Immediately below is a great and conspicuous stratum, twelve feet thick, formed of a vast accumulation of shells of Ostrea distorta (Fig. 300), long familiar to geologists under the local name of “Cinder-bed.” In the uppermost part of this bed Professor Forbes discovered the first echinoderm (Fig. 301) as yet known in the Purbeck series, a species of Hemicidaris, a genus characteristic of the Oolitic period, and scarcely, if at all, distinguishable from a previously known Oolitic fossil. It was accompanied by a species of Perna. Below the Cinder-bed fresh-water strata are again seen, filled in many places with species of Cypris (Fig. 302, a, b, c), and with Valvata, Paludina, Planorbis, Limnæa, Physa (Fig. 303), and Cyclas, all different from any occurring higher in the series. It will be seen that Cypris fasciculata (Fig. 302, b) has tubercles at the end only of each valve, a character by which it can be immediately recognised. In fact, these minute crustaceans, almost as frequent in some of the shales as plates of mica in a micaceous sandstone, enable geologists at once to identify the Middle Purbeck in places far from the Dorsetshire cliffs, as, for example, in the Vale of Wardour in Wiltshire. Thick beds of chert occur in the Middle Purbeck filled with mollusca and cyprides of the genera already enumerated, in a beautiful state of preservation, often converted into chalcedony. Among these Professor Forbes met with gyrogonites (the spore-vessels of Chara), plants never until 1851 discovered in rocks older than the Eocene. About twenty feet below the “Cinder-bed” is a stratum two or three inches thick, in which fossil mammalia presently to be mentioned occur, and beneath this a thin band of greenish shales, with marine shells and impressions of leaves like those of a large Zostera, forming the base of the Middle Purbeck.

Fossil Mammalia of the Middle Purbeck.—In 1852,[[1]] after alluding to the discovery of numerous insects and air-breathing mollusca in the Purbeck strata, I remarked that, although no mammalia had then been found, “it was too soon to infer their non-existence on mere negative evidence.” Only two years after this remark was in print, Mr. W. R. Brodie found in the Middle Purbeck, about twenty feet below the “Cinder-bed” above alluded to, in Durdlestone Bay, portions of several small jaws with teeth, which Professor Owen recognised as belonging to a small mammifer of the insectivorous class, more closely allied in its dentition to the Amphitherium (or Thylacotherium) than to any existing type.

Four years later (in 1856) the remains of several other species of warm-blooded quadrupeds were exhumed by Mr. S. H. Beckles, F.R.S., from the same thin bed of marl near the base of the Middle Purbeck. In this marly stratum many reptiles, several insects, and some fresh-water shells of the genera Paludina, Planorbis, and Cyclas, were found.

Mr. Beckles had determined thoroughly to explore the thin layer of calcareous mud from which in the suburbs of Swanage the bones of the Spalacotherium had already been obtained, and in three weeks he brought to light from an area forty feet long and ten wide, and from a layer the average thickness of which was only five inches, portions of the skeletons of six new species of mammalia, as interpreted by Dr. Falconer, who first examined them. Before these interesting inquiries were brought to a close, the joint labours of Professor Owen and Dr. Falconer had made it clear that twelve or more species of mammalia characterised this portion of the Middle Purbeck, most of them insectivorous or predaceous, varying in size from that of a mole to that of the common polecat, Mustela putorius. While the majority had the character of insectivorous marsupials, Dr. Falconer selected one as differing widely from the rest, and pointed out that in certain characters it was allied to the living Kangaroo-rat, or Hypsiprymnus, ten species of which now inhabit the prairies and scrub-jungle of Australia, feeding on plants, and gnawing scratched-up roots. A striking peculiarity of their dentition, one in which they differ from all other quadrupeds, consists in their having a single large pre-molar, the enamel of which is furrowed with vertical grooves, usually seven in number.

The largest pre-molar (see [Fig. 305]) in the fossil genus exhibits in like manner seven parallel grooves, producing by their termination a similar serrated edge in the crown; but their direction is diagonal—a distinction, says Dr. Falconer, which is “trivial, not typical.” As these oblique furrows form so marked a character of the majority of the teeth, Dr. Falconer gave to the fossil the generic name of Plagiaulax. The shape and relative size of the incisor, a, [Fig. 306,] exhibit a no less striking similarity to Hypsiprymnus. Nevertheless, the more sudden upward curve of this incisor, as well as other characters of the jaw, indicate a great deviation in the form of Plagiaulax from that of the living kangaroo-rats.

There are two fossil specimens of lower jaws of this genus evidently referable to two distinct species extremely unequal in size and otherwise distinguishable. The Plagiaulax Becklesii (Fig. 306) was about as big as the English squirrel or the flying phalanger of Australia (Petaurus Australis, Waterhouse). The smaller fossil, having only half the linear dimensions of the other, was probably only one-twelfth of its bulk. It is of peculiar geological interest, because, as shown by Dr. Falconer, its two back molars bear a decided resemblance to those of the Triassic Microlestes ([Fig. 389]), the most ancient of known mammalia, of which an account will be given in Chapter XXI.