314 R. A. DALY OSCILLATIONS OF LEVEL 



profile." Many vertical pot-holes were found in the walls of the "deeps," 

 enforcing the conclusion that the river current was very powerful during 

 the excavation of these remarkable depressions (see plate 14). 



Mathews believes that the cutting of the "deeps" must be "post- 

 Talbot in age," continuing well into the Eecent period, and that it was 

 induced by increase of the river^s volume or by constriction of its volume. 

 Constriction of the channel seems to have been a necessary condition, 

 because the "deeps" are much narrower than the general rock-floor of 

 the channel and are all located near the east bank of the river. On the 

 other hand, mere increase of volume is not likely to account for the 

 forms and localization of the "deeps." If, however, the river's velocity 

 were considerably greater a short time ago, the "deeps" might have been 

 then excavated, in a way analogous to the cutting of channel grooves by 

 torrents. 



Higher velocity for the stream might have been produced by general 

 lowering of sealevel when the Wisconsin ice-cap was formed ; but, since the 

 rock-floor of the Susquehanna is intact below the most southerly "deeps," 

 this cause alone seems inadequate for the deep excavation. A more 

 probable supposition is that the river gradient was temporarily steepened 

 by crustal uplift. The writer is, therefore, inclined to connect this 

 peculiar drainage history with the general hypothesis of marginal warp- 

 ings connected with Wisconsin glaciation. If the peripheral bulge had 

 its crest in Pennsylvania, the river's volume may have been smaller, but 

 its velocity must have been greater than it is now, after the bulge has 

 subsided. The volumes of rock removed, the recency and great depth 

 of the excavations, and the details of form all seem to accord with the 

 assumption that the "deeps" were made on the southerly slope of such 

 an upwarp. 



DRAINAGE CHANGES IN THE OHIO VALLEY AND FARTHER WEST 



The foregoing suggestions of marginal warping have to do with the 

 Wisconsin stage of glaciation. West of the Hudson Eiver the correspond- 

 ing isobase for zero has not been definitively mapped ; hence local criteria 

 for the truth of Jamieson's hypothesis are not easily devised in the larger 

 part of the peripheral belt. However, in western Pennsylvania and be- 

 yond, there are not wanting evidences of widespread and rather drastic 

 rearrangements of drainage during the pre-Wisconsin part of the Glacial 

 period. Are these changes genetically associated with crustal upwarps 

 marginal to one or more of the pre-AVisconsin ice-caps? 



In the basin of the Ohio the changes are registered in valleys now 

 abandoned by the rivers that cut them and in many detrital terraces 



