274 P. N. Bose—History of the Extinct Carnivora. 
the outer border in front of the anterior pair of cusps in the inter- 
space between them. The point in which they meet is slightly 
elevated; if it were better developed (as it is in Rhynchocyon 
and Centetes), the tooth would assume the form of a typical lower 
carnassial. The two posterior tubercles with the space intervening 
between them and the anterior pair would form the “ heel,” and the 
two anterior cusps with the accessory cusp in front the strictly 
“carnassial” part. It may be observed that the development of 
this cusp is inversely proportional to that of the antero-internal 
cusp and to the tubercular part behind (the ‘“ heel”); and in the 
Felide, in which the anterior accessory cusp is best differentiated, 
the representative of the primitive antero-internal cusp (present in 
the Canidae, Viverride, etc.) as well as the “heel” disappear alto- 
gether. The two hinder molars of Gymnura are similar in form to 
the first molar. The molars of all other Insectivora are constructed 
on the same type as those of Gynmura; and they may be regarded 
as a series of three more or less specialized structurally ‘“ carnas- 
sial ’ teeth, the degree of specialization varying greatly from Condy- 
lura cristata,—in which both the original outer cusps are of the same 
height and equally pointed, so that there is no distinction between 
the ‘carnassial ” and the “tubercular” portion,—to Centetes, in which 
the heel is very short and very low. 
Of the Hocene Carnivora our knowledge of the lower teeth of 
Arctocyon is very imperfect. Palgonictis, as well as Didymictis, 
have four premolars and two molars on each side, as in Viverra. 
Cynodon and Uintacyon have each one molar more, as in Canis. 
Uintacyon has, besides, one premolar in excess. There is one very 
important peculiarity common to these forms. In the existing 
Carnivores, in the lower, as well as in the upper, jaw, only one tooth 
is differentiated as a ‘‘carnassial”; if it is followed by any other 
teeth, these are structurally strictly “tuberculous,” as they are 
functionally true grinders. But in the genera just mentioned, the 
second molar (and in the case of Uintacyon, the third also) is 
the exact counterpart of the first. In Palgonictis, both the molars 
are composed of three pointed cusps in front, and a “heel” behind, 
the back molar, however, being smaller, and, so to say, less ‘ car- 
nassial” than the first. Didymictis of Cope agrees so closely with 
Palgonictis, in the number as well as the form of the teeth, that the 
former may be conveniently affiliated with the latter. In Cynodon, 
the second molar is proportionately much smaller than the corre- 
sponding tooth in Paleonictis; nevertheless, as in the latter, it is a 
miniature of the first molar. M. Gaudry, remarking on this feature 
of the molars in Paleonictis, says, that it is not met with amongst 
the Placentals, and that it is peculiar to the Marsupials.’ He 
compares the teeth in question to those of a species of Dasyure. 
But in this genus the second molar is larger and better differentiated 
than the first; whereas, it is just the reverse in Palgonictis. The 
resemblance of the molars of Palgonictis, or of the first two molars 
of Cynodon, to the corresponding teeth of some Insectivores, as, for 
1 op, cit. p. 19. 
