There also should we find the Sigillaria, its tapering and elegant form sustained on a large and firm basis, enormous matted roots almost as large as the trunk itself being given off in every direction, and shooting out their fibres far into the sand and clay in search of moisture. The stem of this tree would appear like a fluted column, rising simply and gracefully without branches to a great height, and then spreading out a magnificent head of leaves like a noble palm tree. Other trees more or less resembling palms, and others like existing firs, also abounded, giving a richness and variety to the scene; while one gigantic species, strikingly resembling the Altingia, or Norfolk Island pine, might be seen towering a hundred feet or more above the rest of the forest, and exhibiting tier after tier of branches richly clothed with its peculiar pointed and pear-like leaves, the branches gradually diminishing in size as they approach the apex of a lofty pyramid of vegetation.

Tree ferns also in abundance might there be recognized, occupying a prominent place in the physiognomy of vegetation, and dotted at intervals over the distant plains and valleys, the intermediate spaces being clothed with low vegetation of more humble plants of the same kind. These we may imagine exhibiting their rich crests of numerous fronds, each many feet in length, and produced in such quantity as to rival even the palm trees in beauty. Besides all these, other lofty trees of that day, whose stems and branches are now called Calamites, existed chiefly in the midst of swamps, and bore their singular branches and leaves aloft with strange and monotonous uniformity. All these trees, and many others that might be associ

ated with them, were perhaps girt round with innumerable creepers and parasitic plants, climbing to the topmost branches of the most lofty amongst them, and enlivening by the bright and vivid colours of their flowers the dark and gloomy character of the great masses of vegetation.

Forests like these (and I have in this description confined myself to a strict analogy, and have very cautiously abstained from transgressing the bounds of probability) are at the present day remarkable even in islands of large size, for their death-like silence, and for the almost total absence of living beings. A few birds and insects seem to form almost the whole population. In many cases no quadruped exists over extensive districts; and it is manifest that most of such islands have depended upon migrations for their inhabitants, so that they offer no guide whatever to the naturalist, when he wishes to determine from them the indigenous animals or vegetables of the district.

If, however, this is the case with the islands of the Polynesian archipelago, it cannot be doubted that the fragments of organic matter carried down into the mud and sand to form coal, and deposited in the creeks and at the mouths of rivers in ancient times, must be looked upon as offering still less evidence with respect to terrestrial animals; and what evidence exists on this subject is almost confined to about half a dozen isolated specimens of organic remains.

The conditions and contents of the newer formations of the secondary epoch, render it also probable that land animals, if existing at all, were yet more rare in the older period than afterwards, since, in the few localities where fresh-water fossils are found

under similar circumstances in newer beds, (as in the oolites and wealden,) the remains of land reptiles, and even of mammals, have been discovered.

Fig. 32

All that can be made out from the fresh-water limestones and other beds of the coal-measures is, that at this period a few insects were in existence, and were associated with some very minute crustaceans (the annexed diagram will give an idea of one of these) and several shell-fish. No fragment of a quadruped, bird, or reptile has, however, yet been obtained from any of the carboniferous strata in any part of the world, although there are not wanting in some of the sandstones associated with the coal of North America distinct indications of foot-prints referred to birds and perhaps reptiles.

The inhabitants of the sea during the carbonife

CYPRELLA.

The smaller figure is of the natural

size.

rous period are, as might be expected, much more clearly made out than those of the land; and their remains are in many cases very abundant, and sufficiently distinctive to enable us to determine the modifications and changes that had taken place since the deposition of the first fossiliferous bed.

And, first, with regard to the corals, we find indeed new species, but the differences are small and unimportant. The encrinites, so similar to the coral animal in some respects, had also been replaced by

new species, the forms having become on the whole more complicated. Higher groups, too, of radiated animals were introduced, and the sea-urchin and the star-fish, although rare, were not unknown in the sea. Trilobites still remained, and herded together chiefly in shoal water near the muddy and sandy bottoms; and some other small crustaceous animals are known from imperfect fragments of them occasionally found fossil.

But the shells of molluscous animals are too abundant, too varied, and too widely distributed in the rocks of the carboniferous period, to be passed over without some careful notice. Those of Brachiopoda (e. g. Spirifer, Productus, and Terebratula*) chiefly preponderate, although not so much so as in the older rocks, while the cephalopods, also very abundant, are developed in new forms, many of them being intermediate in their structure between the nautilus and the ammonite, and others retaining the simple form of the orthoceratite.

The Productus (fig. 34) is a very remarkable and interesting shell, although the exact nature of the animal inhabitant has not yet been satisfactorily made The shell is of a very fibrous texture, and a

out.

* Spirifer, Latin, spire-bearing; so called because a coil of carbonate of lime, useful probably in keeping the shell partly open, is not unfrequently seen in the shells of this genus. An analogous contrivance is met with in most members of the order of Brachiopoda in some shape or other.

Productus, prolonged, or drawn out in length; one valve of the shell being in most species prolonged beyond the other, and often to a great

extent.

Terebratula, from terebratus, pierced; one valve being pierced at the apex, to admit of the passage of a fibrous bundle proceeding from the other valve, serving to attach the animal to some hard substance, as a stone or rock.

number of fine tubular or hair-like appendages are, in many cases, attached to it, sometimes passing through the fibrous shell, but at other times only extending from the line of junction of the two valves, which are not connected by any hinge, and which were proFig. 33

Fig. 34

bably at once united and fastened to some solid body by this contrivance. As in the other brachiopods, there appears here also to have been a mechanical contrivance for keeping the valves partly asunder; and it would seem, that, at least in some species, the shell was very thin, and readily adapted itself to the shape of the stone on which the animal had fastened itself. The fossil shells of the productus are extremely abundant throughout the carboniferous limestone, and are found in a limestone overlying the coal-measures; but they are rarely found in the millstone grit or the coal-measures themselves, the circumstances of the deposit being probably unfavourable for the existence of such animals.

The singular spire of the Spirifer (fig. 33), often well preserved and generally occupying a considerable portion of the interior of the shell, is characteristic of the genus, and alluded to in its name.

The