, as Sir R. Murchison has suggested, shoals of fish were probably destroyed by the evolution of heat, noxious gases, and tufaceous mud, just as happened when Graham’s Island was thrown up between Sicily and Africa in 1831, at which time the waters of the Mediterranean were seen to be charged with red mud, and covered with dead fish over a wide area.*

Associated with the marls and limestones of Monte Bolca are beds containing lignite and shale with numerous plants, which have been described by Unger and Massalongo, and referred by them to the Eocene period. I have already cited (p. 263) Professor Heer’s remark, that several of the species are common to Monte Bolca and the white clay of Alum Bay, a Middle Eocene deposit; and the same botanist dwells on

* Principles of Geology, chap. xxvi, 9th ed., p. 432.

[ 544 ]

the tropical character of the flora of Monte Bolca and its distinctness from the sub-tropical flora of the Lower Miocene of Switzerland and Italy, in which last there is a far more considerable mixture of forms of a temperate climate, such as the willow, poplar, birch, elm, and others. That scarcely any one of the Monte Bolca fish should have been found in any other locality in Europe, is a striking illustration of the extreme imperfection of the palæontological record. We are in the habit of imagining that our insight into the geology of the Eocene period is more than usually perfect, and we are certainly acquainted with an almost unbroken succession of assemblages of shells passing one into the other from the era of the Thanet sands to that of the Bembridge beds or Paris gypsum. The general dearth, therefore, of fish in the different members of the Eocene series, Upper, Middle, and Lower, might induce a hasty reasoner to conclude that there was a poverty of ichthyic forms during this period; but when a local accident, like the volcanic eruptions of Monte Bolca, occurs, proofs are suddenly revealed to us of the richness and variety of this great class of vertebrata in the Eocene sea. The number of genera of Monte Bolca fish is, according to Agassiz, no less than seventy-five, twenty of them peculiar to that locality, and only eight common to the antecedent Cretaceous period. No less than forty-seven out of the seventy-five genera make their appearance for the first time in the Monte Bolca rocks, none of them having been met with as yet in the antecedent formations. They form a great contrast to the fish of the secondary strata, as, with the exception of the Placoids, they are all Teleosteans, only one genus, Pycnodus, belonging to the order of Ganoids, which form, as before stated, the vast majority of the ichthyolites entombed in the secondary are Mesozoic rocks.

Cretaceous Period.—M. Virlet, in his account of the geology of the Morea, p. 205, has clearly shown that certain traps in Greece are of Cretaceous date; as those, for example, which alternate conformably with cretaceous limestone and greensand between Kastri and Damala, in the Morea. They consist in great part of diallage rocks and serpentine, and of an amygdaloid with calcareous kernels, and a base of serpentine. In certain parts of the Morea, the age of these volcanic rocks is established by the following proofs: first, the lithographic limestones of the Cretaceous era are cut through by trap, and then a conglomerate occurs, at Nauplia and other places, containing in its calcareous cement many well-known fossils of the chalk and greensand, together with pebbles

[ 545 ]

ormed of rolled pieces of the same serpentinous trap, which appear in the dikes above alluded to.

Period of Oolite and Lias.—Although the green and serpentinous trap-rocks of the Morea belong chiefly to the Cretaceous era, as before mentioned, yet it seems that some eruptions of similar rocks began during the Oolitic period;* and it is probable that a large part of the trappean masses, called ophiolites in the Apennines, and associated with the limestone of that chain, are of corresponding age.

Trap of the New Red Sandstone Period.—In the southern part of Devonshire, trappean rocks are associated with New Red Sandstone, and, according to Sir H. De la Beche, have not been intruded subsequently into the sandstone, but were produced by contemporaneous volcanic action. Some beds of grit, mingled with ordinary red marl, resemble sands ejected from a crater; and in the stratified conglomerates occurring near Tiverton are many angular fragments of trap porphyry, some of them one or two tons in weight, intermingled with pebbles of other rocks. These angular fragments were probably thrown out from volcanic vents, and fell upon sedimentary matter then in the course of deposition.†

Trap of the Permian Period.—The recent investigations of Mr. Archibald Geikie in Ayrshire have shown that some of the volcanic rocks in that county are of Permian age, and it appears highly probable that the uppermost portion of Arthur’s Seat in the suburbs of Edinburgh marks the site of an eruption of the same era.

Trap of the Carboniferous Period.—Two classes of contemporaneous trap-rocks occur in the coal-field of the Forth, in Scotland. The newest of these, connected with the higher series of coal-measures, is well exhibited along the shores of the Forth, in Fifeshire, where they consist of basalt with olivine, amygdaloid, greenstone, wacke, and tuff. They appear to have been erupted while the sedimentary strata were in a horizontal position, and to have suffered the same dislocations which those strata have subsequently undergone. In the volcanic tuffs of this age are found not only fragments of limestone, shale, flinty slate, and sandstone, but also pieces of coal. The other or older class of carboniferous traps are traced along the south margin of Stratheden, and constitute a ridge parallel with the Ochils, and extending from Stirling to near St. Andrews. They consist almost exclusively of greenstone, becoming, in a few instances, earthy and amygdaloidal. They are regularly interstratified with the

* Boblaye and Virlet, Morea, p. 23.
† De la Beche, Geol. Proceedings, vol. ii, p. 198.

[ 546 ]

sandstone, shale, and iron-stone of the lower coal-measures, and, on the East Lomond, with Mountain Limestone. I examined these trap-rocks in 1838, in the cliffs south of St. Andrews, where they consist in great part of stratified tuffs, which are curved, vertical, and contorted, like the associated coal-measures. In the tuff I found fragments of carboniferous shale and limestone, and intersecting veins of greenstone.

Fife—Flisk Dike.—A trap dike was pointed out to me by Dr. Fleming, in the parish of Flisk, in the northern part of the county of Fife, which cuts through the grey sandstone and shale, forming the lowest part of the Old Red Sandstone, but which may probably be of carboniferous date. It may be traced for many miles, passing through the amygdaloidal and other traps of the hill called Norman’s Law in that parish. In its course it affords a good exemplification of the passage from the trappean into the Plutonic, or highly crystalline texture. Professor Gustavus Rose, to whom I submitted specimens of this dike, found it to be dolerite, and composed of greenish black augite and Labrador feldspar, the latter being the most abundant ingredient. A small quantity of magnetic iron, perhaps titaniferous, is also present. The result of this analysis is interesting, because both the ancient and modern lavas of Etna consist in like manner of augite, Labradorite, and titaniferous iron.

Erect Trees buried in Volcanic Ash at Arran.—An interesting discovery was made in 1867 by Mr. E. A. Wünsch in the carboniferous strata of the north-eastern part of the island of Arran. In the sea-cliff about five miles north of Corrie, near the village of Laggan, strata of volcanic ash occur, forming a solid rock cemented by carbonate of lime and enveloping trunks of trees, determined by Mr. Binney to belong to the genera Sigillaria and Lepidodendron. Some of these trees are at right angles to the planes of stratification, while others are prostrate and accompanied by leaves and fruits of the same genera. I visited the spot in company with Mr. Wünsch in 1870, and saw that the trees with their roots, of which about fourteen had been observed, occur at two distinct levels in volcanic tuffs parallel to each other, and inclined at an angle of about 40°, having between them beds of shale and coaly matter seven feet thick. It is evident that the trees were overwhelmed by a shower of ashes from some neighbouring volcanic vent, as Pompeii was buried by matter ejected from Vesuvius. The trunks, several of them from three to five feet in circumference, remained with their Stigmarian roots spreading through the stratum below, which had served as a soil. The trees must have continued for

[ 547 ]

years in an upright position after they were killed by the shower of burning ashes, giving time for a partial decay of the interior, so as to afford hollow cylinders into which the spores of plants were wafted. These spores germinated and grew, until finally their stems were petrified by carbonate of lime like some of the remaining portions of the wood of the containing Sigillaria. Mr. Carruthers has discovered that sometimes the plants which had thus grown and become fossil in the inside of a single trunk belonged to several distinct genera. The fact that the tree-bearing deposits now dip at an angle of 40° is the more striking, as they must clearly have remained horizontal and undisturbed during a long period of intermittent and contemporaneous volcanic action.

In some of the associated carboniferous shales, ferns and calamites occur, and all the phenomena of the successive buried forests remind us of the sections in pp. 410 and 411 of the Nova Scotia coal-measures, with this difference only, that in the case of the South Joggins the fossilisation of the trees was effected without the eruption of volcanic matter.

Trap of the Old Red Sandstone Period.—By referring to the section explanatory of the structure of Forfarshire, already given (p. 74), the reader will perceive that beds of conglomerate, No. 3, occur in the middle of the Old Red Sandstone system, 1, 2, 3, 4. The pebbles in these conglomerates are sometimes composed of granitic and quartzose rocks, sometimes exclusively of different varieties of trap, which last, although purposely omitted in the section referred to, is often found either intruding itself in amorphous masses and dikes into the old fossiliferous tilestones, No. 4, or alternating with them in conformable beds. All the different divisions of the red sandstone, 1, 2, 3, 4, are occasionally intersected by dikes, but they are very rare in Nos. 1 and 2, the upper members of the group consisting of red shale and red sandstone. These phenomena, which occur at the foot of the Grampians, are repeated in the Sidlaw Hills; and it appears that in this part of Scotland volcanic eruptions were most frequent in the earlier part of the Old Red Sandstone period. The trap-rocks alluded to consist chiefly of feldspathic porphyry and amygdaloid, the kernels of the latter being sometimes calcareous, often chalcedonic, and forming beautiful agates. We meet also with claystone, greenstone, compact feldspar, and tuff. Some of these rocks look as if they had flowed as lavas over the bottom of the sea, and enveloped quartz pebbles which were lying there, so as to form conglomerates with a base of greenstone, as is seen in Lumley Den, in the Sidlaw Hills. On either side of the axis of this chain of hills

[ 548 ]

(see Fig. 55), the beds of massive trap, and the tuffs composed of volcanic sand and ashes, dip regularly to the south-east or north-west, conformably with the shales and sandstones.

But the geological structure of the Pentland Hills, near Edinburgh, shows that igneous rocks were there formed during the newer part of the Devonian or “Old Red” period. These hills are 1900 feet high above the sea, and consist of conglomerates and sandstones of Upper Devonian age, resting on the inclined edges of grits and slates of Lower Devonian and Upper Silurian date. The contemporaneous volcanic rocks intercalated in this Upper Old Red consist of feldspathic lavas, or feldstones, with associated tuffs or ashy beds. The lavas were some of them originally compact, others vesicular, and these last have been converted into amygdaloids. They consist chiefly of feldstone or compact feldspar. The Pentland Hills, say Messrs. Maclaren and Geikie, afford evidence that at the time of the Upper Old Red Sandstone, the district to the south-west of Edinburgh was for a long while the seat of a powerful volcano, which sent out massive streams of lava and showers of ash, and continued active until well-nigh the dawn of the Carboniferous period.*

Silurian Volcanic Rocks.—It appears from the investigations of Sir R. Murchison in Shropshire, that when the Lower Silurian strata of that country were accumulating, there were frequent volcanic eruptions beneath the sea; and the ashes and scoriæ then ejected gave rise to a peculiar kind of tufaceous sandstone or grit, dissimilar to the other rocks of the Silurian series, and only observable in places where syenitic and other trap-rocks protrude. These tuffs occur on the flanks of the Wrekin and Caer Caradoc, and contain Silurian fossils, such as casts of encrinites, trilobites, and mollusca. Although fossiliferous, the stone resembles a sandy claystone of the trap family.†

Thin layers of trap, only a few inches thick, alternate in some parts of Shropshire and Montgomeryshire with sedimentary strata of the Lower Silurian system. This trap consists of slaty porphyry and granular feldspar rock, the beds being traversed by joints like those in the associated sandstone, limestone, and shale, and having the same strike and dip.‡

In Radnorshire there is an example of twelve bands of stratified trap, alternating with Silurian schists and flagstones,

* Maclaren, Geology of Fife and Lothians. Geikie, Trans. Royal Soc. Edinburgh, 1860-1861.
† Murchison, Silurian System, etc., p. 230.
‡ Ibid., p. 212.

[ 549 ]

in a thickness of 350 feet. The bedded traps consist of feldspar porphyry, and other varieties; and the interposed Llandeilo flags are of sandstone and shale, with trilobites and graptolites.*

The Snowdonian hills in Carnarvonshire consist in great part of volcanic tuffs, the oldest of which are interstratified with the Bala and Llandeilo beds. There are some contemporaneous feldspathic lavas of this era, which, says Professor Ramsay, alter the slates on which they repose, having doubtless been poured out over them, in a melted state, whereas the slates which overlie them having been subsequently deposited after the lava had cooled and consolidated, have entirely escaped alteration. But there are greenstones associated with the same formation, which, although they are often conformable to the slates, are in reality intrusive rocks. They alter the stratified deposits both above and below them, and when traced to great distances are sometimes seen to cut through the slates, and to send off branches. Nevertheless, these greenstones appear to belong, like the lavas, to the Lower Silurian period.

Cambrian Volcanic Rocks.—The Lingula beds in North Wales have been described as 5000 feet in thickness. In the upper portion of these deposits volcanic tuffs or ashy materials are interstratified with ordinary muddy sediment, and here and there associated with thick beds of feldspathic lava. These rocks form the mountains called the Arans and the Arenigs; numerous greenstones are associated with them, which are intrusive, although they often run in the lines of bedding for a space. “Much of the ash,” says Professor Ramsay, “seems to have been subaërial. Islands, like Graham’s Island, may have sometimes raised their craters for various periods above the water, and by the waste of such islands some of the ashy matter became waterworn, whence the ashy conglomerate. Viscous matter seems also to have been shot into the air as volcanic bombs, which fell among the dust and broken crystals (that often form the ashes) before perfect cooling and consolidation had taken place.”†

Laurentian Volcanic Rocks.—The Laurentian rocks in Canada, especially in Ottawa and Argenteuil, are the oldest intrusive masses yet known. They form a set of dikes of a fine-grained dark greenstone or dolerite, composed of feldspar and pyroxene, with occasional scales of mica and grains of pyrites. Their width varies from a few feet to a hundred yards, and they have a columnar structure, the columns

* Murchison, Silurian System, etc., p. 325.
† Quart. Geol. Journ., vol. ix, p. 170, 1852.

[ 550 ]

being truly at right angles to the plane of the dike. Some of the dikes send off branches. These dolerites are cut through by intrusive syenite, and this syenite, in its turn, is again cut and penetrated by feldspar porphyry, the base of which consists of petrosilex, or a mixture of orthoclase and quartz. All these trap-rocks appear to be of Laurentian date, as the Cambrian and Huronian rocks rest unconformably upon them.* Whether some of the various conformable crystalline rocks of the Laurentian series, such as the coarse-grained granitoid and porphyritic varieties of gneiss, exhibiting scarcely any signs of stratification, and some of the serpentines, may not also be of volcanic origin, is a point very difficult to determine in a region which has undergone so much metamorphic action.

* Logan, Geology of Canada, 1863.