Chapter VI: The First Air-Breathers

Were our experience limited to the animals whose remains are found in the earlier PalA|ozoic rocks, we might be unable to conceive the possibility of an animal capable of living and breathing in the thin and apparently uncongenial medium of air. More especially would this appear doubtful if our experience of the atmosphere presented it to us as loaded with carbonic acid, and less rich in vital air than it is at present. Even the mechanical difficulties of the case might strike us as considerable, in our ignorance of the capabilities of limbs. Still, as time wore on, we should find this problem worked out along three distinct lines of advancementaEuro"those of the Mollusk, the Arthropod, and the Vertebrate, and in each of these with different machinery, related to the previous locomotive and water-breathing apparatus of the type.

Respiration under water depends, not on the water itself, but on the small percentage of free oxygen which it contains, and this is utilised for the aA"ration of the blood of animals, by that wonderful and often extremely beautiful apparatus of delicate fibres or laminA| penetrated with blood-vessels, which we call a gill. Except those lowest creatures which aA"rate their blood merely at the general surface of the body, all animals capable of respiration in water are provided with gills in some form, though in many of the humbler types, like that of the familiar Oyster, the gills are used for the double purpose of aA"rating the blood and, by their minute vibrating threads or cilia, drifting food to the mouth.

In the great group of radiated animals, the _Protozoa_, _CA"lenterata_, and _Echinodermata_, no air-breathing creature exists, or, in so far as is known, has existed, so that this vast group of animals is limited altogether to the waters; and this is undoubtedly one mark of its inferiority.

In the sub-kingdom of the Mollusks the highest class, that of the Cuttle-fishes and Nautili, has been, singularly enough, rejected as unfit for this promotion, though it was early introduced, and attains to a high development of muscular energy and nervous power. The group next in order, that of the Snails and their allies, alone ventures in some of its families to assume the _rA'le_ of air-breathing. As might be expected, in creatures of this stamp the simplest means are employed to effect the result. In the sub-aquatic species the gills are contained in a chamber, where they are protected and kept supplied with water. In the air-breathing species, this gill-chamber is merely emptied of its contents and converted into an air-sac or functional lung. Thus a rude and imperfect method of air-breathing is contrived, which scarcely separates the animals that possess it from their aquatic relatives, but which nevertheless gives to us the beautiful and varied groups of the Land-snails and of the air-breathing fresh-water Snails.

In the worms and Crustaceans the gills are placed at the sides of the body, and connected with its several segments. But the Crustaceans, like the Cuttle-fishes, though the highest aquatic type, never become air-breathers. It is true some of them, like the Land-crabs, live in the air; but they retain their gills, and have to carry with them a supply of water to keep these moist.

But in order to elevate the Annulose type to the true dignity of air breathing, three new classes had to be introduced, differing altogether in their details of structure; and all three seem to have been placed on the earth about the same time. They are: First, the Myriapods, or Gallyworms and Centipedes; secondly, the Insects; and thirdly, the Arachnidans, or Spiders and Scorpions.

In the Myriapods a system of air-tubes, kept open by elastic spiral fibres, penetrates the body by lateral pores, thus retaining the resemblance to the lateral respiration of the Crustaceans and worms. In the Insects, where this type of structure rises to its highest mechanical perfection, and where the animal is enabled to be not merely an air-breather, but a flier, the same system of lateral pores and internal air-tubes is adopted, and is so extended and ramified as to give a very perfect respiration. In the Spiders and Scorpions the system is the same, except that in the latter and a part of the former the whole or a part of the tracheal system becomes expanded into air-chambers simulating true lungs.

Among the Vertebrates, the fishes are breathers by gills attached to arches at the sides of the neck. But already in the Devonian we have reason to believe that there were fishes having the swimming-bladder opening into the back of the mouth to receive air, and divided into chambers, so as to constitute an imperfect lung. And here we have not, as in the lower types, an adaptation of the old water-breathing organs, but an entirely new apparatus. In the next grade of Vertebrates we find, as in the Frogs, Water-lizards, etc., that the young are aquatic and breathe by gills, while the adults acquire lungs, sometimes retaining their gills also, but in the higher forms parting with them. Thus in the vertebrates alone we have true lungs, distinct structurally from gills; and these lungs attain to their highest perfection in the birds and mammals.

_a_, _Platephemera antiqua_ (Scudder).
_b_, _Homothetus fossilis_ (Scudder).
_c_, _Lithentomum Harttii_ (Scudder).
_d_, _Xenoneura antiquorum_ (Scudder).]

The oldest air breathers at present known are Scorpions and insects allied to the modern May-flies, which have been found in the Silurian. Next to these, and more important in number and variety, are the insects of the Erian plant beds of New Brunswick. They were discovered by the late lamented Prof. C. F. Hartt in the plant-bearing shales of the Middle Devonian (Fig. 123). The beds containing them hold also a species of _Eurypterus_, an obscure Trilobite, and a Crustacean allied to the modern Stomapods,[42] besides a shell which may possibly be that of a Land-snail, to be mentioned in the sequel. They are also exceedingly rich in beautifully-preserved remains of Devonian plants. The collection made by Prof. Hartt is limited to a few fragments of wings; but these, in the skilful hands of Mr. Scudder, have proved to be rich in geological interest. One is a gigantic _Ephemera_ or May-fly, which must have been five inches in the expanse of the wings, which are more complex in their venation than those of its modern allies (Fig. 123, _a_). Another presents peculiarities between those of the May-flies and Dragon-flies (Fig. 123, _b_). A third is a Neuropter, not belonging to any known family, but allied to some in the Coal-formation (Fig. 123, _c_). A fourth (Fig. 123, _d_) is a small and delicate wing, supposed to have belonged to an animal having some points of resemblance to the modern crickets. Two others are represented by mere fragments of wings, insufficient to determine their affinities with certainty. No other insects of this age have been discovered elsewhere; but it is to be borne in mind that no other locality rich in Devonian plants has probably been so thoroughly explored. The hard slaty ridges containing these fossils are well exposed on the coast near the city of St. John, and Messrs. Hartt and Matthew of that city, acting, I believe, in concert with and aided by the Natural History Society of the place, not only searched superficially, but removed by blasting large portions of the richest beds, and examined every fragment with the greatest care. Their primary object was fossil plants, of which they obtained magnificent collections; and it is scarcely possible that the insects could have been found but for the exhaustive methods of exploration employed.

It is interesting to observe, respecting these oldest insects, that they all belong to those families which have jaws, and not suctorial apparatus, that they are not of those which undergo a complete metamorphosis, and that their modern congeners pass their larval stage in the water. Thus the waters gave birth to the first insects, and their earliest families were not of those which suck honied juices or the blood of animals, or which pass through a worm-like infancy. These groups belong apparently to much later times.

On one of the specimens collected by Messrs. Hartt and Matthew, and placed by them in my hands, is a spiral form which in every particular of external marking resembles a genus of modern West Indian Land-snails.[43] I have hesitated to describe it, as the structure is lost and the form imperfect; but I cannot help regarding it as an indication that this group of land animals also will be traced back to the Devonian age.

Ascending from the Devonian to the Carboniferous, we at once find ourselves in the midst of air-breathers of various types. Here are Myriapods, insects of several orders, Spiders, Scorpions, Land-snails, and Batrachian reptiles, and these of many species, and found in many localities widely separated. We can thus people those dark, luxuriant forests, to which we owe our most valuable beds of coal, with many forms of life; and as most of these belong to tribes likely to multiply abundantly where food was plentiful, we can imagine multitudes of Snails and Millepedes feeding on succulent or decaying vegetable matter, swarms of insects flitting through the air in the sunnier spots, while their larvA| luxuriated in decaying masses of leaves or wood, or peopled the pools and streams. In like manner, in imagination we can render these old woods vocal with the trill of crickets and with the piping or booming of smaller and larger Batrachians. Let us now, in accordance with our plan, inquire as to the nature of these early air-breathers and the fortunes of their families in the geological history.

_a_, Natural size. _b_, Magnified. _c_, Apex. _d_, Sculpture. Enlarged.]

_a_, Shell. Enlarged; the line below shows the natural size. _b_, Sculpture. Enlarged.]

The Land-snails known as yet in the Carboniferous are limited to five or six species, belonging to four genera, all American and related to existing American forms. The two earliest known are represented in Figs. 124 and 125.[44] One of them is a _Pupa_, or elongated Land-snail, so similar to modern forms that it does not merit a generic distinction, and is indeed very near to some existing West Indian species. The other is in like manner a member of the modern genus _Zonites_. These are from the Coal-formation of Nova Scotia, and the Pupa must have been very abundant, as it has been found in considerable numbers in a layer of shale, and in the stumps of erect trees, in beds separated from each other by a thickness of 2,000 feet of strata. The _Zonites_ is much more rare. A second Pupa is found in Nova Scotia, and two species occur in the Coal-field of Illinois. One of these is a Pupa still smaller than _P. vetusta_, and, like some modern species, with a tooth-like process on the inner lip. The other has been placed in a new genus,[45] but is very near to some of the smaller American Snails still living. Its most special character is a plate extending from the inner lip over half the aperture, a contrivance for protection still seen in some modern forms. Thus the Land-snails come on the stage in at least three generic forms, similar to those which still live, but all of small size, indicating perhaps that the conditions were less favourable for such creatures than those of the temperate and warmer climates at present. It may seem a small step in advance for Sea-snails to lose their gills and to become Land-snails, and this without any elevation of their general structure; but it must be borne in mind that we have here not only the dropping of the gills for an air-sac, but profound changes in teeth, mucous glands, shell, and other particulars, to fit them for new food and new habits. It is also singular that the Land-snails at once appear instead of the intermediate forms of the air-breathing fresh-water snails. These last may, however, yet be found.

The Millepedes, like the Land-snails, were first found in the Coal-formation of Nova Scotia, but species have since been discovered not only in Illinois, but also in Great Britain and in Bohemia. In Nova Scotia alone two genera and five distinct species have been found, all in the interior of erect trees, to which these creatures probably resorted for food and shelter (Fig. 126). All the species yet known are allied to the modern Gallyworms, though presenting special features which seem to separate them as a distinct family,[46] and were probably vegetable-feeders. Some of the species have the peculiarity, unknown among their modern successors, of being armed with long spines.[47] The moist, equable climate and exuberant vegetation of the Coal-period would naturally be very favourable to Millepedes, and it is likely that the discoveries made as yet give but a faint idea of their actual abundance. It is not improbable that they subsequently declined, as we know of none between the Carboniferous and the Jurassic, and they do not seem to have improved up to the modern period. The Carnivorous Myriapods, however, or Centipedes proper, a higher and essentially distinct type, are not known until much more recent times.

_a_, _Xylobius sigillariA|_ (Dawson). _b_, _Archiulus xylobioides_ (Scudder). Anterior segments. Enlarged, _c_, _X. farctus_ (Scudder). Caudal portion. Enlarged.]

The insects of the Carboniferous as yet known, belong to three out of the ten or more orders into which the class is divided. One of these is represented by a number of species of Cockroach, another by May-flies and a Dragon-fly, and another by some weevil-like Beetles. The Cockroach is characterised by Huxley as one of the aEurooeoldest, least modified, and in many ways most instructive forms of insects;aEuroŁ and both he and Rolleston take its anatomy as typical of that of the class. That these creatures should have abounded in the Coal-period we need not wonder, when we consider the habits of those that infest our houses, and when we further bear in mind the number of species, some of them two inches in length, that exist in tropical climates. So many species of this family have been found in the Coal-formation on both sides of the Atlantic,[48] that we may fairly regard them as constituting one of its most characteristic features, and as probably the oldest representatives of the order to which they belong[49] (Fig. 127). There were also in the Coal-period insects allied to the Locusts and to the Mantids, a carnivorous group. One of the latter (_Lithomantis_), described by Woodward, is a magnificent insect, not unlike some modern tropical species. It was found in the Coal-formation of Scotland. A still larger species, probably the largest insect known, has been described by Brongniart. The May-flies (_EphemeridA|_) are represented in the Carboniferous by several very large species. That of which the wing is shown in Fig. 128 must have been seven inches in expanse of wings. The habits of the modern May-flies show us how animals of this group, living as larvA| in the streams and lakes, must have afforded large supplies of food to fishes, and when mature must have emerged from the waters in countless myriads, filling the air for the brief term of their existence in the perfect state. The May-flies represent another insect order.[50] The Coal-measures of Saarbruck have afforded several species allied to the white ants (_Termites_), insects which must have found abundant scope for their activity in the dead trees of the carboniferous forests. The occurrence of beetles,[51] especially of the weevil family, which have as yet been found only in Europe, might have been expected, considering the habits and modern distribution of this group. It has been asserted that moths[52] have been found in the Carboniferous; but the proof of this, so far as known to me, is the occurrence of leaves, noticed by Sternberg, with markings similar to those made by the larvA| of minute leaf-mining moths. This, however, is uncertain evidence. If we consider the orders of insects not found in the Coal-formation, we can perceive good reasons for the absence of some of them. Those containing the lice and fleas, and other minute and parasitic insects, we can scarcely expect to find. The bees and wasps, and the butterflies and moths, are little likely to have been present where there were scarcely any flowering plants; but such groups as those of the two-winged flies, the plant-bugs and the ants, we might have expected, but for the fact of their being highly specialised forms, and for that reason likely to have appeared later.[53] There are, indeed, as yet no haustellate or suctorial insects known in this early period. Plausible theories of the phylogeny of insects are not wanting; but they do not well suit the known facts as to their first appearance; and perhaps we may venture without much blame to apply to the insects of the Coal-period the remark made by Wollaston with reference to the rich insect fauna of the isolated rock of St. Helena: aEurooeTo a mind which, like my own, can accept the doctrine of creative acts as not necessarily aEuro~unphilosophical,aEuro(TM) the mysteries [of the existence of these species in an island so remote from other lands], however great, become at least conceivable; but those which are not able to do this may, perhaps, succeed in elaborating some special theory of their own, which, even if it does not satisfy all the requirements of the problem, may at least prove convincing to themselves.aEuroŁ

_a_, _Archimulacris Acadicus_ (Scudder). _b_, _Blattina Bretonensis_ (Scudder). _c_, _B. Hesri_ (Scudder).]

The suctorial insects make their first certain appearance in the Jurassic; and the magnificent Sphinx Moth in Fig. 129 is an example of the magnitude and perfection to which that tribe attained in the age of the Solenhofen slate; though Weyenburgh, who describes it, fancies that he sees evidence that it may, unlike any modern moths, have been provided with a sting. The most perfect and beautiful fossil butterfly known to me is that represented in Fig. 130, from a photograph kindly given to me by Mr. Scudder. It is from the Tertiary rocks of Western America, and is laid out in stone as neatly as if prepared by an entomologist, while its preservation is so perfect that even the microscopic scales on the wings can be made out. It belongs to one of the highest types of modern butterflies, that to which the _VanessA|_ belong, but with some points of structure pointing to the lower group of the aEurooeSkippersaEuroŁ (_HesperiadA|_). Scudder remarks that while the fore-wings resemble those of the former group, the hind-wings look more like those of the latter; and this seems to be a common character of two or three others of the few fossil species known, none of which are older than the Tertiary.

]

We know too little of the spiders and scorpions of the Carboniferous to say more than that they closely resemble modern forms. Two of the scorpions are represented in Figs. 131 and 132; and the only spider certainly known, which is from Silesia, is said to belong to the group of the hunting or trap-door spiders (_Lycosa_).[55]

The Batrachians of the Coal are its most characteristic and remarkable air-breathers,aEuro"especially so as the precursors of the reptiles of the Mesozoic age. Cope in a recent summary enumerates no less than thirty-nine genera and about one hundred species; and to these have to be added at least a dozen more recently discovered in Europe; though it was only in 1841 that the first indications of such creatures were found, and were then regarded by geologists with the same scepticism which some of them still apply to _Eozoon_. The first trace ever observed of batrachians in the Carboniferous consisted of a series of small but well-marked footprints found by the late Sir W. E. Logan in the Lower Carboniferous shales of Horton Bluff, in Nova Scotia. In that year this painstaking geologist had examined the coal-fields of Pennsylvania and Nova Scotia, with the view of following up his important discovery of the _StigmariA|_, or roots of _Sigillaria_, as accompaniments of the coal-underclays. On his return he read a paper, detailing his observations, before the Geological Society of London. In this he mentioned the footprints in question; but the paper was published only in abstract, and the importance of the discovery was overlooked for a time, the anatomists evidently being shy to acknowledge the validity of the evidence for a fact so unexpected. Fig. 133 is a representation of another slab subsequently found in beds of the same age in Nova Scotia, and which may serve to indicate the nature of Sir WilliamaEuro(TM)s discovery. In consequence of the neglect of this first hint by the London geologists, the discovery of bones of a batrachian by von Dechen at Saarbruck in 1844, and that of footprints by King in Pennsylvania in the same year, are usually represented as the first facts of this kind. My own earliest discovery of reptilian bones in Nova Scotia was made in 1844, though not published till some time afterward, and was followed up by further collections in company with Sir Charles Lyell in 1851, at which time also the earliest land-snail was found, and in the following year the first millepede. Since that time the progress of discovery has been astonishingly rapid, and has extended over most of the principal coal-areas on both sides of the Atlantic.

We may, for convenience, call these animals reptiles, but they are regarded as belonging to that lower grade of reptilian animals, the Amphibians or Batrachians, which includes the modern frogs and newts and water-lizards.[56] Still it would be doing great injustice to the carboniferous reptiles not to say, that while related to this low type, they presented a much greater range of organisation than it shows at present, evincing a capability to fill most of the places now occupied by the true reptiles. Some of them were aquatic, and with limbs rudimentary or little developed, but many of them walked on the land, and were powerful and predaceous creatures. They had large and complex teeth, they were protected by external bony plates, and some of them had in addition a beautiful covering of horny plates and spines, and ornamental lappets. Many had well-developed ribs, indicating a condition of respiration much in advance of that in the ribless batrachians. Some of them attained to size and strength rivalling those of the modern alligators, while some of the smallest species exhibit characters approaching in some respects to the lizards.

Perhaps the most fish-like of these animals are those first discovered by von Dechen (_Archegosaurus_, Fig. 134). Their long heads, short necks, supports for gills, feeble limbs and long flat tail, show that they were aquatic creatures presenting many points of resemblance to the Ganoid fishes which must have been their companions. Yet they show what no fish can exhibit, fore and hind limbs with proper toes, and the complete series of bones that appear in our own arms and legs, while they must have had true lungs and breathed through nostrils. So different are they from the fish in details, that a single limb bone, a vertebra, a rib, or a fragment of a skull bone, suffices to distinguish them. Much has been said recently of the genesis of limbs; and here, as far as now known, we have the first true limbs; but it is scarcely too much to say that the feet of _Archegosaurus_ differ more from the fins of any carboniferous fish than they do from the human hand; while it is certain that the feet which made the impressions represented in Fig. 133, on the lowest beds of the Carboniferous, or that from the upper coal-formation represented in Fig. 139, were not less typical or perfectly formed feet than those of modern lizards.

Leaving these fish-like forms, we find the remainder of the carboniferous reptiles to diverge from them along three lines.

The first leads to snake-like creatures, destitute of limbs, and which must have been functionally the representatives of the serpents in the PalA|ozoic, though batrachian in their affinities (Fig. 135). They are found both in Europe and America; and Huxley describes one from Ireland more than twenty-one inches long, and with over one hundred vertebrA|.[57] Some extraordinary traces are found on the sandstones of the coal-formation,[58] which appear to indicate that there may have been species of this type much larger than any represented by skeletons, and with bodies perhaps six inches in diameter. It is not unlikely that they had the habits of the modern water snakes.

_a_, Anterior part of the skull, viewed from beneath. One-sixth natural size, _b_, One of the largest teeth, natural size.]

A second line leads upward to large crocodile-like creatures, with formidable teeth, strong bony armour, and well-developed limbs (_Labyrinthodontia_, Figs. 136, 137). Some of them must have attained a length of ten feet. They were lizard-like in form, could walk well, as is seen from the footprints of some of the species which present a considerable stride, and moved over mud without the belly touching the ground. Their tails were long, and probably useful in swimming. Their heads were flat and massive, and their teeth were strengthened by a remarkable folding inward of the outer plate of enamel (Fig. 137 _b_). The belly was protected by bony plates and closely imbricated scales. In some of the species at least the upper parts were clothed with horny scales, and the throat and sides were ornamented with pendent scaly fringes or lappets.[59] Their general aspect and mode of life must have resembled those of modern alligators; and in the vast swamps of the Coal-period, full of ponds and sluggish streams swarming with fish, they must have found a most suitable abode. While rigid anatomy may ally these animals rather with the batrachians than the true reptiles, it is evident that their great size, their capacity for walking with the body borne well above the ground, their bony and scaly armour, their powerful teeth and their capacious chests, with well-developed ribs, indicate conditions of respiration and general vitality quite comparable with those of the highest modern members of the class Reptilia.

_a_, Fragment of maxillary bone showing sculpture, four outer teeth, and one inner tooth. Natural size. _b_, Section of inner tooth. Magnified, _c_, Dermal scale. Natural size.]

The third line of progress leads to some slender and beautiful creatures (_Microsauria_), chiefly known to us by remains found in erect trees, and which resembled in form and habits the smaller modern lizards. They have simple teeth, a well-developed brain-case, limbs of some length, and bony and scaly armour, the latter in some cases highly ornate.[60] They were probably the most thoroughly terrestrial, and the most active of the coal batrachians, if indeed they were not strictly intermediate between them and the lizards proper. Fig. 138 shows some fragments of one of these animals; and the animal represented in Fig. 139, recently figured by Fritsch, probably belongs to this group.

_a_, Maxillary bone; enlarged.
_b_, Mandible; enlarged.
_c_, Teeth; magnified, showing front and side view of
ordinary tooth and grooved anterior tooth.
_d_, Section of tooth; magnified.
_e_, Scale; natural size and magnified.
_f_, Pelvic bone (?); natural size.
_g_, Rib; natural size.
_h_, Scapular bone (?); natural size.
_i_, Palate; natural size.]

The Labyrinthodonts of the Carboniferous continue upward into the Permian, where they meet with the true reptiles; and in the earlier Mesozoic some of the largest and most typical examples are found.[61] But here their reign ceases, and they give place to reptiles of more elevated type, whose history we must consider in the next chapter.

Nothing can be more remarkable than the apparently sudden and simultaneous incoming of the batrachian reptiles in the Coal-formation. As if at a given signal, they came up like the frogs of Egypt everywhere and in all varieties of form. If, as evolutionists suppose, they were developed from fishes, this must have been by some sudden change, occurring at once all over the world, unless indeed some great and unknown gap separates the Devonian from the CarboniferousaEuro"a supposition which seems quite contrary to factaEuro"or unless in some region yet unexplored this change was proceeding, and at a particular time its products spread themselves over the worldaEuro"a supposition equally improbable. In short, the hypothesis of evolution, as applied to these animals, is surrounded with geological improbabilities.

A remarkable picture of the conditions of PalA|ozoic land life is presented by the occurrence of remains of reptiles, millepedes and land-snails in such erect trees as that represented in Fig. 140. In the now celebrated section of the South Joggins in Nova Scotia, trees of this kind occur at more than sixty different levels; but only in one of these have they as yet been found to be rich in animal remains. Fortunately this bed is so well exposed and so abundant in trees, that I have myself, within a few years, removed from it about twenty of them, the greater number affording remains of land animals.

1. Underclay, with rootlets of _Stigmaria_, resting on gray shale,
with two thin coaly seams.
2. Gray sandstone, with erect trees, _Calamites_, and other stems:
9 feet.
3. Coal, with erect tree on its surface: 6 inches.
4. Underclay with _Stigmaria_ rootlets.

_a_, _Calamites_.
_b_, Stem of plant undetermined.
_c_, _Stigmaria_ roots.
_d_, Erect trunk, 9 feet high.]

The history of one of these trees may be shortly stated thus. It was a _Sigillaria_, perhaps two feet in diameter, and its stem had a dense and imperishable outer bark, a soft cellular inner bark liable to rapid decay, and a slender woody axis not very durable. It grew on the surface of a swamp, now represented by a bed of coal. By inundations and by subsidence, this swamp was exposed to the invasion of muddy and sandy sediment, and this went on accumulating until the stem of the tree was buried to the height of about nine feet, before which time it was no doubt killed. After a time the top decayed and fell, leaving the buried stump imbedded in the sandy soil, which had now become dry, or nearly so. The trunk decayed, its inner bark and axis rotting away and falling in shreds into the bottom of the cylindrical hole, about nine feet deep, once occupied by the stem, and now kept open like a shaft or well by the hard resisting outer bank. The ground around this opening became clothed with ferns and reed-like _Calamites_, partly masking and concealing it. And now millepedes and land snails made the buried trunk a home, or fell into it in their wanderings; and small reptiles sporting around, in pursuit of prey, or themselves pursued, stumbled into the open pitfall, and were unable to extricate themselves, though I have found in some of the layers in these trees trails which show that these imprisoned reptiles had wearily wandered round and round, in the vain search for means of exit, till they died of exhaustion and famine. The bones of these dead reptiles, shells of land-snails and crusts of millepedes, accumulated in these natural coffins, and became mixed with vegetable debris falling into them, and with thin layers of mud washed in by the rains; and this process continued so long that a layer of six inches to a foot in thickness, full of bones, was sometimes produced. At length a new change supervened, the area was again inundated and drifted over with sand, and the hollow trunk was filled to the top and buried under many feet of sediment, never to be re-opened till, after the whole had been hardened into sandstone and elevated to form a part of the modern coast, when the old tree and its forest companions which had shared the same fate with it, are made to yield up their treasures to the geologist. This history is no fancy picture. It represents the results of long and careful study of the beds holding these erect trees, and of the laborious extraction of great numbers of them, and the breaking-up of their contents into thin flakes, to be carefully examined with the lens under a bright light in search of the relics they contained. Fig. 11 in Chap. I. represents the extraction of one of these trees, which happened to be partially exposed by the wasting of the cliff; but many others had to be laboriously mined out of the rock by blasting with gunpowder.

_a_, Mineral charcoal. _b_, Dark-coloured sandstone, with plants, bones, &c. _c_, Gray sandstone, with Calamites and Cordaites.]

It is evident that the combination of circumstances referred to above could not often occur; and it is therefore not wonderful that only in one place and one bed has evidence of it been found, and that even in this some of the trees have been filled up at once by sand and clay, or so crushed by falling in or lateral pressure, that they could receive no animal remains. In one respect this is a striking evidence of the imperfection of the geological record, since, but for what may be called a fortunate accident, many of the most interesting inhabitants of the coal forests might have been altogether unknown to us. On the other hand, it shows how strange and unexpected are the ways in which the relics of the old world have been preserved for our inspection, and that there is probably scarcely any animal or plant that has ever lived of which some fragment does not exist, did we know where to look for it.

It may be well to remark, in closing this chapter, how many new forms of life, air-breathing and otherwise, make their first appearance in the Carboniferous, and have continued to prevail until now. Here we find the first specimens of Amphibians, Spiders, Myriapods, Orthopterous and Coleopterous Insects, and of the Crabs among ten-footed Crustaceans. In the latter group Woodward has recently described the oldest known crab, from the Coal-formation of Belgium.

_Pliosaurus_, _Ichthyosaurus_, _Plesiosaurus_, _Mososaurus_, and _Teleosaurus_.]