in the apparent quantity of coronal cement (ib. fig. 2 c) as well as in the evidence of a hinder talon (ib. fig. 3 t), they are more like St. insignis than St. Cliftii. Yet the two hinder ridges, with the terminal talon of the tooth (ib. figs. 3 & 4), which, in breadth, corresponds with the second upper deciduous molar of St. insignis and St. sinensis, clearly differ from both. The last two ridges run straighter across, are of the same extent, and are divided by more numerous vertical grooves into smaller and correspondingly numerous apical mamillae. The second of these ridges is cleft in the middle.
From the alleged conditions of discovery, and the little-altered condition of the above-described portions of proboscidian molars, one would be led to deem them to be of as comparatively recent geological age as our ordinary British Cave-fossils. The section, however, of Proboscidia to which they indubitably belong has not hitherto been known to be represented by fossils of later age than of an upper miocene or older pliocene period.
I believe the ground to be good for indicating this second kind of Chinese proboscidian as Stegodon orientalis, Ow.
Hyaena sinensis, Ow.
The genus Hyaena, Storr, Cuv., is represented, in the present collection, by an upper premolar, p 3 (Pl. XXVIII. figs. 5 & 6), a lower premolar, p 3 (ib. fig. 7), and by a lower canine.
The upper premolar is from the right side of the jaw: it exceeds in antero-posterior diameter that tooth in Hyaena crocuta, is still larger than that of Hyoena brunnea*, and is nearly double the size of that in the existing Asiatic species, Hyoena striata seu vulgaris. The main cone is relatively lower than that in H. crocuta ; its outer vertical contour is more convex ; and this comparison I have been careful to make with a specimen of the recent Cape species having p 3 worn in precisely the same degree as the Chinese tooth, viz. with the apex of the cone just abraded sufficiently to expose a speck of dentine. The hinder basal talon (fig. 5, t) is larger in Hyoena sinensis ; and a tubercular production abuts upon the hind ridge of the main cone as in Hyoena striata. The antero-internal tubercle (fig. 6, a) is relatively less than in Hyoena striata : the ridge rising from it toward the tip of the main cone is as prominent as in Hyoena crocuta. As in that species, there is no trace of cingulum along the outer side of the base of the crown, which is so well marked in Hyoena striata. In the main, the generic character of this massive bone-cracker is closely held by the Chinese cave-tooth.
The third upper premolar of Hyoena speloea, like the rest of the dentition, closely accords with that in Hyoena crocuta ; consequently the distinctions above noted equally hold in differentiating the Chinese p 3 from that of the Hyaena from our own caverns.
I come next to the comparison with the fossil remains of Hyoena from the Siwalik tertiaries, Hyoena sivalensis of Baker and Durand†. The third upper premolar is smaller in the Siwalik Hyaena than in
- Hyoena fusca, Bl., Osteographie (Hyoena), pl. iii.
† Journal of the Asiatic Society, October 1835, vol. iv. p. 569. See also Falconer, ' Palaeontological Memoirs,' i. p. 548.
