62 2 Bulletin Afnerican Museum of Natural History, [Vol. XIX, 



lion of the writer to discuss, in some detail, only the more 

 important localities with the purpose of correcting some 

 obvious errors made by Cummins, both in the distribution 

 and correlation of these beds, and to present briefly some of 

 the writer's observations and conclusions which are at variance 

 with those of both Cope and Cummins. These conclusions are 

 briefly as follows : 



(i) The fossil-bearing formations are fluviatile, not lacus- 

 trine, in origin. 



(2) The Blanco has a limited distribution. 



(3) The Goodnight (Paloduro) Beds are not a valid divi- 

 sion. 



(4) The principal deposits forming the Staked Plains are of 

 the Miocene epoch but older than the Loup Fork stage. 



(5) The fossil-bearing beds in the locality north of Clarendon 

 and at Mulberry Canon correspond in age with the Loup Fork 

 formation, as shown by the fossils. 



Pleistocene. 



Rock Creek Beds} = Sheridan {Equus) Beds. 



The Sheridan or Equus division of the Pleistocene beds, as 

 reported by Cummins, are extensively exposed, especially 

 along the south side o'f Tule Canon and at the head and on 

 both sides of Rock Creek, a small tributary emptying into 

 Tule Canon from the south. 



Cummins's section of Tule Canon, taken near the head, is as 

 follows : 



6. Fine white sand 25 feet 



5. Coarse sand, with pebbles 20 " 



4. Bluish clay 15 " 



3. Coarse sand 30 " 



2. Reddish clay 60 " 



I. Triassic sandstone 20 " 



The upper four strata of this section (Nos. 3-6) belong to 

 the Pleistocene age. But the section is representative of these 

 beds only in a general way. Other sections taken at different 



* Local name proposed by the writer. 



