1 98 Bulletin American Museum of Natural History. [Vol. XIX, 



faunas as do the Oligocene deposits of the plains. They con- 

 tain an important aquatic contingent, fish, crocodiles, and 

 water-turtles being comparatively abundant. In the White 

 River fauna .all these are absent, except in the sandstone 

 lenses,^ while a large element of it is apparently adapted to 

 open grassy plains; this is not found in the Eocene faunas. 

 But in the three zones of the White River a great part of 

 their respective faunas appears to be in direct and exact gen- 

 etic succession. We can therefore measure the amount and 

 direction of change during the Oligocene epoch in many series. 



The amount of evolution as thus measured appears small, 

 but its direction somewhat constant. The species of the 

 Titanotherium Beds are all distinct from their successors in 

 the Oreodon Beds, but the difference is uniformly small. 

 Between the Oreodon and Leptauchenia faunas the difference 

 is often greater but less uniform, so far as present data go. 

 Some genera run through the three horizons {e. g., Cynodictis , 

 PalcBolagiis, Mesohippus, CcBiwpus, Leptomeryx) . Others have 

 been found only in the two lower zones or in the two upper 

 zones, while many are as yet known from one horizon only. 



Stratigraphy. — Mr. Douglas refers all the Tertiary at this 

 locality to one stage, correlating it with the Titanotherium 

 Beds of South Dakota. We find, however, a lithologic dis-, 

 tinction between the higher beds exposed north of the rail- 

 road, which resemble the Oreodon Beds of South Dakota, 

 Colorado, and elsewhere, and the lower beds exposed south 

 of the railroad, which resemble rather the Titanotherium 

 Beds of some parts of South Dakota. Likewise on Thomp- 

 son's Creek, not far from the Pipestone locality, we were able 

 to distinguish between the Oreodon Beds exposed near the 

 head of a small northerly branch of the creek, and the Titan- 

 otherium Beds exposed on the main western branch. At 



^ Mr. Douglas has recently discovered fish remains in strata which he refers to the 

 White River epoch, so-called, in the Madison vaflley in Montana. But these strata are 

 ■quite different in character from the beds in which White River mammals are found, 

 apparently lacustrine or fluviatile in origin, and a very thorough search on his part 

 failed to reveal any mammals in them except a skull of the beaver Sieneofiber. I do not 

 understand that he considers them as of the same formation or origin as the mammal 

 beds, but merely as of equal age. The discovery of fish in them, therefore, does not at 

 all invalidate the fluviatile-eolian hypothesis of origin of the White River formation 

 maintained by Hatcher and myself. The same explanation probably applies to other 

 reported occurrences of fish in the White River. 



