172 BULLETIN 16 9, UNITED STATES NATIONAL MUSEUM 



even if the supposed groups are separable, they do not all together 

 exceed the usual limits of a single family of carnivores. 



Matthew subdivides the family into Oxyclaeninae, Chriacinae, 

 Arctocyoninae, and Triisodontinae. These are essentially the old 

 families, except that the Chriacidae of Osborn and Earle and (tenta- 

 tively suggested, not adopted) of Matthew was merely the Oxyclaeni- 

 dae with Oxyclaenus excluded, while the Chriacinae of Matthew's 

 last work is based on quite a different concept and includes only 

 Chriacus and Deltatherium. 



Despite the name of the family, its central, most varied, and most 

 typical group is that of the Oxyclaeninae. Separation of the Trii- 

 sodontinae seems justified, since these animals (not yet identified in 

 the present fauna) are of a peculiar adaptive type only the earliest 

 examples of which show close approach to oxyclaenines. The Arcto- 

 cyoninae form a clear-cut group if contrasted with such types as 

 Oxyclaenus or Chriacus, but such genera as Protogonodon, Tricenies, 

 Arctocyonides, or Thryptacodon, each in a different way, tend to 

 bridge the morphological gap and to make clear differentiation difficult 

 or impossible. Perhaps in the fanlike radiation of this potent and 

 extremely varied family several lines approached a beariike, omniv- 

 orous adaptive type and the Arctocyoninae may be a partly artificial 

 concept uniting several of the more extreme adaptive types inde- 

 pendently trending in tliis direction. Despite this possibility, the 

 concept is a practical one and may well be adopted pending a better 

 understanding of the actual phylogeny. 



The idea of separating Chriacus from the Oxyclaeninae seems less 

 fortunate, and it is probably impractical at present. The Arcto- 

 cyonidae -wdth the Triisodontinae and Arctocyoninae removed are 

 a hodgepodge including many different lines each potentially or 

 actually as distinct as that suggested by Chriacus. These numerous 

 minor phjda are so intricately interrelated and most of them are so 

 poorly known that a good subfamily or supergeneric arrangement is 

 not now attainable. It does not seem helpful to separate one genus, 

 Chriacus, which is probably no more distinctive than each of a dozen 

 others. From another viewpoint the inadequacy of such an arrange- 

 ment is also shown by the discovery of such a type as Metachriacus , 

 which might roughly be characterized as "chriacine" in premolars 

 and "oxyclaenine" in molars. With the greatest respect for Mat- 

 thew's incomparable knowledge of these faunas and clearness of 

 judgment it further seems to me that his collocation of Deltatherium 

 is based on superficial characters and that this genus differs more from 

 Chriacus than does any of several genera not placed in the Chriacinae. 



The present fauna contains certainly six and possibly seven genera 

 of Arctocyonidae. Five of these were defined from this fauna and 

 have not definitely been recognized elsewhere. Of the Crazy Moun- 



