columns of which it is composed being pressed together so as to allow of no interval between them, whereas in the last two it is of the ordinary Arvicoline character; but they have four columns both externally and internally. The penultimate molar has three columns both externally and internally, and the anterior three internally and four externally; and L. torquatus is further distinguished by a minute accessory posterior column to these two teeth. In all these characters the fossil closely agrees with L. torquatus, as well as in the form of the palatine aspect of the maxillaries and palatine bones, and of the naso-palatine foramen, whereas it differs from the other three species in these respects. The agreement in size between this upper jaw and the lower jaws last described is singular, and indicates the possibility that they may have belonged to the same animal; but as I have shown that L. torquatus and L. norvegicus belong to different sections of the genus, it is hardly probable that we have in the fossil an animal uniting the characteristics of the lower jaw of the one to the upper jaw of the other. We think it safer, for the present, to describe one as a small variety of L. norvegicus, and the other as a large variety of L. torquatus (Pl. VIII.) fig. 4, a.
8. Genus Lagomys. — Lagomys speloeus (Owen) has yielded two lower jaws — one from Hutton, found by Mr. Beard, and a second, found by Mr. Williams, which exactly corresponds in condition with it. Both are stouter than the specimens of Lagomys pusillus with which they have been compared; but the cave-Pika was certainly variable in this respect, and some jaws we have recently met with in Kent's Hole do not appear to be larger or stouter than the small recent Siberian animal; in other respects they coincide: we therefore consider it most probable that the cave animal was a variety of Lagomys pusillus.
9. Genus Lepus. — In examining very numerous bones of Lepus which we have met with, which are found side by side with the bones of Elephas primigenius and its associated fauna, and comparing them with those found in more recent beds, and with the bones of Lepus timidus, we have been much struck by the almost uniformly larger size of all the bones, particularly of the skulls of the more ancient animal. The same thing appears also to have struck the French palaeontologists; they have consequently named the larger and stouter form Lepus diluvianus (Pictet).
A complete skull (Pl. VIII. fig. 5), many and considerable fragments of others, and a large number of lower jaws and of other bones have enabled us to make a comparison for which no previous opportunity appears to have existed. These bones are from Hutton, Banwell, and several other Mendip caves, as well as from Kent's Hole and some other localities out of the county of Somerset.
In the examination of the skulls of hares generally, when a large series of individuals, as well as of species, is compared, we have found that what at first might appear to be a specific difference is so frequently obliterated by individual variation, that it is extremely difficult to seize characters by which even entire skulls can be diffe-
