FORT UNION OF CRAZY MOUNTAIN FIELD, MONT. 171 



characterized by some strangely ungulatelike limb characters. The 

 Miacidae had typical carnassials as in fissipedes and were generally 

 progressive and adaptive. 



All five groups are rather abundantly represented in the Torrejon 

 in individuals, although only the Oxyclaeninae there show much 

 variety in genera and species. In the present fauna the Oxyclaeninae 

 are also abundant and varied, relatively about as in the Torrejon, 

 although all the species and most of or all the genera are here different. 

 The Arctocyoninae are not abundant but are apparently more varied 

 than in the Torrejon. Triisodonts, common Torrejon fossils, are 

 absent in this fauna as now known, and the mesonychids are repre- 

 sented only by extremely rare fragments. Miacids, on the other hand, 

 are present and are more varied than in the Torrejon. 



Family ARCTOCYONIDAE Murray, 1866 '' 



This is one of the groups so largely and adequately defined and dis- 

 cussed in Matthew's memoir that redefinition here is quite unneces- 

 sary. The genera placed in the Arctocyonidae have commonly been 

 distributed in the Oxyclaenidae, Triisodontidae, and Arctocyonidae 

 since Scott (1892) defined the first two families. 



Osborn and Earle (1895) also proposed a family Chriacidae, but this 

 was rather a substitution for Oxyclaenidae (because they considered 

 Oxyclaenus, proper, as incertae sedis) than a separation from it. 

 Matthew (1897) provisionally proposed the use of Chriacidae for 

 Chriacus, "Protochriacus" {Loxolophus) , and Tricentes if, as he then 

 suspected, Oxyclaenus were referable to the Triisodontidae. The 

 latter step was not taken. Oxyclaenus and Triisodon were eventually 

 placed by Matthew in the same family, Arctocyonidae, but Chriacus 

 was also placed there. The earlier work adumbrated a fourfold 

 division, with groups typified by Oxyclaenus, Chriacus, Arctocyon, and 

 Triisodon. Various of these were at times separated mdely, but the 

 way in which some genera were shifted from one to another and all 

 sorts of combinations made shows how hard it really is to tell these 

 groups, or supposed groups, apart. 



In his latest work, Matthew (Pale. Mem.) took the logical step of 

 reuniting all these genera under the oldest family name, Arctocyonidae. 

 It seems to be demonstrated that all are rather closely related and 



» I would prefer to give, and in some earlier publications have given, as author of a family the first writer 

 who recognized the group and gave it a name based on a valid genus, even if he did not follow the family 

 form now maintained. This would make Giebel, who named the Arctocyoninae in 1855, author of the 

 Arctocyonidae. In fact he was, aside from quibbling, for his group Arctocyoninae was distinguished from 

 nonarctocyonids, not from other arctocyonids (none of which were then known) and was, as far as then possible, 

 the group we now call Arctocyonidae. In reality, then, Giebel is the author of this family, but the bibliog- 

 raphers will not have it so and, of course, they are correct in the letter of the law, if not in a spirit of justice. 

 On the grounds of literal correctness and largely of feeling that the purpose of quoting authority is not to 

 honor but only to define, I have abandoned my former practice. This statement applies to a number of 

 other groups as well as to this. 



