TORRENTIAL ORIGIN OF SANDSTONES AND CONGLOMERATES 293 



must figure largely in the formations of the district when these have 

 been exposed upon a later erosion surface. Such a formation should be 

 bordered by a fringe of coarse breccia-like conglomerate giving the im- 

 pression that it is a basement layer, and perhaps also having a basement 

 layer of the same or similar material. It should be characterized by in- 

 cluded lenticular areas of coarser or finer material and by layers of fine 

 material in sharply defined films and plates which reveal its bedding. 

 If undisturbed from its original attitude, its bedding should be generally 

 horizontal, though with a gentle centrally directed inclination upon its 

 borders, resembling in this the initial dip of a marine formation. Lo- 

 cally, at least, it may show a type of cross-bedding like that represented 

 in figure 3, which, though often observed in ancient sandstones, is not 

 adequately explained by the changing currents along a marine shore.* 

 In its finer-textured portions a rectangular jointing will be likely to be 

 found characteristic, and minor faulting dating from a period when the 

 material was only slightly compacted is possible (see figure 3). The 

 great variety of rock type, the range in dimensions of included pebbles 

 and boulders, and the frequency of faceted forms among them may sug- 

 gest an origin of the formation through glaciation, as has been true of 

 the massive conglomerates of the original Huronian of Canada. 



Having in mind the fact that arid conditions prevail today over about 

 three-fifths of the earth surface, the study of the torrential, play a, and 

 eolian deposits, which are characteristic of desert regions, must be 

 brought into consideration before an adequate explanation can be found 

 for the masses of sandstone and conglomerate, 1,000 to 1,500 feet and 

 more in thickness, which' exist within the ancient rock formations of the 

 globe. It is but natural that the first explanation of these formations 

 should have been based upon geological processes which are most famil- 

 iar — those concerned in the erosion of the land areas under humid 

 conditions with deposition along the ocean littoral. Marine sandstones 

 should, however, be relatively thin; for it would appear that during a 

 transgression of the sea on the land the formation of sand should be 

 limited to a depth not far below the wave base — a depth measured in tens 

 rather than in hundreds or thousands of feet. To meet this difficulty, 

 the theory of Hall, that depression is in areas of deposition adjusted to 

 the material deposited, has been greatly strained. 



The dominance of ripple marks and the paucity of marine fossils just 

 where marine life should have been most abundant are facts difficult to 



* Such a structure is shown in great perfection over considerable areas by the 

 Cambrian sandstone exposed in the picturesque "Dalles" of the Wisconsin river near 

 Kilbourn City, Wisconsin. 



XXVI — Bull. Geol. Soc. Am., Vol. 17, 1905. 



