146 EUGENE WESLEY SHAW 



old divides ; (c) material brought after the ice melted, by tributaries 

 as they cut new valleys. Second, a decrease in velocity and carry- 

 ing power, produced by (a) the attraction of the ice mass; within 

 a degree of the ice front this may so have changed water level that 

 in a stream flowing away from the ice a gradient of if feet per mile 

 might have been reduced to i^ or ij feet per mile; (b) crustal 

 deformation, due to the weight of the ice; (c) the divides crossed; 

 each of these would check the velocity and cause deposits for a 

 short distance upstream; and ice jams operate in a similar way. 

 Third, a possible but not probable decrease in volume, arising from 

 a change in climate. It is probable that during Kansan time the 

 river had a larger volume than now because it was carrying the 

 run-off from a much larger territory. 



TERRACES OF TRIBUTARY RIVERS 



The second group of high-terrace deposits is found on streams 

 tributary to the Allegheny and Ohio. Those along the Clarion 

 River may be taken as typical and described in detail. 1 At Fox- 

 burg the high gravels of the Clarion connect and mingle with those 

 of the Allegheny, both the rock floors and the upper surfaces of the 

 deposits connecting, without interruption (see Fig. 3). The Clarion 

 gravels are much like those of the Allegheny, but differ from them in 

 the following respects: First, the material is of local, not glacial, 

 origin. Second, the thickness decreases upstream. Third, the 

 gravels are as a whole much finer, only the base being as coarse 

 as the glacial gravels. There are some minor points of dissimilarity, 

 but these are the important ones in the present discussion. 



In present distribution the gravels are as continuous as those 

 of the Allegheny. There is scarcely a half-mile of the lower part 

 of the valley where they are absent or even approximately so. 



That the Clarion terraces are of stream origin is shown by char- 

 acters similar to and as decisive as those of the Allegheny terraces 

 mentioned above, and certain important features indicate the imme- 

 diate cause of the accumulation of gravel. First, at the confluence 

 of the two rivers the high-terrace gravels correspond exactly in 

 elevation and thickness. Second, the Clarion gravel rises and 



1 See Foxburg-Clarion folio, U.S. Geol. Surv. (in press). 



