632 Bulletin American Museum of Natural History. [Vol. XIX, 



The species does not indicate a later phase than the Upper 

 Miocene, for the writer found, in the Clarendon locality, 

 several specimens, including a skull with complete dentition, 

 which are indistinguishable from Protohippus lenticularis Cope. 

 The second species, as already mentioned, is indistinguishable 

 from some species of Protohippus. The third species of this 

 group, Hipparion eiiry stylus Cope, was founded on lower teeth 

 and is not distinguishable from specimens found in the Clar- 

 endon locality. It is quite possible that Cope's type of this 

 species represents the lower dentition of Neohipparion lenti- 

 cularis, with which it corresponds in size. 



Thus it will be seen that the palaeontological evidence at 

 hand not only fails to prove a new horizon for the so-called 

 Goodnight beds, but, on the contrary, seems to prove con- 

 clusively that they are identical in age with the beds in the 

 vicinity of Clarendon, which Cope recognized as Loup Fork 

 (Upper Miocene) deposits. 



There is little doubt, then, that there is no break at Mulberry 

 Canon, either in strata or fauna. Hence the Goodnight beds, 

 as a new horizon, should be abandoned. 



Upper Miocene. 



Clarendon Beds.^ ''Loup Fork " Stage. 



The deposits at the locality north of Clarendon belong un- 

 doubtedly to the Miocene epoch. The fauna indicates a close 

 relationship with the Loup Fork formation. 



This locality is east of the Staked Plains proper, but con- 

 nected with them through long, low divides. The fossil-bear- 

 ing strata do not, however, follow these divides back to the 

 Plains, and is it impossible to say, owing to so much of the 

 country in that direction being covered by recent deposits 

 which are now more or less grass-covered, whether they ex- 

 tended any great distance to the westward or not. The 



' Local name proposed by the writer. The name Loup Fork was first proposed as 

 a formation name, and its subsequent extended use has given rise to so much con- 

 fusion that it seems better not to employ it in the sense of a time division, but to limit 

 its use to the formation occurring in the Loup River, Niobrara, and White River val- 

 leys for which it was originally used. The Clarendon beds are of approximately the 

 same age as the Loup Fork beds, as judged by the known fauna, but cannot be re- 

 garded as a part of the same terrane, and they differ considerably in structure and 

 composition. 



