ON BRITISH FOSSIL MAMMALIA. 61 



racteristic of the inner surface of the ramus of the lower jaw. The posterior 

 grinder in the present jaw is fortunately more complete than in the first ex- 

 ample, and shows a small, middle, internal cusp, with part of a large external 

 cusp, both projecting from the crown of the tooth in nearly the same transverse 

 line. The enamel covering the internal cusp, which is vertically fractured, is 

 beautifully distinct from the ivory, and considerably thicker in proportion to 

 the size of the tooth than is the enamel or its analogue in the teeth of any spe- 

 cies of reptile, recent or fossil. The six molars anterior to the one in place, are 

 broken off close to the sockets ; both the fifth and fourth false molars are en- 

 tire : the anterior cusp presents the same superior size as in the first specimen. 

 The thick external enamel, and the silky, iridescent lustre of the compact ivory, 

 are beautifully shown in these teeth. The third and second grinders are more 

 fractured than in the first specimen, but sufficient remains to show that they 

 possess the same form and relative size ; but the most interesting evidence as 

 regards the teeth, which the present jaw affords, is the existence of the sockets 

 of not less than seven te'eth, anterior to those above described. Of these 

 sockets the four anterior ones are small and simple, like those of the mole, 

 being more equal in their size and interspaces than in the Didelphys. The 

 fifth socket contained a small premolar with double fangs, and so likewise 

 did the sixth and seventh sockets. Thus the two false molars, with perfect 

 crowns in the present specimen, are the eighth and ninth teeth, counting 

 backwards, or the fourth and fifth of their class, viz. premolares or false 

 molars. This fossil, therefore, gives evidence that the dental formula of the 

 Amphitherium must have included thirty-two teeth in the lower jaw (sixteen 

 on each side) ; that these, instead of presenting an uniform, compressed, tri- 

 cuspid structure and being all of one kind, were divided into three series as 

 regards their form : five, if not six, of the posterior teeth are quinque-cuspi- 

 date, and must be regarded as molares veri. Some of the molares spurii are 

 tricuspid and some bicuspid, as in the opossums ; but these are six, if not 

 seven in number. Anterior to the molars are four simple teeth, of which the 

 fourth may be regarded as the representative of the canine, and the anterior 

 three as incisors. Thus the Amphitherium differs considerably from the ge- 

 nus Didelphys in the number of its teeth. Indeed at the time when Cuvier 

 wrote respecting it, believing it to have had ten molars, no mammiferous 

 ferine quadruped was known to possess a greater number of these teeth 

 than the Chrysochlore, which has nine molars on each side of the upper jaw, 

 and eight molars on each side of the lower jaw. The Chrysochlore, however, 

 is not the only mammal in which the molars exceed the number usually 

 found in the unguiculate Mammalia. The marsupial genus, Myrmecobius, 

 has nine molars on each side of the lower jaw, besides one small canine and 

 three conical incisors. 



The teeth of Amphitherium, moreover, differ from those of Didelphys not 

 only in number but also in size, being relatively smaller. The teeth of Myr- 

 mecobius, besides their approximation in number to those of Amphitherium, 

 resemble them in their small relative size more than do those of Didelphys, 

 but they are still smaller than in Amphitherium, which in this respect, as 

 well as in the structure of the teeth, appears to hold an intermediate posi- 

 tion between Didelphys and Myrmecobius. The Didelphys Prevostii being 

 evidently, as Cuvier states, a distinct genus from Didelphys properly so 

 called, a distinct generic name was no doubt desirable for it, and the term 

 Amphitherium fulfils all the requisite conditions. In my memoir of 1838 I 

 ventured to observe in reference to the new name proposed by M. Valen- 

 ciennes, that it would have been more prudent to have chosen one less de- 

 scriptive than Thylacotherium, since the affinities of the fossil insectivore to 



