(FIGURE 447. Diagram section from north, through Minudie, S. Joggins, Shoulie R. and Cobequid Mountains, south, showing the curvature and supposed denudation of the Carboniferous strata in Nova Scotia. A. Anticlinal axis of Minudie. B. Synclinal of Shoulie River. 1. Coal-measures. 2. Lower Carboniferous.)

The amount to which the bed of the sea sank down in order to allow of the formation of so vast a thickness of rock of sedimentary and organic origin is expressed by the total thickness of the Carboniferous strata, including the coal-measures, No. 1, and the rocks which underlie them, No. 2, Figure 447.

After the strata No. 2 had been elaborated, the conditions proper to a great delta exclusively prevailed, the subsidence still continuing so that one forest after another grew and was submerged until their under-clays with roots, and usually seams of coal, were left at more than eighty distinct levels. Here and there, also, deposits bearing testimony to the existence of fresh or brackish- water lagoons, filled with calcareo-bituminous mud, were formed. In these beds (h and i, Figure 439) are found fresh-water bivalves or mussels allied to Anodon, though not identical with that or any living genus, and called Naiadites carbonarius by Dawson. They are associated with small entomostracous crustaceans of the genus Cythere, and scales of small fishes. Occasionally some of the calamite brakes and forests of Sigillariae and Coniferae were exposed in the flood season, or sometimes, perhaps, by slight elevatory movements to the denuding action of the river or the sea.

In order to interpret the great coast section exposed to view on the shores of the Bay of Fundy, the student must, in the first place, understand that the newest or last-mentioned coal formations would have been the only ones known to us (for they would have covered all the others), had there not been two great movements in opposite directions, the first consisting of a general sinking of three miles, which took place during the Carboniferous Period, and the second an upheaval of more limited horizontal extent, by which the anticlinal axis A was formed. That the first great change of level was one of subsidence is proved by the fact that there are shallow-water deposits at the base of the Carboniferous series, or in the lowest beds of No. 2.

Subsequent movements produced in the Nova Scotia and the adjoining New Brunswick coal-fields the usual anticlinal and synclinal flexures. In order to follow these, we must survey the country for about thirty miles round the South Joggins, or the region where the erect trees described in the foregoing pages are seen. As we pass along the cliffs for miles in a southerly direction, the beds containing these fossil trees, which were mentioned as dipping about 18 degrees south, are less and less inclined, until they become nearly horizontal in the valley of a small river called the Shoulie, as ascertained by Dr. Dawson. After passing this synclinal line the beds begin to dip in an opposite or north- easterly direction, acquiring a steep dip where they rest unconformably on the edges of the Upper Silurian strata of the Cobequid Hills, as shown in Figure 447. But if we travel northward towards Minudie from the region of the coal- seams and buried forests, we find the dip of the coal-strata increasing from an angle of 18 degrees to one of more than 40 degrees, lower beds being continually exposed to view until we reach the anticlinal axis A and see the lower Carboniferous formation, No. 2, at the surface. The missing rocks removed by denudation are expressed by the faint lines at A, and thus the student will see that, according to the principles laid down in the seventh chapter, we are enabled, by the joint operations of upheaval and denudation, to look, as it were, about three miles into the interior of the earth without passing beyond the limits of a single formation.

CHAPTER XXIV.

FLORA AND FAUNA OF THE CARBONIFEROUS PERIOD.

Vegetation of the Coal Period. Ferns, Lycopodiaceae, Equisetaceae, Sigillariae, Stigmariae, Coniferae. Angiosperms. Climate of the Coal Period. Mountain Limestone. Marine Fauna of the Carboniferous Period. Corals. Bryozoa, Crinoidea. Mollusca. Great Number of fossil Fish. Foraminifera.

VEGETATION OF THE COAL PERIOD.

In the last chapter we have seen that the seams of coal, whether bituminous or anthracitic, are derived from the same species of plants, and Goppert has ascertained that the remains of every family of plants scattered through the shales and sandstones of the coal-measures are sometimes met with in the pure coal itself— a fact which adds greatly to the geological interest of this flora.

The coal-period was called by Adolphe Brongniart the age of Acrogens, so great appears to have been the numerical preponderance of flowerless or cryptogamic plants of the families of ferns, club-mosses, and horse-tails. (For botanical nomenclature see Chapter 17.) He reckoned the known species in 1849 at 500, and the number has been largely increased by recent research in spite of reductions owing to the discovery that different parts of even the same plants had been taken for distinct species. Notwithstanding these changes, Brongniart's generalisation concerning this flora still holds true, namely, that the state of the vegetable world was then extremely different from that now prevailing, not only because the cryptogamous plants constituted nearly the whole flora, but also because they were, on the whole, more highly developed than any belonging to the same class now existing, and united some forms of structure now only found separately and in distinct orders. The only phaenogamous plants were constitute any feature in the coal are the coniferae; monocotyledonous angiosperms appear to have been very rare, and the dicotyledonous, with one or two doubtful exceptions, were wanting. For this we are in some measure prepared by what we have seen of the Secondary or Mesozoic floras if, consistently with the belief in the theory of evolution, we expect to find the prevalence of simpler and less specialised organisms in older rocks.

FERNS.

(FIGURE 448. Pecopteris elliptica, Bunbury. (Sir C. Bunbury Quarterly Geological Journal volume 2 1845.) Frostburg.)

We are struck at the first glance with the similarity of the ferns to those now living. In the fossil genus Pecopteris, for example (Figure 448), it is not easy to decide whether the fossils might not be referred to the same genera as those established for living ferns; whereas, in regard to some of the other contemporary families of plants, with the exception of the fir tribe, it is not easy to guess even the class to which they belong. The ferns of the Carboniferous period are generally without organs of fructification, but in the few instances in which these do occur in a fit state for microscopical investigations they agree with those of the living ferns.

(FIGURE 449. Caulopteris primaeva, Lindley.)

When collecting fossil specimens from the coal-measures of Frostburg, in Maryland, I found in the iron-shales several species with well-preserved rounded spots or marks of the sori (see Figure 448). In the general absence of such characters they have been divided into genera distinguished chiefly by the branching of the fronds and the way in which the veins of the leaves are disposed. The larger portion are supposed to have been of the size of ordinary European ferns, but some were decidedly arborescent, especially the group called Caulopteris (see Figure 449) by Lindley, and the Psaronius of the upper or newest coal-measures, before alluded to (Chapter 22). All the recent tree-ferns belong to one tribe (Polypodiaceae), and to a small number only of genera in that tribe, in which the surface of the trunk is marked with scars, or cicatrices, left after the fall of the fronds. These scars resemble those of Caulopteris.

(FIGURES 450, 451 and 452. Living tree-ferns of different genera. (Ad. Brong.)

(FIGURE 450. Tree-fern from Isle of Bourbon.)

(FIGURE 451. Cyathea glauca, Mauritius.)

(FIGURE 452. Tree-fern from Brazil.))

No less than 130 species of ferns are enumerated as having been obtained from the British coal-strata, and this number is more than doubled if we include the Continental and American species. Even if we make some reduction on the ground of varieties which have been mistaken, in the absence of their fructification, for species, still the result is singular, because the whole of Europe affords at present no more than sixty-seven indigenous species.

LYCOPODIACEAE— LEPIDODENDRON.

(FIGURES 453, 454 and 455. Lepidodendron Sternbergii. Coal-measures, near Newcastle.

(FIGURE 453. Branching trunk, 49 feet long, supposed to have belonged to L. Sternbergii. (Foss. Flo. 203.))

(FIGURE 454. Branching stem with bark and leaves of L. Sternbergii. (Foss. Flo. 4.)

(FIGURE 455. Portion of same nearer the root. Natural size. (Ibid.)))

(FIGURE 456. Lycopodium densum. a. Living species. New Zealand. b. Branch; natural size. c. Part of same, magnified.)

About forty species of fossil plants of the Coal have been referred to this genus, more than half of which are found in the British coal-measures. They consist of cylindrical stems or trunks, covered with leaf-scars. In their mode of branching, they are always dichotomous (see Figure 454). They belong to the Lycopodiaceae, bearing sporangia and spores similar to those of the living representatives of this family (Figure 457); and although most of the Carboniferous species grew to the size of large trees, Mr. Carruthers has found by careful measurement that the volume of the fossil spores did not exceed that of the recent club-moss, a fact of some geological importance, as it may help to explain the facility with which these seeds may have been transported by the wind, causing the same wide distribution of the species of the fossil forests in Europe and America which we now observe in the geographical distribution of so many living families of cryptogamous plants. The Figures 453-455 represent a fossil Lepidodendron, 49 feet long, found in Jarrow Colliery, near Newcastle, lying in shale parallel to the planes of stratification. Fragments of others, found in the same shale, indicate, by the size of the rhomboidal scars which cover them, a still greater magnitude. The living club-mosses, of which there are about 200 species, are most abundant in tropical climates. They usually creep on the ground, but some stand erect, as the Lycopodium densum from New Zealand (see Figure 456), which attains a height of three feet.

(FIGURE 457. Lepidostrobus ornatus, Brong. Shropshire. a. (Body) half natural size. b. Portion of a section, showing the large sporangia in their natural position, and each supported by its bract or scale. c. Spores in these sporangia, highly magnified. (Hooker Mem. Geological Survey volume 2 part 2 page 440.)

In the Carboniferous strata of Coalbrook Dale, and in many other coal-fields, elongated cylindrical bodies, called fossil cones, named Lepidostrobus by M. Adolphe Brongniart, are met with. (See Figure 457.) They often form the nucleus of concretionary balls of clay-ironstone, and are well preserved, exhibiting a conical axis, around which a great quantity of scales were compactly imbricated. The opinion of M. Brongniart that the Lepidostrobus is the fruit of Lepidodendron has been confirmed, for these strobili or fruits have been found terminating the tip of a branch of a well-characterised Lepidodendron in Coalbrook Dale and elsewhere.

EQUISETACEAE.

(FIGURE 458. Calamites Sucowii, Brong.; natural size. Common in coal throughout Europe.)

(FIGURE 459. Stem of Figure 458, as restored by Dr. Dawson.)

(FIGURE 460. Radical termination of a Calamite. Nova Scotia.)

To this family belong two fossil genera of the coal, Equisetites and Calamites. The Calamites were evidently closely related to the modern horse-tails (Equiseta) differing principally in their great size, the want of sheaths at the joints, and some details of fructification. They grew in dense brakes on sandy and muddy flats in the manner of modern Equisetaceae, and their remains are frequent in the coal. Seven species of this plant occur in the great Nova Scotia section before described, where the stems of some of them five inches in diameter, and sometimes eight feet high, may be seen terminating downward in a tapering root (see Figure 460).

(FIGURE 461. Asterophillites foliosus. (Foss. Flo. 25.) Coal-measures, Newcastle.)

(FIGURE 462. Annularia sphenophylloides, Dawson.)

(FIGURE 463. Sphenophyllum erosum, Dawson.)

Botanists are not yet agreed whether the Asterophyllites, a species of which is represented in Figure 461, can form a separate genus from the Calamite, from which, however, according to Dr. Dawson, its foliage is distinguished by a true mid-rib, which is wanting in the leaves known to belong to some Calamites. Figures 462 and 463 represent leaves of Annularia and Sphenophyllum, common in the coal, and believed by Mr. Carruthers to be leaves of Calamites. Dr. Williamson, who has carefully studied the Calamites, thinks that they had a fistular pith, exogenous woody stem, and thick smooth bark, which last having always disappeared, leaves a fluted stem, as represented in Figure 459.

SIGILLARIA.

(FIGURE 464. Sigillaria laevigata, Brong.)

A large portion of the trees of the Carboniferous period belonged to this genus, of which as many as 28 species are enumerated as British. The structure, both internal and external, was very peculiar, and, with reference to existing types, very anomalous. They were formerly referred, by M. Ad. Brongniart, to ferns, which they resemble in the scalariform texture of their vessels and, in some degree, in the form of the cicatrices left by the base of the leaf-stalks which have fallen off (see Figure 464). But some of them are ascertained to have had long linear leaves, quite unlike those of ferns. They grew to a great height, from 30 to 60, or even 70 feet, with regular cylindrical stems, and without branches, although some species were dichotomous towards the top. Their fluted trunks, from one to five feet in diameter, appear to have decayed more rapidly in the interior than externally, so that they became hollow when standing; and when thrown prostrate, they were squeezed down and flattened. Hence, we find the bark of the two opposite sides (now converted into bright shining coal) constitute two horizontal layers, one upon the other, half an inch, or an inch, in their united thickness. These same trunks, when they are placed obliquely or vertically to the planes of stratification, retain their original rounded form, and are uncompressed, the cylinder of bark having been filled with sand, which now affords a cast of the interior.

Dr. Hooker inclined to the belief that the Sigillariae may have been cryptogamous, though more highly developed than any flowerless plants now living. Dr. Dawson having found in some species what he regards as medullary rays, thinks with Brongniart that they have some relation to gymnogens, while Mr. Carruthers leans to the opinion that they belong to the Lycopodiaceae.

STIGMARIA.

(FIGURE 465. Stigmaria attached to a trunk of Sigillaria.)

This fossil, the importance of which has already been pointed out in Chapter 23, was originally conjectured to be an aquatic plant. It is now ascertained to be the root of Sigillaria. The connection of the roots with the stem, previously suspected, on botanical grounds, by Brongniart, was first proved, by actual contact, in the Lancashire coal-field, by Mr. Binney. The fact has lately been shown, even more distinctly, by Mr. Richard Brown, in his description of the Stigmariae occurring in the under-clays of the coal-seams of the Island of Cape Breton, in Nova Scotia. In a specimen of one of these, represented in Figure 465, the spread of the roots was sixteen feet, and some of them sent out rootlets, in all directions, into the surrounding clay.

(FIGURE 466. Stigmaria ficoides, Brong. 1/4 natural size. (Foss. Flo. 32.))

(FIGURE 467. Stigmaria ficoides, Brong. Surface of another individual of same species, showing form of tubercles. (Foss. Flo. 34.))

In the sea-cliffs of the South Joggins in Nova Scotia, I examined several erect Sigillariae, in company with Dr. Dawson, and we found that from the lower extremities of the trunk they sent out Stigmariae as roots. All the stools of the fossil trees dug out by us divided into four parts, and these again bifurcated, forming eight roots, which were also dichotomous when traceable far enough. The cylindrical rootlets formerly regarded as leaves are now shown by more perfect specimens to have been attached to the root by fitting into deep cylindrical pits. In the fossil there is rarely any trace of the form of these cavities, in consequence of the shrinkage of the surrounding tissues. Where the rootlets are removed, nothing remains on the surface of the Stigmaria but rows of mammillated tubercles (see Figures 466, 467), which have formed the base of each rootlet. These protuberances may possibly indicate the place of a joint at the lower extremity of the rootlet. Rows of these tubercles are arranged spirally round each root, which have always a medullary axis and woody system much resembling that of Sigillaria, the structure of the vessels being, like it, scalariform.

CONIFERAE.

(FIGURE 468. Fragment of coniferous wood, Dadoxylon, of Endlicher, fractured longitudinally; from Coalbrook Dale. W.C. Williamson. (Manchester Philosophical Mem. volume 9 1851.) a. Bark. b. Woody zone or fibre (pleurenchyma). c. Medulla or pith. d. Cast of hollow pith or "Sternbergia.")

(FIGURE 469. Fragment of coniferous wood, Dadoxylon, of Endlicher. Magnified portion of Figure 468; transverse section. b-b. Woody fibre. c. Pith. e, e, e. Medullary rays.)

The coniferous trees of this period are referred to five genera; the woody structure of some of them showing that they were allied to the Araucarian division of pines, more than to any of our common European firs. Some of their trunks exceeded forty-four feet in height. Many, if not all of them, seem to have differed from living Coniferae in having large piths; for Professor Williamson has demonstrated the fossil of the coal-measures called Sternbergia to be the pith of these trees, or rather the cast of cavities formed by the shrinking or partial absorption of the original medullary axis (see Figures 468, 469). This peculiar type of pith is observed in living plants of very different families, such as the common Walnut and the White Jasmine, in which the pith becomes so reduced as simply to form a thin lining of the medullary cavity, across which transverse plates of pith extend horizontally, so as to divide the cylindrical hollow into discoid interspaces. When these interspaces have been filled up with inorganic matter, they constitute an axis to which, before their true nature was known, the provisional name of Sternbergia (d, d, Figure 468) was given. In the above specimen the structure of the wood (b, Figures 468 and 469) is coniferous, and the fossil is referable to Endlicher's fossil genus Dadoxylon.

(FIGURE 470. Trigonocarpum ovatum, Lindley and Hutton. Peel Quarry, Lancashire.)

(FIGURE 471. Trigonocarpum olivaeforme, Lindley, with its fleshy envelope. Felling Colliery, Newcastle.)

The fossil named Trigonocarpon (Figures 470 and 471), formerly supposed to be the fruit of a palm, may now, according to Dr. Hooker, be referred, like the Sternbergia, to the Coniferae. Its geological importance is great, for so abundant is it in the coal-measures, that in certain localities the fruit of some species may be procured by the bushel; nor is there any part of the formation where they do not occur, except the under-clays and limestone. The sandstone, ironstone, shales, and coal itself, all contain them. Mr. Binney has at length found in the clay-ironstone of Lancashire several specimens displaying structure, and from these, says Dr. Hooker, we learn that the Trigonocarpon belonged to that large section of existing coniferous plants which bear fleshy solitary fruits, and not cones. It resembled very closely the fruit of the Chinese genus Salisburia, one of the Yew tribe, or Taxoid conifers.

ANGIOSPERMS.

(FIGURE 472. Antholithes. Felling Colliery, Newcastle.)

The curious fossils called Antholithes by Lindley have usually been considered to be flower spikes, having what seems a calyx and linear petals (see Figure 472). Dr. Hooker, after seeing very perfect specimens, also thought that they resembled the spike of a highly-organised plant in full flower, such as one of the Bromeliaceae, to which Professor Lindley had at first compared them. Mr. Carruthers, who has lately examined a large series in different museums, considers it to be a dicotyledonous angiosperm allied to Orobanche (broom-rape), which grew, not on the soil, but parasitically on the trees of the coal forests.

(FIGURE 473. Pothocites Grantonii, Pat. Coal-measures, Edinburgh. c. Stem and spike; 1/2 natural size. b. Remains of the spathe magnified. c. Portion of spike magnified. d. One of the calyces magnified.)

In the coal-measures of Granton, near Edinburgh, a remarkable fossil (Figure 473) was found and described in 1840, by Dr. Robert Paterson. (Transactions of the Botanical Society of Edinburgh volume 1 1844.) It was compressed between layers of bituminous shale, and consists of a stem bearing a cylindrical spike, a, which in the portion preserved in the slate exhibits two subdivisions and part of a third. The spike is covered on the exposed surface with the four-cleft calyces of the flowers arranged in parallel rows. The stem shows, at b, a little below the spike, remains of a lateral appendage, which is supposed to indicate the beginning of the spathe. The fossil has been referred to the Aroidiae, and there is every probability that it is a true member of this order. There can at least be no doubt as to the high grade of its organisation, and that it belongs to the monocotyledonous angiosperms. Mr. Carruthers has carefully examined the original specimen in the Botanical Museum, Edinburgh, and thinks it may have been an epiphyte.

CLIMATE OF THE COAL PERIOD.

As to the climate of the Coal, the Ferns and the Coniferae are perhaps the two classes of plants which may be most relied upon as leading us to safe conclusions, as the genera are nearly allied to living types. All botanists admit that the abundance of ferns implies a moist atmosphere. But the coniferae, says Hooker, are of more doubtful import, as they are found in hot and dry, and in cold and dry climates; in hot and moist, and in cold and moist regions. In New Zealand the coniferae attain their maximum in numbers, constituting 1/62 part of all the flowering plants; whereas in a wide district around the Cape of Good Hope they do not form 1/1600 of the phenogamic flora. Besides the conifers, many species of ferns flourish in New Zealand, some of them arborescent, together with many lycopodiums; so that a forest in that country may make a nearer approach to the carboniferous vegetation than any other now existing on the globe.

MARINE FAUNA OF THE CARBONIFEROUS PERIOD.

It has already been stated that the Carboniferous or Mountain Limestone underlies the coal-measures in the South of England and Wales, whereas in the North, and in Scotland, marine calcareous rocks partly of the age of the Mountain Limestone alternate with shales and sandstones, containing seams of coal. In its most calcareous form the Mountain Limestone is destitute of land- plants, and is loaded with marine remains— the greater part, indeed, of the rock being made up bodily of crinoids, corals, and bryozoa with interspersed mollusca.

CORALS.

(FIGURE 474. Palaeozoic type of lamelliferous cup-shaped Coral. Order ZOANTHARIA RUGOSA, Milne Edwards and Jules Haime. a. Vertical section of Campophyllum flexuosum, (Cyathophyllum, Goldfuss); 1/2 natural size: from the Devonian of the Eifel. The lamellae are seen around the inside of the cup; the walls consist of cellular tissue; and large transverse plates, called tubulae, divide the interior into chambers. b. Arrangement of the lamellae in Polycoelia profunda, Germar, sp.; natural size: from the Magnesian Limestone, Durham. This diagram shows the quadripartite arrangement of the primary septa, characteristic of palaeozoic corals, there being four principal and eight intermediate lamellae, the whole number in this type being always a multiple of four. c. Stauria astraeiformis, Milne Edwards. Young group, natural size. Upper Silurian, Gothland. The lamellae or septal system in each cup are divided by four prominent ridges into four groups.)

(FIGURE 475. Neozoic type of lamelliferous cup-shaped Coral. Order ZOANTHARIA APOROSA, M. Edwards and J. Haime. a. Parasmilia centralis, Mantell, sp. Vertical section; natural size. Upper Chalk, Gravesend. In this type the lamellae are massive, and extend to the axis or columella composed of loose cellular tissue, without any transverse plates like those in Figure 474, a. b. Cyathina Bowerbankii, Ed. and H. Transverse section, enlarged. Gault, Folkestone. In this coral the primary septa are a multiple of six. The twelve principal plates reach the columella, and between each pair there are three secondaries, in all forty-eight. The short intermediate plates which proceed from the columella are not counted. They are called pali. c. Fungia patellaris, Lamarck. Recent; very young state. Diagram of its six primary and six secondary septa, magnified. The sextuple arrangement is always more manifest in the young than in the adult state.)

The corals deserve especial notice, as the cup-and-star corals, which have the most massive and stony skeletons, display peculiarities of structure by which they may be distinguished generally, as MM. Milne Edwards and Haime first pointed out, from all species found in strata newer than the Permian. There is, in short, an ancient or PALAEOZOIC, and a modern or NEOZOIC type, if, by the latter term, we designate (as proposed by Professor E. Forbes) all strata from the triassic to the most modern, inclusive. The accompanying diagrams (Figures 474, 475) may illustrate these types.

It will be seen that the more ancient corals have what is called a quadripartite arrangement of the chief plates or LAMELLAE— parts of the skeleton which support the organs of reproduction. The number of these lamellae in the Palaeozoic type is 4, 8, 16, etc.; while in the Neozoic type the number is 6, 12, 24, or some other multiple of six; and this holds good, whether they be simple forms, as in Figures 474, a, and 475, a, or aggregate clusters of corallites, as in 474, c. But further investigations have shown in this, as in all similar grand generalisations in natural history, that there are exceptions to the rule. Thus in the Lower Greensand Holocystis elegans (Ed. and H.) and other forms have the Palaeozoic type, and Dr. Duncan has shown to what extent the Neozoic forms penetrate downward into the Carboniferous and Devonian rocks.

(FIGURE 476. Lithostrotion basaltiforme, Phil. sp. (Lithostrotion striatum, Fleming; Astraea basaltiformis, Conyb. and Phill.). England, Ireland, Russia, Iowa, and westward of the Mississippi, United States. (D.D. Owen.)

(FIGURE 477. Lonsdaleia floriformis, Martin, sp., M. Edwards. (Lithostrotion floriforme, Fleming. Strombodes.) a. Young specimen, with buds or corallites on the disk, illustrating calicular gemmation. b. Part of a full-grown compound mass. Bristol, etc.; Russia.)

From a great number of lamelliferous corals met with in the Mountain Limestone, two species (Figures 476, 477) have been selected, as having a very wide range, extending from the eastern borders of Russia to the British Isles, and being found almost everywhere in each country. These fossils, together with numerous species of Zaphrentis, Amplexus, Cyathophyllum, Clisiophyllum, Syringopora, and Michelinia, form a group of rugose corals widely different from any that followed them. (For figures of these corals, see Palaeontographical Society's Monographs 1852.)

BRYOZOA AND CRINOIDEA.

(FIGURE 478. Cyathocrinus planus, Miller. Body and arms. Mountain Limestone.)

(FIGURE 479. Cyathocrinus caryocrinoides, M'Coy. a. Surface of one of the joints of the stem. b. Pelvis or body; called also calyx or cup. c. One of the pelvic plates.)

Of the Bryozoa, the prevailing forms are Fenestella, Hemitrypa, and Polypora, and these often form considerable beds. Their net-like fronds are easily recognised. Crinoidea are also numerous in the Mountain Limestone (see Figures 478, 479), two genera, Pentremites and Codonaster, being peculiar to this formation in Europe and North America.

(FIGURE 480. Palaechinus gigas, M'Coy. Reduced one-third. Mountain Limestone. Ireland.)

In the greater part of them, the cup or pelvis, Figure 479, b, is greatly developed in size in proportion to the arms, although this is not the case in Figure 478. The genera Poteriocrinus, Cyathocrinus, Pentremites, Actinocrinus, and Platycrinus, are all of them characteristic of this formation. Other Echinoderms are rare, a few Sea-Urchins only being known: these have a complex structure, with many more plates on their surface than are seen in the modern genera of the same group. One genus, the Palaechinus (Figure 480), is the analogue of the modern Echinus, but has four, five, or six rows of plates in the interambulacral region or area, whereas the modern genera have only two. The other, Archaeocidaris, represents, in like manner, the Cidaris of the present seas.

MOLLUSCA.

(FIGURE 481. Productus semireticulatus, Martin, sp. (P. antiquatus, Sowerby.) Mountain Limestone. England, Russia, the Andes, etc.)

(FIGURE 482. Spirifera trigonalis, Martin, sp. Mountain Limestone. Derbyshire, etc.)

(FIGURE 483. Spirifera glabra, Martin, sp. Mountain Limestone.)

The British Carboniferous mollusca enumerated by Mr. Etheridge comprise 653 species referable to 86 genera, occurring chiefly in the Mountain Limestone. (Quarterly Geological Journal volume 23 page 674 1867.) Of this large number only 40 species are common to the underlying Devonian rocks, 9 of them being Cephalopods, 7 Gasteropods, and the rest bivalves, chiefly Brachiopoda (or Palliobranchiates). This latter group constitutes the larger part of the Carboniferous Mollusca, 157 species being known in Great Britain alone, and it will be found to increase in importance in the fauna of the primary rocks the lower we descend in the series. Perhaps the most characteristic shells of the formation are large species of Productus, such as P. giganteus, p. hemisphericus, P. semireticulatus (Figure 481), and P. scabriculus. Large plaited spirifers, as Spirifera striata, S. rotundata, and S. trigonalis (Figure 482), also abound; and smooth species, such as Spirifera glabra (Figure 483), with its numerous varieties.

(FIGURE 484. Terebratula hastata, Sowerby, with radiating bands of colour. Mountain Limestone. Derbyshire, Ireland, Russia, etc.)

(FIGURE 485. Aviculopecten sublobatus, Phill. Mountain Limestone. Derbyshire, Yorkshire.)

(FIGURE 486. Pleurotomaria carinata, Sowerby. (P. flammigera, Phillips). Mountain Limestone. Derbyshire, etc.)

Among the brachiopoda, Terebratula hastata (Figure 484) deserves mention, not only for its wide range, but because it often retains the pattern of the original coloured stripes which ornamented the living shell. These coloured bands are also preserved in several lamellibranchiate bivalves, as in Aviculopecten (Figure 485), in which dark stripes alternate with a light ground. In some also of the spiral univalves the pattern of the original painting is distinctly retained, as in Pleurotomaria (Figure 486), which displays wavy blotches, resembling the colouring in many recent trochidae.

(FIGURE 487. Euomphalus pentangulatus, Sowerby. Mountain Limestone. a. Upper side. b. Lower or umbilical side. c. View showing mouth, which is less pentagonal in older individuals. d. View of polished section, showing internal chambers.)

Some few of the carboniferous mollusca, such as Avicula, Nucula (sub-genus Ctenodonta), Solemya, and Lithodomus, belong no doubt to existing genera; but the majority, though often referred to as living types, such as Isocardia, Turritella, and Buccinum, belong really to forms which appear to have become extinct at the close of the Palaeozoic epoch. Euomphalus is a characteristic univalve shell of this period. In the interior it is divided into chambers (Figure 487, d), the septa or partitions not being perforated as in foraminiferous shells, or in those having siphuncles, like the Nautilus. The animal appears to have retreated at different periods of its growth from the internal cavity previously formed, and to have closed all communication with it by a septum. The number of chambers is irregular, and they are generally wanting in the innermost whorl. The animal of the recent Turritella communis partitions off in like manner as it advances in age a part of its spire, forming a shelly septum.

(FIGURE 488. Bellerophon costatus, Sowerby. Mountain Limestone.)

More than twenty species of the genus Bellerophon (see Figure 488), a shell like the living Argonaut without chambers, occur in the Mountain Limestone. The genus is not met with in strata of later date. It is most generally regarded as belonging to the pelagic Nucleobranchiata and the family Atlantidae, partly allied to the Glass-Shell, Carinaria; but by some few it is thought to be a simple form of Cephalopod.

(FIGURE 489. Portion of Orthoceras laterale. Phill. Mountain Limestone.)

(FIGURE 490. Goniatites crenistra, Phillips. Mountain Limestone. North America, Britain, Germany, etc. a. Lateral view. b. Front view, showing the mouth.)

The carboniferous Cephalopoda do not depart so widely from the living type (the Nautilus) as do the more ancient Silurian representatives of the same order; yet they offer some remarkable forms. Among these is Orthoceras, a siphuncled and chambered shell, like a Nautilus uncoiled and straightened (Figure 489). Some species of this genus are several feet long. The Goniatite is another genus, nearly allied to the Ammonite, from which it differs in having the lobes of the septa free from lateral denticulations, or crenatures; so that the outline of these is angular, continuous, and uninterrupted. The species represented in Figure 490 is found in most localities, and presents the zigzag character of the septal lobes in perfection. The dorsal position of the siphuncle, however, clearly distinguishes the Goniatite from the Nautilus, and proves it to have belonged to the family of the Ammonites, from which, indeed, some authors do not believe it to be generically distinct.

FOSSIL FISH.

(FIGURE 491. Psammodus porosus, Agassiz. Bone-bed, Mountain Limestone. Bristol, Armagh.)

(FIGURE 492. Cochliodus contortus, Agassiz. Bone-bed, Mountain Limestone. Bristol, Armagh.)

The distribution of these is singularly partial; so much so, that M. De Koninck of Liege, the eminent palaeontologist, once stated to me that, in making his extensive collection of the fossils of the Mountain Limestone of Belgium, he had found no more than four or five examples of the bones or teeth of fishes. Judging from Belgian data, he might have concluded that this class of vertebrata was of extreme rarity in the Carboniferous seas; whereas the investigation of other countries has led to quite a different result. Thus, near Clifton, on the Avon, as well as at numerous places around the Bristol basin from the Mendip Hills to Tortworth, there is a celebrated "bone-bed," almost entirely made up of ichthyolites. It occurs at the base of the Lower Limestone shales immediately resting upon the passage beds of the Old Red Sandstone. Similar bone-beds occur in the Carboniferous Limestone of Armagh, in Ireland, where they are made up chiefly of the teeth of fishes of the Placoid order, nearly all of them rolled as if drifted from a distance. Some teeth are sharp and pointed, as in ordinary sharks, of which the genus Cladodus afford an illustration; but the majority, as in Psammodus and Cochliodus, are, like the teeth of the Cestracion of Port Jackson (see Figure 261), massive palatal teeth fitted for grinding. (See Figures 491, 492.)

There are upward of seventy other species of fossil fish known in the Mountain Limestone of the British Islands. The defensive fin-bones of these creatures are not infrequent at Armagh and Bristol; those known as Oracanthus, Ctenocanthus, and Onchus are often of a very large size. Ganoid fish, such as Holoptychius, also occur; but these are far less numerous. The great Megalichthys Hibberti appears to range from the Upper Coal-measures to the lowest Carboniferous strata.

FORAMINIFERA.

(FIGURE 493. Fusulina cylindrica, d'Orbigny. Magnified 3 diameters. Mountain Limestone.)

In the upper part of the Mountain Limestone group in the south-west of England, near Bristol, limestones having a distinct oolitic structure alternate with shales. In these rocks the nucleus of every minute spherule is seen, under the microscope, to consist of a small rhizopod or foraminifer. This division of the lower animals, which is represented so fully at later epochs by the Nummulites and their numerous minute allies, appears in the Mountain Limestone to be restricted to a very few species, among which Textularia, Nodosaria, Endothyra, and Fusulina (Figure 493), have been recognised. The first two genera are common to this and all the after periods; the third has been found in the Upper Silurian, but is not known above the Carboniferous strata; the fourth (Figure 493) is characteristic of the Mountain Limestone in the United States, Arctic America, Russia, and Asia Minor, but is also known in the Permian.

CHAPTER XXV.

DEVONIAN OR OLD RED SANDSTONE GROUP.

Classification of the Old Red Sandstone in Scotland and in Devonshire. Upper Old Red Sandstone in Scotland, with Fish and Plants. Middle Old Red Sandstone. Classification of the Ichthyolites of the Old Red, and their Relation to Living Types. Lower Old Red Sandstone, with Cephalaspis and Pterygotus. Marine or Devonian Type of Old Red Sandstone. Table of Devonian Series. Upper Devonian Rocks and Fossils. Middle. Lower. Eifel Limestone of Germany. Devonian of Russia. Devonian Strata of the United States and Canada. Devonian Plants and Insects of Canada.

CLASSIFICATION OF THE TWO TYPES OF OLD RED SANDSTONE.

We have seen that the Carboniferous strata are surmounted by the Permian and Trias, both originally included in England under the name "New Red Sandstone," from the prevailing red colour of the strata. Under the coal came other red sandstones and shales which were distinguished by the title of "Old Red Sandstone." Afterwards the name of "Devonian" was given by Sir R. Murchison and Professor Sedgwick to marine fossiliferous strata which, in the south of England, occupy a similar position between the overlying coal and the underlying Silurian formations.

It may be truly said that in the British Isles the rocks of this age present themselves in their mineral aspect, and even to some extent in their fossil contents, under two very different forms; the one as distinct from the other as are often lacustrine or fluviatile from marine strata. It has indeed been suggested that by far the greater part of the deposits belonging to what may be termed the Old Red Sandstone type are of fresh-water origin. The number of land- plants, the character of the fishes, and the fact that the only shell yet discovered belongs to the genus Anodonta, must be allowed to lend no small countenance to this opinion. In this case the difficulty of classification when the strata of this type are compared in different regions, even where they are contiguous, may arise partly from their having been formed in distinct hydrographical basins, or in the neighbourhood of the land in shallow parts of the sea into which large bodies of fresh-water entered, and where no marine mollusca or corals could flourish. Under such geographical conditions the limited extent of some kinds of sediment, as well as the absence of those marine forms by which we are able to identify or contrast marine formations, may be explained, while the great thickness of the rocks, which might seem at first sight to require a corresponding depth of water, can often be shown to have been due to the gradual sinking down of the bottom of the estuary or sea where the sediment was accumulated.

Another active cause of local variation in Scotland was the frequency of contemporaneous volcanic eruptions; some of the rocks derived from this source, as between the Grampians and the Tay, having formed islands in the sea, and having been converted into shingle and conglomerate, before the upper portions of the red shales and sandstones were superimposed.

The dearth of calcareous matter over wide areas is characteristic of the Old Red Sandstone. This is, no doubt, in great part due to the absence of shells and corals; but why should these be so generally wanting in all sedimentary rocks the colour of which is determined by the red oxide of iron? Some geologists are of opinion that the waters impregnated with this oxide were prejudicial to living beings, others that strata permeated with this oxide would not preserve such fossil remains.

In regard to the two types, the Old Red Sandstone and the Devonian, I shall first treat of them separately, and then allude to the proofs of their having been to a great extent contemporaneous. That they constitute a series of rocks intermediate in date between the lowest Carboniferous and the uppermost Silurian is not disputed by the ablest geologists; and it can no longer be contended that the Upper, Middle, and Lower Old Red Sandstone preceded in date the three divisions to which, by aid of the marine shells, the Devonian rocks have been referred, while, on the other hand, we have not yet data for enabling us to affirm to what extent the subdivisions of the one series may be the equivalents in time of those of the other.

UPPER OLD RED SANDSTONE.

(FIGURE 494. Anodonta Jukesii, Forbes. Upper Devonian, Kiltorkan, Ireland.)

(FIGURE 495. Bifurcating branch of Lepidodendron Griffithsii, Brongn. Upper Devonian, Kilkenny.)

(FIGURE 496. Palaeopteris Hibernica, Schimp. (Cyclopteris Hibernica), Edward Forbes (Adiantites, Gop.). Upper Devonian, Kilkenny.)

The highest beds of the series in Scotland, lying immediately below the coal in Fife, are composed of yellow sandstone well seen at Dura Den, near Coupar, in Fife, where, although the strata contain no mollusca, fish have been found abundantly, and have been referred to the genera Holoptychius, Pamphractus, Glyptopomus, and many others. In the county of Cork, in Ireland, a similar yellow sandstone occurs containing fish of genera characteristic of the Scotch Old Red Sandstone, as for example Coccosteus (a form represented by many species in the Old Red Sandstone and by one only in the Carboniferous group), and Glytolepis and Asterolepis, both exclusively confined to the "Old Red." In the same Irish sandstone at Kiltorkan has been found an Anodonta or fresh-water mussel, the only shell hitherto discovered in the Old Red Sandstone of the British Isles (see Figure 494). In the same formation are found the fern (Figure 496) and the Lepidodendron (Figure 495), and other species of plants, some of which, Professor Heer remarks, agree specifically with species from the lower carboniferous beds. This induces him to lean to the opinion long ago advocated by Sir Richard Griffiths, that the yellow sandstone, in spite of its fish remains, should be classed as Lower Carboniferous, an opinion which I am not yet prepared to adopt. Between the Mountain Limestone and the yellow sandstone in the south-west of Ireland there intervenes a formation no less than 5000 feet thick, called the "Carboniferous slate," and at the base of this, in some places, are local deposits, such as the Glengariff Grits, which appear to be beds of passage between the Carboniferous and Old Red Sandstone groups.

It is a remarkable result of the recent examination of the fossil flora of Bear Island, latitude 74 degrees 30' N., that Professor Heer has described as occurring in that part of the Arctic region (nearly twenty-six degrees to the north of the Irish locality) a flora agreeing in several of its species with that of the yellow sandstones of Ireland. This Bear Island flora is believed by Professor Heer to comprise species of plants some of which ascend even to the higher stages of the European Carboniferous formation, or as high as the Mountain Limestone and Millstone Grit. Palaeontologists have long maintained that the same species which have a wide range in space are also the most persistent in time, which may prepare us to find that some plants having a vast geographical range may also have endured from the period of the Upper Devonian to that of the Millstone Grit.

(FIGURE 497. Scale of Holoptychius nobilissimus, Agassiz. Clashbinnie. 1/2 natural size.)

(FIGURE 498. Holoptychius, as restored by Professor Huxley. a. The fringed pectoral fins. b. The fringed ventral fins. c. Anal fin. d, e. Dorsal fins.)

Outliers of the Upper "Old Red" occur unconformably on older members of the group, and the formation represented at Whiteness, near Arbroath, a, Figure 55, may probably be one of these outliers, though the want of organic remains renders this uncertain. It is not improbable that the beds given in this section as Nos. 1, 2, and 3, may all belong to the early part of the period of the Upper Old Red, as some scales of Holoptychius nobilissimus have been found scattered through these beds, No. 2, in Strathmore. Another nearly allied Holoptychius occurs in Dura Den, see Figure 498 of this fish and also Figure 497 of one of its scales, as these last are often the only parts met with; being scattered in Forfarshire through red-coloured shales and sandstones, as are scales of a large species of the same genus in a corresponding matrix in Herefordshire. (Siluria 4th edition page 265.) The number of fish obtained from the British Upper Old Red Sandstone amounts to fifteen species referred to eleven genera.

Sir R. Murchison groups with this upper division of the Old Red of Scotland certain light-red and yellow sandstones and grits which occur in the northernmost part of the mainland, and extend also into the Orkney and Shetland Islands. They contain Calamites and other plants which agree generically with Carboniferous forms.

MIDDLE OLD RED SANDSTONE.

In the northern part of Scotland there occur a great series of bituminous schists and flagstones, to the fossil fish of which attention was first called by the late Hugh Miller. They were afterwards described by Agassiz, and the rocks containing them were examined by Sir R. Murchison and Professor Sedgwick, in Caithness, Cromarty, Moray, Nairn, Gamrie in Banff, and the Orkneys and Shetlands, in which great numbers of fossil fish have been found. These were at first supposed to be the oldest known vertebrate animals, as in Cromarty the beds in which they occur seem to form the base of the Old Red system resting almost immediately on the crystalline or metamorphic rocks. But in fact these fish-bearing beds, when they are traced from north to south, or to the central parts of Scotland, thin out, so that their relative age to the Lower Old Red Sandstone, presently to be mentioned, was not at first detected, the two formations not appearing in superposition in the same district. In Caithness, however, many hundred feet below the fish-zone of the middle division, remains of Pteraspis were found by Mr. Peach in 1861. This genus has never yet been found in either of the two higher divisions of the Old Red Sandstone, and confirms Sir R. Murchison's previous suspicion that the rocks in which it occurs belong to the Lower "Old Red," or agree in age with the Arbroath paving-stone. (Siluria 4th edition page 258.)

FOSSIL FISH OF THE MIDDLE OLD RED SANDSTONE.

The Devonian fish were referred by Agassiz to two of his great orders, namely, the Placoids and Ganoids. Of the first of these, which in the Recent period comprise the shark, the dog-fish, and the ray, no entire skeletons are preserved, but fin-spines, called ichthyodorulites, and teeth occur. On such remains the genera Onchus, Odontacanthus, and Ctenodus, a supposed cestraciont, and some others, have been established.

(FIGURE 499. Polypterus. See Agassiz, "Recherces sur les Poissons Fossiles." Living in the Nile and other African rivers. a. One of the fringed pectoral fins. b. One of the ventral fins. c. Anal fin. d. Dorsal fin, or row of finlets.)

(FIGURE 500. Restoration of Osteolepis. Pander. Old Red Sandstone, or Devonian. a. One of the fringed pectoral fins. b. One of the ventral fins. c. Anal fin. d, e. Dorsal fins.)

By far the greater number of the Old Red Sandstone fishes belong to a sub-order of Ganoids instituted by Huxley in 1861, and for which he has proposed the name of Crossopterygidae (Abridged from crossotos, a fringe, and pteryx, a fin.), or the fringe-finned, in consideration of the peculiar manner in which the fin-rays of the paired fins are arranged so as to form a fringe round a central lobe, as in the Polypterus (see a, Figure 499), a genus of which there are several species now inhabiting the Nile and other African rivers. The reader will at once recognise in Osteolepis (Figure 500), one of the common fishes of the Old Red Sandstone, many points of analogy with Polypterus. They not only agree in the structure of the fin, at first pointed out by Huxley, but also in the position of the pectoral, ventral, and anal fins, and in having an elongated body and rhomboidal scales. On the other hand, the tail is more symmetrical in the recent fish, which has also an apparatus of dorsal finlets of a very abnormal character, both as to number and structure. As to the dorsals of Osteolepis, they are regular in structure and position, having nothing remarkable about them, except that there are two of them, which is comparatively unusual in living fish.

Among the "fringe-finned" Ganoids we find some with rhomboidal scales, such as Osteolepis, Figure 500; others with cycloidal scales, as Holoptychius, before mentioned (see Figure 498). In the genera Dipterus and Diplopterus, as Hugh Miller pointed out, and in several other of the fringe-finned genera, as in Gyroptychius and Glyptolepis, the two dorsals are placed far backward, or directly over the ventral and anal fins. The Asterolepis was a ganoid fish of gigantic dimensions. A. Asmusii, Eichwald, a species characteristic of the Old Red Sandstone of Russia, as well as that of Scotland, attained the length of between twenty and thirty feet. It was clothed with strong bony armour, embossed with star-like tubercles, but it had only a cartilaginous skeleton. The mouth was furnished with two rows of teeth, the outer ones small and fish-like, the inner larger and with a reptilian character. The Asterolepis occurs also in the Devonian rocks of North America.

If we except the Placoids already alluded to, and a few other families of doubtful affinities, all the Old Red Sandstone fishes are Ganoids, an order so named by Agassiz from the shining outer surface of their scales; but Professor Huxley has also called our attention to the fact that, while a few of the primary and the great majority of the secondary Ganoids resemble the living bony pike, Lepidosteus, or the Amia, genera now found in North American rivers, and one of them, Lepidosteus, extending as far south as Guatemala, the Crossopterygii, or fringe-finned Ichthyolites, of the Old Red are closely related to the African Polypterus, which is represented by five or six species now inhabiting the Nile and the rivers of Senegal. These North American and African Ganoids are quite exceptional in the living creation; they are entirely confined to the northern hemisphere, unless some species of Polypterus range to the south of the line in Africa; and, out of about 9000 living species of fish known to M. Gunther, and of which more than 6000 are now preserved in the British Museum, they probably constitute no more than nine.

If many circumstances favour the theory of the fresh-water origin of the Old Red Sandstone, this view of its nature is not a little confirmed by our finding that it is in Llake Superior and the other inland Canadian seas of fresh water, and in the Mississippi and African rivers, that we at present find those fish which have the nearest affinity to the fossil forms of this ancient formation.

(FIGURE 501. Pterichthys, Agassiz; Upper side, showing mouth; as restored by H. Miller.)

Among the anomalous forms of Old Red fishes not referable to Huxley's Crossopterygii is the Pterichthys, of which five species have been found in the middle division of the Old Red of Scotland. Some writers have compared their shelly covering to that of Crustaceans, with which, however, they have no real affinity. The wing-like appendages, whence the genus is named, were first supposed by Hugh Miller to be paddles, like those of the turtle; and there can now be no doubt that they do really correspond with the pectoral fins.

The number of species of fish already obtained from the middle division of the Old Red Sandstone in Great Britain is about 70, and the principal genera, besides Osteolepis and Pterichthys, already mentioned, are Glyptolepis, Diplacanthus, Dendrodus, Coccosteus, Cheirancanthus, and Acanthoides.

LOWER OLD RED SANDSTONE.

(FIGURE 502. Cephalaspis Lyellii, Agassiz. Length 6 3/4 inches. From a specimen in my collection found at Glammiss, in Forfarshire. (See other figures, Agassiz, volume 2 table 1 a and 1 b. a. One of the peculiar scales with which the head is covered when perfect. These scales are generally removed, as in the specimen above figured. b, c. Scales from different parts of the body and tail.)

The third or lowest division south of the Grampians consists of grey paving- stone and roofing-slate, with associated red and grey shales; these strata underlie a dense mass of conglomerate. In these grey beds several remarkable fish have been found of the genus named by Agassiz Cephalaspis, or "buckler- headed," from the extraordinary shield which covers the head (see Figure 502), and which has often been mistaken for that of a trilobite, such as Asaphus. A species of Pteraspis, of the same family, has also been found by the Reverend Hugh Mitchell in beds of corresponding age in Perthshire; and Mr. Powrie enumerates no less than five genera of the family Acanthodidae, the spines, scales, and other remains of which have been detected in the grey flaggy sandstones. (Powrie Geological Quarterly Journal volume 20 page 417.)

(FIGURE 503. Pterygotus anglicus, Agassiz. Middle portion of the back of the head called the seraphim.)

(FIGURE 504. Pterygotus anglicus, Agassiz. Forfarshire. Ventral aspect. Restored by H. Wodward, F.G.S. a. Carapace, showing the large sessile eyes at the anterior angles. b. The metastoma or post-oral plate (serving the office of a lower lip). c, c. Chelate appendages (antennules). d. First pair of simple palpi (antennae). e. Second pair of simple palpi (mandibles). f. Third pair of simple palpi (first maxillae). g. Pair of swimming feet with their broad basal joints, whose serrated edges serve the office of maxillae. h. Thoracic plate covering the first two thoracic segments, which are indicated by the figures 1, 2, and a dotted line. 1-6. Thoracic segments. 7-12. Abdominal segments. 13. Telson, or tail-plate.)

In the same formation at Carmylie, in Forfarshire, commonly known as the Arbroath paving-stone, fragments of a huge crustacean have been met with from time to time. They are called by the Scotch quarrymen the "Seraphim," from the wing-like form and feather-like ornament of the thoracic appendage, the part most usually met with. Agassiz, having previously referred some of these fragments to the class of fishes, was the first to recognise their crustacean character, and, although at the time unable correctly to determine the true relation of the several parts, he figured the portions on which he founded his opinion, in the first plate of his "Poissons Fossiles du Vieux Gres Rouge."

A restoration in correct proportion to the size of the fragments of P. anglicus (Figure 504), from the Lower Old Red Sandstone of Perthshire and Forfarshire, would give us a creature measuring from five to six feet in length, and more than one foot across.

The largest crustaceans living at the present day are the Inachus Kaempferi, of De Haan, from Japan (a brachyurous or short-tailed crab), chiefly remarkable for the extraordinary length of its limbs; the fore-arm measuring four feet in length, and the others in proportion, so that it covers about 25 square feet of ground; and the Limulus Moluccanus, the great King Crab of China and the Eastern seas, which, when adult, measures 1 1/2 foot across its carapace, and is three feet in length.

(FIGURE 505. Parka decipiens, Fleming. In sandstone of lower beds of Old Red, Ley's Mill, Forfarshire.)

(FIGURE 506. Parka decipiens, Fleming. In shale of Lower Old Red, Park Hill, Fife.)

(FIGURE 507. Shale of Old Red Sandstone. Forfarshire. With impression of plants and eggs of Crustaceans. a. Two pair of ova? resembling those of large Salamanders or Tritons— on the same leaf. b, b. Detached ova.)

Besides some species of Pterygotus, several of the allied genus Eurypterus occur in the Lower Old Red Sandstone, and with them the remains of grass-like plants so abundant in Forfarshire and Kincardineshire as to be useful to the geologist by enabling him to identify the inferior strata at distant points. Some botanists have suggested that these plants may be of the family Fluviales, and of fresh-water genera. They are accompanied by fossils, called "berries" by the quarrymen, which they compared to a compressed blackberry (see Figures 505, 506), and which were called "Parka" by Dr. Fleming. They are now considered by Mr. Powrie to be the eggs of crustaceans, which is highly probable, for they have not only been found with Pterygotus anglicus in Forfarshire and Perthshire, but also in the Upper Silurian strata of England, in which species of the same genus, Pterygotus, occur.

The grandest exhibitions, says Sir R. Murchison, of the Old Red Sandstone in England and Wales appear in the escarpments of the Black Mountains and in the Fans of Brecon and Carmarthen, the one 2862, and the other 2590 feet above the sea. The mass of red and brown sandstone in these mountains is estimated at not less than 10,000 feet, clearly intercalated between the Carboniferous and Silurian strata. No shells or corals have ever been found in the whole series, not even where the beds are calcareous, forming irregular courses of concretionary lumps called "corn-stones," which may be described as mottled red and green earthy limestones. The fishes of this lowest English Old Red are Cephalaspis and Pteraspis, specifically different from species of the same genera which occur in the uppermost Ludlow or Silurian tilestones. Crustaceans also of the genus Eurypterus are met with.

MARINE OR DEVONIAN TYPE.

We may now speak of the marine type of the British strata intermediate between the Carboniferous and Silurian, in treating of which we shall find it much more easy to identify the Upper, Middle, and Lower divisions with strata of the same age in other countries. It was not until the year 1836 that Sir R. Murchison and Professor Sedgwick discovered that the culmiferous or anthracitic shales and sandstones of North Devon, several thousand feet thick, belonged to the coal, and that the beds below them, which are of still greater thickness, and which, like the carboniferous strata, had been confounded under the general name "graywacke," occupied a geological position corresponding to that of the Old Red Sandstone already described. In this reform they were aided by a suggestion of Mr. Lonsdale, who, after studying the Devonshire fossils, perceived that they belonged to a peculiar palaeontological type of intermediate character between the Carboniferous and Silurian.

It is in the north of Devon that these formations may best be studied, where they have been divided into an Upper, Middle, and Lower Group, and where, although much contorted and folded, they have for the most part escaped being altered by intrusive trap-rocks and by granite, which in Dartmoor and the more southern parts of the same county have often reduced them to a crystalline or metamorphic state.

TABLE 25.1 DEVONIAN SERIES IN NORTH DEVON.

UPPER DEVONIAN OR PILTON GROUP.

a. Sandy slates and schists with fossils, 36 species out of 110 common to the Carboniferous group (Pilton, Barnstaple, etc.), resting on soft schists in which fossils are very abundant (Croyde, etc.), and which pass down into

b. Yellow, brown, and red sandstone, with land plants (Cyclopteris, etc.) and marine shells. One zone, characterised by the abundance of cucullaea (Baggy Point, Marwood, Sloly, etc.) resting on hard grey and reddish sandstone and micaceous flags, no fossils yet found (Dulverton, Pickwell, Down, etc.)

MIDDLE DEVONIAN OR ILFRACOMBE GROUP.

a. Green glossy slates of considerable thickness, no fossils yet recorded from these beds (Mortenoe, Lee Bay, etc.).

b. Slates and schists, with several irregular courses of limestone containing shells and corals like those of the Plymouth Limestone (Combe Martin, Ilfracombe, etc.).

LOWER DEVONIAN OR LYNTON GROUP.

a. Hard, greenish, red, and purple sandstone— no fossils yet found (Hangman Hill, etc.).

b. Soft slates with subordinate sandstones— fossils numerous at various horizons— Orthis, Corals, Encrinites, etc. (Valley of Rocks, Lynmouth, etc.).

Table 25.1 exhibits the sequence of the strata or subdivisions as seen both on the sea-coast of the British Channel and in the interior of Devon. It will be seen that in all main points it agrees with the table drawn up in 1864 for the sixth edition of my "Elements." Mr. Etheridge has since published an excellent account of the different subdivisions of the rocks and their fossils, and has also pointed out their relation to the corresponding marine strata of the Continent. (Quarterly Geological Journal volume 23 1867.) The slight modifications introduced in my table since 1864 are the result of a tour made in 1870 in company with Mr. T. Mck. Hughes, when we had the advantage of Mr. Etheridge's memoir as our guide.

The place of the sandstones of the Foreland is not yet clearly made out, as they are cut off by a great fault and disturbance.

UPPER DEVONIAN ROCKS.

(FIGURE 508. Spirifera disjuncta, Sowerby. Syn. Sp. Verneuilii, Murch. Upper Devonian, Boulogne.)

(FIGURE 509. Phacops latifrons, Bronn. Characteristic of the Devonian in Europe, Asia, and N. and S. America.)

(FIGURE 510. Clymenia linearis, Munster. Petherwyn, Cornwall; Elbersreuth, Bavaria.)

(FIGURE 511. Cypridina serrato-striata, Sandberger, Weilburg, etc.; Cornwall, Nassau, Saxony, Belgium.)

The slates and sandstones of Barnstaple (a and b of the preceding section) contain the shell Spirifera disjuncta, Sowerby (S. Verneuilii, Murch.), (see Figure 508), which has a very wide range in Europe, Asia Minor, and even China; also Strophalosia caperata, together with the large trilobite Phacops latifrons, Bronn. (See Figure 509), which is all but world-wide in its distribution. The fossils are numerous, and comprise about 150 species of mollusca, a fifth of which pass up into the overlying Carboniferous rocks. To this Upper Devonian belong a series of limestones and slates well developed at Petherwyn, in Cornwall, where they have yielded 75 species of fossils. The genus of Cephalopoda called Clymenia (Figure 510) is represented by no less than eleven species, and strata occupying the same position in Germany are called Clymenien- Kalk, or sometimes Cypridinen-Schiefer, on account of the number of minute bivalve shells of the crustacean called Cypridina serrato-striata (Figure 511), which is found in these beds, in the Rhenish provinces, the Harz, Saxony, and Silesia, as well as in Cornwall and Belgium.

MIDDLE DEVONIAN ROCKS.

(FIGURE 512. Heliolites porosa, Goldf. sp. (Porites pyriformis, Lonsd.) a. Portion of the same magnified. Middle Devonian, Torquay, Plymouth; Eifel.)

(FIGURE 513. Favosites cervicornis, Blainv. S. Devon, from a polished specimen. a. Portion of the same magnified, to show the pores.)

(FIGURE 514. Cyathophyllum caespitosum, Goldf.; Plymouth and Ilfracombe. b. A terminal star. c. Vertical section, exhibiting transverse plates, and part of another branch.)

We come next to the most typical portion of the Devonian system, including the great limestones of Plymouth and Torbay, replete with shells, trilobites, and corals. Of the corals 51 species are enumerated by Mr. Etheridge, none of which pass into the Carboniferous formation. Among the genera we find Favosites, Heliolites, and Cyathophyllum. The two former genera are very frequent in Silurian rocks: some few even of the species are said to be common to the Devonian and Silurian groups, as, for example, Favosites cervicornis (Figure 513), one of the commonest of all the Devonshire fossils. The Cyathophyllum caespitosum (Figure 514) and Heliolites pyriformis (Figure 512) are species peculiar to this formation.

(FIGURE 515. Stringocephalus Burtini, Def. a. Valves united. b. Interior of ventral or large valve, showing thick partition and portion of a large process which projects from the dorsal valve across the shell.)

(FIGURE 516. Uncites Gryphus, Def. Middle Devonian. S. Devon and the Continent.)

With the above are found no less than eleven genera of stone-lilies or crinoids, some of them, such as Cupressocrinites, distinct from any Carboniferous forms. The mollusks, also, are no less characteristic; of 68 species of Brachiopoda, ten only are common to the Carboniferous Limestone. The Stringocephalus Burtini (Figure 515) and Uncites Gryphus (Figure 516) may be mentioned as exclusively Middle Devonian genera, and extremely characteristic of the same division in Belgium. The Stringocephalus is also so abundant in the Middle Devonian of the banks of the Rhine as to have suggested the name of Stringocephalus Limestone. The only two species of Brachiopoda common to the Silurian and Devonian formations are Atrypa reticularis (Figure 532), which seems to have been a cosmopolite species, and Strophomena rhomboidalis.

(FIGURE 517. Megalodon cucullatus, Sowerby. Eifel; also Bradley, S. Devon. a. The valves united. b. Interior of valve, showing the large cardinal tooth.)

(FIGURE 518. Conularia ornata, D'Arch. and De Vern. (Geological Transactions Sec. Ser. volume 6. Plate 29.) Refrath, near Cologne.)

(FIGURE 519. Bronteus flabellifer, Goldf. Mid. Devon; S. Devon; and the Eifel.)

Among the peculiar lamellibranchiate bivalves common to the Plymouth limestone of Devonshire and the Continent, we find the Megalodon (Figure 517). There are also twelve genera of Gasteropods which have yielded 36 species, four of which pass to the Carboniferous group, namely Macrocheilus, Acroculia, Euomphalus, and Murchisonia. Pteropods occur, such as Conularia (Figure 518), and Cephalopods, such as Cyrtoceras, Gyroceras, Orthoceras, and others, nearly all of genera distinct from those prevailing in the Upper Devonian Limestone, or Clymenien- kalk of the Germans already mentioned. Although but few species of Trilobites occur, the characteristic Bronteus flabellifer (Figure 519) is far from rare, and all collectors are familiar with its fan-like tail. In this same group, called, as before stated, the Stringocephalus, or Eifel Limestone, in Germany, several fish remains have been detected, and among others the remarkable genus Coccosteus, covered with its tuberculated bony armour; and these ichthyolites serve, as Sir R. Murchison observes (Siluria page 362), to identify this middle marine Devonian with the Old Red Sandstone of Britain and Russia.

(FIGURE 520. Calceola sandalina, Lam. Eifel; also South Devon. a. Ventral valve. b. Inner side of dorsal valve.)

Beneath the Eifel Limestone (the great central and typical member of "the Devonian" on the Continent) lie certain schists called by German writers "Calceola-schiefer," because they contain in abundance a fossil body of very curious structure, Calceola sandalina (Figure 520), which has been usually considered a brachiopod, but which some naturalists have lately referred to a Goniophyllum, supposing it to be an abnormal form of the order Zoantharia rugosa (see Figure 474), differing from all other corals in being furnished with a strong operculum. This is by no means a rare fossil in the slaty limestone of South Devon, and, like the Eifel form, is confined to the middle group of this country.

LOWER DEVONIAN ROCKS.

(FIGURE 521. Spirifera mucronata, Hall. Devonian of Pennsylvania.)

A great series of sandstones and glossy slates, with Crinoids, Brachiopods, and some corals, occurring on the coast at Lynmouth and the neighbourhood, and called the Lynton Group (see Table 25.1), form the lowest member of the Devonian in North Devon. Among the 18 species of all classes enumerated by Mr. Etheridge, two-thirds are common to the Middle Devonian, but only one, the ubiquitous Atrypa reticularis, can with certainty be identified with Silurian species. Among the characteristic forms are Alveolites suborbicularis, also common to this formation in the Rhine, and Orthis arcuata, very widely spread in the North Devon localities. But we may expect a large addition to the number of fossils whenever these strata shall have been carefully searched. The Spirifer Sandstone of Sandberger, as exhibited in the rocks bordering the Rhine between Coblentz and Caub, belong to this Lower division, and the same broad-winged Spirifers distinguish the Devonian strata of North America.

(FIGURE 522. Homalonotus armatus, Burmeister. Lower Devonian; Daun, in the Eifel; and S. Devon. Obs. The two rows of spines down the body give an appearance of more distinct trilobation than really occurs in this or most other species of the genus.)

Among the Trilobites of this era several large species of Homalonotus (Figure 522) are conspicuous. The genus is still better known as a Silurian form, but the spinose species appear to belong exclusively to the "Lower Devonian," and are found in Britain, Europe, and the Cape of Good Hope.

DEVONIAN OF RUSSIA.

The Devonian strata of Russia extend, according to Sir R. Murchison, over a region more spacious than the British Isles; and it is remarkable that, where they consist of sandstone like the "Old Red" of Scotland and Central England, they are tenanted by fossil fishes often of the same species and still oftener of the same genera as the British, whereas when they consist of limestone they contain shells similar to those of Devonshire, thus confirming, as Sir Roderick has pointed out, the contemporaneous origin which had been previously assigned to formations exhibiting two very distinct mineral types in different parts of Britain. (Murchison's Siluria page 329.) The calcareous and the arenaceous rocks of Russia above alluded to alternate in such a manner as to leave no doubt of their having been deposited in different parts of the same great period.

DEVONIAN STRATA IN THE UNITED STATES AND CANADA.

(FIGURE 523. Psilophyton princeps, Dawson, Quarterly Geological Journal volume 15 1863; and Canada Survey 1863. Species characteristic of the whole Devonian series in North America. a. Fruit; natural size. b. Stem; natural size. c. Scalariform tissue of the axis highly magnified.)

Between the Carboniferous and Silurian strata there intervenes, in the United States and Canada, a great series of formations referable to the Devonian group, comprising some strata of marine origin abounding in shells and corals, and others of shallow-water and littoral origin in which terrestrial plants abound. The fossils, both of the deep and shallow water strata, are very analogous to those of Europe, the species being in some cases the same. In Eastern Canada Sir W. Logan has pointed out that in the peninsula of Gaspe, south of the estuary of St. Lawrence, a mass of sandstone, conglomerate, and shale referable to this period occurs, rich in vegetable remains, together with some fish-spines. Far down in the sandstones of Gaspe, Dr. Dawson found, in 1869, an entire specimen of the genus Cephalaspis, a form so characteristic, as we have already seen, of the Scotch Lower Old Red Sandstone. Some of the sandstones are ripple-marked, and towards the upper part of the whole series a thin seam of coal has been observed, measuring, together with some associated carbonaceous shale, about three inches in thickness. It rests on an under-clay in which are the roots of Psilophyton (see Figure 523). At many other levels rootlets of this same plant have been shown by Principal Dawson to penetrate the clays, and to play the same part as do the rootlets of Stigmaria in the coal formation.

We had already learnt from the works of Goppert, Unger, and Bronn that the European plants of the Devonian epoch resemble generically, with few exceptions, those already known as Carboniferous; and Dr. Dawson, in 1859, enumerated 32 genera and 69 species which he had then obtained from the State of New York and Canada. A perusal of his catalogue (Quarterly Geological Journal volume 15 page 477 1859; also volume 18 page 296 1862.), comprising Coniferae, Sigillariae, Calamites, Asterophyllites, Lepidodendra, and ferns of the genera Cyclopteris, Neuropteris, Sphenopteris, and others, together with fruits, such as Cardiocarpum and Trigonocarpum, might dispose geologists to believe that they were presented with a list of Carboniferous fossils, the difference of the species from those of the coal-measures, and even a slight admixture of genera unknown in Europe, being naturally ascribed to geographical distribution and the distance of the New from the Old World. But fortunately the coal formation is fully developed on the other side of the Atlantic, and is singularly like that of Europe, both lithologically and in the species of its fossil plants. There is also the most unequivocal evidence of relative age afforded by superposition, for the Devonian strata in the United States are seen to crop out from beneath the Carboniferous on the borders of Pennsylvania and New York, where both formations are of great thickness.

The number of American Devonian plants has now been raised by Dr. Dawson to 120, to which we may add about 80 from the European flora of the same age, so that already the vegetation of this period is beginning to be nearly half as rich as that of the coal-measures which have been studied for so much longer a time and over so much wider an area. The Psilophyton above alluded to is believed by Dr. Dawson to be a lycopodiaceous plant, branching dichotomously (see P. princeps, Figure 523), with stems springing from a rhizome, which last has circular areoles, much resembling those of Stigmaria, and like it sending forth cylindrical rootlets. The extreme points of some of the branchlets are rolled up so as to resemble the croziers or circinate vernation of ferns; the leaves or bracts, a, supposed to belong to the same plant, are described by Dawson as having inclosed the fructification. The remains of Psilophyton princeps have been traced through all the members of the Devonian series in America, and Dr. Dawson has lately recognised it in specimens of Old Red Sandstone from the north of Scotland.

The monotonous character of the Carboniferous flora might be explained by imagining that we have only the vegetation handed down to us of one set of stations, consisting of wide swampy flats. But Dr. Dawson supposes that the geographical conditions under which the Devonian plants grew were more varied, and had more of an upland character. If so, the limitation of this more ancient flora, represented by so many genera and species, to the gymnospermous and cryptogamous orders, and the absence or extreme rarity of plants of higher grade, lead us naturally to speculate on the theory of progressive development, however difficult it may be to avail ourselves of this explanation, so long as we meet with even a few exceptional cases of what may seem to be monocotyledonous or dicotyledonous exogens.

DEVONIAN INSECTS OF CANADA.

The earliest known insects were brought to light in 1865 in the Devonian strata of St. John's, New Brunswick, and are referred by Mr. Scudder to four species of Neuroptera. One of them is a gigantic Ephemera, and measured five inches in expanse of wing.

Like many other ancient animals, says Dr. Dawson, they show a remarkable union of characters now found in distinct orders of insects, or constitute what have been named "synthetic types." Of this kind is a stridulating or musical apparatus like that of the cricket in an insect otherwise allied to the Neuroptera. This structure, as Dr. Dawson observes, if rightly interpreted by Mr. Scudder, introduces us to the sounds of the Devonian woods, bringing before our imagination the trill and hum of insect life that enlivened the solitudes of these strange old forests.

CHAPTER XXVI.

SILURIAN GROUP.

Classification of the Silurian Rocks. Ludlow Formation and Fossils. Bone-bed of the Upper Ludlow. Lower Ludlow Shales with Pentamerus. Oldest known Remains of fossil Fish. Table of the progressive Discovery of Vertebrata in older Rocks. Wenlock Formation, Corals, Cystideans and Trilobites. Llandovery Group or Beds of Passage. Lower Silurian Rocks. Caradoc and Bala Beds. Brachiopoda. Trilobites. Cystideae. Graptolites. Llandeilo Flags. Arenig or Stiper-stones Group. Foreign Silurian Equivalents in Europe. Silurian Strata of the United States. Canadian Equivalents. Amount of specific Agreement of Fossils with those of Europe.

CLASSIFICATION OF THE SILURIAN ROCKS.

We come next in descending order to that division of Primary or Palaeozoic rocks which immediately underlie the Devonian group or Old Red Sandstone. For these strata Sir Roderick Murchison first proposed the name of Silurian when he had studied and classified them in that part of Wales and some of the contiguous counties of England which once constituted the kingdom of the Silures, a tribe of ancient Britons. Table 26.1 will explain the two principal divisions, Upper and Lower, of the Silurian rocks, and the minor subdivisions usually adopted, comprehending all the strata originally embraced in the Silurian system by Sir Roderick Murchison. The formations below the Arenig or Stiper-stones group are treated of in the next chapter, when the "Primordial" or Cambrian group is described.

TABLE 26.1. SILURIAN ROCKS (THICKNESS GIVEN IN FEET).

UPPER SILURIAN ROCKS.

1. LUDLOW FORMATION:

a. Upper Ludlow beds: 780. b. Lower Ludlow beds: 1,050.

2. WENLOCK FORMATION:

a. Wenlock limestone and shale and b. Woolhope limestone and shale, and Denbighshire grits: above 4,000.

3. LLANDOVERY FORMATION (Beds of passage between Upper and Lower Silurian):

a. Upper Llandovery (May-Hill beds): 800. b. Lower Llandovery: 600-1,000.

LOWER SILURIAN ROCKS.

1. BALA AND CARADOC BEDS, including volcanic rocks: 12,000.

2. LLANDEILO FLAGS, including volcanic rocks: 4,500.

3. ARENIG OR STIPER-STONES GROUP, including volcanic rocks: above 10,000.

UPPER SILURIAN ROCKS.

1. LUDLOW FORMATION.

This member of the Upper Silurian group, as will be seen by Table 26.1, is of great thickness, and subdivided into two parts— the Upper Ludlow and the Lower Ludlow. Each of these may be distinguished near the town of Ludlow, and at other places in Shropshire and Herefordshire, by peculiar organic remains; but out of more than 500 species found in the Ludlow formation as a whole, not more than five species per hundred are common to the overlying Devonian. The student may refer to the excellent tables given in the last edition of Sir R. Murchison's Siluria for a list of the organic remains of all classes distributed through the different subdivisions of the Upper and Lower Silurian.

A. UPPER LUDLOW: DOWNTON SANDSTONE.

At the top of this subdivision there occur beds of fine-grained yellowish sandstone and hard reddish grits which were formerly referred by Sir R. Murchison to the Old Red Sandstone, under the name of "Tilestones." In mineral character this group forms a transition from the Silurian to the Old Red Sandstone, the strata of both being conformable; but it is now ascertained that the fossils agree in great part specifically, and in general character entirely, with those of the underlying Upper Ludlow rocks. Among these are Orthoceras bullatum, Platyschisma helicites, Bellerophon trilobatus, Chonetes lata, etc., with numerous defenses of fishes.

These beds, therefore, now generally called the "Downton Sandstone," are classed as the newest member of the Upper Silurian. They are well seen at Downton Castle, near Ludlow, where they are quarried for building, and at Kington, in Herefordshire. In the latter place, as well as at Ludlow, crustaceans of the genera Pterygotus (for genus see Figure 504) and Eurypterus are met with.

BONE-BED OF THE UPPER LUDLOW.

At the base of the Downton sandstones there occurs a bone-bed which deserves especial notice as affording the most ancient example of fossil fish occurring in any considerable quantity. It usually consists of one or two thin layers of brown bony fragments near the junction of the Old Red Sandstone and the Ludlow rocks, and was first observed by Sir R. Murchison near the town of Ludlow, where it is three or four inches thick. It has since been traced to a distance of 45 miles from that point into Gloucestershire and other counties, and is commonly not more than an inch thick, but varies to nearly a foot. Near Ludlow two bone- beds are observable, with 14 feet of intervening strata full of Upper Ludlow fossils. (Murchison's Siluria page 140.) At that point immediately above the upper fish-bed numerous small globular bodies have been found, which were determined by Dr. Hooker to be the sporangia of a cryptogamic land-plant, probably lycopodiaceous.

(FIGURE 524. Onchus tenuistriatus, Agassiz. Bone-bed. Upper Silurian. Ludlow.)

(FIGURE 525. Shagreen-scales of a placoid fish, Thelodus parvidens, Agassiz. Bone-bed, Upper Ludlow.)

(FIGURE 526. Plectrodus mirabilis, Agassiz. Bone-bed, Upper Ludlow.)

Most of the fish have been referred by Agassiz to his placoid order, some of them to the genus Onchus, to which the spine (Figure 524) and the minute scales (Figure 525) are supposed to belong. It has been suggested, however, that Onchus may be one of those Acanthodian fish referred by Agassiz to his Ganoid order, which are so characteristic of the base of the Old Red Sandstone in Forfarshire, although the species of the Old Red are all different from these of the Silurian beds now under consideration. The jaw and teeth of another predaceous genus (Figure 526) have also been detected, together with some specimens of Pteraspis Ludensis. As usual in bone-beds, the teeth and bones are, for the most part, fragmentary and rolled.

GREY SANDSTONE AND MUDSTONE, ETC.

(FIGURE 527. Orthis elegantula, Dalm. Var. Orbicularis, Sowerby. Upper Ludlow.)

(FIGURE 528. Rhynchonella navicula, Sowerby. Ludlow Beds.)

The next subdivision of the Upper Ludlow consists of grey calcareous sandstone, or very commonly a micaceous stone, decomposing into soft mud, and contains, besides the shells mentioned above, Lingula cornea, Orthis orbicularis, a round variety of O. elegantula, Modiolopsis platyphylla, Grammysia cingulata, all characteristic of the Upper Ludlow. The lowest or mud-stone beds contain Rhynchonella navicula (Figure 528), which is common to this bed and the Lower Ludlow. As usual in Palaeozoic strata older than the coal, the brachiopodous mollusca greatly outnumber the lamellibranchiate (see below); but the latter are by no means unrepresented. Among other genera, for example, we observe Avicula and Pterinea, Cardiola, Ctenodonta (sub-genus of Nucula), Orthonota, Modiolopsis, and Palaearca.

Some of the Upper Ludlow sandstones are ripple-marked, thus affording evidence of gradual deposition; and the same may be said of the accompanying fine argillaceous shales, which are of great thickness, and have been provincially named "mud-stones." In some of these shales stems of crinoidea are found in an erect position, having evidently become fossil on the spots where they grew at the bottom of the sea. The facility with which these rocks, when exposed to the weather, are resolved into mud, proves that, notwithstanding their antiquity, they are nearly in the state in which they were first thrown down.

b. LOWER LUDLOW BEDS.

(FIGURE 529. Pentamerus Knightii, Sowerby. Aymestry. One-half natural size. a. View of both valves united. b. Longitudinal section through both valves, showing the central plates or septa.)

The chief mass of this formation consists of a dark grey argillaceous shale with calcareous concretions, having a maximum thickness of 1000 feet. In some places, and especially at Aymestry, in Herefordshire, a subcrystalline and argillaceous limestone, sometimes 50 feet thick, overlies the shale. Sir R. Murchison therefore classes this Aymestry limestone as holding an intermediate position between the Upper and Lower Ludlow, but Mr. Lightbody remarks that at Mocktrie, near Leintwardine, the Lower Ludlow shales, with their characteristic fossils, occur both above and below a similar limestone. This limestone around Aymestry and Sedgeley is distinguished by the abundance of Pentamerus Knightii, Sowerby (Figure 529), also found in the Lower Ludlow and Wenlock shale. This genus of brachiopoda was first found in Silurian strata, and is exclusively a palaeozoic form. The name was derived from pente, five, and meros, a part, because both valves are divided by a central septum, making four chambers, and in one valve the septum itself contains a small chamber, making five. The size of these septa is enormous compared with those of any other brachiopod shell; and they must nearly have divided the animal into two equal halves; but they are, nevertheless, of the same nature as the septa or plates which are found in the interior of Spirifera, Terebratula, and many other shells of this order. Messrs. Murchison and De Verneuil discovered this species dispersed in myriads through a white limestone of Upper Silurian age, on the banks of the Is, on the eastern flank of the Urals in Russia, and a similar species is frequent in Sweden.

(FIGURE 530. Lingula Lewisii, J. Sowb. Abberley Hills.)

(FIGURE 531. Rhynchonella (Terebratula) Wilsoni, Sowerby. Aymestry.)

(FIGURE 532. Atrypa reticularis, Linn. (Terebratula affinis, Min. Con.) Aymestry. a. Upper valve. b. Lower valve. c. Anterior margin of the valves.)

Three other abundant shells in the Aymestry limestone are, 1st, Lingula Lewisii (Figure 530); second, Rhynchonella Wilsoni, Sowerby. (Figure 531), which is also common to the Lower Ludlow and Wenlock limestone; third, Atrypa reticularis, Linn. (Figure 532), which has a very wide range, being found in every part of the Upper Silurian system, and even ranging up into the Middle Devonian series.

The Aymestry Limestone contains many shells, especially brachiopoda, corals, trilobites, and other fossils, amounting on the whole to 74 species, all except three or four being common to the beds either above or below.

(FIGURE 533. Phragmoceras ventricosum, J. Sowerby. (Orthoceras ventricosum, Stein.) Aymestry; one-quarter natural size.)

(FIGURE 534. Lituites (Trochoceras) giganteus, J. Sowerby. Near Ludlow; also in the Aymestry and Wenlock Limestones; 1/4 natural size.)

(FIGURE 535. Fragment of Orthoceras Ludense, J. Sowerby. Leintwardine, Shropshire.)

The Lower Ludlow Shale contains, among other fossils, many large cephalopoda not known in newer rocks, as the Phragmoceras of Broderip, and the Lituites of Breynius (see Figures 533, 534). The latter is partly straight and partly convoluted in a very flat spire. The Orthoceras Ludense (Figure 535), as well as the cephalopod last mentioned, occurs in this member of the species.

A species of Graptolite, G. priodon, Bronn (Figure 545), occurs plentifully in the Lower Ludlow. This fossil, referred, though somewhat doubtfully, to a form of hydrozoid or sertularian polyp, has not yet been met with in strata above the Silurian.

Star-fish, as Sir R. Murchison points out, are by no means rare in the Lower Ludlow rock. These fossils, of which six extinct genera are now known in the Ludlow series, represented by 18 species, remind us of various living forms now found in our British seas, both of the families Asteriadae and Ophiuridae.

OLDEST KNOWN FOSSIL FISH.

Until 1859 there was no example of a fossil fish older than the bone-bed of the Upper Ludlow, but in that year a specimen of Pteraspis was found at Church Hill, near Leintwardine, in Shropshire, by Mr. J.E. Lee of Caerleon, F.G.S., in shale below the Aymestry limestone, associated with fossil shells of the Lower Ludlow formation— shells which differ considerably from those characterising the Upper Ludlow already described. This discovery is of no small interest as bearing on the theory of progressive development, because, according to Professor Huxley, the genus Pteraspis is allied to the sturgeon, and therefore by no means of low grade in the piscine class.

It is a fact well worthy of notice that no remains of vertebrata have yet been met with in any strata older than the Lower Ludlow.

When we reflect on the hundreds of Mollusks, Echinoderms, Trilobites, Corals, and other fossils already obtained from more ancient Silurian formations, Upper, Middle, and Lower, we may well ask whether any set of fossiliferous rocks newer in the series were ever studied with equal diligence, and over so vast an area, without yielding a single ichthyolite. Yet we must hesitate before we accept, even on such evidence, so sweeping a conclusion, as that the globe, for ages after it was inhabited by all the great classes of invertebrata, remained wholly untenanted by vertebrate animals.

TABLE 26.2. DATES OF THE DISCOVERY OF DIFFERENT CLASSES OF FOSSIL VERTEBRATA; SHOWING THE GRADUAL PROGRESS MADE IN TRACING THEM TO ROCKS OF HIGHER ANTIQUITY.

COLUMN 1: YEAR.

COLUMN 2: FORMATIONS.

COLUMN 3: GEOGRAPHICAL LOCALITIES.

MAMMALIA:

1798: Upper Eocene: Paris (Gypsum of Montmartre). (George Cuvier, Bulletin Soc. Philom. 20.)

1818: Lower Oolite: Stonesfield. (In 1818, Cuvier, visiting the Museum of Oxford, decided on the mammalian character of a jaw from Stonesfield. See also above Chapter 19.)

1847: Upper Trias: Stuttgart. (Professor Plieninger. See above Chapter 21.)

AVES:

1782: Upper Eocene: Paris (Gypsum of Montmartre). (Cuvier, Ossemens Foss. Art. "Oiseaux.")

1839: Lower Eocene: Isle of Sheppey (London Clay). (Professor Owen Geological Transactions second series volume 6 page 203 1839.)

1854: Lower Eocene: Woolwich Beds. (Upper part of the Woolwich beds. Prestwich Quarterly Geological Journal volume 10 page 157.)

1855: Lower Eocene: Meudon (Plastic Clay). (Gastornis Parisiensis. Owen Quarterly Geological Journal volume 12 page 204 1856.)

1858: Chloritic Series, or Upper Greensand: Cambridge. (Coprolitic bed, in the Upper Greensand. See above Chapter 17.)