790 THE WHITE RIVER FAUNA. 



The general osteology of this genus is unknown, as the only portions of 

 the skeleton yet discovered are mandibular rami and the cranium anterior 

 to the middle of the orbits. From these it appears that the cranial sutures 

 are distinct at maturity, and that the nasal bones are elongate and well 

 developed Either the maxillary bounds the orbit in front or there is a 

 large lachrymal bone. A pit on the anterior rim of the orbit is probably 

 the lachrymal canal. The foramen infraorhitale is well anterior to the orbit. 

 The symphysis mandibuli is loose. 



In determining the affinities of this and other questionable genera of 

 this epoch it is first necessarj^ to ascertain the homologies of the cusps of the 

 molar teeth. The opossums are characterized by the presence of three 

 longitudinal series of tubercles on the superior molars. The homologies 

 of these cusps are rendered clear by the character presented by the fourth 

 superior premolar, where the anterior intermediate cusp is wanting. The 

 external cusps are really such, and are not developed from a cingulum external 

 to the true external cusps, as appears at first sight to be the case with such 

 animals as the Talpidce. The intermediate cusps are really such, although 

 the posterior looks like the apex of a V-shaped external cusp. In Perathe- 

 rium the external cusps are smaller than in Didelpliys, and the intermediate 

 Vs so much better developed, that the type is much like that of the Talpidce, 

 in whose neighborhood I originally referred it. 



This leads to a consideration of the question of the homologies of the 

 cusps in the genera of the old order of Insectivora proper, and of the Creo- 

 donta. Mr. St. George Mivart has briefly discussed the question so far 

 as relates to the former group.'' He commences with the primitive quadri- 

 tuberculate type presented by Gynmura and Erinaceus, and believes that 

 the external cusps occupy a successivel}'^ more and more internal position, 

 till they come to be represented by the apices of well-developed Vs, as in 

 the ungulate tj'pes. The Vs are well developed in several families, and in 

 Chrysochloris the two Vs are supposed to be united and to constitute almost 

 the entire apex of the crown, while in Centetes the same kind of a V forms a 

 still larger part of the crown. 



I believe that tliese conclusions must be modified, in the light of the 

 characters of various extinct genera, and of the genus Didelpliys. In the 



'Journal of Anatomy and Physiology, II, 138 figures. 



