WANLESS— LITHOLOGY OF WHITE RIVER SEDIMENTS. 195 



times in the form of algal ball levels made up of a series of flattened, 

 oval-shaped balls which are often quite persistent. Elsewhere thin 

 sheets of silicified limestones occur in which the original organic char- 

 acter has been almost destroyed by replacement by secondary silica. 

 Such a sheet as this is sometimes quite persistent at the contact of the 

 Oreodon and Titanotherium beds and, as elsewhere indicated, was 

 used by Darton in defining the contact plane of these two horizons. 

 A third type of limestone forms a lens-shaped solid sheet a few acres 

 in extent. This type develops a limestone as much as 3 feet thick 

 and is quite rich in organic remains, especially the shells of cyprids. 

 This latter type undoubtedly represents a small pond on the surface 

 of the flood-plain. Small pelecypods and gastropods are visible in 

 this, but the fauna has not yet been described. It is believed that one 

 reason for the silicification of the limestone is the solution of the 

 siliceous shells of diatoms and replacement of the lime by this dis- 

 solved silica and also the solution of the limestone by ground water 

 and replacement by colloidal silica. The fact that the limestone was 

 in large measure formed by the action of fresh-water algae was first 

 suggested by Sinclair in 1920 for the algal ball type of limestone.^" 

 The presence of algae has since been recognized in the other types of 

 limestone mentioned above. The amount of lime in these beds is very 

 variable, from 5 per cent, in the silicified cherty layer at the base of 

 the Oreodon beds to 95 per cent, in one of the massive lenses of 

 limestone. 



The detrital material in the limestones showed the presence of 

 fragments of quartz, biotite and tourmaline, and of pumice needles 

 and angular volcanic glass fragments, suggesting that the detrital 

 deposition in the ponds was probably by wind. It may be that the 

 thinner sheets of algal limestone do not represent pond deposits, but 

 were formed in moist meadow lands under prairie conditions. 



Nodular Layers. 

 The presence of nodular layers is of great interest to the palae- 

 ontologist, as it is in them that vertebrate remains are best preserved. 

 In some localities the nodules occur more or less isolated, but more 

 often they are found as broad sheets and as such have a remarkable 

 continuity. 



10 Amer. Philos. Soc, Vol. 40, 1921, p. 460. 



