472 GEOLOGICAL SURVEY OF THE TEEEITOEIES. 



lobes are wanting, but were clearly attached at the extremity of the 

 hemispheres. The superior face of the hemispheres and cerebellum 

 together have a subquadrate outline, a little wider than long. The 

 cerebellum is completely exposed behind the hemispheres, and is 

 strongly angulate at its upper posterior border to fit the inion. The 

 vermis is nearly as wide as each lateral lobe. The surface of the hemis- 

 pheres is smooth, and the sylvian fissure distinctly indicated. 



In determining the affinities of this and the two allied genera already 

 named, it is first necessary to ascertain the homologies of the cusps of the 

 molar teeth. Insectivora and some genera of liarsupialiasbve characterized 

 by the presence of three longitudinal series of tubercles on the molar teeth. 

 With the exception of the posterior molars of some Gariiivora, the 

 arrangement usual among Mammalia is in two longitudinal series, with 

 frequently but two in each row. In most Insectivora, the cusp or cusps 

 of the median series are the most prominent, and those of the outer series 

 sometimes entirely wanting. Hence, Mivart* homologizes the middle pair 

 with those of the external series of the other Jfammaim, the inner with the 

 internal, and the outer are regarded as representing cusps of abasalcinga- 

 lum. Now, in the Isacidce, (including Leptictis and Ictops,) the external 

 cusps very are largely developed; there is one well-developed median and 

 a rudimental internal cusp. Are the external cusps only cingular, or 

 homologous with those that occupy the same position in other Mammalia ; 

 and what are their relations to the corresponding ones of the premolars? 



In the true molars of the opossum and tubercular molars of the dog, 

 there are three rows of tubercles. In both, it is evident that the two of 

 the outer series correspond with the outer tubercles of the teeth which 

 precede them in the jaw, and in which the inner tubercles are reduced 

 in size and number. This is notably clear in comparison with the sec- 

 torial molar of the dog. It might, however, be asserted that the single 

 outer tubercle of the last premolar in Isacis {and Leptictis) is homologous 

 with that of the middle series of the true molars instead of the outer, 

 and some color is given to this view from the internal position of 

 the last premolar in Leptictis, so that its outer cusps range with the 

 median of the true molars. This opinion is, however, readily corrected 

 by a consideration of the arrangement in Ictops dalcotensis, t where the 

 last premolar exhibits both the median and internal cusps, so that the 

 homology of the outer pair with those of the true molars is assured. 



In Didelphys, in passing to the anterior molar, where the three series 

 of cusps are not well defined, we observe that it is the middle tubercles 

 are the ones which disappear, the internal remaining. In the same man- 

 ner, in tracing the series of forms from the horse to the tapir, we find 

 the cusps of the middle series disappear, leaving the internal and 

 external to represent those of the original quadritubercular molar. In 

 Ictops and probably Isacis, the median cusp is preserved at the interval 

 of the last premolar. The case is quite different in Talpa and Scalops, 

 where it is evident that, as we advance along the dental series forward, 

 the cusps of the outer row disappear, and the external ones remain- 

 ing represent those of the middle series of the true molars. As already 

 indicated, the arrangement in Herpetotherium is as in Isacis, and the 

 premolars have, therefore, an entirely distinct structure from that 

 observed in the Talpidce. 



From the above considerations, it appears that the external, often mi- 

 nute, cusps of the teeth of Insectivora are the homologues of those of 



* See his valuable Memoir on Osteology of Insectivora, Journ. Anatomy and Physi- 

 ology, vol. ii, p. 138. 

 t See Leidy's Extinct Fauna of Dakota and Nebraska, p. 351, PI. xxvi. Fig. 29. 



