372 BULLETIN 120, UNITED STATES NATIONAL MUSEUM. 



Oxyglossus in Permian times in India and its spread from India 

 through Africa and Europe to Greenland and on to western North 

 America, and the persistence of the genus in India and Malaysia 

 until to-day would be most remarkable and hardly believable. The 

 fossil form in Wyoming, assigned to the genus Oxyglossus is said 

 by Schuchert (1915?) to be of the Comanchian period, though it was 

 earlier assigned to the Eocene. Accepting the idea of land connec- 

 tion between all continents in the Permian and the existence of North 

 Atlantis and South Atlantis both in the early Triassic and the early 

 Jurassic, we must ask why the Baninae, if already evolved in any 

 of these periods, did not make more general use of the bridges, and 

 how it happens that only Oxyglossus succeeded in crossing from one 

 hemisphere to the other. In this connection, however, we should re- 

 member that Raninae have lived in northern South America ap- 

 parently since early Cretaceous times, but have not spread to the rest 

 of that continent. 



The presence of Oxyglossus in North America at any time would 

 be remarkable, for to-day there are in all Arctogea (omitting Asia 

 south of the Himalayas, and Africa, both originally portions of 

 South Atlantis) no Raninae except of the two genera Rana and 

 Polypedates. It seems difficult to accept the identification of the 

 Wyoming fossil as an Oxyglossus or even as a Ranid. 



If the Wyoming fossil were a Miocene Rana there would be no 

 difficulty, or we might possibly reconcile the presence of a representa- 

 tive of the Gastrophrynidae, but a Comanchian Oxyglossus seems in 

 itself most strange and also seems far astray in Wyoming. I make 

 no attempt to resolve the puzzle, but must express doubt of the 

 identification of the fossil as an Oxyglossus^ and I may be allowed to 

 confess to some doubt of the Comanchian reference of the strata if 

 the fossil is a Ranid of any genus. 



The South American genera of Raninae {Phyllohates, Prosthe- 

 rapis, Phyllodromus, Calosthetis, Hylixalus) probably reached 

 northern South America along with the Dendrohatinae in the early 

 Cretaceous (fig. 234). 



The Dendrobatinae (fig. 231, p. 295). 



The distribution of this subfamil}^ suggests origin in South At- 

 lantis in early Cretaceous times (fig. 234), making its evolution about 

 contemporaneous with that of another branch of the frog family, 

 the Raninae. 



GENERAL CONSIDERATIONS. 



Having in mind now the chief indications as to the dates and 

 places of origin and the dates and routes of the spreading of the 

 several groups of Opalinidae and their Anuran hosts, we may profit- 

 ably discuss a few further points. 



