146 Bulletin American Aluseiim of Natural History. [Vol. XII, 



symphysis is of much less extent and the jaw is deeper. The 

 facial portion of the skull in Oxycenodon is longer. 



I have suggested ' that this family has some affinities with the 

 Seals. The skull Oxycenodon strengthens this view in certain 

 particulars. The most important resemblances to the Sea Lion 

 are seen in (i) the great postorbital construction of the -skull ; 

 (2) the pressure of a distinct lachrymal tubercle ; (3) the ex- 

 clusion of the fro-ntal from any share in the anterior boundary 

 of the orbit ; (4) early union of lachrymal with maxillary ; (5) 

 forward extension of sagittal crest ; (6) absence of anterior 

 process of frontal between nasal and maxillary, and (7) an 

 elongated mandibular symphysis. While the skull is not strik- 

 ingly seal-like in its general appearance yet when we know the 

 immediate ancestors of such a form as the Sea Lion it will not be 

 surprising if it is found to connect with such a form as 

 Oxycenodo7i. 



Family MESONYCHID^. 



The Mesonychidffi undoubtedly represent a distinct line of 

 the primitive Carnivora whose later representatives came to be 

 characterized by elongated and specialized limbs for a running 

 habit, as well as by a peculiar reduction and simplification of cer- 

 tain cusps of the molar teeth. Somewhat contrary to the generally 

 accepted arrangement of this group, I unite with them the so- 

 called Triisodontid^ as a well marked subfamily. This latter 

 series is the older and displays the same tendency towards the 

 pecu'liar reduction and rounded, conical form of the cusps of the 

 molars as is seen in the Dissacus-Pachyana-Mesonyx line. The 

 differences between the subfamilies are not great and, as observed 

 by Matthew,' consist in the deep, heavy jaws with powerful sym- 

 physis, and the much wider, more distinctively tubercular charac- 

 ter of the molars in the Triisodontidae. 



The single genus Triisodon, with three species, comes from the 

 Puerco and is succeeded in the Torrejon by the _two genera 

 Sarcothraustes and Goniacodon. Matthew has pointed out that 

 Sarcothratistes antiquus is without doubt the direct descendant of 

 Triisodon and in this opinion I entirely agree. As noted by this 



' ' Osteology of Patrio/eliSy' Bull. Amer. Mus. Nat. Hist., Vol. IV, 18(54, P- iS7- 

 Bull. Am. Mus. Nat. Hist., Vol. IX, i8g7, p. 278. 



