196 WANLESS— LITHOLOGY OF WHITE RIVER SEDIMENTS. 



The essential difference between the composition of the caliche 

 levels and that of the other parts of the clay series is in the proportion 

 of calcareous cement. Samples of the upper and lower zones of rusty 

 nodules analyzed gave, respectively, 51 and 37 per cent, soluble cal- 

 careous cement. A nodular layer (not caliche) of more sandy char- 

 acter, but of uncertain origin, in the upper 50 feet of the Oreodon 

 beds, exposed only in the Sheep Mountain sections, contains 30 per 

 cent, of soluble cement. The average amount of soluble cement of 

 the clays of the White River series is from about 5 per cent, in the 

 Titanothere beds to 15 per cent, in parts of the Oreodon series. The 

 caliche nodular layers were doubtless cemented at or near the surface 

 by ground water rising by capillary attraction and depositing the cal- 

 cium carbonate when it evaporated. Thus the bones of animals which 

 happened to be near the surface at this time were protected from dis- 

 integration. The clay forming the nodular layers was originally a 

 fine-grained flood-plain deposit which persisted as a land surface for 

 a considerable period of time and probably indicates a period of 

 greater aridity, by reason of the breaks in deposition, evidence of 

 evaporation, etc. The nodular layer has in the presence of coprolites 

 of carnivorous animals and rodent-gnawed bones absolute proof of 

 subaerial deposition. Many of the bones found are also of a decid- 

 edly weathered appearance much as are bones of modern animals 

 weathering on the surface of the Badlands today and occasionally 

 buried in silt. Further proof of subaerial deposition was recently 

 found during the preparation of a Ccenopus skull collected in the 

 lower zone of rusty nodules, a large number of casts of insect larval 

 burrows being found within and around the skull, suggesting that it 

 had lain on the surface for some time before burial and the meat had 

 been devoured by the scavenging insects, which subsequently bur- 

 rowed for pupation in the mud investing the skull. It has not been 

 possible yet to determine to what form these insects belonged. The 

 great continuity of the nodular layers, which has been mentioned 

 above, implies a very level surface controlling evaporation, which was 

 cut here and there by shallow winding stream channels, which may 

 have carried running water only in flood time. A uniform climatic 

 factor, probably a period of greater aridity, was one of the most im- 



