184 TEN YEARS^ PROGRESS IN VERTEBRATE PALEONTOLOGY 



RESULTS GENERALLY ACCEPTED 



1. The Eocene Carnivora comprise a number of groups of divergent 

 specialization analogous in large part to the modern Carnivora. From 

 one of these groups (Miacidae) are derived all the later Tertiary and 

 modern Carnivora by a second adaptive radiation. This confirms the 

 views set forth by Scott and Osborn twenty years ago. 



2. The Eocene Carnivora are not intermediate between Placental and 

 Marsupial Carnivora, but are true placentals. The marsupial resem- 

 blances are due in part to inheritance of common primitive characters 

 from a very remote common ancestry, but chiefly to parallel, or even 

 convergent, evolution. The opposite view held fifty years ago by Gau- 

 dry, and more recently revived by Ameghino, Lydekker, and Wortman, 

 is not substantiated by the evidence. 



3. The Felidas are not derived from Palceonictis, but from some un- 

 known Eocene genus of Miacidse. 



4. The various families of Carnivora, both Creodonta and Fissipedia, 

 are mostly common to the Palgearctic and Nearctic regions, but some are 

 more abundant and varied in one, some in the other division. The 

 Viverridae are exclusively Palsearctic, the Procyonidae Nearctic, until the 

 later Tertiary, when each family invaded the regions to the south of it. 



5. The African Oligocene Carnivora are all Hvienodontida^. Xo evi- 

 dence of other Creodont families or of modernized Carnivora. They 

 are more primitive than their European contemporaries, but some are of 

 divergent specialization. 



6. The middle and later Tertiary Carnivora are nearly allied in the 

 Palaearctic and Xearctic regions, and their evolution is partly upon par- 

 allel lines, partly by successive new t}^es appearing from some common 

 center of dispersal. The families had already become established in the 

 Miocene, but in the Oligocene it is difficult to distinguish the phyla 

 except upon evidence of the complete skeleton. The Oligocene Car- 

 nivora are mostly synthetic types, mingling characters which have since 

 become peculiar to different families. The modern genera were estab- 

 lished in the Pliocene and Pleistocene. 



HYPOTHESES UNDER CONSIDERATION 



1. Many of the basal Eocene Creodonts ( Oxyclienidae, Triisodontidae) 

 are of doubtful affinities. They may be Insectivora or of other orders. 



2. The Mesonychidae are not closely related to the rest of the Creo- 

 donta and show a considerable fundamental resemblance to the Eocene 

 Artiodactyla. A revision of their relations is necessarv. 



