716 G. SHERBURNE ROGERS 
their polish. Finally, while relatively little detailed study of the 
mineral composition has been made, we have a fairly definite idea 
of the minerals which commonly make up clastic rocks, and, in the 
ease of feldspar at least, some conception of its genetic significance. 
It is the belief of the writer that the three lines of research men- 
uioned in the preceding paragraph, viz., upon the broad structural 
features, the shape of the grains, and the mineral character of the 
grains, may profitably be combined, and that the value of the 
results obtained by any one is greatly enhanced when considered in 
relation to the other two. The importance of a close study of the 
structural features of a clastic rock has been amply demonstrated 
by Barrell and others. The criteria given by Scherzer, when it is 
possible to apply them, are of similar value in determining the 
manner of deposition of the sediment and something of its history 
prior to deposition. Finally, the mineralogical composition of the 
rock, though in many cases having no apparent significance, is in 
others decidedly important. Wnen a sandstone is made up 
practically entirely of quartz, as is the St. Peter sandstone, the 
significance of this fact is immediately recognized and taken into 
account in considering its history and derivation.t Thus, C. P. 
Berkey considers the St. Peter sandstone of eolian (sand dune) 
origin, because of the fact that it contains 98 to 99 per cent of 
quartz, largely in rounded grains of a nearly uniform size. Certain 
formations in the east, such as the Newark group, have been 
assigned a continental origin partly because of their arkosic charac- 
ter and red color, and this also illustrates a partial application of 
these principles. When the composition is less apparent to the 
naked eye, however, the importance of its determination may not 
be so evident; and very little is known of the exact mineral con- 
stituents of the great thickness of Cretaceous and Tertiary strata 
in the Great Plains and Rocky Mountain provinces. 
Aside from the importance of such data in determining the 
history and origin of a rock they may in some cases have a distinct 
correlative value. Thus, in the instance of the Lebo shale member 
of the Fort Union formation described below, the writer was able to 
«See “‘Paleogeography of St. Peter Time,” Bull. Geol. Soc. Amer., XVII (1906), 
229-50. 
